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. PPv2 support was
29 updated based on HAProxy spec change in May 2014.
31 JH/02 Add ${listextract {number}{list}{success}{fail}}.
33 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
34 Properly escape header and check for NULL return.
36 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
39 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
41 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
42 characters in header names, implemented as a verify condition.
43 Contributed by Michael Fischer v. Mollard.
45 TL/05 Rename SPF condition results err_perm and err_temp to standardized
46 results permerror and temperror. Previous values are deprecated but
47 still accepted. In a future release, err_perm and err_temp will be
48 completely removed, which will be a backward incompatibility if the
49 ACL tests for either of these two old results. Patch contributed by
50 user bes-internal on the mailing list.
52 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
54 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
55 selectors, in both main and reject logs.
57 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
60 JH/07 Add malware type "sock" for talking to simple daemon.
62 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
64 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
65 routers/transports under cutthrough routing.
67 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
68 numbers. Touch up "bool" conditional to keep the same definition.
70 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
72 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
74 JH/12 Expand items in router/transport headers_add or headers_remove lists
75 individually rather than the list as a whole. Bug 1452.
77 Required for reasonable handling of multiple headers_ options when
78 they may be empty; requires that headers_remove items with embedded
79 colons must have them doubled (or the list-separator changed).
81 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
82 view the policy declared in the DMARC record. Currently, $dmarc_status
83 is a combined value of both the record presence and the result of the
86 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
88 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
89 dnslookup router and the smtp transport (applying to the forward
92 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
93 of ldap servers used for a specific lookup. Patch provided by Heiko
96 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
97 New variable $lookup_dnssec_authenticated for observability.
99 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
100 Patch submitted by Lars Timman.
102 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
104 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
105 Requires trusted mode and valid format message id, aborts otherwise.
106 Patch contributed by Heiko Schlichting.
108 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
109 certextract with support for various fields. Bug 1358.
111 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
112 is requested by default, modifiable by smtp transport option
115 JH/22 Expansion operators ${md5:string} and ${sha1::string} can now
116 operate on certificate variables to give certificate fingerprints
117 Also new ${sha256:cert_variable}.
119 JH/23 The PRDR feature is moved from being Experimental into the mainline.
121 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
122 Christian Aistleitner.
124 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
126 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
127 file. Patch from Wolfgang Breyha.
133 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
135 PP/02 Make -n do something, by making it not do something.
136 When combined with -bP, the name of an option is not output.
138 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
141 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
142 $sender_host_name and config options to manage this, and basic check
145 PP/05 DSCP support for outbound connections and control modifier for inbound.
147 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
148 (Only plugin which currently uses this is kerberos4, which nobody should
149 be using, but we should make it available and other future plugins might
150 conceivably use it, even though it would break NAT; stuff *should* be
151 using channel bindings instead).
153 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
154 name; added for Sendmail compatibility; requires admin caller.
155 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
156 just ignore it); requires trusted caller.
157 Also parse but ignore: -Ac -Am -X<logfile>
160 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
162 TL/02 Add +smtp_confirmation as a default logging option.
164 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
165 Patch by Magnus Holmgren from 2007-02-20.
167 TL/04 Bugzilla 1281 - Spec typo.
168 Bugzilla 1283 - Spec typo.
169 Bugzilla 1290 - Spec grammar fixes.
171 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
173 TL/06 Add Experimental DMARC support using libopendmarc libraries.
175 TL/07 Fix an out of order global option causing a segfault. Reported to dev
176 mailing list by by Dmitry Isaikin.
178 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
180 JH/02 Support "G" suffix to numbers in ${if comparisons.
182 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
184 NM/01 Bugzilla 1197 - Spec typo
185 Bugzilla 1196 - Spec examples corrections
187 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
189 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
190 gnutls_enable_pkcs11, but renamed to more accurately indicate its
193 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
194 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
196 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
197 "acl {{name}{arg}...}", and optional args on acl condition
200 JH/05 Permit multiple router/transport headers_add/remove lines.
202 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
204 JH/07 Avoid using a waiting database for a single-message-only transport.
205 Performance patch from Paul Fisher. Bugzilla 1262.
207 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
210 JH/09 Add $headers_added variable, with content from use of ACL modifier
211 add_header (but not yet added to the message). Bugzilla 199.
213 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
214 Pulled from Bugzilla 817 by Wolfgang Breyha.
216 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
218 (nb: this is the same fix as in Exim 4.80.1)
220 JH/11 Add A= logging on delivery lines, and a client_set_id option on
223 JH/12 Add optional authenticated_sender logging to A= and a log_selector
226 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
228 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
229 advertise SMTP AUTH mechanism to us, instead of a generic
230 protocol violation error. Also, make Exim more robust to bad
231 data from the Dovecot auth socket.
233 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
235 When a queue runner is handling a message, Exim first routes the
236 recipient addresses, during which it prunes them based on the retry
237 hints database. After that it attempts to deliver the message to
238 any remaining recipients. It then updates the hints database using
241 So if a recipient address works intermittently, it can get repeatedly
242 deferred at routing time. The retry hints record remains fresh so the
243 address never reaches the final cutoff time.
245 This is a fairly common occurrence when a user is bumping up against
246 their storage quota. Exim had some logic in its local delivery code
247 to deal with this. However it did not apply to per-recipient defers
248 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
250 This change adds a proper retry rule check during routing so that the
251 final cutoff time is checked against the message's age. We only do
252 this check if there is an address retry record and there is not a
253 domain retry record; this implies that previous attempts to handle
254 the address had the retry_use_local_parts option turned on. We use
255 this as an approximation for the destination being like a local
256 delivery, as in LMTP.
258 I suspect this new check makes the old local delivery cutoff check
259 redundant, but I have not verified this so I left the code in place.
261 TF/02 Correct gecos expansion when From: is a prefix of the username.
263 Test 0254 submits a message to Exim with the header
267 When I ran the test suite under the user fanf2, Exim expanded
268 the header to contain my full name, whereas it should have added
269 a Resent-Sender: header. It erroneously treats any prefix of the
270 username as equal to the username.
272 This change corrects that bug.
274 GF/01 DCC debug and logging tidyup
275 Error conditions log to paniclog rather than rejectlog.
276 Debug lines prefixed by "DCC: " to remove any ambiguity.
278 TF/03 Avoid unnecessary rebuilds of lookup-related code.
280 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
281 Bug spotted by Jeremy Harris; was flawed since initial commit.
282 Would have resulted in OCSP responses post-SNI triggering an Exim
283 NULL dereference and crash.
285 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
287 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
288 Bug detection, analysis and fix by Samuel Thibault.
289 Bugzilla 1331, Debian bug #698092.
291 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
293 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
294 Server implementation by Todd Lyons, client by JH.
295 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
296 config variable "prdr_enable" controls whether the server
297 advertises the facility. If the client requests PRDR a new
298 acl_data_smtp_prdr ACL is called once for each recipient, after
299 the body content is received and before the acl_smtp_data ACL.
300 The client is controlled by bolth of: a hosts_try_prdr option
301 on the smtp transport, and the server advertisement.
302 Default client logging of deliveries and rejections involving
303 PRDR are flagged with the string "PRDR".
305 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
306 fclose(). Diagnosis by Todd Lyons.
308 PP/17 Update configure.default to handle IPv6 localhost better.
309 Patch by Alain Williams (plus minor tweaks).
312 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
313 This is now consistent with GnuTLS, and is now documented: the
314 previous undocumented portable approach to treating the option as
315 unset was to force an expansion failure. That still works, and
316 an empty string is now equivalent.
318 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
319 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
320 not performing validation itself.
322 PP/20 Added force_command boolean option to pipe transport.
323 Patch from Nick Koston, of cPanel Inc.
325 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
328 TF/04 Added udpsend ACL modifer and hexquote expansion operator
330 PP/21 Fix eximon continuous updating with timestamped log-files.
331 Broken in a format-string cleanup in 4.80, missed when I repaired the
332 other false fix of the same issue.
333 Report and fix from Heiko Schlichting.
336 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
337 Report from Prashanth Katuri.
339 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
340 It's SecureTransport, so affects any MacOS clients which use the
341 system-integrated TLS libraries, including email clients.
343 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
344 using a MIME ACL for non-SMTP local injection.
345 Report and assistance in diagnosis by Warren Baker.
347 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
349 JH/16 Fix comparisons for 64b. Bugzilla 1385.
351 TL/09 Add expansion variable $authenticated_fail_id to keep track of
352 last id that failed so it may be referenced in subsequent ACL's.
354 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
357 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
358 ldap library initialization, allowing self-signed CA's to be
359 used. Also properly sets require_cert option later in code by
360 using NULL (global ldap config) instead of ldap handle (per
361 session). Bug diagnosis and testing by alxgomz.
363 TL/12 Enhanced documentation in the ratelimit.pl script provided in
364 the src/util/ subdirectory.
366 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
367 renamed to Transport Post Delivery Action by Jeremy Harris, as
370 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
371 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
372 redis_servers = needs to be configured which will be used by the redis
373 lookup. Patch from Warren Baker, of The Packet Hub.
375 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
377 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
378 hostname or reverse DNS when processing a host list. Used suggestions
379 from multiple comments on this bug.
381 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
383 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
384 Missed a few lines, added it to make the runtest require no keyboard
387 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
388 contains upper case chars. Make router use caseful_local_part.
390 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
391 support when GnuTLS has been built with p11-kit.
397 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
399 This, or similar/improved, will also be change PP/11 of 4.82.
405 PP/01 Handle short writes when writing local log-files.
406 In practice, only affects FreeBSD (8 onwards).
407 Bugzilla 1053, with thanks to Dmitry Isaikin.
409 NM/01 Bugzilla 949 - Documentation tweak
411 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
414 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
416 PP/02 Implemented gsasl authenticator.
418 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
420 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
421 `pkg-config foo` for cflags/libs.
423 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
424 with rest of GSASL and with heimdal_gssapi.
426 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
427 `pkg-config foo` for cflags/libs for the TLS implementation.
429 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
430 properties get this fed in as external SSF. A number of robustness
431 and debugging improvements to the cyrus_sasl authenticator.
433 PP/08 cyrus_sasl server now expands the server_realm option.
435 PP/09 Bugzilla 1214 - Log authentication information in reject log.
436 Patch by Jeremy Harris.
438 PP/10 Added dbmjz lookup type.
440 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
442 PP/12 MAIL args handles TAB as well as SP, for better interop with
443 non-compliant senders.
444 Analysis and variant patch by Todd Lyons.
446 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
447 Bug report from Lars Müller <lars@samba.org> (via SUSE),
448 Patch from Dirk Mueller <dmueller@suse.com>
450 PP/13 tls_peerdn now print-escaped for spool files.
451 Observed some $tls_peerdn in wild which contained \n, which resulted
452 in spool file corruption.
454 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
455 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
456 or write after TLS renegotiation, which otherwise led to messages
459 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
460 as a tracking header (ie: a signed header comes before the signature).
461 Patch from Wolfgang Breyha.
463 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
464 comma-sep list; embedded commas doubled.
466 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
468 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
470 Report and patch from Dmitry Banschikov.
472 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
473 Removed SSL_clear() after SSL_new() which led to protocol negotiation
474 failures. We appear to now support TLS1.1+ with Exim.
476 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
477 lets Exim select keys and certificates based upon TLS SNI from client.
478 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
479 before an outbound SMTP session. New log_selector, +tls_sni.
481 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
482 NULL dereference. Report and patch from Alun Jones.
484 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
485 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
486 Not seeing resolver debug output on NetBSD, but suspect this is a
487 resolver implementation change.
489 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
490 Left warnings. Added "eximon gdb" invocation mode.
492 PP/21 Defaulting "accept_8bitmime" to true, not false.
494 PP/22 Added -bw for inetd wait mode support.
496 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
497 locate the relevant includes and libraries. Made this the default.
499 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
500 Bugzilla 1246, report and most of solution from Tomasz Kusy.
502 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
503 This may cause build issues on older platforms.
505 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
506 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
507 gnutls_require_mac & gnutls_require_protocols (no longer supported).
508 Added SNI support via GnuTLS too.
509 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
511 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
513 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
514 Applied second patch from Janne, implementing suggestion to default
515 multiple-strings-in-record handling to match SPF spec.
517 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
519 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
520 read-only, out of scope).
521 Patch from Wolfgang Breyha, report from Stuart Northfield.
523 PP/29 Fix three issues highlighted by clang analyser static analysis.
524 Only crash-plausible issue would require the Cambridge-specific
525 iplookup router and a misconfiguration.
526 Report from Marcin Mirosław.
528 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
530 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
531 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
532 As part of this, removing so much warning spew let me fix some minor
533 real issues in debug logging.
535 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
536 assignment on my part. Fixed.
538 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
539 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
540 Janne Snabb (who went above and beyond: thank you).
542 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
543 string otherwise requires a connection and a bunch more work and it's
544 relatively easy to get wrong. Should also expose TLS library linkage
547 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
548 64-bit ${eval} (JH/03).
550 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
551 GNU libc to support some of the 64-bit stuff, should not lead to
552 conflicts. Defined before os.h is pulled in, so if a given platform
553 needs to override this, it can.
555 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
556 protection layer was required, which is not implemented.
557 Bugzilla 1254, patch from Wolfgang Breyha.
559 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
560 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
561 tls_dhparam take prime identifiers. Also unbreak combination of
562 OpenSSL+DH_params+TLSSNI.
564 PP/39 Disable SSLv2 by default in OpenSSL support.
570 PP/01 Solaris build fix for Oracle's LDAP libraries.
571 Bugzilla 1109, patch from Stephen Usher.
573 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
575 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
578 TF/02 Fix a couple more cases where we did not log the error message
579 when unlink() failed. See also change 4.74-TF/03.
581 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
582 lock up or crash if it happened to be inside a call to libc when it
583 got a SIGUSR1 from exiwhat.
585 The SIGUSR1 handler appends the current process status to the process
586 log which is later printed by exiwhat. It used to use the general
587 purpose logging code to do this, but several functions it calls are
588 not safe for signals.
590 The new output code in the SIGUSR1 handler is specific to the process
591 log, and simple enough that it's easy to inspect for signal safety.
592 Removing some special cases also simplifies the general logging code.
593 Removing the spurious timestamps from the process log simplifies
596 TF/04 Improved ratelimit ACL condition.
598 The /noupdate option has been deprecated in favour of /readonly which
599 has clearer semantics. The /leaky, /strict, and /readonly update modes
600 are mutually exclusive. The update mode is no longer included in the
601 database key; it just determines when the database is updated. (This
602 means that when you upgrde Exim will forget old rate measurements.)
604 Exim now checks that the per_* options are used with an update mode that
605 makes sense for the current ACL. For example, when Exim is processing a
606 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
607 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
608 must specify per_mail/readonly. If you omit the update mode it defaults to
609 /leaky where that makes sense (as before) or /readonly where required.
611 The /noupdate option is now undocumented but still supported for
612 backwards compatibility. It is equivalent to /readonly except that in
613 ACLs where /readonly is required you may specify /leaky/noupdate or
614 /strict/noupdate which are treated the same as /readonly.
616 A useful new feature is the /count= option. This is a generalization
617 of the per_byte option, so that you can measure the throughput of other
618 aggregate values. For example, the per_byte option is now equivalent
619 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
621 The per_rcpt option has been generalized using the /count= mechanism
622 (though it's more complicated than the per_byte equivalence). When it is
623 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
624 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
625 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
626 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
627 Note that using per_rcpt with a non-readonly update mode in more than
628 one ACL will cause the recipients to be double-counted. (The per_mail
629 and per_byte options don't have this problem.)
631 The handling of very low rates has changed slightly. If the computed rate
632 is less than the event's count (usually one) then this event is the first
633 after a long gap. In this case the rate is set to the same as this event's
634 count, so that the first message of a spam run is counted properly.
636 The major new feature is a mechanism for counting the rate of unique
637 events. The new per_addr option counts the number of different
638 recipients that someone has sent messages to in the last time period. It
639 behaves like per_rcpt if all the recipient addresses are different, but
640 duplicate recipient addresses do not increase the measured rate. Like
641 the /count= option this is a general mechanism, so the per_addr option
642 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
643 example, measure the rate that a client uses different sender addresses
644 with the options per_mail/unique=$sender_address. There are further
645 details in the main documentation.
647 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
649 TF/06 Removed a few PCRE remnants.
651 TF/07 Automatically extract Exim's version number from tags in the git
652 repository when doing development or release builds.
654 PP/02 Raise smtp_cmd_buffer_size to 16kB.
655 Bugzilla 879. Patch from Paul Fisher.
657 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
658 Heavily based on revision 40f9a89a from Simon Arlott's tree.
661 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
663 PP/05 Variable $av_failed, true if the AV scanner deferred.
664 Bugzilla 1078. Patch from John Horne.
666 PP/06 Stop make process more reliably on build failure.
667 Bugzilla 1087. Patch from Heiko Schlittermann.
669 PP/07 Make maildir_use_size_file an _expandable_ boolean.
670 Bugzilla 1089. Patch from Heiko Schlittermann.
672 PP/08 Handle ${run} returning more data than OS pipe buffer size.
673 Bugzilla 1131. Patch from Holger Weiß.
675 PP/09 Handle IPv6 addresses with SPF.
676 Bugzilla 860. Patch from Wolfgang Breyha.
678 PP/10 GnuTLS: support TLS 1.2 & 1.1.
680 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
683 PP/11 match_* no longer expand right-hand-side by default.
684 New compile-time build option, EXPAND_LISTMATCH_RHS.
685 New expansion conditions, "inlist", "inlisti".
687 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
689 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
691 PP/14 fix log_write() format string regression from TF/03.
692 Bugzilla 1152. Patch from Dmitry Isaikin.
698 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
700 PP/02 Harmonised TLS library version reporting; only show if debugging.
701 Layout now matches that introduced for other libraries in 4.74 PP/03.
703 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
705 PP/04 New "dns_use_edns0" global option.
707 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
710 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
711 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
713 TK/01 Updated PolarSSL code to 0.14.2.
714 Bugzilla 1097. Patch from Andreas Metzler.
716 PP/07 Catch divide-by-zero in ${eval:...}.
719 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
722 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
723 format-string attack -- SECURITY: remote arbitrary code execution.
725 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
726 time unintentionally subject to list matching rules, letting the header
727 cause arbitrary Exim lookups (of items which can occur in lists, *not*
728 arbitrary string expansion). This allowed for information disclosure.
730 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
731 INT_MIN/-1 -- value coerced to INT_MAX.
737 NM/01 Workround for PCRE version dependency in version reporting
740 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
741 This fixes portability to compilers other than gcc, notably
742 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
744 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
745 makefiles for portability to HP-UX and POSIX correctness.
747 PP/01 Permit LOOKUP_foo enabling on the make command-line.
748 Also via indented variable definition in the Makefile.
749 (Debugging by Oliver Heesakkers).
751 PP/02 Restore caching of spamd results with expanded spamd_address.
752 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
754 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
755 Improves build reliability. Fix from: Frank Elsner
757 NM/02 Fix wide character breakage in the rfc2047 coding
758 Fixes bug 1064. Patch from Andrey N. Oktyabrski
760 NM/03 Allow underscore in dnslist lookups
761 Fixes bug 1026. Patch from Graeme Fowler
763 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
764 Code patches from Adam Ciarcinski of NetBSD.
766 NM/04 Fixed exiqgrep to cope with mailq missing size issue
769 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
770 is logged, to avoid truncation. Patch from John Horne.
772 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
773 Patch from Jakob Hirsch.
775 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
776 SQL string expansion failure details.
777 Patch from Andrey Oktyabrski.
779 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
780 Patch from Simon Arlott.
782 PP/09 New lookups functionality failed to compile on old gcc which rejects
783 extern declarations in function scope.
784 Patch from Oliver Fleischmann
786 PP/10 Use sig_atomic_t for flags set from signal handlers.
787 Check getgroups() return and improve debugging.
788 Fixed developed for diagnosis in bug 927 (which turned out to be
791 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
792 Patch from Mark Zealey.
794 PP/12 Bugzilla 1056: Improved spamd server selection.
795 Patch from Mark Zealey.
797 PP/13 Bugzilla 1086: Deal with maildir quota file races.
798 Based on patch from Heiko Schlittermann.
800 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
801 Patch from Uwe Doering, sign-off by Michael Haardt.
803 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
804 variable declaration deep within a block. Bug and patch from
807 PP/15 lookups-Makefile IRIX compatibilty coercion.
809 PP/16 Make DISABLE_DKIM build knob functional.
811 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
812 Patch by Simon Arlott
814 TF/03 Fix valgrind.h portability to C89 compilers that do not support
815 variable argument macros. Our copy now differs from upstream.
821 TF/01 Failure to get a lock on a hints database can have serious
822 consequences so log it to the panic log.
824 TF/02 Log LMTP confirmation messages in the same way as SMTP,
825 controlled using the smtp_confirmation log selector.
827 TF/03 Include the error message when we fail to unlink a spool file.
829 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
830 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
831 for maintaining out-of-tree patches for some time.
833 PP/01 Bugzilla 139: Documentation and portability issues.
834 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
835 Handle per-OS dynamic-module compilation flags.
837 PP/02 Let /dev/null have normal permissions.
838 The 4.73 fixes were a little too stringent and complained about the
839 permissions on /dev/null. Exempt it from some checks.
840 Reported by Andreas M. Kirchwitz.
842 PP/03 Report version information for many libraries, including
843 Exim version information for dynamically loaded libraries. Created
844 version.h, now support a version extension string for distributors
845 who patch heavily. Dynamic module ABI change.
847 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
848 privilege escalation vulnerability whereby the Exim run-time user
849 can cause root to append content of the attacker's choosing to
852 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
855 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
856 If dropping privileges for untrusted macros, we disabled normal logging
857 on the basis that it would fail; for the Exim run-time user, this is not
858 the case, and it resulted in successful deliveries going unlogged.
859 Fixed. Reported by Andreas Metzler.
865 PP/01 Date: & Message-Id: revert to normally being appended to a message,
866 only prepend for the Resent-* case. Fixes regression introduced in
867 Exim 4.70 by NM/22 for Bugzilla 607.
869 PP/02 Include check_rfc2047_length in configure.default because we're seeing
870 increasing numbers of administrators be bitten by this.
872 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
874 PP/03 Bugzilla 994: added openssl_options main configuration option.
876 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
878 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
880 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
882 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
883 without a peer certificate, leading to a segfault because of an
884 assumption that peers always have certificates. Be a little more
885 paranoid. Problem reported by Martin Tscholak.
887 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
888 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
889 NB: ClamAV planning to remove STREAM in "middle of 2010".
890 CL also introduces -bmalware, various -d+acl logging additions and
891 more caution in buffer sizes.
893 PP/09 Implemented reverse_ip expansion operator.
895 PP/10 Bugzilla 937: provide a "debug" ACL control.
897 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
899 PP/12 Bugzilla 973: Implement --version.
901 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
903 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
905 PP/15 Bugzilla 816: support multiple condition rules on Routers.
907 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
908 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
909 ignore trailing whitespace.
911 JJ/02 prevent non-panic DKIM error from being sent to paniclog
913 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
916 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
917 Notification from Dr Andrew Aitchison.
919 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
920 ExtendedDetectionInfo response format.
921 Notification from John Horne.
923 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
926 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
927 XSL and documented dependency on system catalogs, with examples of how
930 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
933 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
934 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
935 configuration file which is writeable by the Exim user or group.
937 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
938 of configuration files to cover files specified with the -C option if
939 they are going to be used with root privileges, not just the default
942 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
943 option (effectively making it always true).
945 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
946 files to be used while preserving root privileges.
948 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
949 that rogue child processes cannot use them.
951 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
952 run-time user, instead of root.
954 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
955 Exim run-time user without dropping privileges.
957 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
958 result string, instead of calling string_vformat() twice with the same
961 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
962 for other users. Others should always drop root privileges if they use
963 -C on the command line, even for a whitelisted configure file.
965 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
967 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
973 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
974 $data_path, and $header_path variables; fixed documentation bugs and
977 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
978 exipick to access non-standard spools, including the "frozen" queue
981 NM/01 Bugzilla 965: Support mysql stored procedures.
982 Patch from Alain Williams
984 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
986 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
987 Patch from Andreas Metzler
989 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
990 Patch from Kirill Miazine
992 NM/05 Bugzilla 671: Added umask to procmail example.
994 JJ/03 installed exipick 20100323.0, fixing doc bug
996 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
997 directory. Notification and patch from Dan Rosenberg.
999 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1001 TK/02 Improve log output when DKIM signing operation fails.
1003 MH/01 Treat the transport option dkim_domain as a colon separated
1004 list, not as a single string, and sign the message with each element,
1005 omitting multiple occurences of the same signer.
1007 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1008 Bugzilla 985, 986. Patch by Simon Arlott
1010 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1011 Patch by Simon Arlott
1013 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1014 MBX locking. Notification from Dan Rosenberg.
1020 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1022 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1024 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1026 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1028 NM/04 Bugzilla 915: Flag broken perl installation during build.
1034 TK/01 Added patch by Johannes Berg that expands the main option
1035 "spamd_address" if it starts with a dollar sign.
1037 TK/02 Write list of recipients to X-Envelope-Sender header when building
1038 the mbox-format spool file for content scanning (suggested by Jakob
1041 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1042 (http://www.dcc-servers.net/) support via dccifd. Activated by
1043 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1045 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1046 by Mark Daniel Reidel <mr@df.eu>.
1048 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1049 When building exim an external PCRE library is now needed -
1050 PCRE is a system library on the majority of modern systems.
1051 See entry on PCRE_LIBS in EDITME file.
1053 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1054 conversation. Added nologin parameter to request.
1055 Patch contributed by Kirill Miazine.
1057 TF/01 Do not log submission mode rewrites if they do not change the address.
1059 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1061 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1062 log files in place. Contributed by Roberto Lima.
1064 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1066 TF/03 Bugzilla 615: When checking the local_parts router precondition
1067 after a local_part_suffix or local_part_prefix option, Exim now
1068 does not use the address's named list lookup cache, since this
1069 contains cached lookups for the whole local part.
1071 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1072 Robert Millan. Documentation is in experimental-spec.txt.
1074 TF/04 Bugzilla 668: Fix parallel build (make -j).
1076 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1078 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1079 Patch provided by Jan Srzednicki.
1081 TF/05 Leading white space used to be stripped from $spam_report which
1082 wrecked the formatting. Now it is preserved.
1084 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1085 that they are available at delivery time.
1087 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1089 TF/08 TLS error reporting now respects the incoming_interface and
1090 incoming_port log selectors.
1092 TF/09 Produce a more useful error message if an SMTP transport's hosts
1093 setting expands to an empty string.
1095 NM/06 Bugzilla 744: EXPN did not work under TLS.
1096 Patch provided by Phil Pennock.
1098 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1099 Patch provided by Richard Godbee.
1101 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1102 acl_smtp_notquit, added index entry.
1104 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1105 Patch provided by Eugene Bujak.
1107 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1108 accept(). Patch provided by Maxim Dounin.
1110 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1111 Patch provided by Phil Pennock.
1113 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1115 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1116 Patch provided by Brad "anomie" Jorsch.
1118 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1119 Patch provided by Dean Brooks.
1121 TK/05 Add native DKIM support (does not depend on external libraries).
1123 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1124 Patch provided by Graeme Fowler.
1126 NM/16 Bugzilla 851: Documentation example syntax fix.
1128 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1130 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1133 NM/19 Bugzilla 745: TLS version reporting.
1134 Patch provided by Phil Pennock.
1136 NM/20 Bugzilla 167: bool: condition support.
1137 Patch provided by Phil Pennock.
1139 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1140 clients. Patch provided by Phil Pennock.
1142 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1143 Patch provided by Brad "anomie" Jorsch.
1145 NM/23 Bugzilla 687: Fix misparses in eximstats.
1146 Patch provided by Heiko Schlittermann.
1148 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1149 Patch provided by Heiko Schlittermann.
1151 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1152 plus update to original patch.
1154 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1156 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1157 Patch provided by David Brownlee.
1159 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1161 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1163 NM/30 Bugzilla 888: TLS documentation bugfixes.
1165 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1167 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1168 Unlike the original bugzilla I have changed all shell scripts in src tree.
1170 NM/33 Bugzilla 898: Transport filter timeout fix.
1171 Patch by Todd Rinaldo.
1173 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1174 Patch by Serge Demonchaux.
1176 NM/35 Bugzilla 39: Base64 decode bug fixes.
1177 Patch by Jakob Hirsch.
1179 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1181 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1183 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1185 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1191 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1192 ALT-N's libdkim that I have put here:
1193 http://duncanthrax.net/exim-experimental/
1195 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1196 were called 'true' and it seems that C99 defines that as a reserved
1197 keyword to be used with 'bool' variable types. That means you could
1198 not include C99-style headers which use bools without triggering
1199 build errors in sieve.c.
1201 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1202 as mailq or other aliases. Changed the --help handling significantly
1203 to do whats expected. exim_usage() emits usage/help information.
1205 SC/01 Added the -bylocaldomain option to eximstats.
1207 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1209 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1211 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1217 PH/01 Another patch from the Sieve maintainer.
1219 PH/02 When an IPv6 address is converted to a string for single-key lookup
1220 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1221 dots are used instead of colons so that keys in lsearch files need not
1222 contain colons. This was done some time before quoting was made available
1223 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1224 (notated using the quote facility) so as to distinguish them from IPv4
1225 keys. This meant that lookups for IP addresses in host lists did not work
1226 for iplsearch lookups.
1228 This has been fixed by arranging for IPv6 addresses to be expressed with
1229 colons if the lookup type is iplsearch. This is not incompatible, because
1230 previously such lookups could never work.
1232 The situation is now rather anomolous, since one *can* have colons in
1233 ordinary lsearch keys. However, making the change in all cases is
1234 incompatible and would probably break a number of configurations.
1236 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1239 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1240 conversion specification without a maximum field width, thereby enabling
1241 a rogue spamd server to cause a buffer overflow. While nobody in their
1242 right mind would setup Exim to query an untrusted spamd server, an
1243 attacker that gains access to a server running spamd could potentially
1244 exploit this vulnerability to run arbitrary code as the Exim user.
1246 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1247 $primary_hostname instead of what libspf2 thinks the hosts name is.
1249 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1250 a directory entry by the name of the lookup key. Previously, if a
1251 symlink pointed to a non-existing file or a file in a directory that
1252 Exim lacked permissions to read, a lookup for a key matching that
1253 symlink would fail. Now it is enough that a matching directory entry
1254 exists, symlink or not. (Bugzilla 503.)
1256 PH/03 The body_linecount and body_zerocount variables are now exported in the
1259 PH/04 Added the $dnslist_matched variable.
1261 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1262 This means they are set thereafter only if the connection becomes
1265 PH/06 Added the client_condition to authenticators so that some can be skipped
1266 by clients under certain conditions.
1268 PH/07 The error message for a badly-placed control=no_multiline_responses left
1269 "_responses" off the end of the name.
1271 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1273 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1274 (without spaces) instead of just copying the configuration text.
1276 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1278 PH/11 Added $max_received_linelength.
1280 PH/12 Added +ignore_defer and +include_defer to host lists.
1282 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1283 way in which PCRE > 7.0 is built.
1285 PH/14 Implemented queue_only_load_latch.
1287 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1288 MAIL command. The effect was to mangle the value on 64-bit systems.
1290 PH/16 Another patch from the Sieve maintainer.
1292 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1294 PH/18 If a system quota error occurred while trying to create the file for
1295 a maildir delivery, the message "Mailbox is full" was not appended to the
1296 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1297 only to a quota excession during the actual writing of the file.
1299 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1300 characters?) which causes problems in log lines. The DN values are now
1301 passed through string_printing() before being added to log lines.
1303 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1304 and InterBase are left for another time.)
1306 PH/21 Added message_body_newlines option.
1308 PH/22 Guard against possible overflow in moan_check_errorcopy().
1310 PH/23 POSIX allows open() to be a macro; guard against that.
1312 PH/24 If the recipient of an error message contained an @ in the local part
1313 (suitably quoted, of course), incorrect values were put in $domain and
1314 $local_part during the evaluation of errors_copy.
1320 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1321 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1324 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1325 issue a MAIL command.
1327 PH/02 In an ACL statement such as
1329 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1331 if a client was not listed at all, or was listed with a value other than
1332 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1333 the condition was not true (as it should be), so access was not denied.
1334 The bug was that the ! inversion was incorrectly passed on to the second
1335 item. This has been fixed.
1337 PH/03 Added additional dnslists conditions == and =& which are different from
1338 = and & when the dns lookup returns more than one IP address.
1340 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1341 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1343 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1344 FSYNC, which compiles an option called disable_fsync that allows for
1345 bypassing fsync(). The documentation is heavily laced with warnings.
1347 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1349 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1350 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1351 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1352 including adding "make clean"; (3) Added -fPIC when compiling the test
1353 dynamically loaded module, to get rid of a warning.
1355 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1356 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1357 = 0s. The bug is otherwise harmless.
1359 PH/07 There was a bug in the dovecot authenticator such that the value of
1360 $auth1 could be overwritten, and so not correctly preserved, after a
1361 successful authentication. This usually meant that the value preserved by
1362 the server_setid option was incorrect.
1364 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1366 PH/09 Installed PCRE release 7.0.
1368 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1369 run for batched SMTP input. It is now run at the start of every message
1370 in the batch. While fixing this I discovered that the process information
1371 (output by running exiwhat) was not always getting set for -bs and -bS
1372 input. This is fixed, and it now also says "batched" for BSMTP.
1374 PH/11 Added control=no_pipelining.
1376 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1377 patch, slightly modified), and move the expansion of helo_data till after
1378 the connection is made in the smtp transport (so it can use these
1381 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1383 PH/14 Added log_selector = +pid.
1385 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1387 PH/16 Add ${if forany and ${if forall.
1389 PH/17 Added dsn_from option to vary the From: line in DSNs.
1391 PH/18 Flush SMTP output before performing a callout, unless control =
1392 no_callout_flush is set.
1394 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1395 was true (the default) a successful delivery failed to delete the retry
1396 item, thus causing premature timeout of the address. The bug is now
1399 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1401 PH/21 Long custom messages for fakedefer and fakereject are now split up
1402 into multiline reponses in the same way that messages for "deny" and
1403 other ACL rejections are.
1405 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1406 with slight modification.
1408 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1409 draft, changing the syntax and factoring some duplicate code".
1411 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1412 for deliveries of the second and subsequent messages over the same SMTP
1415 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1416 ${reduce, with only minor "tidies".
1418 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1420 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1421 expansion side effects.
1423 PH/27 When a message times out after an over-quota error from an Exim-imposed
1424 quota, the bounce message says "mailbox is full". This message was not
1425 being given when it was a system quota that was exceeded. It now should
1428 MH/03 Made $recipients available in local_scan(). local_scan() already has
1429 better access to the recipient list through recipients_list[], but
1430 $recipients can be useful in postmaster-provided expansion strings.
1432 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1433 in the case of a MAIL command with additional options following the
1434 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1435 were accidentally chopped off.
1437 PH/29 SMTP synchronization checks are implemented when a command is read -
1438 there is a check that no more input is waiting when there shouldn't be
1439 any. However, for some commands, a delay in an ACL can mean that it is
1440 some time before the response is written. In this time, more input might
1441 arrive, invalidly. So now there are extra checks after an ACL has run for
1442 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1443 pipelining has not been advertised.
1445 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1447 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1448 returned in the SMTP response when smtp_return_error_details was false.
1449 This has been fixed.
1451 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1452 instead of the C I/O that was originally supplied, because problems were
1453 reported on Solaris.
1455 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1456 Exim which did not show up earlier: it was assuming that a call to
1457 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1458 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1459 was a macro that became an assignment, so it seemed to work. This has
1460 changed to a proper function call with a void return, hence the compile
1461 error. Exim's code has been fixed.
1463 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1466 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1468 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1470 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1471 because it turns out to be too big or there is a timeout) there is no
1472 "Completed" line in the log. When some messages of this type were
1473 selected by exigrep, they were listed as "not completed". Others were
1474 picked up by some special patterns. I have improved the selection
1475 criteria to be more general.
1477 PH/38 The host_find_failed option in the manualroute router can now be set
1478 to "ignore", to completely ignore a host whose IP address cannot be
1479 found. If all hosts are ignored, the behaviour is controlled by the new
1480 host_all_ignored option.
1482 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1483 homing or because of multiple MX records with /mx) generated more than
1484 one IP address, and the following item turned out to be the local host,
1485 all the secondary addresses of the first item were incorrectly removed
1486 from the list, along with the local host and any following hosts (which
1487 is what is supposed to happen).
1489 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1490 whoever called Exim into the -H file. In the case of the daemon it was
1491 behaving confusingly. When first started, it used values for whoever
1492 started the daemon, but after a SIGHUP it used the Exim user (because it
1493 calls itself on a restart). I have changed the code so that it now always
1496 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1497 message are rejected with the same error (e.g. no authentication or bad
1498 sender address), and a DATA command is nevertheless sent (as can happen
1499 with PIPELINING or a stupid MUA), the error message that was given to the
1500 RCPT commands is included in the rejection of the DATA command. This is
1501 intended to be helpful for MUAs that show only the final error to their
1504 PH/42 Another patch from the Sieve maintainer.
1506 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1507 Eximstats - Fixed some broken HTML links and added missing column headers
1509 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1510 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1512 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1514 PH/43 Yet another patch from the Sieve maintainer.
1516 PH/44 I found a way to check for a TCP/IP connection going away before sending
1517 the response to the final '.' that terminates a message, but only in the
1518 case where the client has not sent further data following the '.'
1519 (unfortunately, this is allowed). However, in many cases there won't be
1520 any further data because there won't be any more messages to send. A call
1521 to select() can be used: if it shows that the input is "ready", there is
1522 either input waiting, or the socket has been closed. An attempt to read
1523 the next input character can distinguish the two cases. Previously, Exim
1524 would have sent an OK response which the client would never have see.
1525 This could lead to message repetition. This fix should cure that, at
1526 least in a lot of common cases.
1528 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1529 advertised in response to EHLO.
1535 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1536 fixed by 4.65/MH/01 (is this a record?) are fixed:
1538 (i) An empty string was always treated as zero by the numeric comparison
1539 operators. This behaviour has been restored.
1541 (ii) It is documented that the numeric comparison operators always treat
1542 their arguments as decimal numbers. This was broken in that numbers
1543 starting with 0 were being interpreted as octal.
1545 While fixing these problems I realized that there was another issue that
1546 hadn't been noticed. Values of message_size_limit (both the global option
1547 and the transport option) were treated as octal if they started with 0.
1548 The documentation was vague. These values are now always treated as
1549 decimal, and I will make that clear in the documentation.
1555 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1556 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1559 MH/01 Don't check that the operands of numeric comparison operators are
1560 integers when their expansion is in "skipping" mode (fixes bug
1561 introduced by 4.64-PH/07).
1563 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1564 child addresses, Exim now panics and dies. Previously, because the count
1565 is held in a short int, deliveries were likely to be lost. As such a
1566 large number of recipients for a single message is ridiculous
1567 (performance will be very, very poor), I have chosen to impose a limit
1568 rather than extend the field.
1574 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1575 leftover -K file (the existence of which was triggered by #402).
1576 While we were at it, introduced process PID as part of the -K
1577 filename. This should rule out race conditions when creating
1580 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1581 processing considerably. Previous code took too long for large mails,
1582 triggering a timeout which in turn triggers #401.
1584 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1585 in the DK code in transports.c. sendfile() is not really portable,
1586 hence the _LINUX specificness.
1588 TF/01 In the add_headers option to the mail command in an Exim filter,
1589 there was a bug that Exim would claim a syntax error in any
1590 header after the first one which had an odd number of characters
1593 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1594 callout verification, Exim cached a "reject" for the entire domain. This
1595 is correct for most verifications, but it is not correct for a recipient
1596 verification with use_sender or use_postmaster set, because in that case
1597 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1598 case of MAIL FROM:<> rejection from other early rejections (e.g.
1599 rejection of HELO). When verifying a recipient using a non-null MAIL
1600 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1601 Whatever the result of the callout, the value of the domain cache is
1602 left unchanged (for any other kind of callout, getting as far as trying
1603 RCPT means that the domain itself is ok).
1605 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1608 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1609 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1610 session, the daemon ignores ECONNECT errors and logs others; it now
1611 ignores EPIPE as well.
1613 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1614 (quoted-printable decoding).
1616 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1617 later the small subsequent patch to fix an introduced bug.
1619 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1621 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1623 PH/08 An error is now given if message_size_limit is specified negative.
1625 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1626 to be given (somewhat) arbitrary names.
1628 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1631 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1632 miscellaneous code fixes
1634 PH/10 Added the log_reject_target ACL modifier to specify where to log
1637 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1638 hostname. This is wrong, because it relates to the incoming message (and
1639 probably the interface on which it is arriving) and not to the outgoing
1640 callout (which could be using a different interface). This has been
1641 changed to use the value of the helo_data option from the smtp transport
1642 instead - this is what is used when a message is actually being sent. If
1643 there is no remote transport (possible with a router that sets up host
1644 addresses), $smtp_active_hostname is used.
1646 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1647 tweaks were necessary in order to get it to work (see also 21 below):
1648 (a) The code assumed that strncpy() returns a negative number on buffer
1649 overflow, which isn't the case. Replaced with Exim's string_format()
1651 (b) There were several signed/unsigned issues. I just did the minimum
1652 hacking in of casts. There is scope for a larger refactoring.
1653 (c) The code used strcasecmp() which is not a standard C function.
1654 Replaced with Exim's strcmpic() function.
1655 (d) The code set only $1; it now sets $auth1 as well.
1656 (e) A simple test gave the error "authentication client didn't specify
1657 service in request". It would seem that Dovecot has changed its
1658 interface. Fortunately there's a specification; I followed it and
1659 changed what the client sends and it appears to be working now.
1661 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1664 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1665 address A is aliased to B and C, where B exists and C does not. Without
1666 -v the output is "A verified" because verification stops after a
1667 successful redirection if more than one address is generated. However,
1668 with -v the child addresses are also verified. Exim was outputting "A
1669 failed to verify" and then showing the successful verification for C,
1670 with its parentage. It now outputs "B failed to verify", showing B's
1671 parentage before showing the successful verification of C.
1673 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1674 look up a TXT record in a specific list after matching in a combined
1677 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1678 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1679 they consult the DNS. I had assumed they would set it the way they
1680 wanted; and indeed my experiments on Linux seem to show that in some
1681 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1682 To be on the safe side, however, I have now made the interface to
1683 host_find_byname() similar to host_find_bydns(), with an argument
1684 containing the DNS resolver options. The host_find_byname() function now
1685 sets these options at its start, just as host_find_bydns() does. The smtp
1686 transport options dns_qualify_single and dns_search_parents are passed to
1687 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1688 of host_find_byname() use the default settings of RES_DEFNAMES
1689 (qualify_single) but not RES_DNSRCH (search_parents).
1691 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1692 spool_read_header() do less string testing, by means of a preliminary
1693 switch on the second character of optional "-foo" lines. (This is
1694 overdue, caused by the large number of possibilities that now exist.
1695 Originally there were few.) While I was there, I also converted the
1696 str(n)cmp tests so they don't re-test the leading "-" and the first
1697 character, in the hope this might squeeze out yet more improvement.
1699 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1700 flag allowing group syntax was set by the header_syntax check but not
1701 turned off, possible causing trouble later; (2) The flag was not being
1702 set at all for the header_verify test, causing "group"-style headers to
1703 be rejected. I have now set it in this case, and also caused header_
1704 verify to ignore an empty address taken from a group. While doing this, I
1705 came across some other cases where the code for allowing group syntax
1706 while scanning a header line wasn't quite right (mostly, not resetting
1707 the flag correctly in the right place). These bugs could have caused
1708 trouble for malformed header lines. I hope it is now all correct.
1710 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1711 with the "reply" argument non-NULL. The code, however (which originally
1712 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1713 but it didn't always do it. This confused somebody who was copying the
1714 code for some other use. I have removed all the tests.
1716 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1717 feature that was used to support insecure browsers during the U.S. crypto
1718 embargo. It requires special client support, and Exim is probably the
1719 only MTA that supported it -- and would never use it because real RSA is
1720 always available. This code has been removed, because it had the bad
1721 effect of slowing Exim down by computing (never used) parameters for the
1722 RSA_EXPORT functionality.
1724 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1725 authenticator to fail if there's a tab character in the incoming data
1726 (there should never be unless someone is messing about, as it's supposed
1727 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1728 if the connection is using TLS or if the remote IP is the same as the
1729 local IP, and the "valid-client-cert option" if a client certificate has
1732 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1733 authenticators. This can be used for authorization after authentication
1734 succeeds. (In the case of plaintext, it servers for both authentication
1737 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1738 if any retry times were supplied.
1740 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1741 connection, where there is no client IP address to check. In this
1742 situation, the verify now always succeeds.
1744 PH/25 Applied John Jetmore's -Mset patch.
1746 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1748 PH/27 In a string expansion for a processed (not raw) header when multiple
1749 headers of the same name were present, leading whitespace was being
1750 removed from all of them, but trailing whitespace was being removed only
1751 from the last one. Now trailing whitespace is removed from each header
1752 before concatenation. Completely empty headers in a concatenation (as
1753 before) are ignored.
1755 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1756 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1758 PH/29 [Removed. This was a change that I later backed out, and forgot to
1759 correct the ChangeLog entry (that I had efficiently created) before
1760 committing the later change.]
1762 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1763 address errors (4xx response to RCPT) over the same connection as other
1764 messages routed to the same hosts. Such deliveries are always "forced",
1765 so retry times are not inspected. This resulted in far too many retries
1766 for the affected addresses. The effect occurred only when there were more
1767 hosts than the hosts_max_try setting in the smtp transport when it had
1768 the 4xx errors. Those hosts that it had tried were not added to the list
1769 of hosts for which the message was waiting, so if all were tried, there
1770 was no problem. Two fixes have been applied:
1772 (i) If there are any address or message errors in an SMTP delivery, none
1773 of the hosts (tried or untried) are now added to the list of hosts
1774 for which the message is waiting, so the message should not be a
1775 candidate for sending over the same connection that was used for a
1776 successful delivery of some other message. This seems entirely
1777 reasonable: after all the message is NOT "waiting for some host".
1778 This is so "obvious" that I'm not sure why it wasn't done
1779 previously. Hope I haven't missed anything, but it can't do any
1780 harm, as the worst effect is to miss an optimization.
1782 (ii) If, despite (i), such a delivery is accidentally attempted, the
1783 routing retry time is respected, so at least it doesn't keep
1784 hammering the server.
1786 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1787 in ${readsocket because some servers need this prod.
1789 PH/32 Added some extra debug output when updating a wait-xxx database.
1791 PH/33 The hint "could be header name not terminated by colon", which has been
1792 given for certain expansion errors for a long time, was not being given
1793 for the ${if def:h_colon_omitted{... case.
1795 PH/34 The spec says: "With one important exception, whenever a domain list is
1796 being scanned, $domain contains the subject domain." There was at least
1797 one case where this was not true.
1799 PH/35 The error "getsockname() failed: connection reset by peer" was being
1800 written to the panic log as well as the main log, but it isn't really
1801 panic-worthy as it just means the connection died rather early on. I have
1802 removed the panic log writing for the ECONNRESET error when getsockname()
1805 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1806 runs only) independently of the message's sender address. This meant
1807 that, if the 4xx error was in fact related to the sender, a different
1808 message to the same recipient with a different sender could confuse
1809 things. In particualar, this can happen when sending to a greylisting
1810 server, but other circumstances could also provoke similar problems.
1811 I have changed the default so that the retry time for these errors is now
1812 based a combination of the sender and recipient addresses. This change
1813 can be overridden by setting address_retry_include_sender=false in the
1816 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1817 remote server are returned as part of bounce messages. This was not
1818 happening for LMTP over a pipe (the lmtp transport), but now it is the
1819 same for both kinds of LMTP.
1821 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1822 in header lines that were in fact CNAMEs. This is no longer the case.
1824 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1825 and queue runs started by the daemon processed all messages. This has
1826 been fixed so that -R and -S can now usefully be given with -q<time>.
1828 PH/40 Import PCRE release 6.7 (fixes some bugs).
1830 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1832 PH/42 Give an error if -q is specified more than once.
1834 PH/43 Renamed the variables $interface_address and $interface_port as
1835 $received_ip_address and $received_port, to make it clear that these
1836 values apply to message reception, and not to the outgoing interface when
1837 a message is delivered. (The old names remain recognized, of course.)
1839 PH/44 There was no timeout on the connect() call when using a Unix domain
1840 socket in the ${readsocket expansion. There now is.
1842 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1843 be meaningful with "accept".
1845 SC/01 Eximstats V1.43
1846 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1848 SC/02 Eximstats V1.44
1849 Use a glob alias rather than an array ref in the generated
1850 parser. This improves both readability and performance.
1852 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1853 Collect SpamAssassin and rejection statistics.
1854 Don't display local sender or destination tables unless
1855 there is data to show.
1856 Added average volumes into the top table text output.
1858 SC/04 Eximstats V1.46
1859 Collect data on the number of addresses (recipients)
1860 as well as the number of messages.
1862 SC/05 Eximstats V1.47
1863 Added 'Message too big' to the list of mail rejection
1864 reasons (thanks to Marco Gaiarin).
1866 SC/06 Eximstats V1.48
1867 Mainlog lines which have GMT offsets and are too short to
1868 have a flag are now skipped.
1870 SC/07 Eximstats V1.49 (Alain Williams)
1871 Added the -emptyok flag.
1873 SC/08 Eximstats V1.50
1874 Fixes for obtaining the IP address from reject messages.
1876 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1877 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1878 whitesspace changes from 4.64-PH/27
1880 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1883 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1884 are found, allow negative numbers in numeric criteria)
1886 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1888 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1889 to match changes made in 4.64-PH/43
1891 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1893 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1894 "domain" is actually an IP address, and force a failure. This locks out
1895 those revolvers/nameservers that support "A-for-A" lookups, in
1896 contravention of the specifications.
1898 PH/48 When a host name was looked up from an IP address, and the subsequent
1899 forward lookup of the name timed out, the host name was left in
1900 $sender_host_name, contrary to the specification.
1902 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1903 restricted to single-key lookups, Exim was not diagnosing an error if
1904 * or *@ was used with a query-style lookup.
1906 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1908 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1909 long ago, but noone interested enough thought of it. Let's just say that
1910 the "1.1" means that there are some new functions that weren't there at
1911 some point in the past.
1913 PH/51 Error processing for expansion failure of helo_data from an smtp
1914 transport during callout processing was broken.
1916 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1917 tested/used via the -bh/-bhc/-bs options.
1919 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1920 bug, fixed in subsequent PCRE releases).
1922 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1923 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1925 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1931 SC/01 Use a glob alias rather than an array ref in eximstats generated
1932 parser. This improves both readability and performance.
1934 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1935 Don't display local sender or destination tables in eximstats unless
1936 there is data to show.
1937 Added average volumes into the eximstats top table text output.
1939 SC/03 Collect data on the number of addresses (recipients) as well
1940 as the number of messages in eximstats.
1942 TF/01 Correct an error in the documentation for the redirect router. Exim
1943 does (usually) call initgroups() when daemonizing.
1945 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1946 with consistent privilege compared to when running as a daemon.
1948 TF/03 Note in the spec that $authenticated_id is not set for local
1949 submissions from trusted users.
1951 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1952 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1954 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1955 by adding some example configuration directives to the default
1956 configuration file. A little bit of work is required to uncomment the
1957 directives and define how usernames and passwords are checked, but
1958 there is now a framework to start from.
1960 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1961 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1962 without this. I don't know how relevant this is to other LDAP libraries.
1964 PH/02 Add the verb name to the "unknown ACL verb" error.
1966 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1968 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1970 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1971 directory not expanded when it should be if an expanded home directory
1972 was set for the address (which is overridden by the transport).
1974 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1977 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1978 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1979 because it is too late at that time, and has no effect.
1981 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1982 security issue with \' (bugzilla #107). I could not use the
1983 PQescapeStringConn() function, because it needs a PGconn value as one of
1986 PH/08 When testing addresses using -bt, indicate those final addresses that
1987 are duplicates that would not cause an additional delivery. At least one
1988 person was confused, thinking that -bt output corresponded to deliveries.
1989 (Suppressing duplicates isn't a good idea as you lose the information
1990 about possibly different redirections that led to the duplicates.)
1992 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1993 systems where poll() doesn't work, in particular OS X.
1995 PH/10 Added more information to debugging output for retry time not reached.
1997 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1998 operations in malware.c.
2000 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2003 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2004 syslog_duplication set false, log lines that would normally be written
2005 both the the main log and to the reject log were not written to syslog at
2008 PH/14 In the default configuration, change the use of "message" in ACL warn
2009 statements to "add_header".
2011 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2012 not followed by a command (e.g. "seen endif").
2014 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2015 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2018 PH/17 Added extra conditions to the default value of delay_warning_condition
2022 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2023 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2024 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2027 The Auto-Submitted: and various List- headers are standardised, whereas I
2028 don't think Precedence: ever was.
2030 PH/18 Refactored debugging code in route_finduser() to show more information,
2031 in particular, the error code if getpwnam() issues one.
2033 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2034 This is apparently needed in addition to the PH/07 change above to avoid
2035 any possible encoding problems.
2037 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2038 but not after initializing Perl.
2040 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2041 output them only if debugging. By default they are written stderr,
2042 apparently, which is not desirable.
2044 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2047 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2050 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2052 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2053 authenticated or an ident call has been made. Suppress the default
2054 values for $authenticated_id and $authenticated_sender (but permit -oMai
2055 and -oMas) when testing with -bh.
2057 PH/24 Re-jigged the order of the tests in the default configuration so that the
2058 tests for valid domains and recipients precede the DNS black list and CSA
2059 tests, on the grounds that those ones are more expensive.
2061 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2062 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2063 This bug exists in every version of Exim that I still have, right back to
2066 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2067 However, an attempt to turn on case-sensitivity in a regex key by
2068 including (?-i) didn't work because the subject string was already
2069 lowercased, and the effects were non-intuitive. It turns out that a
2070 one-line patch can be used to allow (?-i) to work as expected.
2076 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2077 other effects) broke the use of negated acl sub-conditions.
2079 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2082 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2083 "Deny" causes Exim to reject the incoming connection with a 554 error.
2084 Unfortunately, if there is a major crisis, such as a disk failure,
2085 tcp-wrappers gives "deny", whereas what one would like would be some
2086 kind of temporary error. A kludge has been added to help with this.
2087 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2088 554 error is used if errno is still zero or contains ENOENT (which occurs
2089 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2092 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2094 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2095 errors. Otherwise a message that provokes a temporary error (when other
2096 messages do not) can cause a whole host to time out.
2098 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2099 addresses were routed directly to files or pipes from a redirect router.
2100 File deliveries just didn't batch; pipe deliveries might have suffered
2103 PH/06 A failure to get a lock for a hints database would erroneously always say
2104 "Failed to get write lock", even when it was really a read lock.
2106 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2107 of 0600. This has been changed to use the value of the lockfile_mode
2108 option (which defaults to 0600).
2110 PH/08 Applied small patch from the Sieve maintainer.
2112 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2113 folder from quota calculations, a direct delivery into this folder messed
2114 up the contents of the maildirsize file. This was because the regex was
2115 used only to exclude .Trash (or whatever) when the size of the mailbox
2116 was calculated. There was no check that a delivery was happening into an
2117 excluded directory. This bug has been fixed by ignoring all quota
2118 processing for deliveries into excluded directories.
2120 PH/10 Added the maildirfolder_create_regex option to appendfile.
2126 PH/01 The code for finding all the local interface addresses on a FreeBSD
2127 system running IPv6 was broken. This may well have applied to all BSD
2128 systems, as well as to others that have similar system calls. The broken
2129 code found IPv4 interfaces correctly, but gave incorrect values for the
2130 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2131 that it would not match correctly against @[] and not recognize the IPv6
2134 PH/02 The ipliteral router was not recognizing addresses of the form user@
2135 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2137 PH/03 Added disable_ipv6.
2139 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2140 decoded form, because it is most often used to construct To: headers
2141 lines in autoreplies, and the decoded form may well be syntactically
2142 invalid. However, $reply_address has leading white space removed, and all
2143 newlines turned into spaces so that the autoreply transport does not
2146 PH/05 If group was specified without a user on a router, and no group or user
2147 was specified on a transport, the group from the router was ignored.
2149 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2150 for visible compile-time settings that can be used to change these
2151 numbers, for those that want even more. Backwards compatibility with old
2152 spool files has been maintained. However, going back to a previous Exim
2153 release will lost any variables that are in spool files.
2155 PH/07 Two small changes when running in the test harness: increase delay when
2156 passing a TCP/IP connection to a new process, in case the original
2157 process has to generate a bounce, and remove special handling of
2158 127.0.0.2 (sic), which is no longer necessary.
2160 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2161 be the same on different OS.
2163 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2166 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2167 whether --show-vars was specified or not
2169 JJ/02 exipick: Added support for new ACL variable spool format introduced
2172 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2173 syntactically invalid From: or Reply-to: line, and a filter used this to
2174 generate an autoreply, and therefore failed to obtain an address for the
2175 autoreply, Exim could try to deliver to a non-existent relative file
2176 name, causing unrelated and misleading errors. What now happens is that
2177 it logs this as a hard delivery error, but does not attempt to create a
2180 PH/11 The exinext utility has a -C option for testing purposes, but although
2181 the given file was scanned by exinext itself; it wasn't being passed on
2182 when Exim was called.
2184 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2185 an end-of-file indication when reading a command response.
2187 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2188 compiled. In many other places in Exim, IPv6 addresses are always
2189 recognized, so I have changed this. It also means that IPv4 domain
2190 literals of the form [IPV4:n.n.n.n] are now always recognized.
2192 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2193 used if the router is not running as root, for example, when verifying at
2194 ACL time, or when using -bh. The debugging output from this situation was
2195 non-existent - all you got was a failure to exec. I have made two
2198 (a) Failures to set uid/gid, the current directory, or a process leader
2199 in a subprocess such as that created by queryprogram now generate
2200 suitable debugging ouput when -d is set.
2202 (b) The queryprogram router detects when it is not running as root,
2203 outputs suitable debugging information if -d is set, and then runs
2204 the subprocess without attempting to change uid/gid.
2206 PH/15 Minor change to Makefile for building test_host (undocumented testing
2209 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2210 additional section of a DNS packet that returns MX or SRV records.
2211 Instead, it always explicitly searches for A/AAAA records. This avoids
2212 major problems that occur when a DNS server includes only records of one
2213 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2214 fixed another bug: if SRV records were looked up and the corresponding
2215 address records were *not* found in the additional section, the port
2216 values from the SRV records were lost.
2218 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2219 using the correct key (the original address) when searching the retry
2220 rules in order to find which one to use for generating the retry hint.
2222 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2223 adding the default one. Similarly, if it contains a Reply-To: header, the
2224 errors_reply_to option, if set, is not used.
2226 PH/19 When calculating a retry time, Exim used to measure the "time since
2227 failure" by looking at the "first failed" field in the retry record. Now
2228 it does not use this if it is later than than the arrival time of the
2229 message. Instead it uses the arrival time. This makes for better
2230 behaviour in cases where some deliveries succeed, thus re-setting the
2231 "first failed" field. An example is a quota failure for a huge message
2232 when small messages continue to be delivered. Without this change, the
2233 "time since failure" will always be short, possible causing more frequent
2234 delivery attempts for the huge message than are intended.
2235 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2237 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2238 $1, $2, $3) because the numerical variables can be reset during some
2239 expansion items (e.g. "match"), thereby losing the authentication data.
2241 PH/21 Make -bV show the size of off_t variables so that the test suite can
2242 decide whether to run tests for quotas > 2G.
2244 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2245 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2246 filecount value is greater than 2G or if a quota value is greater than 2G
2247 on a system where the size of off_t is not greater than 4, a panic error
2250 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2251 never match. The debug and -bh output now contains an explicit error
2252 message indicating a malformed IPv4 address or mask.
2254 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2255 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2256 PH/23 above applies.
2258 PH/25 Do not write to syslog when running in the test harness. The only
2259 occasion when this arises is a failure to open the main or panic logs
2260 (for which there is an explicit test).
2262 PH/26 Added the /no_tell option to "control=freeze".
2264 PH/27 If a host name lookup failed very early in a connection, for example, if
2265 the IP address matched host_lookup and the reverse lookup yielded a name
2266 that did not have a forward lookup, an error message of the form "no IP
2267 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2268 could be logged. Now it outputs the IP address instead of "NULL".
2270 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2271 allows the sender and the authenticated sender to be set when
2272 submitting a message from within Exim. Since child_open_exim() is
2273 documented for local_scan(), the new function should be too.
2275 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2276 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2277 results in an empty string is now treated as unset.
2279 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2281 PH/31 Added sender_verify_fail logging option.
2283 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2284 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2285 tidied the source and removed it altogether.
2287 PH/33 When a queue run was abandoned because the load average was too high, a
2288 log line was always written; now it is written only if the queue_run log
2289 selector is set. In addition, the log line for abandonment now contains
2290 information about the queue run such as the pid. This is always present
2291 in "start" and "stop" lines but was omitted from the "abandon" line.
2293 PH/34 Omit spaces between a header name and the colon in the error message that
2294 is given when verify = headers_syntax fails (if there are lots of them,
2295 the message gets confusing).
2297 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2298 names, as there are now some PTR records that contain slashes. This check
2299 is only to protect against broken name servers that fall over on strange
2300 characters, so the fact that it applies to all lookups doesn't matter.
2302 PH/36 Now that the new test suite is complete, we can remove some of the
2303 special code in Exim that was needed for the old test suite. For example,
2304 sorting DNS records because real resolvers return them in an arbitrary
2305 order. The new test suite's fake resolver always returns records in the
2308 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2309 bounces) except when queue_only is set, to avoid logging races between
2310 the different processes.
2312 PH/38 Panic-die if .include specifies a non-absolute path.
2314 PH/39 A tweak to the "H" retry rule from its user.
2316 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2317 a label. They prevented compilation on older perls.
2319 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2320 a warning to be raised on newish perls.
2322 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2323 on queue. Changes to match documented behaviour of showing count of
2324 messages matching specified criteria.
2326 PH/40 Changed the default ident timeout from 30s to 5s.
2328 PH/41 Added support for the use of login_cap features, on those BSD systems
2329 that have them, for controlling the resources used by pipe deliveries.
2331 PH/42 The content-scanning code uses fopen() to create files in which to put
2332 message data. Previously it was not paying any attention to the mode of
2333 the files. Exim runs with umask(0) because the rest of the code creates
2334 files with open(), and sets the required mode explicitly. Thus, these
2335 files were ending up world-writeable. This was not a big issue, because,
2336 being within the spool directory, they were not world-accessible. I have
2337 created a function called modefopen, which takes an additional mode
2338 argument. It sets umask(777), creates the file, chmods it to the required
2339 mode, then resets the umask. All the relevant calls to fopen() in the
2340 content scanning code have been changed to use this function.
2342 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2343 to 24 hours. This avoids potential overflow problems when processing G
2344 and H retry rules. I suspect nobody ever tinkers with this value.
2346 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2348 PH/45 When the plaintext authenticator is running as a client, the server's
2349 challenges are checked to ensure they are valid base64 strings. By
2350 default, the authentication attempt is cancelled if an invalid string is
2351 received. Setting client_ignore_invalid_base64 true ignores these errors.
2352 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2353 they are received. Thus, the responses can be made to depend on the
2354 challenges. If an invalid string is ignored, an empty string is placed in
2357 PH/46 Messages that are created by the autoreply transport now contains a
2358 References: header, in accordance with RFCs 2822 and 3834.
2360 PH/47 Added authenticated_sender_force to the smtp transport.
2362 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2364 PH/49 Installed latest patch from the Sieve maintainer.
2366 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2367 was also set, the appendfile transport was unnecessarily scanning a
2368 directory of message files (e.g. for maildir delivery) to find the count
2369 of files (along with the size), even though it did not need this
2370 information. It now does the scan only if it needs to find either the
2371 size of the count of files.
2373 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2375 PH/52 Two bugs concerned with error handling when the smtp transport is
2378 (i) Exim was not creating retry information for temporary errors given
2379 for individual recipients after the DATA command when the smtp transport
2380 was used in LMTP mode. This meant that they could be retried too
2381 frequently, and not timed out correctly.
2383 (ii) Exim was setting the flag that allows error details to be returned
2384 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2385 recipients that were returned after the DATA command.
2387 PH/53 This is related to PH/52, but is more general: for any failing address,
2388 when detailed error information was permitted to be returned to the
2389 sender, but the error was temporary, then after the final timeout, only
2390 "retry timeout exceeded" was returned. Now it returns the full error as
2391 well as "retry timeout exceeded".
2393 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2394 do this, and (what is worse) MTAs that accept it.
2396 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2397 will now be deprecated.
2399 PH/56 New os.c-cygwin from the Cygwin maintainer.
2401 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2402 formats (previously only available in exim formats via -bpr, -bpru,
2403 and -bpra. Now also available in native and exiqgrep formats)
2405 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2406 with very large, slow to parse queues
2408 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2410 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2412 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2413 responses to authentication challenges, though it was showing the
2414 challenges; (ii) I've removed the CR characters from the debug output for
2417 PH/58 Allow for the insertion of a newline as well as a space when a string
2418 is turned into more than one encoded-word during RFC 2047 encoding. The
2419 Sieve code now uses this.
2421 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2422 data_4xx, lost_connection, tls_required.
2424 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2425 message was being sent as an SMTP response.
2427 PH/61 Add -l and -k options to exicyclog.
2429 PH/62 When verifying, if an address was redirected to one new address, so that
2430 verification continued, and the new address failed or deferred after
2431 having set something in $address_data, the value of $address_data was not
2432 passed back to the ACL. This was different to the case when no
2433 redirection occurred. The value is now passed back in both cases.
2435 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2436 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2437 use login_cap.h, so on its own it isn't the distinguishing feature. The
2438 new name refers directly to the setclassresources() function.
2440 PH/65 Added configuration files for NetBSD3.
2442 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2444 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2445 is preferred over IPv4.
2447 PH/68 The bounce_return_message and bounce_return_body options were not being
2448 honoured for bounces generated during the reception of non-SMTP messages.
2449 In particular, this applied to messages rejected by the ACL. This bug has
2450 been fixed. However, if bounce_return_message is true and bounce_return_
2451 body is false, the headers that are returned for a non-SMTP message
2452 include only those that have been read before the error was detected.
2453 (In the case of an ACL rejection, they have all been read.)
2455 PH/69 The HTML version of the specification is now built in a directory called
2456 spec_html instead of spec.html, because the latter looks like a path with
2457 a MIME-type, and this confuses some software.
2459 PH/70 Catch two compiler warnings in sieve.c.
2461 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2462 function verify_get_ident() calls ip_connect() to connect a socket, but
2463 if the "connect()" function timed out, ip_connect() used to close the
2464 socket. However, verify_get_ident() also closes the socket later, and in
2465 between Exim writes to the log, which may get opened at this point. When
2466 the socket was closed in ip_connect(), the log could get the same file
2467 descriptor number as the socket. This naturally causes chaos. The fix is
2468 not to close the socket in ip_connect(); the socket should be closed by
2469 the function that creates it. There was only one place in the code where
2470 this was missing, in the iplookup router, which I don't think anybody now
2471 uses, but I've fixed it anyway.
2473 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2474 well as to direct DNS lookups. Otherwise the handling of names in host
2475 lists is inconsistent and therefore confusing.
2481 PH/01 Two changes to the default runtime configuration:
2483 (1) Move the checks for relay_from_hosts and authenticated clients from
2484 after to before the (commented out) DNS black list checks.
2486 (2) Add control=submission to the relay_from_hosts and authenticated
2487 clients checks, on the grounds that messages accepted by these
2488 statements are most likely to be submissions.
2490 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2492 (1) Generate an error if the third argument for the ${prvs expansion is
2495 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2498 (3) Reset the variables that are obtained from the first argument of
2499 ${prvscheck and used in the second argument before leaving the code,
2500 because their memory is reclaimed, so using them afterwards may do
2503 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2504 one (it's much easier than Tom thought :-).
2506 (5) Because of (4), we can now allow for the use of $prvscheck_result
2507 inside the third argument.
2509 PH/03 For some reason, the default setting of PATH when running a command from
2510 a pipe transport was just "/usr/bin". I have changed it to
2513 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2514 anything to be listed in the output from -bV.
2516 PH/05 When a filter generated an autoreply, the entire To: header line was
2517 quoted in the delivery log line, like this:
2519 => >A.N.Other <ano@some.domain> <original@ddress> ...
2521 This has been changed so that it extracts the operative address. There
2522 may be more than one such address. If so, they are comma-separated, like
2525 => >ano@some.domain,ona@other.domain <original@ddress> ...
2527 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2528 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2529 not being looked up in the rDNS to get a host name, Exim was showing the
2530 IP address twice in Received: lines, even though the IP addresses were
2531 identical. For example:
2533 Received: from [1.2.3.4] (helo=[1.2.3.4])
2535 However, if the real host name was known, it was omitting the HELO data
2536 if it matched the actual IP address. This has been tidied up so that it
2537 doesn't show the same IP address twice.
2539 PH/07 When both +timestamp and +memory debugging was on, the value given by
2540 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2541 called by the debug printing, thereby overwriting the timestamp buffer.
2542 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2544 PH/08 When the original message was included in an autoreply transport, it
2545 always said "this is a copy of the message, including all the headers",
2546 even if body_only or headers_only was set. It now gives an appropriate
2549 PH/09 Applied a patch from the Sieve maintainer which:
2551 o fixes some comments
2552 o adds the (disabled) notify extension core
2553 o adds some debug output for the result of if/elsif tests
2554 o points to the current vacation draft in the documentation
2555 and documents the missing references header update
2559 o fixes a bug in processing the envelope test (when testing
2560 multiple envelope elements, the last element determinted the
2563 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2564 Electronic Mail") by including:
2566 Auto-submitted: auto-generated
2568 in the messages that it generates (bounce messages and others, such as
2569 warnings). In the case of bounce messages for non-SMTP mesages, there was
2570 also a typo: it was using "Auto_submitted" (underscore instead of
2571 hyphen). Since every message generated by Exim is necessarily in response
2572 to another message, thes have all been changed to:
2574 Auto-Submitted: auto-replied
2576 in accordance with these statements in the RFC:
2578 The auto-replied keyword:
2580 - SHOULD be used on messages sent in direct response to another
2581 message by an automatic process,
2583 - MUST NOT be used on manually-generated messages,
2585 - MAY be used on Delivery Status Notifications (DSNs) and Message
2586 Disposition Notifications (MDNs),
2588 - MUST NOT be used on messages generated by automatic or periodic
2589 processes, except for messages which are automatic responses to
2592 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2593 to the default Received: header definition.
2595 PH/12 Added log selector acl_warn_skipped (default on).
2597 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2598 variables because (a) they are in the wrong storage pool and (b) even if
2599 they were copied, it wouldn't work properly because of the caching.
2601 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2602 checking when decoding. Apparently there are clients that generate
2603 overlong encoded strings. Why am I not surprised?
2605 PH/15 If the first argument of "${if match_address" was not empty, but did not
2606 contain an "@" character, Exim crashed. Now it writes a panic log message
2607 and treats the condition as false.
2609 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2611 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2612 extension "envelope-auth". The code is finished and in agreement with
2613 other implementations, but there is no documentation so far and in fact,
2614 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2615 not changing the active code.
2617 Print executed "if" and "elsif" statements when debugging is used. This
2618 helps a great deal to understand what a filter does.
2620 Document more things not specified clearly in RFC3028. I had all this
2621 sorted out, when out of a sudden new issues came to my mind. Oops."
2623 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2626 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2627 canonical form (as documented). However, after a host name lookup from
2628 the IP address, check_host() was doing a simple string comparison with
2629 addresses acquired from the DNS when checking that the found name did
2630 have the original IP as one of its addresses. Since any found IPv6
2631 addresses are likely to be in abbreviated form, the comparison could
2632 fail. Luckily, there already exists a function for doing the comparison
2633 by converting both addresses to binary, so now that is used instead of
2634 the text comparison.
2636 PH/20 There was another similar case to PH/19, when a complete host name was
2637 given in a host list; looking up its IP address could give an abbreviated
2638 form, whereas the current host's name might or might not be abbreviated.
2639 The same fix has been applied.
2645 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2646 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2649 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2650 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2652 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2654 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2655 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2656 Also, for "Auto-Submitted", treat anything other than "no" as
2657 non-personal, in accordance with RFC 3834. (Previously it treated
2658 anything starting "auto-" as non-personal.)
2660 TF/01 The control=submission/name=... option had a problem with syntax
2661 errors if the name included a slash character. The /name= option
2662 now slurps the rest of the string, so it can include any characters
2663 but it must come last in the list of options (after /sender_retain
2666 PH/05 Some modifications to the interface to the fake nameserver for the new
2674 TK/01 Added the "success_on_redirect" address verification option. See
2675 NewStuff for rationale and an example.
2677 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2679 PH/02 Patch to exigrep to allow it to work on syslog lines.
2681 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2682 fread() to skip over the body file's header line, because in Cygwin the
2683 header line is locked and is inaccessible.
2685 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2686 co-exist for some time) to make it clear that it is the Exim ID that is
2687 referenced, not the Message-ID: header line.
2689 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2690 string_format() function, because snprintf() does not exist on all
2693 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2694 ${stat: expansion item.
2696 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2697 protocol synchronization error", to keep the pedants happy.
2699 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2700 well as for IRIX systems, when gcc is being used. See the host.c source
2703 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2705 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2708 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2709 transport and to the smtp transport in LMTP mode.
2711 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2713 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2714 run a filter in a subprocess. This could lead to confusion in subsequent
2715 lookups in the parent process. There should also be a search_tidyup() at
2716 the end of the subprocess.
2718 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2719 only if the host matched helo_try_verify_hosts, which caused the
2720 verification to occur when the EHLO/HELO command was issued. The ACL just
2721 tested the remembered result. Now, if a previous verification attempt has
2722 not happened, "verify = helo" does it there and then.
2724 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2726 TK/03 Fix log output including CR from clamd.
2728 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2729 exist provoked a memory error which could cause a segfault.
2731 PH/15 Installed PCRE 6.2
2733 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2735 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2736 of the problem. Specifically, suggested +O2 rather than +O1 for the
2739 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2741 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2742 record, the port was not correctly logged when the outgoing_port log
2743 selector was set (it logged the transort's default port).
2745 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2746 fallback_hosts, and "hosts" in the smtp transport.
2748 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2749 host errors such as "Connection refused".
2751 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2752 authentication with radiusclient 0.4.9:
2754 - Error returned from rc_read_config was caught wrongly
2755 - Username/password not passed on to radius server due to wrong length.
2757 The presumption is that some radiusclient API changes for 4.51/PH/17
2758 were not taken care of correctly. The code is still untested by me (my
2759 Linux distribution still has 0.3.2 of radiusclient), but it was
2760 contributed by a Radius user.
2762 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2763 expanding the "port" option of the smtp transport.
2765 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2766 while reading a MIME header. Thanks to Tom Hughes for a patch.
2768 PH/24 Include config.h inside local_scan.h so that configuration settings are
2771 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2772 that in an ACL for RCPT (for example), you can examine exactly what was
2775 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2776 commands, but it was not correctly comparing the address with the actual
2777 client host address. Thus, it would show the EHLO address in Received:
2778 header lines when this was not necessary.
2780 PH/27 Added the % operator to ${eval:}.
2782 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2783 it should be ignoring failures (because with -C, for example, it has lost
2784 privilege). It wasn't ignoring creation failures other than "already
2787 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2790 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2791 because an input line was too long, either on its own, or by virtue of
2792 too many continuations, the temporary file was not being removed, and the
2793 return code was incorrect.
2795 PH/31 Missing "BOOL" in function definition in filtertest.c.
2797 PH/32 Applied Sieve patches from the maintainer.
2799 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2801 PH/33 Added "verify = not_blind".
2803 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2804 Local/Makefile (with some defaults set). These are used in built scripts
2805 such as exicyclog, but they have never been used in the exim_install
2806 script (though there are many overriding facilities there). I have
2807 arranged that the exim_install script now takes note of these two
2810 PH/35 Installed configuration files for Dragonfly.
2812 PH/36 When a locally submitted message by a trusted user did not contain a
2813 From: header, and the sender address was obtained from -f or from an SMTP
2814 MAIL command, and the trusted user did not use -F to supply a sender
2815 name, $originator_name was incorrectly used when constructing a From:
2816 header. Furthermore, $originator_name was used for submission mode
2817 messages from external hosts without From: headers in a similar way,
2818 which is clearly wrong.
2820 PH/37 Added control=suppress_local_fixups.
2822 PH/38 When log_selector = +received_sender was set, and the addition of the
2823 sender made the log line's construction buffer exactly full, or one byte
2824 less than full, an overflow happened when the terminating "\n" was
2827 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2828 when the result of a list match is failure because a DNS lookup failed.
2830 PH/40 RM_COMMAND is now used in the building process.
2832 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2833 the "build-* directories that it finds.
2835 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2836 address in a domain literal was a prefix of an interface address.
2838 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2839 when verifying a sender address, unless rewrite_headers is false.
2841 PH/44 Wrote a long comment about why errors_to addresses are verified as
2842 recipients, not senders.
2844 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2845 the ratelimit ACL was added.
2847 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2849 PH/46 Added extra information about PostgreSQL errors to the error string.
2851 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2852 avoiding the need to install special zones in a real server. This is
2853 backwards compatible; if it can't find the fake resolver, it drops back.
2854 Thus, both old and new test suites can be run.
2856 TF/02 Added util/ratelimit.pl
2858 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2859 clock is set back in time.
2861 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2862 Candler <B.Candler@pobox.com>.
2864 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2865 OK for addresses that are the result of redirections.
2867 PH/48 A number of further additions for the benefit of the new test suite,
2868 including a fake gethostbyname() that interfaces to the fake DNS resolver
2871 TF/06 The fix for widen_domains has also been applied to qualify_single and
2872 search_parents which are the other dnslookup options that can cause
2875 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2878 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2880 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2881 service" if it encountered an MX record with an empty target hostname.
2882 The message is now "an MX or SRV record indicated no SMTP service".
2884 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2885 if the DNS of the sending site is misconfigured. This is quite a
2886 common situation. This change restores the behaviour of treating a
2887 helo verification defer as a failure.
2889 PH/51 If self=fail was set on a router, the bounce message did not include the
2890 actual error message.
2896 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2898 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2899 command itself ended in error, the underlying message about the transport
2900 filter timeout was being overwritten with the pipe command error. Now the
2901 underlying error message should be appended to the second error message.
2903 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2905 PH/02 Reduce the amount of output that "make" produces by default. Full output
2906 can still be requested.
2908 PH/03 The warning log line about a condition test deferring for a "warn" verb
2909 was being output only once per connection, rather than after each
2910 occurrence (because it was using the same function as for successful
2911 "warn" verbs). This seems wrong, so I have changed it.
2913 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2914 it should not have, which might have caused a crash in the right
2915 circumstances, but probably never did.
2917 PH/04 Installed a modified version of Tony Finch's patch to make submission
2918 mode fix the return path as well as the Sender: header line, and to
2919 add a /name= option so that you can make the user's friendly name appear
2922 TF/03 Added the control = fakedefer ACL modifier.
2924 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2925 Mark Lowes for thorough testing.
2927 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2929 TK/03 Merged latest SRS patch from Miles Wilton.
2931 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2932 with the definition in sysexits.h (which is #included earlier).
2933 Fortunately, Exim does not actually use EX_OK. The code used to try to
2934 preserve the sysexits.h value, by assumimg that macro definitions were
2935 scanned for macro replacements. I have been disabused of this notion,
2936 so now the code just undefines EX_OK before #including unistd.h.
2938 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2939 in the smtp transport. When a block could not be written in a single
2940 write() function, the timeout was being re-applied to each part-write.
2941 This seems wrong - if the receiver was accepting one byte at a time it
2942 would take for ever. The timeout is now adjusted when this happens. It
2943 doesn't have to be particularly precise.
2945 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2946 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2948 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2949 without a domain if the check to <postmaster@domain> fails.
2951 SC/01 Eximstats: added -xls and the ability to specify output files
2952 (patch written by Frank Heydlauf).
2954 SC/02 Eximstats: use FileHandles for outputing results.
2956 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2958 SC/04 Eximstats: fixed display of large numbers with -nvr option
2960 SC/05 Eximstats: fixed merging of reports with empty tables.
2962 SC/06 Eximstats: added the -include_original_destination flag
2964 SC/07 Eximstats: removed tabs and trailing whitespace.
2966 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2968 TK/06 MBOX spool code: Add real "From " MBOX separator line
2969 so the .eml file is really in mbox format (even though
2970 most programs do not really care). Patch from Alex Miller.
2972 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2973 The latter is generated from $received_to and is only set if the
2974 message has one envelope recipient. SA can use these headers,
2975 obviously out-of-the-box. Patch from Alex Miller.
2977 PH/08 The ${def test on a variable was returning false if the variable's
2978 value was "0", contrary to what the specification has always said!
2979 The result should be true unless the variable is empty.
2981 PH/09 The syntax error of a character other than { following "${if
2982 def:variable_name" (after optional whitespace) was not being diagnosed.
2983 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2984 accidental colon was present, for example, could give incorrect results.
2986 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2987 result is used (not including appendfile, where other changes are about
2990 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2991 This involved changing a lot of size variables from int to off_t. It
2992 should work with maildirs and everything.
2994 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2995 spamd dying while we are connected to it.
2997 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3000 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3001 function for BDB 4.3.
3003 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3005 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3006 the sources for PCRE have been re-arranged and more files are now
3009 PH/15 The code I had for printing potentially long long variables in PH/11
3010 above was not the best (it lost precision). The length of off_t variables
3011 is now inspected at build time, and an appropriate printing format (%ld
3012 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3013 to be "long long int" or "long int". This is needed for the internal
3014 formatting function string_vformat().
3016 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3017 the configuration file to be ":syslog", then the script "guesses" where
3018 the logs files are, rather than using the compiled in default. In our
3019 case the guess is not the same as the compiled default, so the script
3020 suddenly stopped working when I started to use syslog. The patch checks
3021 to see if log_file_path is "". If so, it attempts to read it from exim
3022 with no configuration file to get the compiled in version, before it
3023 falls back to the previous guessing code."
3025 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3026 implementing BATV in an Exim configuration. See NewStuff for the gory
3029 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3030 Makefile that are specific to HP-UX.
3032 PH/18 If the "use_postmaster" option was set for a recipient callout together
3033 with the "random" option, the postmaster address was used as the MAIL
3034 FROM address for the random test, but not for the subsequent recipient
3035 test. It is now used for both.
3037 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3038 patch removes a few documentation additions to RFC 3028, because the
3039 latest draft now contains them. It adds the new en;ascii-case comparator
3040 and a new error check for 8bit text in MIME parts. Comparator and
3041 require names are now matched exactly. I enabled the subaddress
3042 extension, but it is not well tested yet (read: it works for me)."
3044 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3045 rework some of the code of TK/09 above to avoid the hardwired use of
3046 "%lld" and "long long". Replaced the call to snprintf() with a call to
3049 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3050 records point to non-existent hosts", "retry timeout exceeded", and
3051 "retry time not reached for any host after a long failure period".
3053 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3054 experimental DomainKeys support:
3056 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3057 (2) On an error such as an illegally used "control", the wrong name for
3058 the control was given.
3060 These problems did NOT occur unless DomainKeys support was compiled.
3062 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3064 PH/24 Added ${if match_ip condition.
3066 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3067 databases so that it will be absolutely obvious if a crash occurs in the
3068 DB library. This is a regular occurrence (often caused by mis-matched
3071 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3072 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3073 picked up on a user's system that detects such things. There doesn't seem
3074 to be a gcc warning option for this - only an attribute that has to be
3075 put on the function's prototype. It seems that in Fedora Core 4 they have
3076 set this on a number of new functions. No doubt there will be more in due
3079 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3080 specify a transport. However, if an address that was verified by such a
3081 router was the subject of a callout, Exim crashed because it tried to
3082 read the rcpt_include_affixes from the non-existent transport. Now it
3083 just assumes that the setting of that option is false. This bug was
3084 introduced by 4.51/PH/31.
3086 PH/28 Changed -d+all to exclude +memory, because that information is very
3087 rarely of interest, but it makes the output a lot bigger. People tend to
3088 do -d+all out of habit.
3090 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3091 code in os-type was giving problems when libc.so lives in lib64, like on
3094 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3095 aren't the modern standard, and it seems that some systems' include files
3096 don't always have them. Exim was already checking for some of the newer
3097 ones like T_AAAA, and defining it itself. I've added checks for all the
3098 record types that Exim uses.
3100 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3101 not automatically generating a new one, as it is supposed to. This
3102 prevented TLS from working. If the file did exist, but contained invalid
3103 data, a new version was generated, as expected. It was only the case of a
3104 non-existent file that was broken.
3106 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3107 with a change in libdomainkeys > 0.64.
3109 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3110 from DNS. If the selector record carries the flag, it now has
3111 precedence over the domain-wide flag.
3113 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3115 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3116 the use of a port name that isn't defined in /etc/services) occurred, the
3117 message was deferred as in a normal delivery, and thus remained on the
3118 spool, instead of being failed because of the mua_wrapper setting. This
3119 is now fixed, and I tidied up some of the mua_wrapper messages at the
3122 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3123 the messages in a hash of arrays rather than using individual hashes.
3124 This is a bit cleaner and results in dramatic memory savings, albeit
3125 at a slight CPU cost.
3127 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3128 as requested by Marc Sherman.
3130 SC/10 Eximstats: added histograms for user specified patterns as requested
3133 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3135 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3136 fopen() in the content-scanning modules that did not already have it.
3142 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3143 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3145 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3147 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3149 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3150 to test Sieve filters that use "vacation".
3152 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3153 that changes the way the GnuTLS parameters are stored in the cache file.
3154 The new format can be generated externally. For backward compatibility,
3155 if the data in the cache doesn't make sense, Exim assumes it has read an
3156 old-format file, and it generates new data and writes a new file. This
3157 means that you can't go back to an older release without removing the
3160 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3161 work if there are any delivery delays because "one_time" forces the
3162 parent to be marked "delivered", so its unseen clone is never tried
3163 again. For this reason, Exim now forbids the simultaneous setting of
3166 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3167 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3168 entry if you want to know the details. The fix, however, neglected to
3169 consider the case when local delivery batching is involved. The test for
3170 "previously delivered" was not happening when checking to see if an
3171 address could be batched with a previous (undelivered) one; under
3172 certain circumstances this could lead to multiple deliveries to the same
3175 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3176 in its include files, and this causes problems building Exim.
3178 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3179 header_syntax) but Exim was just ignoring anything given after a slash.
3180 In particular, this caused confusion with an attempt to use "verify =
3181 reverse_host_lookup/defer_ok". An error is now given when options are
3182 supplied for verify items that do not have them. (Maybe reverse_host_
3183 lookup should have a defer_ok option, but that's a different point.)
3185 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3186 defined by RFC 821) to 2048, because there were problems with some AUTH
3187 commands, and RFC 1869 says the size should be increased for extended
3188 SMTP commands that take arguments.
3190 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3193 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3194 "unknown" error; now it says that the functionality isn't in the binary.
3196 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3197 an address' error message when a string expansion fails (syntax or
3198 whatever). Otherwise the password may appear in the log. Following change
3199 PH/42 below, there is no longer a chance of it appearing in a bounce
3202 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3204 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3205 of its addresses was ever tried. (Bugzilla bug #2.)
3207 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3208 the result incorrectly in the debug output. (It correctly added a newline
3209 to what was transported.)
3211 TF/01 Added $received_time.
3213 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3214 commented out examples of how to interface to a virus scanner and to
3215 SpamAssassin. Also added commented examples of av_scanner and
3216 spamd_address settings.
3218 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3219 and controls are allowed in which ACLs. There were a couple of minor
3220 errors. Some of the entries in the conditions table (which is a table of
3221 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3222 negation of where the condition IS allowed.
3224 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3226 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3227 header file does not have a version number, so I've had to invent a new
3228 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3229 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3230 radiusclient), but it was contributed by a Radius user.
3232 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3233 files or directories, for OpenSSL.
3235 PH/19 When an Exim process that is running as root has to create an Exim log
3236 file, it does so in a subprocess that runs as exim:exim so as to get the
3237 ownership right at creation (otherwise, other Exim processes might see
3238 the file with the wrong ownership). There was no test for failure of this
3239 fork() call, which would lead to the process getting stuck as it waited
3240 for a non-existent subprocess. Forks do occasionally fail when resources
3241 run out. I reviewed all the other calls to fork(); they all seem to check
3244 PH/20 When checking for unexpected SMTP input at connect time (before writing
3245 the banner), Exim was not dealing correctly with a non-positive return
3246 from the read() function. If the client had disconnected by this time,
3247 the result was a log entry for a synchronization error with an empty
3248 string after "input=" when read() returned zero. If read() returned -1
3249 (an event I could not check), uninitialized data bytes were printed.
3250 There were reports of junk text (parts of files, etc) appearing after
3253 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3255 PH/22 Added support for macro redefinition, and (re)definition in between
3256 driver and ACL definitions.
3258 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3259 forgetting to use the resulting value; it was using the unexpanded value.
3261 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3262 hadn't been configured. The fix is from Juergen Kreileder, who
3263 understands it better than I do:
3265 "Here's what I see happening with three configured cyrus_sasl
3266 authenticators configured (plain, login, cram-md5):
3268 On startup auth_cyrus_sasl_init() gets called for each of these.
3269 This means three calls to sasl_listmech() without a specified mech_list.
3270 => SASL tests which mechs of all available mechs actually work
3271 => three warnings about OTP not working
3272 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3274 With the patch, sasl_listmech() also gets called three times. But now
3275 SASL's mech_list option is set to the server_mech specified in the the
3276 authenticator. Or in other words, the answer from sasl_listmech()
3277 gets limited to just the mech you're testing for (which is different
3279 => the return list contains just 'plain' or 'login', 'cram-md5' or
3280 nothing depending on the value of ob->server_mech.
3282 I've just tested the patch: Authentication still works fine,
3283 unavailable mechs specified in the exim configuration are still
3284 caught, and the auth.log warnings about OTP are gone."
3286 PH/25 When debugging is enabled, the contents of the command line are added
3287 to the debugging output, even when log_selector=+arguments is not
3290 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3291 answer is "GNU", and only if the return is "GNU/something" is the answer
3294 PH/27 $acl_verify_message is now set immediately after the failure of a
3295 verification in an ACL, and so is available in subsequent modifiers. In
3296 particular, the message can be preserved by coding like this:
3298 warn !verify = sender
3299 set acl_m0 = $acl_verify_message
3301 Previously, $acl_verify_message was set only while expanding "message"
3302 and "log_message" when a very denied access.
3304 PH/28 Modified OS/os.c-Linux with
3306 -#ifndef OS_LOAD_AVERAGE
3307 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3309 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3310 nomenclature these days.)
3312 PH/29 Installed patch from the Sieve maintainer that adds the options
3313 sieve_useraddress and sieve_subaddress to the redirect router.
3315 PH/30 In these circumstances:
3316 . Two addresses routed to the same list of hosts;
3317 . First host does not offer TLS;
3318 . First host accepts first address;
3319 . First host gives temporary error to second address;
3320 . Second host offers TLS and a TLS session is established;
3321 . Second host accepts second address.
3322 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3323 and peerdn, if requested) that were in fact used only for the second
3326 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3327 attention to any local part prefix or suffix that was matched by the
3328 router that accepted the address. It now behaves in the same way as it
3329 does for delivery: the affixes are removed from the local part unless
3330 rcpt_include_affixes is set on the transport.
3332 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3333 timeout during the DATA phase of an incoming message.
3335 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3336 applied a patch sanctioned by the Sieve maintainer.
3338 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3339 the uid or gid is negative. A case of a negative gid caused this to be
3340 noticed. The fix allows for either to be negative.
3342 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3343 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3346 PH/36 The change PH/12 above was broken. Fixed it.
3348 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3349 the grounds that routing the same address twice would always produce the
3350 same answer. This might have been true once, but it is certainly no
3351 longer true now. Routing a child address may depend on the previous
3352 routing that produced that child. Some complicated redirection strategies
3353 went wrong when messages had multiple recipients, and made Exim's
3354 behaviour dependent on the order in which the addresses were given.
3356 I have moved the duplicate checking until after the routing is complete.
3357 Exim scans the addresses that are assigned to local and remote
3358 transports, and removes any duplicates. This means that more work will be
3359 done, as duplicates will always all be routed, but duplicates are
3360 presumably rare, so I don't expect this is of any significance.
3362 For deliveries to pipes, files, and autoreplies, the duplicate checking
3363 still happens during the routing process, since they are not going to be
3366 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3367 It corrects a timeout issue with spamd. This is Ian's comment: "The
3368 background is that sometimes spamd either never reads data from a
3369 connection it has accepted, or it never writes response data. The exiscan
3370 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3371 blindly assumes that writes won't block so it may never time out."
3373 PH/39 Allow G after quota size as well as K and M.
3375 PH/40 The value set for $authenticated_id in an authenticator may not contain
3376 binary zeroes or newlines because the value is written to log lines and
3377 to spool files. There was no check on this. Now the value is run through
3378 the string_printing() function so that such characters are converted to
3379 printable escape sequences.
3381 PH/41 $message_linecount is a new variable that contains the total number of
3382 lines in the message. Compare $body_linecount, which is the count for the
3385 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3386 bounce and delay warning messages, except in certain special cases, which
3389 (a) An SMTP error message from a remote host;
3390 (b) A message specified in a :fail: redirection;
3391 (c) A message specified in a "fail" command in a system filter;
3392 (d) A message specified in a FAIL return from the queryprogram router;
3393 (e) A message specified by the cannot_route_message router option.
3395 In these cases only, Exim does include the error details in bounce and
3396 warning messages. There are also a few cases where bland messages such
3397 as "unrouteable address" or "local delivery error" are given.
3399 PH/43 $value is now also set for the "else" part of a ${run expansion.
3401 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3402 being worked on, but at least Exim now implements the latest version to
3405 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3406 process to complete was treated as a delivery failure, a timeout while
3407 writing the message to the pipe was logged, but erroneously treated as a
3408 successful delivery. Such timeouts include transport filter timeouts. For
3409 consistency with the overall process timeout, these timeouts are now
3410 treated as errors, giving rise to delivery failures by default. However,
3411 there is now a new Boolean option for the pipe transport called
3412 timeout_defer, which, if set TRUE, converts the failures into defers for
3413 both kinds of timeout. A transport filter timeout is now identified in
3416 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3417 systems where "make" and "gmake" are different, calling "gmake" at top
3418 level broke things. I've arranged for the value of $(MAKE) to be passed
3419 from the Makefile to this script so that it can call the same version of
3423 A note about Exim versions 4.44 and 4.50
3424 ----------------------------------------
3426 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3427 changes of various kinds. As a consequence, a big documentation update was
3428 needed. This delayed the release for rather longer than seemed good, especially
3429 in the light of a couple of (minor) security issues. Therefore, the changes
3430 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3431 release. So 4.44 and 4.50 are in effect two different branches that both start
3434 I have left the 4.50 change log unchanged; it contains all the changes since
3435 4.43. The change log for 4.44 is below; many of its items are identical to
3436 those for 4.50. This seems to be the most sensible way to preserve the
3437 historical information.
3443 1. Minor wording change to the doc/README.SIEVE file.
3445 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3446 computation of the current number of files was incorrect.
3448 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3449 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3452 4. Give more explanation in the error message when the command for a transport
3453 filter fails to execute.
3455 5. There are several places where Exim runs a non-Exim command in a
3456 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3457 was being done only for the command run by the queryprogram router. It is
3458 now done for all such subprocesses. The other cases are: ${run, transport
3459 filters, and the commands run by the lmtp and pipe transports.
3461 6. Added CONFIGURE_GROUP build-time option.
3463 7. Some older OS have a limit of 256 on the maximum number of file
3464 descriptors. Exim was using setrlimit() to set 1000 as a large value
3465 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3466 systems. I've change it so that if it can't get 1000, it tries for 256.
3468 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3469 was an oversight, and furthermore, ever since the addition of extra
3470 controls (e.g. 4.43/32), the checks on when to allow different forms of
3471 "control" were broken. There should now be diagnostics for all cases when a
3472 control that does not make sense is encountered.
3474 9. Added the /retain_sender option to "control=submission".
3476 10. $recipients is now available in the predata ACL (oversight).
3478 11. Tidy the search cache before the fork to do a delivery from a message
3479 received from the command line. Otherwise the child will trigger a lookup
3480 failure and thereby defer the delivery if it tries to use (for example) a
3481 cached ldap connection that the parent has called unbind on.
3483 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3484 of $address_data from the recipient verification was clobbered by the
3485 sender verification.
3487 13. The value of address_data from a sender verification is now available in
3488 $sender_address_data in subsequent conditions in the ACL statement.
3490 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3492 15. Added a new option "connect=<time>" to callout options, to set a different
3495 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3496 was its contents. (It was OK if the option was not defined at all.)
3498 17. A "Completed" log line is now written for messages that are removed from
3499 the spool by the -Mrm option.
3501 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3502 information about exactly what failed.
3504 19. Added -dd to debug only the daemon process.
3506 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3507 handles timeouts, both on the server side and network timeouts. Renamed the
3508 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3510 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3511 It is now set to "smtps".
3513 22. $host_address is now set to the target address during the checking of
3514 ignore_target_hosts.
3516 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3517 being passed; this would have caused $sender_host_name to have been used if
3518 matching the list had actually called for a host name (not very likely,
3519 since this list is usually IP addresses). A host name is now passed as
3522 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3523 code that specifies a non-restarting handler (typically sigaction() in
3524 modern systems) in an attempt to fix a rare and obscure crash bug.
3526 25. Narrowed the window for a race in the daemon that could cause it to ignore
3527 SIGCHLD signals. This is not a major problem, because they are used only to
3528 wake it up if nothing else does.
3530 26. A malformed maildirsize file could cause Exim to calculate negative values
3531 for the mailbox size or file count. Odd effects could occur as a result.
3532 The maildirsize information is now recalculated if the size or filecount
3535 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3536 support for a long time. Removed HAVE_SYS_VFS_H.
3538 28. Installed the latest version of exipick from John Jetmore.
3540 29. In an address list, if the pattern was not a regular expression, an empty
3541 subject address (from a bounce message) matched only if the pattern was an
3542 empty string. Non-empty patterns were not even tested. This was the wrong
3543 because it is perfectly reasonable to use an empty address as part of a
3544 database query. An empty address is now tested by patterns that are
3545 lookups. However, all the other forms of pattern expect the subject to
3546 contain a local part and a domain, and therefore, for them, an empty
3547 address still always fails if the pattern is not itself empty.
3549 30. Exim went into a mad DNS loop when attempting to do a callout where the
3550 host was specified on an smtp transport, and looking it up yielded more
3551 than one IP address.
3553 31. Re-factored the code for checking spool and log partition space into a
3554 function that finds that data and another that does the check. The former
3555 is then used to implement four new variables: $spool_space, $log_space,
3556 $spool_inodes, and $log_inodes.
3558 32. The RFC2047 encoding function was originally intended for short strings
3559 such as real names; it was not keeping to the 75-character limit for
3560 encoded words that the RFC imposes. It now respects the limit, and
3561 generates multiple encoded words if necessary. To be on the safe side, I
3562 have increased the buffer size for the ${rfc2047: expansion operator from
3565 33. It is now permitted to omit both strings after an "if" condition; if the
3566 condition is true, the result is "true". As before, when the second string
3567 is omitted, a false condition yields an empty string. This makes it less
3568 cumbersome to write custom ACL and router conditions.
3570 34. Failure to deliver a bounce message always caused it to be frozen, even if
3571 there was an errors_to setting on the router. The errors_to setting is now
3574 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3575 canonical form (fully expanded) before being placed in
3576 $sender_host_address.
3578 36. The table in the code that translates DNS record types into text (T_A to
3579 "A" for instance) was missing entries for NS and CNAME. It is just possible
3580 that this could have caused confusion if both these types were looked up
3581 for the same domain, because the text type is used as part of Exim's
3582 per-process caching. But the chance of anyone hitting this buglet seems
3585 37. The dnsdb lookup has been extended in a number of ways.
3587 (1) There is a new type, "zns", which walks up the domain tree until it
3588 finds some nameserver records. It should be used with care.
3590 (2) There is a new type, "mxh", which is like "mx" except that it returns
3591 just the host names, not the priorities.
3593 (3) It is now possible to give a list of domains (or IP addresses) to be
3594 looked up. The behaviour when one of the lookups defers can be
3595 controlled by a keyword.
3597 (4) It is now possible to specify the separator character for use when
3598 multiple records are returned.
3600 38. The dnslists ACL condition has been extended: it is now possible to supply
3601 a list of IP addresses and/or domains to be looked up in a particular DNS
3604 39. Added log_selector=+queue_time_overall.
3606 40. When running the queue in the test harness, wait just a tad after forking a
3607 delivery process, to get repeatability of debugging output.
3609 41. Include certificate and key file names in error message when GnuTLS fails
3610 to set them up, because the GnuTLS error message doesn't include the name
3611 of the failing file when there is a problem reading it.
3613 42. Allow both -bf and -bF in the same test run.
3615 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3617 44. The "Exiscan patch" is now merged into the mainline Exim source.
3619 45. Sometimes the final signoff response after QUIT could fail to get
3620 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3621 < 0 before doing a fflush(). This bug looks as though it goes back to the
3622 introduction of TLS in release 3.20, but "sometimes" must have been rare
3623 because the tests only now provoked it.
3625 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3626 (this can affect the format of dates).
3628 47. exim_tidydb, when checking for the continued existence of a message for
3629 which it has found a message-specific retry record, was not finding
3630 messages that were in split spool directories. Consequently, it was
3631 deleting retry records that should have stayed in existence.
3633 48. Steve fixed some bugs in eximstats.
3635 49. The SPA authentication driver was not abandoning authentication and moving
3636 on to the next authenticator when an expansion was forced to fail,
3637 contradicting the general specification for all authenticators. Instead it
3638 was generating a temporary error. It now behaves as specified.
3640 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3641 (the order specifies the preference for clients). The order is now AES256,
3642 AES128, 3DES, ARCFOUR128.
3644 51. Small patch to Sieve code - explicitly set From: when generating an
3647 52. Exim crashed if a remote delivery caused a very long error message to be
3648 recorded - for instance if somebody sent an entire SpamAssassin report back
3649 as a large number of 550 error lines. This bug was coincidentally fixed by
3650 increasing the size of one of Exim's internal buffers (big_buffer) that
3651 happened as part of the Exiscan merge. However, to be on the safe side, I
3652 have made the code more robust (and fixed the comments that describe what
3655 53. Now that there can be additional text after "Completed" in log lines (if
3656 the queue_time_overall log selector is set), a one-byte patch to exigrep
3657 was needed to allow it to recognize "Completed" as not the last thing in
3660 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3661 patch that reportedly fixes this has been added. I am not expert enough to
3662 create a test for it. This is what the patch creator wrote:
3664 "I found a little strange behaviour of ldap code when working with
3665 Windows 2003 AD Domain, where users was placed in more than one
3666 Organization Units. When I tried to give exim partial DN, the exit code
3667 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3668 But simultaneously result of request was absolutely normal ldap result,
3669 so I produce this patch..."
3671 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3672 REFERENCE, so I have modified the code to exclude the patch when that macro
3675 55. Some experimental protocols are using DNS PTR records for new purposes. The
3676 keys for these records are domain names, not reversed IP addresses. The
3677 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3678 leaves it alone. Component reversal etc. now happens only for IP addresses.
3681 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3683 57. Double the size of the debug message buffer (to 2048) so that more of very
3684 long debug lines gets shown.
3686 58. The exicyclog utility now does better if the number of log files to keep
3687 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3689 59. Two changes related to the smtp_active_hostname option:
3691 (1) $smtp_active_hostname is now available as a variable.
3692 (2) The default for smtp_banner uses $smtp_active_hostname instead
3693 of $primary_hostname.
3695 60. The host_aton() function is supposed to be passed a string that is known
3696 to be a valid IP address. However, in the case of IPv6 addresses, it was
3697 not checking this. This is a hostage to fortune. Exim now panics and dies
3698 if the condition is not met. A case was found where this could be provoked
3699 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3700 components; fortuitously, this particular loophole had already been fixed
3701 by change 4.50/55 above.
3703 If there are any other similar loopholes, the new check in host_aton()
3704 itself should stop them being exploited. The report I received stated that
3705 data on the command line could provoke the exploit when Exim was running as
3706 exim, but did not say which command line option was involved. All I could
3707 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3708 running as the user.
3711 61. There was a buffer overflow vulnerability in the SPA authentication code
3712 (which came originally from the Samba project). I have added a test to the
3713 spa_base64_to_bits() function which I hope fixes it.
3716 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3717 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3719 63. The daemon start-up calls getloadavg() while still root for those OS that
3720 need the first call to be done as root, but it missed one case: when
3721 deliver_queue_load_max is set with deliver_drop_privilege. This is
3722 necessary for the benefit of the queue runner, because there is no re-exec
3723 when deliver_drop_privilege is set.
3725 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3726 This has been fixed.
3728 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3729 was in use, was not putting the data itself into the right store pool;
3730 consequently, it could be overwritten for a subsequent message in the same
3731 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3734 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3736 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3737 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3738 the function were treating the return as a boolean value, which happened to
3739 work because 0=false and not-0=true, but is not correct code.
3741 68. The host_aton() function was not handling scoped IPv6 addresses (those
3742 with, for example, "%eth0" on the end) correctly.
3744 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3745 negated items (that is, ~something) in unsigned ints. Some compilers
3746 apparently mutter when there is no cast.
3748 70. If an address verification called from an ACL failed, and did not produce a
3749 user-specific message (i.e. there was only a "system" message), nothing was
3750 put in $acl_verify_message. In this situation, it now puts the system
3753 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3754 session; change 4.31/43 added the unwanted input to the log line - except
3755 that it did not do this in the start of session case. It now does.
3757 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3758 This is wrong and can cause the other end to generate a synchronization
3759 error if it is another Exim or anything else that does the synchronization
3760 check. A QUIT command is no longer sent after a timeout.
3762 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3763 during host lookups.
3765 74. The defer_ok option of callout verification was not working if it was used
3766 when verifying addresses in header lines, that is, for this case:
3768 verify = header_sender/callout=defer_ok
3770 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3771 those file descriptors could be used for SMTP connections. If anything
3772 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3773 could be sent to the SMTP client, causing chaos. The daemon now opens
3774 stdin, stdout, and stderr to /dev/null when it puts itself into the
3777 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3778 log by default. The user can override this with suitable Perl magic.
3780 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3781 the log message when discard triggers, was not working for the DATA ACL or
3782 for the non-SMTP ACL.
3784 78. Error message wording change in sieve.c.
3786 79. If smtp_accept_max_per_host was set, the number of connections could be
3787 restricted to fewer than expected, because the daemon was trying to set up
3788 a new connection before checking whether the processes handling previous
3789 connections had finished. The check for completed processes is now done
3790 earlier. On busy systems, this bug wouldn't be noticed because something
3791 else would have woken the daemon, and it would have reaped the completed
3794 80. If a message was submitted locally by a user whose login name contained one
3795 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3796 It caused a spool format error. I have fixed the spool reading code. A
3797 related problem was that the "from" clause in the Received: line became
3798 illegal because of the space(s). It is now covered by ${quote_local_part.
3800 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3803 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3805 83. Patch from Sieve maintainer to fix unterminated string problem in
3806 "vacation" handling.
3808 84. Some minor changes to the Linux configuration files to help with other
3809 OS variants using glibc.
3811 85. One more patch for Sieve to update vacation handling to latest spec.
3814 ----------------------------------------------------
3815 See the note above about the 4.44 and 4.50 releases.
3816 ----------------------------------------------------
3822 1. Change 4.43/35 introduced a bug that caused file counts to be
3823 incorrectly computed when quota_filecount was set in an appendfile
3826 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3827 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3830 3. Give more explanation in the error message when the command for a transport
3831 filter fails to execute.
3833 4. There are several places where Exim runs a non-Exim command in a
3834 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3835 was being done only for the command run by the queryprogram router. It is
3836 now done for all such subprocesses. The other cases are: ${run, transport
3837 filters, and the commands run by the lmtp and pipe transports.
3839 5. Some older OS have a limit of 256 on the maximum number of file
3840 descriptors. Exim was using setrlimit() to set 1000 as a large value
3841 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3842 systems. I've change it so that if it can't get 1000, it tries for 256.
3844 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3845 was an oversight, and furthermore, ever since the addition of extra
3846 controls (e.g. 4.43/32), the checks on when to allow different forms of
3847 "control" were broken. There should now be diagnostics for all cases when a
3848 control that does not make sense is encountered.
3850 7. $recipients is now available in the predata ACL (oversight).
3852 8. Tidy the search cache before the fork to do a delivery from a message
3853 received from the command line. Otherwise the child will trigger a lookup
3854 failure and thereby defer the delivery if it tries to use (for example) a
3855 cached ldap connection that the parent has called unbind on.
3857 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3858 of $address_data from the recipient verification was clobbered by the
3859 sender verification.
3861 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3862 was its contents. (It was OK if the option was not defined at all.)
3864 11. A "Completed" log line is now written for messages that are removed from
3865 the spool by the -Mrm option.
3867 12. $host_address is now set to the target address during the checking of
3868 ignore_target_hosts.
3870 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3871 being passed; this would have caused $sender_host_name to have been used if
3872 matching the list had actually called for a host name (not very likely,
3873 since this list is usually IP addresses). A host name is now passed as
3876 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3877 code that specifies a non-restarting handler (typically sigaction() in
3878 modern systems) in an attempt to fix a rare and obscure crash bug.
3880 15. Narrowed the window for a race in the daemon that could cause it to ignore
3881 SIGCHLD signals. This is not a major problem, because they are used only to
3882 wake it up if nothing else does.
3884 16. A malformed maildirsize file could cause Exim to calculate negative values
3885 for the mailbox size or file count. Odd effects could occur as a result.
3886 The maildirsize information is now recalculated if the size or filecount
3889 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3890 support for a long time. Removed HAVE_SYS_VFS_H.
3892 18. Updated exipick to current release from John Jetmore.
3894 19. Allow an empty sender to be matched against a lookup in an address list.
3895 Previously the only cases considered were a regular expression, or an
3898 20. Exim went into a mad DNS lookup loop when doing a callout where the
3899 host was specified on the transport, if the DNS lookup yielded more than
3902 21. The RFC2047 encoding function was originally intended for short strings
3903 such as real names; it was not keeping to the 75-character limit for
3904 encoded words that the RFC imposes. It now respects the limit, and
3905 generates multiple encoded words if necessary. To be on the safe side, I
3906 have increased the buffer size for the ${rfc2047: expansion operator from
3909 22. Failure to deliver a bounce message always caused it to be frozen, even if
3910 there was an errors_to setting on the router. The errors_to setting is now
3913 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3914 canonical form (fully expanded) before being placed in
3915 $sender_host_address.
3917 24. Updated eximstats to version 1.33
3919 25. Include certificate and key file names in error message when GnuTLS fails
3920 to set them up, because the GnuTLS error message doesn't include the name
3921 of the failing file when there is a problem reading it.
3923 26. Expand error message when OpenSSL has problems setting up cert/key files.
3926 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3927 (this can affect the format of dates).
3929 28. exim_tidydb, when checking for the continued existence of a message for
3930 which it has found a message-specific retry record, was not finding
3931 messages that were in split spool directories. Consequently, it was
3932 deleting retry records that should have stayed in existence.
3934 29. eximstats updated to version 1.35
3935 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3936 1.35 - bugfix such that pie charts by volume are generated correctly
3938 30. The SPA authentication driver was not abandoning authentication and moving
3939 on to the next authenticator when an expansion was forced to fail,
3940 contradicting the general specification for all authenticators. Instead it
3941 was generating a temporary error. It now behaves as specified.
3943 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3944 (the order specifies the preference for clients). The order is now AES256,
3945 AES128, 3DES, ARCFOUR128.
3947 31. Small patch to Sieve code - explicitly set From: when generating an
3950 32. Exim crashed if a remote delivery caused a very long error message to be
3951 recorded - for instance if somebody sent an entire SpamAssassin report back
3952 as a large number of 550 error lines. This bug was coincidentally fixed by
3953 increasing the size of one of Exim's internal buffers (big_buffer) that
3954 happened as part of the Exiscan merge. However, to be on the safe side, I
3955 have made the code more robust (and fixed the comments that describe what
3958 33. Some experimental protocols are using DNS PTR records for new purposes. The
3959 keys for these records are domain names, not reversed IP addresses. The
3960 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3961 leaves it alone. Component reversal etc. now happens only for IP addresses.
3964 34. The host_aton() function is supposed to be passed a string that is known
3965 to be a valid IP address. However, in the case of IPv6 addresses, it was
3966 not checking this. This is a hostage to fortune. Exim now panics and dies
3967 if the condition is not met. A case was found where this could be provoked
3968 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3969 components; fortuitously, this particular loophole had already been fixed
3970 by change 4.50/55 or 4.44/33 above.
3972 If there are any other similar loopholes, the new check in host_aton()
3973 itself should stop them being exploited. The report I received stated that
3974 data on the command line could provoke the exploit when Exim was running as
3975 exim, but did not say which command line option was involved. All I could
3976 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3977 running as the user.
3980 35. There was a buffer overflow vulnerability in the SPA authentication code
3981 (which came originally from the Samba project). I have added a test to the
3982 spa_base64_to_bits() function which I hope fixes it.
3985 36. The daemon start-up calls getloadavg() while still root for those OS that
3986 need the first call to be done as root, but it missed one case: when
3987 deliver_queue_load_max is set with deliver_drop_privilege. This is
3988 necessary for the benefit of the queue runner, because there is no re-exec
3989 when deliver_drop_privilege is set.
3991 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3992 was in use, was not putting the data itself into the right store pool;
3993 consequently, it could be overwritten for a subsequent message in the same
3994 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3997 38. Sometimes the final signoff response after QUIT could fail to get
3998 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3999 < 0 before doing a fflush(). This bug looks as though it goes back to the
4000 introduction of TLS in release 3.20, but "sometimes" must have been rare
4001 because the tests only now provoked it.
4007 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4008 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4009 Now it returns an integer. A number of calls were still expecting a T/F
4010 return. Fortuitously, in all cases, the tests worked in OK situations,
4011 which is the norm. However, things would have gone wrong on any write
4012 failures on the smtp file descriptor. This function is used when sending
4013 messages over SMTP and also when doing verify callouts.
4015 2. When Exim is called to do synchronous delivery of a locally submitted
4016 message (the -odf or -odi options), it no longer closes stderr before doing
4019 3. Implemented the mua_wrapper option.
4021 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4023 5. Implemented the functions header_remove(), header_testname(),
4024 header_add_at_position(), and receive_remove_recipient(), and exported them
4027 6. If an ACL "warn" statement specified the addition of headers, Exim already
4028 inserted X-ACL-Warn: at the start if there was no header name. However, it
4029 was not making this test for the second and subsequent header lines if
4030 there were newlines in the string. This meant that an invalid header could
4031 be inserted if Exim was badly configured.
4033 7. Allow an ACL "warn" statement to add header lines at the start or after all
4034 the Received: headers, as well as at the end.
4036 8. Added the rcpt_4xx retry error code.
4038 9. Added postmaster_mailfrom=xxx to callout verification option.
4040 10. Added mailfrom=xxxx to the callout verification option, for verify=
4043 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4044 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4046 12. Inserted some casts to stop certain compilers warning when using pointer
4047 differences as field lengths or precisions in printf-type calls (mostly
4048 affecting debugging statements).
4050 13. Added optional readline() support for -be (dynamically loaded).
4052 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4053 same clock tick as a message's arrival, so that its received time was the
4054 same as the "first fail" time on the retry record, and that message
4055 remained on the queue past the ultimate address timeout, every queue runner
4056 would try a delivery (because it was past the ultimate address timeout) but
4057 after another failure, the ultimate address timeout, which should have then
4058 bounced the address, did not kick in. This was a "< instead of <=" error;
4059 in most cases the first failure would have been in the next clock tick
4060 after the received time, and all would be well.
4062 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4063 being recognized when the domain list was tested by the match_domain
4064 condition in an expansion string.
4066 16. Added the ${str2b64: operator.
4068 17. Exim was always calling setrlimit() to set a large limit for the number of
4069 processes, without checking whether the existing limit was already
4070 adequate. (It did check for the limit on file descriptors.) Furthermore,
4071 errors from getrlimit() and setrlimit() were being ignored. Now they are
4072 logged to the main and panic logs, but Exim does carry on, to try to do its
4073 job under whatever limits there are.
4075 18. Imported PCRE 5.0.
4077 19. Trivial typo in log message " temporarily refused connection" (the leading
4080 20. If the log selector return_path_on_delivery was set and an address was
4081 redirected to /dev/null, the delivery process crashed because it assumed
4082 that a return path would always be set for a "successful" delivery. In this
4083 case, the whole delivery is bypassed as an optimization, and therefore no
4086 21. Internal re-arrangement: the function for sending a challenge and reading
4087 a response while authentication was assuming a zero-terminated challenge
4088 string. It's now changed to take a pointer and a length, to allow for
4089 binary data in such strings.
4091 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4093 23. Exim was not respecting finduser_retries when seeking the login of the
4094 uid under which it was called; it was always trying 10 times. (The default
4095 setting of finduser_retries is zero.) Also, it was sleeping after the final
4096 failure, which is pointless.
4098 24. Implemented tls_on_connect_ports.
4100 25. Implemented acl_smtp_predata.
4102 26. If the domain in control=submission is set empty, Exim assumes that the
4103 authenticated id is a complete email address when it generates From: or
4104 Sender: header lines.
4106 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4107 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4108 chown and chgrp in /bin and hostname in /usr/bin.
4110 28. Exim was keeping the "process log" file open after each use, just as it
4111 does for the main log. This opens the possibility of it remaining open for
4112 long periods when the USR1 signal hits a daemon. Occasional processlog
4113 errors were reported, that could have been caused by this. Anyway, it seems
4114 much more sensible not to leave this file open at all, so that is what now
4117 29. The long-running daemon process does not normally write to the log once it
4118 has entered its main loop, and it closes the log before doing so. This is
4119 so that log files can straightforwardly be renamed and moved. However,
4120 there are a couple of unusual error situations where the daemon does write
4121 log entries, and I had neglected to close the log afterwards.
4123 30. The text of an SMTP error response that was received during a remote
4124 delivery was being truncated at 512 bytes. This is too short for some of
4125 the long messages that one sometimes sees. I've increased the limit to
4128 31. It is now possible to make retry rules that apply only when a message has a
4129 specific sender, in particular, an empty sender.
4131 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4132 it possible to be selective about when SMTP synchronization is enforced.
4134 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4136 32. Implemented hosts_connection_nolog.
4138 33. Added an ACL for QUIT.
4140 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4143 35. Added mailbox_size and mailbox_filecount to appendfile.
4145 36. Added control = no_multiline_responses to ACLs.
4147 37. There was a bug in the logic of the code that waits for the clock to tick
4148 in the case where the clock went backwards by a substantial amount such
4149 that the microsecond fraction of "now" was more than the microsecond
4150 fraction of "then" (but the whole seconds number was less).
4152 38. Added support for the libradius Radius client library this is found on
4153 FreeBSD (previously only the radiusclient library was supported).
4159 1. When certain lookups returned multiple values in the form name=value, the
4160 quoting of the values was not always being done properly. Specifically:
4161 (a) If the value started with a double quote, but contained no whitespace,
4163 (b) If the value contained whitespace other than a space character (i.e.
4164 tabs or newlines or carriage returns) it was not quoted.
4165 This fix has been applied to the mysql and pgsql lookups by writing a
4166 separate quoting function and calling it from the lookup code. The fix
4167 should probably also be applied to nisplus, ibase and oracle lookups, but
4168 since I cannot test any of those, I have not disturbed their existing code.
4170 2. A hit in the callout cache for a specific address caused a log line with no
4171 reason for rejecting RCPT. Now it says "Previous (cached) callout
4172 verification failure".
4174 3. There was an off-by-one bug in the queryprogram router. An over-long
4175 return line was truncated at 256 instead of 255 characters, thereby
4176 overflowing its buffer with the terminating zero. As well as fixing this, I
4177 have increased the buffer size to 1024 (and made a note to document this).
4179 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4180 when Exim is waiting for an SMTP response from a remote server, Exim
4181 restarts its select() call on the socket, thereby resetting its timeout.
4182 This is not a problem when such interrupts are rare. Somebody set up a cron
4183 job to run exiwhat every 2 minutes, which is less than the normal select()
4184 timeout (5 or 10 minutes). This meant that the select() timeout never
4185 kicked in because it was always reset. I have fixed this by comparing the
4186 time when an interrupt arrives with the time at the start of the first call
4187 to select(). If more time than the timeout has elapsed, the interrupt is
4188 treated as a timeout.
4190 5. Some internal re-factoring in preparation for the addition of Sieve
4191 extensions (by MH). In particular, the "personal" test is moved to a
4192 separate function, and given an option for scanning Cc: and Bcc: (which is
4193 not set for Exim filters).
4195 6. When Exim created an email address using the login of the caller as the
4196 local part (e.g. when creating a From: or Sender: header line), it was not
4197 quoting the local part when it contained special characters such as @.
4199 7. Installed new OpenBSD configuration files.
4201 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4202 try to make them clearer.
4204 9. Callout options, other than the timeout value, were being ignored when
4205 verifying sender addresses in header lines. For example, when using
4207 verify = header_sender/callout=no_cache
4209 the cache was (incorrectly) being used.
4211 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4212 only the Cygwin environment.
4214 11. When return_path_on_delivery was set as a log selector, if different remote
4215 addresses in the same message used different return paths and parallel
4216 remote delivery occurred, the wrong values would sometimes be logged.
4217 (Whenever a remote delivery process finished, the return path value from
4218 the most recently started remote delivery process was logged.)
4220 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4221 lines when AUTH and/or TLS are in use. This is the "received protocol"
4222 field. Exim used to use "asmtp" for authenticated SMTP, without any
4223 indication (in the protocol name) for TLS use. Now it follows the RFC and
4224 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4225 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4226 names appear in log lines as well as in Received: header lines.
4228 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4229 extensions, and comparison tests, and to fix some bugs.
4231 14. Changes to the "personal" filter test:
4233 (1) The test was buggy in that it was just doing the equivalent of
4234 "contains" tests on header lines. For example, if a user's address was
4235 anne@some.where, the "personal" test would incorrectly be true for
4237 To: susanne@some.where
4239 This test is now done by extracting each address from the header in turn,
4240 and checking the entire address. Other tests that are part of "personal"
4241 are now done using regular expressions (for example, to check local parts
4242 of addresses in From: header lines).
4244 (2) The list of non-personal local parts in From: addresses has been
4245 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4246 taken from the Sieve specification recommendations.
4248 (3) If the message contains any header line starting with "List-" it is
4249 treated as non-personal.
4251 (4) The test for "circular" in the Subject: header line has been removed
4252 because it now seems ill-conceived.
4254 15. Minor typos in src/EDITME comments corrected.
4256 16. Installed latest exipick from John Jetmore.
4258 17. If headers_add on a router specified a text string that was too long for
4259 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4260 of string_sprintf() is now avoided.
4262 18. $message_body_size was not set (it was always zero) when running the DATA
4263 ACL and the local_scan() function.
4265 19. For the "mail" command in an Exim filter, no default was being set for
4266 the once_repeat time, causing a random time value to be used if "once" was
4267 specified. (If the value happened to be <= 0, no repeat happened.) The
4268 default is now 0s, meaning "never repeat". The "vacation" command was OK
4269 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4270 (I found it when inspecting the code).
4272 20. There is now an overall timeout for performing a callout verification. It
4273 defaults to 4 times the callout timeout, which applies to individual SMTP
4274 commands during the callout. The overall timeout applies when there is more
4275 than one host that can be tried. The timeout is checked before trying the
4276 next host. This prevents very long delays if there are a large number of
4277 hosts and all are timing out (e.g. when the network connections are timing
4278 out). The value of the overall timeout can be changed by specifying an
4279 additional sub-option for "callout", called "maxwait". For example:
4281 verify = sender/callout=5s,maxwait=20s
4283 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4284 to the end before writing, but this should make it even safer).
4286 22. Exim was forgetting that it had advertised PIPELINING for the second and
4287 subsequent messages on an SMTP connection. It was also not resetting its
4288 memory on STARTTLS and an internal HELO.
4290 23. When Exim logs an SMTP synchronization error within a session, it now
4291 records whether PIPELINING has been advertised or not.
4293 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4294 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4295 rather than long int.
4297 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4299 26. Added the never_mail option to autoreply.
4305 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4306 crash if the getsockname() call failed; this can happen if a connection is
4307 closed very soon after it is established. The problem was simply in the
4308 order in which certain operations were done, causing Exim to try to write
4309 to the SMTP stream before it had set up the file descriptor. The bug has
4310 been fixed by making things happen in the correct order.
4316 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4317 before the connection was closed, thus losing the rejection response.
4319 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4320 some early Solaris releases, but causes trouble in current releases where
4321 socklen_t is defined.
4323 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4326 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4329 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4330 that releasing the top of it at the end releases what was used for sub-
4331 expansions (unless the block got too big). However, discard this block if
4332 the first thing is a variable or header, so that we can use its block when
4333 it is dynamic (useful for very large $message_headers, for example).
4335 6. Lookups now cache *every* query, not just the most recent. A new, separate
4336 store pool is used for this. It can be recovered when all lookup caches are
4337 flushed. Lookups now release memory at the end of their result strings.
4338 This has involved some general refactoring of the lookup sources.
4340 7. Some code has been added to the store_xxx() functions to reduce the amount
4341 of flapping under certain conditions.
4343 8. log_incoming_interface used to affect only the <= reception log lines. Now
4344 it causes the local interface and port to be added to several more SMTP log
4345 lines, for example "SMTP connection from", and rejection lines.
4347 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4349 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4351 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4352 could be overwritten at the end of the current message (or the start of a
4353 new message if it was set in a HELO ACL). The value is now preserved for
4354 the duration of the SMTP connection.
4356 12. If a transport had a headers_rewrite setting, and a matching header line
4357 contained an unqualified address, that address was qualified, even if it
4358 did not match any rewriting rules. The underlying bug was that the values
4359 of the flags that permit the existence of unqualified sender and recipient
4360 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4361 non-local messages, and by -bnq for local messages) were not being
4362 preserved with the message after it was received.
4364 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4365 "next input=" as part of the text comprising the host identity instead of
4366 the correct text. The code was using the same buffer for two different
4367 strings. However, depending on which order the printing function evaluated
4368 its arguments, the bug did not always show up. Under Linux, for example, my
4369 test suite worked just fine.
4371 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4372 doesn't work with some older versions of Perl. It has been changed to "my",
4373 which in any case is probably the better facility to use.
4375 15. A really picky compiler found some instances of statements for creating
4376 error messages that either had too many or two few arguments for the format
4379 16. The size of the buffer for calls to the DNS resolver has been increased
4380 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4381 for addresses that have a lot of PTR records. This alleviates a problem; it
4382 does not fully solve it.
4384 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4385 buffer now truncates the list and logs the incident, which is the same
4386 action as happens when Exim is looking up a host name and its aliases.
4387 Previously in this situation something unpredictable would happen;
4388 sometimes it was "internal error: store_reset failed".
4390 18. If a server dropped the connection unexpectedly when an Exim client was
4391 using GnuTLS and trying to read a response, the client delivery process
4392 crashed while trying to generate an error log message.
4394 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4395 string, for example:
4397 warn message = H1: something\nH2: something
4399 the text was added as a single header line from Exim's point of view
4400 though it ended up OK in the delivered message. However, searching for the
4401 second and subsequent header lines using $h_h2: did not work. This has been
4402 fixed. Similarly, if a system filter added multiple headers in this way,
4403 the routers could not see them.
4405 20. Expanded the error message when iplsearch is called with an invalid key to
4406 suggest using net-iplsearch in a host list.
4408 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4409 ACLs are no longer actually imposed (and a message to that effect is
4412 22. If a "gecos" field in a passwd entry contained escaped characters, in
4413 particular, if it contained a \" sequence, Exim got it wrong when building
4414 a From: or a Sender: header from that name. A second bug also caused
4415 incorrect handling when an unquoted " was present following a character
4416 that needed quoting.
4418 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4419 was not being matched caselessly.
4421 24. Arranged for all hyphens in the exim.8 source to be escaped with
4424 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4425 properly. Recipient callout cache records were still being keyed to include
4426 the sender, even when use_sender was set false. This led to far more
4427 callouts that were necessary. The sender is no longer included in the key
4428 when use_sender is false.
4430 26. Added "control = submission" modifier to ACLs.
4432 27. Added the ${base62d: operator to decode base 62 numbers.
4434 28. dnsdb lookups can now access SRV records.
4436 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4437 the configuration file.
4439 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4440 (-v) mode. This makes the output for a verbose queue run more intelligible.
4442 31. Added a use_postmaster feature to recipient callouts.
4444 32. Added the $body_zerocount variable, containing the number of binary zero
4445 bytes in the message body.
4447 33. The time of last modification of the "new" subdirectory is now used as the
4448 "mailbox time last read" when there is a quota error for a maildir
4451 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4453 35. Added +ignore_unknown as a special item in host lists.
4455 36. Code for decoding IPv6 addresses in host lists is now included, even if
4456 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4457 address was recognized as an IP address, but was then not correctly decoded
4458 into binary, causing unexpected and incorrect effects when compared with
4465 1. Very minor rewording of debugging text in manualroute to say "list of
4466 hosts" instead of "hostlist".
4468 2. If verify=header_syntax was set, and a header line with an unqualified
4469 address (no domain) and a large number of spaces between the end of the
4470 name and the colon was received, the reception process suffered a buffer
4471 overflow, and (when I tested it) crashed. This was caused by some obsolete
4472 code that should have been removed. The fix is to remove it!
4474 3. When running in the test harness, delay a bit after writing a bounce
4475 message to get a bit more predictability in the log output.
4477 4. Added a call to search_tidyup() just before forking a reception process. In
4478 theory, someone could use a lookup in the expansion of smtp_accept_max_
4479 per_host which, without the tidyup, could leave open a database connection.
4481 5. Added the variables $recipient_data and $sender_data which get set from a
4482 lookup success in an ACL "recipients" or "senders" condition, or a router
4483 "senders" option, similar to $domain_data and $local_part_data.
4485 6. Moved the writing of debug_print from before to after the "senders" test
4488 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4489 problems for message scanning, either using a data ACL, or using
4490 local_scan() because the Received: header was not generated till after they
4491 were called (in order to set the time as the time of reception completion).
4492 I have revised the way this works. The header is now generated after the
4493 body is received, but before the ACL or local_scan() are called. After they
4494 are run, the timestamp in the header is updated.
4500 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4501 before starting a queue runner without re-exec. This happened only when
4502 deliver_drop_privilege was set or when the Exim user was set to root. The
4503 effect of the bug was that timeouts during subsequent deliveries caused
4504 crashes instead of being properly handled. The handler is now left at its
4505 default (and expected) setting.
4507 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4508 message, again when deliver_drop_privilege is set or Exim is run as root.
4509 The bug described in (1) was not present in this case, but the tidying up
4510 of the other signals was missing. I have made the two cases consistent.
4512 3. The ignore_target_hosts setting on a manualroute router was being ignored
4513 for hosts that were looked up using the /MX notation.
4515 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4518 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4519 operated on the sender address. After changing the $sender_address to <>
4520 for the sender address verify, Exim was re-instated it as the original
4521 (before rewriting) address, but remembering that it had rewritten it, so it
4522 wasn't rewriting it again. This bug also had the effect of breaking the
4523 sender address verification caching when the sender address was rewritten.
4525 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4526 This has been changed so that if the ip literal address matches
4527 ignore_target_hosts, the router declines.
4529 7. Added expansion conditions match_domain, match_address, and match_local_
4530 part (NOT match_host).
4532 8. The placeholder for the Received: header didn't have a length field set.
4534 9. Added code to Exim itself and to exim_lock to test for a specific race
4535 condition that could lead to file corruption when using MBX delivery. The
4536 issue is with the lockfile that is created in /tmp. If this file is removed
4537 after a process has opened it but before that process has acquired a lock,
4538 there is the potential for a second process to recreate the file and also
4539 acquire a lock. This could lead to two Exim processes writing to the file
4540 at the same time. The added code performs the same test as UW imapd; it
4541 checks after acquiring the lock that its file descriptor still refers to
4542 the same named file.
4544 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4545 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4548 11. Added the smtp_active_hostname option. If used, this will typically be made
4549 to depend on the incoming interface address. Because $interface_address is
4550 not set up until the daemon has forked a reception process, error responses
4551 that can happen earlier (such as "too many connections") no longer contain
4554 12. If an expansion in a condition on a "warn" statement fails because a lookup
4555 defers, the "warn" statement is abandoned, and the next ACL statement is
4556 processed. Previously this caused the whole ACL to be aborted.
4558 13. Added the iplsearch lookup type.
4560 14. Added ident_timeout as a log selector.
4562 15. Added tls_certificate_verified as a log selector.
4564 16. Added a global option tls_require_ciphers (compare the smtp transport
4565 option of the same name). This controls incoming TLS connections.
4567 17. I finally figured out how to make tls_require_ciphers do a similar thing
4568 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4569 before starting the TLS session.
4571 18. Tabs are now shown as \t in -bP output.
4573 19. If the log selector return_path_on_delivery was set, Exim crashed when
4574 bouncing a message because it had too many Received: header lines.
4576 20. If two routers both had headers_remove settings, and the first one included
4577 a superfluous trailing colon, the final name in the first list and the
4578 first name in the second list were incorrectly joined into one item (with a
4579 colon in the middle).
4585 1. Added -C and -D options to the exinext utility, mainly to make it easier
4586 to include in the automated testing, but these could be helpful when
4587 multiple configurations are in use.
4589 2. The exinext utility was not formatting the output nicely when there was
4590 an alternate port involved in the retry record key, nor when there was a
4591 message id as well (for retries that were specific to a specific message
4592 and a specific host). It was also confused by IPv6 addresses, because of
4593 the additional colons they contain. I have fixed the IPv4 problem, and
4594 patched it up to do a reasonable job for IPv6.
4596 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4597 delivery, the log line now contains "pipelined" if PIPELINING was used.
4599 4. An SMTP transport process used to panic and die if the bind() call to set
4600 an explicit outgoing interface failed. This has been changed; it is now
4601 treated in the same way as a connect() failure.
4603 5. A reference to $sender_host_name in the part of a conditional expansion
4604 that was being skipped was still causing a DNS lookup. This no longer
4607 6. The def: expansion condition was not recognizing references to header lines
4608 that used bh_ and bheader_.
4610 7. Added the _cache feature to named lists.
4612 8. The code for checking quota_filecount in the appendfile transport was
4613 allowing one more file than it should have been.
4615 9. For compatibility with Sendmail, the command line option
4623 and sets the incoming protocol and host name (for trusted callers). The
4624 host name and its colon can be omitted when only the protocol is to be set.
4625 Note the Exim already has two private options, -pd and -ps, that refer to
4626 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4627 "s", but I don't think that's a major issue.
4629 10. A number of refactoring changes to the code, none of which should affect
4632 (a) The number of logging options was getting close to filling up the
4633 32-bit word that was used as a bit map. I have split them into two classes:
4634 those that are passed in the argument to log_write(), and those that are
4635 only ever tested independently outside of that function. These are now in
4636 separate 32-bit words, so there is plenty of room for expansion again.
4637 There is no change in the user interface or the logging behaviour.
4639 (b) When building, for example, log lines, the code previously used a
4640 macro that called string_cat() twice, in order to add two strings. This is
4641 not really sufficiently general. Furthermore, there was one instance where
4642 it was actually wrong because one of the argument was used twice, and in
4643 one call a function was used. (As it happened, calling the function twice
4644 did not affect the overall behaviour.) The macro has been replaced by a
4645 function that can join an arbitrary number of extra strings onto a growing
4648 (c) The code for expansion conditions now uses a table and a binary chop
4649 instead of a serial search (which was left over from when there were very
4650 few conditions). Also, it now recognizes conditions like "pam" even when
4651 the relevant support is not compiled in: a suitably worded error message is
4652 given if an attempt is made to use such a condition.
4654 11. Added ${time_interval:xxxxx}.
4656 12. A bug was causing one of the ddress fields not to be passed back correctly
4657 from remote delivery subprocesses. The field in question was not being
4658 subsequently used, so this caused to problems in practice.
4660 13. Added new log selectors queue_time and deliver_time.
4662 14. Might have fixed a bug in maildirsizefile handling that threw up
4663 "unexpected character" debug warnings, and recalculated the data
4664 unnecessarily. In any case, I expanded the warning message to give more
4667 15. Added the message "Restricted characters in address" to the statements in
4668 the default ACL that block characters like @ and % in local parts.
4670 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4671 Three changes have been made:
4673 (a) There was a serious bug; a negative response to MAIL caused the whole
4674 recipient domain to be cached as invalid, thereby blocking all messages
4675 to all local parts at the same domain, from all senders. This bug has
4676 been fixed. The domain is no longer cached after a negative response to
4677 MAIL if the sender used is not empty.
4679 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4682 (c) A new callout option, "use_sender" has been added for people who want
4683 the modified behaviour.
4689 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4692 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4693 indeed breaks things for older releases.
4695 3. Added additional logging to the case where there is a problem reading data
4696 from a filter that is running in a subprocess using a pipe, in order to
4697 try to track down a specific problem.
4699 4. Testing facility fudge: when running in the test harness and attempting
4700 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4701 getting "No route to host". Convert this to a timeout.
4703 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4706 6. Some OS don't have socklen_t but use size_t instead. This affects the
4707 fifth argument of getsockopt() amongst other things. This is now
4708 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4709 can be set for individual OS. I have set it for SunOS5, OSF1, and
4710 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4711 some earlier ones do not.
4713 7. Change 4.30/15 was not doing the test caselessly.
4715 8. The standard form for an IPv6 address literal was being rejected by address
4716 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4717 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4718 this, as well as the form without the "IPv6" on the front (but only when
4719 address literals are enabled, of course).
4721 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4723 10. Exim crashed if a message with an empty sender address specified by -f
4724 encountered a router with an errors_to setting. This could be provoked only
4725 by a command such as
4729 where an empty string was supplied; "<>" did not hit this bug.
4731 11. Installed PCRE release 4.5.
4733 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4734 remained set. It is now erased.
4736 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4737 times from message ids (which are base 36 rather than the normal 62).
4739 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4740 were being counted as actual protocol errors, and logged if the log
4741 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4742 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4743 and DATA following a set of rejected RCPTs do not count as protocol errors.
4744 In other words, Exim assumes they were pipelined, though this may not
4745 actually be the case. Of course, in all cases the client gets an
4746 appropriate error code.
4748 15. If a lookup fails in an ACL condition, a message about the failure may
4749 be available; it is used if testing the ACL cannot continue, because most
4750 such messages specify what the cause of the deferral is. However, some
4751 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4752 that caused an old message to be retained and used if a later statement
4753 caused a defer, replacing the real cause of the deferral.
4755 16. If an IP address had so many PTR records that the DNS lookup buffer
4756 was not large enough to hold them, Exim could crash while trying to process
4757 the truncated data. It now detects and logs this case.
4759 17. Further to 4.21/58, another change has been made: if (and only if) the
4760 first line of a message (the first header line) ends with CRLF, a bare LF
4761 in a subsequent header line has a space inserted after it, so as not to
4762 terminate the header.
4764 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4765 unnecessarily, used atoi() instead of strtol(), and didn't check the
4766 termination when getting file sizes from file names by regex.
4768 19. Completely re-implemented the support for maildirsize files, in the light
4769 of a number of problems with the previous contributed implementation
4770 (4.30/29). In particular:
4772 . If the quota is zero, the maildirsize file is maintained, but no quota is
4775 . If the maildir directory does not exist, it is created before any attempt
4776 to write a maildirsize file.
4778 . The quota value in the file is just a cache; if the quota is changed in
4779 the transport, the new value overrides.
4781 . A regular expression is available for excluding directories from the
4784 20. The autoreply transport checks the characters in options that define the
4785 message's headers; it allows continued headers, but it was checking with
4786 isspace() after an embedded newline instead of explicitly looking for a
4789 21. If all the "regular" hosts to which an address was routed had passed their
4790 expiry times, and had not reached their retry times, the address was
4791 bounced, even if fallback hosts were defined. Now Exim should go on to try
4794 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4795 equivalent code in the SMTP transport. Some hosts send humungous responses
4796 to HELO/EHLO, more than 1024 it seems.
4798 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4799 gives compiler warnings in some environments. I've now done it "properly",
4802 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4803 (because of problems with the built-in one) was declared to return uschar *
4804 instead of char *, causing compiler failure.
4806 25. Fixed a file descriptor leak when processing alias/forward files.
4808 26. Fixed a minor format string issue in dbfn.c.
4810 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4812 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4813 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4816 29. When a sender address is verified, it is cached, to save repeating the test
4817 when there is more than one recipient in a message. However, when the
4818 verification involves a callout, it is possible for different callout
4819 options to be set for different recipients. It is too complicated to keep
4820 track of this in the cache, so now Exim always runs a verification when a
4821 callout is required, relying on the callout cache for the optimization.
4822 The overhead is duplication of the address routing, but this should not be
4825 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4826 to be verified with callout=postmaster, and the main callout worked but the
4827 postmaster check failed, the verification correctly failed. However, if a
4828 subsequent RCPT command asked for sender verification *without* the
4829 postmaster check, incorrect caching caused this verification also to fail,
4832 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4833 it was not caching the DNS options (qualify_single, search_parents) that
4834 were used when the lookup failed. A subsequent lookup with different
4835 options therefore always gave the same answer, though there were cases
4836 where it should not have. (Example: a "domains = !$mx_any" option on a
4837 dnslookup router: the "domains" option is always processed without any
4838 widening, but the router might have qualify_single set.) Now Exim uses the
4839 cached value only when the same options are set.
4841 32. Added John Jetmore's "exipick" utility to the distribution.
4843 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4844 than a timeout (e.g. a certificate is required, and is not provided), an
4845 Exim server now closes the connection immediately. Previously it waited for
4846 the client to close - but if the client is SSL, it seems that they each
4847 wait for each other, leading to a delay before one of them times out.
4849 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4850 maintained 0.8.x compatibility because I don't think many are using it, and
4851 it is clearly obsolete.
4853 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4856 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4857 client certificate was expired. A simple patch fixes this, though I don't
4858 understand the full logic of why the verify callback is called multiple
4861 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4862 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4863 which causes problems with some clients (such as the Certicom SSL Plus
4864 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4865 disables the coutermeasure allowing Eudora to connect."
4867 38. Exim was not checking that a write() to a log file succeeded. This could
4868 lead to Bad Things if a log got too big, in particular if it hit a file
4869 size limit. Exim now panics and dies if it cannot write to a log file, just
4870 as it does if it cannot open a log file.
4872 39. Modified OS/Makefile-Linux so that it now contains
4874 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4876 The two -D definitions ensure that Exim is compiled with large file
4877 support, which makes it possible to handle log files that are bigger than
4880 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4881 instance) a domain was checked against a named list that involved a lookup,
4882 causing $domain_data to be set, then another domain was checked against the
4883 same list, then the first domain was re-checked, the value of $domain_data
4884 after the final check could be wrong. In particular, if the second check
4885 failed, it could be set empty. This bug probably also applied to
4888 41. The strip_trailing_dot option was not being applied to the address given
4889 with the -f command-line option.
4891 42. The code for reading a message's header from the spool was incrementing
4892 $received_count, but never initializing it. This meant that the value was
4893 incorrect (doubled) while delivering a message in the same process in which
4894 it was received. In the most common configuration of Exim, this never
4895 happens - a fresh exec is done - but it can happen when
4896 deliver_drop_privilege is set.
4898 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4899 it now includes up to 150 characters of the unexpected data in the log
4902 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4903 and building data strings. The size of both of these buffers was 10 000
4904 bytes - far larger than anybody would *ever* want, thought I. Needless to
4905 say, somebody hit the limit. I have increased the maximum line length to
4906 20 000 and the maximum data length of concatenated lines to 100 000. I have
4907 also fixed two bugs, because there was no checking on these buffers. Tsk,
4908 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4909 buffer is too small.
4911 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4912 lsearch lookups. Now it does.
4914 46. When parsing a route_list item in a manualroute router, a fixed-length
4915 buffer was used for the list of hosts. I made this 1024 bytes long,
4916 thinking that nobody would ever have a list of hosts that long. Wrong.
4917 Somebody had a whole pile of complicated expansion conditions, and the
4918 string was silently truncated, leading to an expansion error. It turns out
4919 that it is easier to change to an unlimited length (owing to other changes
4920 that have happened since this code was originally written) than to build
4921 structure for giving a limitation error. The length of the item that
4922 expands into the list of hosts is now unlimited.
4924 47. The lsearch lookup could not handle data where the length of text line was
4925 more than 4095 characters. Such lines were truncated, leading to shortened
4926 data being returned. It should now handle lines of any length.
4928 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4929 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4932 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4933 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4934 no longer happens in comment lines. A list of replacements is now placed
4935 at the head of all of the source files, except those whose only change is
4936 to replace PERL_COMMAND in the very first #! line.
4938 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4939 messages on the queue, with a bottom-up merge sort, using code contributed
4940 by Michael Haardt. This should make operations like -bp somewhat faster on
4941 large queues. It won't affect queue runners, except when queue_run_in_order
4944 51. Installed eximstats 1.31 in the distribution.
4946 52. Added support for SRV lookups to the dnslookup router.
4948 53. If an ACL referred to $message_body or $message_body_end, the value was not
4949 reset for any messages that followed in the same SMTP session.
4951 54. The store-handling optimization for building very long strings was not
4952 differentiating between the different store pools. I don't think this
4953 actually made any difference in practice, but I've tidied it.
4955 55. While running the routers to verify a sender address, $sender_address
4956 was still set to the sender address. This is wrong, because when routing to
4957 send a bounce to the sender, it would be empty. Therefore, I have changed
4958 it so that, while verifying a sender address, $sender_address is set to <>.
4959 (There is no change to what happens when verifying a recipient address.)
4961 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4962 target A or AAAA records (if not already returned) without resetting the
4963 qualify_single or search_parents options of the DNS resolver. These are
4964 inappropriate in this case because the targets of MX and SRV records must
4965 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4966 target that, when qualified, matched something in the local domain. These
4967 two options are now turned off when doing these lookups.
4969 57. It seems that at least some releases of Reiserfs (which does not have the
4970 concept of a fixed number of inodes) returns zero and not -1 for the
4971 number of available inodes. This interacted badly with check_spool_inodes,
4972 which assumed that -1 was the "no such thing" setting. What I have done is
4973 to check that the total number of inodes is greater than zero before doing
4974 the test of how many are available.
4976 58. When a "warn" ACL statement has a log_message modifier, the message is
4977 remembered, and not repeated. This is to avoid a lot of repetition when a
4978 message has many recipients that cause the same warning to be written.
4979 Howewer, Exim was preserving the list of already written lines for an
4980 entire SMTP session, which doesn't seem right. The memory is now reset if a
4981 new message is started.
4983 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4984 debugging output unless log_rewrite was also set.
4986 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4987 of (int)(handle) when we know that handle contains (void *)(-1).
4989 61. The Exim daemon panic-logs an error return when it closes the incoming
4990 connection. However "connection reset by peer" seems to be common, and
4991 isn't really an error worthy of noting specially, so that particular error
4994 62. When Exim is trying to find all the local interfaces, it used to panic and
4995 die if the ioctl to get the interface flags failed. However, it seems that
4996 on at least one OS (Solaris 9) it is possible to have an interface that is
4997 included in the list of interfaces, but for which you get a failure error
4998 for this call. This happens when the interface is not "plumbed" into a
4999 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5000 failure of the "get flags" call assumes that the interface is down.
5002 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5003 makes life easier for people who are doing arithmetic on fields extracted
5004 from dates, where you often get leading zeros that should not be
5005 interpreted as octal.
5007 64. Added qualify_domain to the redirect router, to override the global
5010 65. If a pathologically long header line contained very many addresses (the
5011 report of this problem mentioned 10 000) and each of them was rewritten,
5012 Exim could use up a very large amount of memory. (It kept on making new
5013 copies of the header line as it rewrote, and never released the old ones.)
5014 At the expense of a bit more processing, the header rewriting function has
5015 been changed so that it no longer eats memory in this way.
5017 66. The generation of the Received: header has been moved from the time that a
5018 message starts to be received, to the time that it finishes. The timestamp
5019 in the Received: header should now be very close to that of the <= log
5020 line. There are two side-effects of this change:
5022 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5023 logged header lines no longer include the local Received: line, because
5024 it has not yet been created. The same applies to a copy of the message
5025 that is returned to a non-SMTP sender when a message is rejected.
5027 (b) When a filter file is tested using -bf, no additional Received: header
5028 is added to the test message. After some thought, I decided that this
5031 This change does not affect the value of $received_for. It is still set
5032 after address rewriting, but before local_scan() is called.
5034 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5036 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5037 gave an unhelpful panic error message, and a defer error. I have managed to
5038 change this behaviour so that it now rejects any supplied certificate,
5039 which seems right, as the list of acceptable certificates is empty.
5041 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5042 gave an unhelpful defer error. I have not managed to make this reject any
5043 supplied certificates, but the error message it gives is "no certificate
5044 supplied", which is not helpful.
5046 70. exigrep's output now also includes lines that are not associated with any
5047 message, but which match the given pattern. Implemented by a patch from
5048 Martin Sluka, which also tidied up the Perl a bit.
5050 71. Recipient callout verification, like sender verification, was using <> in
5051 the MAIL FROM command. This isn't really the right thing, since the actual
5052 sender may affect whether the remote host accepts the recipient or not. I
5053 have changed it to use the actual sender in the callout; this means that
5054 the cache record is now keyed on a recipient/sender pair, not just the
5055 recipient address. There doesn't seem to be a real danger of callout loops,
5056 since a callout by the remote host to check the sender would use <>.
5057 [SEE ABOVE: changed after hitting problems.]
5059 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5060 temporary errors. However, in the case of such a code being given after
5061 the end of a data transmission (i.e. after ".") Exim was failing to write
5062 a retry record for the message. (Yes, there was some broken host that was
5063 actually sending 8xx at this point.)
5065 73. An unknown lookup type in a host list could cause Exim to panic-die when
5066 the list was checked. (An example that provoked this was putting <; in the
5067 middle of a list instead of at the start.) If this happened during a DATA
5068 ACL check, a -D file could be left lying around. This kind of configuration
5069 error no longer causes Exim to die; instead it causes a defer errror. The
5070 incident is still logged to the main and panic logs.
5072 74. Buglet left over from Exim 3 conversion. The message "too many messages
5073 in one connection" was written to the rejectlog but not the mainlog, except
5074 when address rewriting (yes!) was being logged.
5076 75. Added write_rejectlog option.
5078 76. When a system filter was run not as root (that is, when system_filter_user
5079 was set), the values of the $n variables were not being returned to the
5080 main process; thus, they were not subsequently available in the $sn
5083 77. Added +return_path_on_delivery log selector.
5085 78. A connection timeout was being treated differently from recipients deferred
5086 when testing hosts_max_try with a message that was older than the host's
5087 retry timeout. (The host should not be counted, thus allowing all hosts to
5088 be tried at least once before bouncing.) This may have been the cause of an
5089 occasionally reported bug whereby a message would remain on the queue
5090 longer than the retry timeout, but would be bounced if a delivery was
5091 forced. I say "may" because I never totally pinned down the problem;
5092 setting up timeout/retry tests is difficult. See also the next item.
5094 79. The ultimate address timeout was not being applied to errors that involved
5095 a combination of host plus message (for example, a timeout on a MAIL
5096 command). When an address resolved to a number of possible hosts, and they
5097 were not all tried for each delivery (e.g. because of hosts_max_try), a
5098 message could remain on the queue longer than the retry timeout.
5100 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5103 81. Fixed an obscure SMTP outgoing bug which required at least the following
5104 conditions: (a) there was another message waiting for the same server;
5105 (b) the server returned 5xx to all RCPT commands in the first message so
5106 that the message was not completed; (c) the server dropped the connection
5107 or gave a negative response to the RSET that Exim sends to abort the
5108 transaction. The observed case was a dropped connection after DATA that had
5109 been sent in pipelining mode. That is, the server had advertised PIPELINING
5110 but was not implementing it correctly. The effect of the bug was incorrect
5111 behaviour, such as trying another host, and this could lead to a crash.
5117 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5118 and daemon.c were passed as pointers to ints; they should have been
5119 pointers to socklen_t variables (which are typically unsigned ints).
5121 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5124 3. Fixed a really odd bug that affected only the testing scheme; patching a
5125 certain fixed string in the binary changed the value of another string that
5126 happened to be identical to the end of the original first string.
5128 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5129 name", it returns that address as the IP address. On some operating
5130 systems (e.g. Solaris), it also passes back the IP address string as the
5131 "host name". However, on others (e.g. Linux), it passes back an empty
5132 string. Exim wasn't checking for this, and was changing the host name to an
5133 empty string, assuming it had been canonicized.
5135 5. Although rare, it is permitted to have more than one PTR record for a given
5136 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5137 all the names associated with an address, because they do in Solaris.
5138 However, it seems that they do not in Linux for data that comes from the
5139 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5140 I found this out when I moved to a new Linux workstation and tried to run
5141 the Exim test suite.
5143 To get round this problem I have changed the code so that it now does its
5144 own call to the DNS to look up PTR records when searching for a host name.
5145 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5146 addresses that are only in /etc/hosts are still found.
5148 This behaviour is, however, controlled by an option called host_lookup_
5149 order, which defaults to "bydns:byaddr". If people want to use the other
5150 order, or indeed, just use one or the other means of lookup, they can
5151 specify it in this variable.
5153 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5154 some operating systems, this comes back from gethostbyaddr() as an empty
5155 string, and this is what Exim used to test for. However, it seems that in
5156 other systems, "." is yielded. Exim now tests for this case too.
5158 7. The values of check_spool_space and check_log_space are now held internally
5159 as a number of kilobytes instead of an absolute number of bytes. If a
5160 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5161 kilobyte. This means that much larger values can be stored.
5163 8. Exim monitor: an attempt to get the action menu when not actually pointing
5164 at a message produces an empty menu entitled "No message selected". This
5165 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5166 no entries in it ("Shell widget menu has zero width and/or height"). So I
5167 have added a single, blank menu entry in this case.
5169 9. Added ${quote_local_part.
5171 10. MIME decoding is now applied to the contents of Subject: header lines when
5174 11. Now that a reference to $sender_host_address automatically causes a reverse
5175 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5176 host lookup before query-style lookups in lists that might use this
5177 variable. This has therefore been abolished, and the "net-" prefix is no
5178 longer necessary for query-style lookups.
5180 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5181 have been a typo for LFLAGS, so it has been changed.
5183 13. The install script calls Exim with "-C /dev/null" in order to find the
5184 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5185 to be output. Howeve, since Exim outputs its version number before the
5186 error, it didn't break the script. It just looked ugly. I fixed this by
5187 always allowing "-C /dev/null" if the caller is root.
5189 14. Ignore overlarge ACL variable number when reading spool file - insurance
5190 against a later release with more variables having written the file.
5192 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5193 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5194 this, as well as the form without the "IPv6" on the front.
5196 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5197 OS/Makefile-Darwin file.
5199 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5200 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5202 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5203 variable. However, some OS use ints for inodes. I've added cast to long int
5204 to get rid of the compiler warning.
5206 19. I had forgotten to lock out "/../" in configuration file names when
5207 ALT_CONFIG_PREFIX was set.
5209 20. Routers used for verification do not need to specify transports. However,
5210 if such a router generated a host list, and callout was configured, Exim
5211 crashed, because it could not find a port number from the (non-existent)
5212 transport. It now assumes port 25 in this circumstance.
5214 21. Added the -t option to exigrep.
5216 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5217 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5218 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5219 them is set, LOOKUP_LSEARCH is forced.
5221 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5223 24. Added sender and host information to the "rejected by local_scan()" log
5224 line; previously there was no indication of these.
5226 25. Added .include_if_exists.
5228 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5229 Linux. It turns out that not all file systems support this. Apparently some
5230 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5231 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5232 not supported on the file descriptor, is now ignored when Exim is trying to
5233 sync a directory. This applies only to Linux.
5235 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5237 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5238 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5241 29. Added support for maildirsize files from supplied patch (modified a bit).
5243 30. The use of :fail: followed by an empty string could lead Exim to respond to
5244 sender verification failures with (e.g.):
5246 550 Verification failed for <xxx>
5247 550 Sender verify failed
5249 where the first response line was missing the '-' that indicates it is not
5250 the final line of the response.
5252 31. The loop for finding the name of the user that called Exim had a hardwired
5253 limit of 10; it now uses the value of finduser_retries, which is used for
5254 all other user lookups.
5256 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5259 33. Exim was neglecting to zero errno before one call of strtol() when
5260 expanding a string and expecting an integer value. On some systems this
5261 resulted in spurious "integer overflow" errors. Also, it was casting the
5262 result into an int without checking.
5264 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5265 did not work. The code looks as if it has *never* worked, though it appears
5266 to have been documented since at least releast 1.62. I have made it work.
5268 35. The "timeout_DNS" error in retry rules, also documented since at least
5269 1.62, also never worked. As it isn't clear exactly what this means, and
5270 clearly it isn't a major issue, I have abolished the feature by treating it
5271 as "timeout", and writing a warning to the main and panic logs.
5273 36. The display of retry rules for -brt wasn't always showing the error code
5276 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5277 timeout_connect_A, timeout_connect_MX.
5279 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5280 to the empty sender.
5282 39. The daemon was not analysing the content of -oX till after it had closed
5283 stderr and disconnected from the controlling terminal. This meant that any
5284 syntax errors were only noted on the panic log, and the return code from
5285 the command was 0. By re-arranging the code a little, I've made the
5286 decoding happen first, so such errors now appear on stderr, and the return
5287 code is 1. However, the actual setting up of the sockets still happens in
5288 the disconnected process, so errors there are still only recorded on the
5291 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5292 connections (as happens on some IP stacks) was logged at start up time as
5293 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5294 it from "IPv6 and IPv4", which means that two separate sockets are being
5297 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5298 says whether AF_INET or AF_INET6 was passed as an argument.
5300 42. Exiwhat output was messed up when time zones were included in log
5303 43. Exiwhat now gives more information about the daemon's listening ports,
5304 and whether -tls-on-connect was used.
5306 44. The "port" option of the smtp transport is now expanded.
5308 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5309 silently ignored. Now an error message is written to the main and panic
5312 46. There's a new ACL modifier called "logwrite" which writes to a log file
5313 as soon as it is encountered.
5315 47. Added $local_user_uid and $local_user_gid at routing time.
5317 48. Exim crashed when trying to verify a sender address that was being
5320 49. Exim was recognizing only a space character after ".include". It now also
5321 recognizes a tab character.
5323 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5324 extracting the relevant information from the specification. The man page no
5325 longer contains scrambled data for the -d option, and I've added a section
5326 at the front about calling Exim under different names.
5328 51. Added "extra_headers" argument to the "mail" command in filter files.
5330 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5333 53. Installed eximstats 1.29.
5335 54. Added transport_filter_timeout as a generic transport option.
5337 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5338 Cc: header lines. This was required by RFC 822, but it not required by RFC
5341 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5342 incoming messages that did not have them. Now it does so only if the
5343 message originates locally, that is, if there is no associated remote host
5344 address. When Resent- header lines are present, this applies to the Resent-
5345 lines rather than the non-Resent- lines.
5347 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5348 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5350 58. Messages for configuration errors now include the name of the main
5351 configuration file - useful now that there may be more than one file in a
5352 list (.included file names were always shown).
5354 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5355 for those rare installations that do not start the daemon as root or run it
5356 setuid root. I've cut out the call to initgroups() if the daemon is not
5359 60. The Exim user and group can now be bound into the binary as text strings
5360 that are looked up at the start of Exim's processing.
5362 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5364 62. Added $mailstore_basename variable.
5366 63. Installed patch to sieve.c from Michael Haardt.
5368 64. When Exim failed to open the panic log after failing to open the main log,
5369 the original message it was trying to log was written to stderr and debug
5370 output, but if they were not available (the usual case in production), it
5371 was lost. Now it is written to syslog before the two lines that record the
5372 failures to open the logs.
5374 65. Users' Exim filters run in subprocesses under the user's uid. It is
5375 possible for a "deliver" command or an alias in a "personal" command to
5376 provoke an address rewrite. If logging of address rewriting is configured,
5377 this fails because the process is not running as root or exim. There may be
5378 a better way of dealing with this, but for the moment (because 4.30 needs
5379 to be released), I have disabled address rewrite logging when running a
5380 filter in a non-root, non-exim process.
5386 1. The buildconfig auxiliary program wasn't quoting the value set for
5387 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5388 not defined. This bug was masked in 4.22 by the effect that was fixed in
5391 2. Some messages that were rejected after a message id was allocated were
5392 shown as "incomplete" by exigrep. It no longer does this for messages that
5393 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5395 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5396 have allow_domain_literals set, the ID did not get logged in the <= line.
5397 Domain literals are now always recognized in Message-ID: header lines.
5399 4. The first argument for a ${extract expansion item is the key name or field
5400 number. Leading and trailing spaces in this item were not being ignored,
5401 causing some misleading effects.
5403 5. When deliver_drop_privilege was set, single queue runner processes started
5404 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5405 same command in the process it spins off) were not dropping privilege.
5407 6. When the daemon running as "exim" started a queue runner, it always
5408 re-executed Exim in the spun-off process. This is a waste of effort when
5409 deliver_drop_privilege is set. The new process now just calls the
5410 queue-runner function directly.
5416 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5419 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5420 ignored. Though the use of -oP was forcing the writing of a pid file, it
5421 was always written to the default place.
5423 3. If the message "no IP address found for host xxxx" is generated during
5424 incoming verification, it is now followed by identification of the incoming
5425 connection (so you can more easily find what provoked it).
5427 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5429 5. Added some features to "harden" Exim a bit more against certain attacks:
5431 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5432 be put in Local/Makefile. This is like the never_users runtime option,
5433 but it cannot be overridden. The default setting is "root".
5435 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5436 prefix string with which any file named in a -C command line option
5439 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5440 is retained for -C and -D only if the caller of Exim is root. Without
5441 it, the exim user may also use -C and -D and retain privilege.
5443 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5444 command line option is disabled.
5446 6. Macro names set by the -D option must start with an upper case letter, just
5447 like macro names defined in the configuration file.
5449 7. Added "dereference=" facility to LDAP.
5451 8. Two instances of the typo "uknown" in the source files are fixed.
5453 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5454 the Configure-Makefile script screwed up while processing it.
5456 10. Incorporated PCRE 4.4.
5458 11. The SMTP synchronization check was not operating right at the start of an
5459 SMTP session. For example, it could not catch a HELO sent before the client
5460 waited for the greeting. There is now a check for outstanding input at the
5461 point when the greeting is written. Because of the duplex, asynchronous
5462 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5463 way, but not yet received, when the check is performed.
5465 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5466 on TCP/IP sockets, because this apparently causes some broken clients to
5469 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5470 unchanged) from the Cygwin maintainer.
5472 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5473 was supported instead of testing for mbx. Effectively a typo.
5475 15. The spa authenticator server code was not checking that the input it
5476 received was valid base64.
5478 16. The debug output line for the "set" modifier in ACLs was not showing the
5479 name of the variable that was being set.
5481 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5483 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5484 lookup on its own. Something else was needed to trigger the lookup. For
5485 example, a match in host_lookup or the need for a host name in a host list.
5486 Now, if $sender_host_name is referenced and the host name has not yet been
5487 looked up, a lookup is performed. If the lookup fails, the variable remains
5488 empty, and $host_lookup_failed is set to "1".
5490 19. Added "eqi" as a case-independent comparison operator.
5492 20. The saslauthd authentication condition could segfault if neither service
5493 nor realm was specified.
5495 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5496 error message that was logged was misleading, and incoming SMTP
5497 connections were dropped. The message is now more accurate, and temporary
5498 errors are given to SMTP connections.
5500 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5501 (except RSET) with a 421 error, until QUIT is received. However, it was
5502 failing to send a response to QUIT.
5504 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5505 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5506 ACL is now run after the helo_try_verify_hosts code.
5508 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5509 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5510 case-independent, so other case variants are also recognized). Apparently
5511 some systems use these upper case variants.
5513 25. If more than two messages were waiting for the same host, and a transport
5514 filter was specified for the transport, Exim sent two messages over the
5515 same TCP/IP connection, and then failed with "socket operation on non-
5516 socket" when it tried to send the third.
5518 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5520 27. The extern definition of crypt16() in expand.c was not being excluded when
5521 the OS had its own crypt16() function.
5523 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5524 as a preferred synonym for return_size_limit, both as an option and as an
5527 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5529 30. Changed the default configuration ACL to relax the local part checking rule
5530 for addresses that are not in any local domains. For these addresses,
5531 slashes and pipe symbols are allowed within local parts, but the sequence
5532 /../ is explicitly forbidden.
5534 31. SPA server authentication was not clearing the challenge buffer before
5537 32. log_message in a "warn" ACL statement was writing to the reject log as
5538 well as to the main log, which contradicts the documentation and doesn't
5539 seem right (because no rejection is happening). So I have stopped it.
5541 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5542 However, I am unable to do any testing of this.
5544 34. Fixed an infelicity in the appendfile transport. When checking directories
5545 for a mailbox, to see if any needed to be created, it was accidentally
5546 using path names with one or more superfluous leading slashes; tracing
5547 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5549 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5550 contents are added to the log line that is written for every discarded
5551 recipient. (Previously a log_message setting was ignored.)
5553 36. The ${quote: operator now quotes the string if it is empty.
5555 37. The install script runs exim in order to find its version number. If for
5556 some reason other than non-existence or emptiness, which it checks, it
5557 could not run './exim', it was installing it with an empty version number,
5558 i.e. as "exim-". This error state is now caught, and the installation is
5561 38. An argument was missing from the function that creates an error message
5562 when Exim fails to connect to the socket for saslauthd authentication.
5563 This could cause Exim to crash, or give a corrupted message.
5565 39. Added isip, isip4, and isip6 to ${if conditions.
5567 40. The ACL variables $acl_xx are now saved with the message, and can be
5568 accessed later in routers, transports, and filters.
5570 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5571 strings in the file are not string-expanded.
5573 42. If a MAIL command specified a SIZE value that was too large to fit into an
5574 int variable, the check against message_size_limit failed. Such values are
5575 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5576 day this will have to be increased, but I don't think I want to be around
5577 when emails are that large.
5584 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5585 iconv() is not standard in FreeBSD.
5587 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5588 IPv6 enabled. The observed symptom was a segmentation fault on return from
5589 the function os_common_find_running_interfaces() in src/os.c.
5591 3. In the check_special_case() function in daemon.c I had used "errno" as an
5592 argument name, which causes warnings on some systems. This was basically a
5593 typo, since it was named "eno" in the comments!
5595 4. The code that waits for the clock to tick (at a resolution of some fraction
5596 of a second) so as to ensure message-id uniqueness was always waiting for
5597 at least one whole tick, when it could have waited for less. [This is
5598 almost certainly not relevant at current processor speeds, where it is
5599 unlikely to ever wait at all. But we try to future-proof.]
5601 5. The function that sleeps for a time interval that includes fractions of a
5602 second contained a race. It did not block SIGALRM between setting the
5603 timer, and suspending (a couple of lines later). If the interval was short
5604 and the sigsuspend() was delayed until after it had expired, the suspension
5605 never ended. On busy systems this could lead to processes getting stuck for
5608 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5609 process, before it forks any delivery processes. The open lookup caching
5610 mechanism meant that the open file or database connection was passed into
5611 the delivery process. The problem was that delivery processes always tidy
5612 up cached lookup data. This could cause a problem for the next delivery
5613 process started by the queue runner, because the external queue runner
5614 process does not know about the closure. So the next delivery process
5615 still has data in the lookup cache. In the case of a file lookup, there was
5616 no problem because closing a file descriptor in a subprocess doesn't affect
5617 the parent. However, if the lookup was caching a connection to a database,
5618 the connection was closed, and the second delivery process was likely to
5619 see errors such as "PGSQL: query failed: server closed the connection
5620 unexpectedly". The problem has been fixed by closing all cached lookups
5621 in a queue runner before running a delivery process.
5623 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5624 seem to have the "const" qualifier which it has on other OS. I've
5627 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5628 *of the same type* (client or server) with the same public name that an
5629 error should be diagnosed.
5631 9. When Exim looked up a host name for an IP address, but failed to find the
5632 original IP address when looking up the host name (a safety check), it
5633 output the message "<ip address> does not match any IP for NULL", which was
5634 confusing, to say the least. The bug was that the host name should have
5635 appeared instead of "NULL".
5637 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5638 user that is built into the binary, and the -C or -D options is used, root
5639 privilege is dropped before the configuration file is read. In addition,
5640 logging is switched to stderr instead of the normal log files. If the
5641 configuration then re-defines the Exim user, the unprivileged environment
5642 is probably not what is expected, so Exim logs a panic warning message (but
5645 However, if deliver_drop_privilege is set, the unprivileged state may well
5646 be exactly what is intended, so the warning has been cut out in that case,
5647 and Exim is allowed to try to write to its normal log files.
5653 1. smtp_return_error_details was not giving details for temporary sender
5654 or receiver verification errors.
5656 2. Diagnose a configuration error if two authenticators have the same public
5659 3. Exim used not to create the message log file for a message until the first
5660 delivery attempt. This could be confusing when incoming messages were held
5661 for policy or load reasons. The message log file is now created at the time
5662 the message is received, and an initial "Received" line is written to it.
5664 4. The automatically generated man page for command line options had a minor
5665 bug that caused no ill effects; however, a more serious problem was that
5666 the procedure for building the man page automatically didn't always
5667 operate. Consequently, release 4.20 contains an out-of-date version. This
5668 shouldn't happen again.
5670 5. When building Exim with embedded Perl support, the script that builds the
5671 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5672 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5674 6. The freeze_tell option was not being used for messages that were frozen on
5675 arrival, either by an ACL or by local_scan().
5677 7. Added the smtp_incomplete_transaction log selector.
5679 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5680 was accepting AUTH without a new EHLO.
5682 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5683 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5686 10. Logging of TCP/IP connections (when configured) now happens in the main
5687 daemon process instead of the child process, so that the TCP/IP connection
5688 count is more accurate (but it can never be perfect).
5690 11. The use of "drop" in a nested ACL was not being handled correctly in the
5691 outer ACL. Now, if condition failure induced by the nested "drop" causes
5692 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5693 or "require"), the connection is dropped.
5695 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5696 that yield "discard" can now be used with an "accept" or a "discard" verb,
5697 but an error is generated for any others (because I can't see a useful way
5698 to define what should happen).
5700 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5701 are now processed in the same way as lines in the Exim configuration file.
5702 In particular, continuation lines are supported.
5704 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5706 15. Added -ti meaning -t -i.
5708 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5709 domains, and warn by logging if others are found.
5711 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5712 returned from GIFCONF when Exim is trying to find the list of interfaces on
5713 a host. The code in os.c has been modified to copy each ifreq to an aligned
5714 structure in all cases.
5716 Also, in some cases, the returned ifreq's were being copied to a 'struct
5717 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5718 means the last couple of bytes of an IPv6 address could be chopped if the
5719 ifreq contained only a normal sockaddr (14 bytes storage).
5721 18. Named domain lists were not supported in the hosts_treat_as_local option.
5722 An entry such as +xxxx was not recognized, and was treated as a literal
5725 19. Ensure that header lines added by a DATA ACL are included in the reject log
5726 if the ACL subsequently rejects the message.
5728 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5729 MD5 (which is deprecated).
5731 21. When testing a filter file using -bf, Exim was writing a message when it
5732 took the sender from a "From " line in the message, but it was not doing so
5733 when it took $return_path from a Return-Path: header line. It now does.
5735 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5736 with a valid header line field name (a series of printing characters
5737 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5739 23. Changed "disc" in the source to "disk" to conform to the documentation and
5740 the book and for uniformity.
5742 24. Ignore Sendmail's -Ooption=value command line item.
5744 25. When execve() failed while trying to run a command in a pipe transport,
5745 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5746 could be confused with a return value of 69 from the command itself. This
5747 has been changed to 127, the value the shell returns if it is asked to run
5748 a non-existent command. The wording for the related log line suggests a
5749 non-existent command as the problem.
5751 26. If received_header_text expands to an empty string, do not add a Received:
5752 header line to the message. (Well, it adds a token one on the spool, but
5753 marks it "old" so that it doesn't get used or transmitted.)
5755 27. Installed eximstats 1.28 (addition of -nt option).
5757 28. There was no check for failure on the call to getsockname() in the daemon
5758 code. This can fail if there is a shortage of resources on the system, with
5759 ENOMEM, for example. A temporary error is now given on failure.
5761 29. Contrary to the C standard, it seems that in some environments, the
5762 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5763 program. Exim now does this explicitly; it affects the formatting of
5764 timestamps using strftime().
5766 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5767 complaints. I've now initialized all the variables, to avoid this.
5769 32. Header lines added by a system filter were not being "seen" during
5770 transport-time rewrites.
5772 33. The info_callback() function passed to OpenSSL is set up with type void
5773 (*)(SSL *, int, int), as described somewhere. However, when calling the
5774 function (actually a macro) that sets it up, the type void(*)() is
5775 expected. I've put in a cast to prevent warnings from picky compilers.
5777 34. If a DNS black list lookup found a CNAME record, but there were no A
5778 records associated with the domain it pointed at, Exim crashed.
5780 35. If a DNS black list lookup returned more than one A record, Exim ignored
5781 all but the first. It now scans all returned addresses if a particular IP
5782 value is being sought. In this situation, the contents of the
5783 $dnslist_value variable are a list of all the addresses, separated by a
5786 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5787 to accept a host name and all its aliases if the forward lookup for any of
5788 them yielded the IP address of the incoming connection. Now it accepts only
5789 those names whose forward lookup yields the correct IP address. Any other
5790 names are discarded. This closes a loophole whereby a rogue DNS
5791 administrator could create reverse DNS records to break through a
5792 wildcarded host restriction in an ACL.
5794 37. If a user filter or a system filter that ran in a subprocess used any of
5795 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5796 the wrong values were passed to the pipe command ($thisaddress had the
5797 value of $0, $0 had the value of $1, etc). This bug was introduced by
5798 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5800 38. Improved the line breaking for long SMTP error messages from ACLs.
5801 Previously, if there was no break point between 40 and 75 characters, Exim
5802 left the rest of the message alone. Two changes have been made: (a) I've
5803 reduced the minimum length to 35 characters; (b) if it can't find a break
5804 point between 35 and 75 characters, it looks ahead and uses the first one
5805 that it finds. This may give the occasional overlong line, but at least the
5806 remaining text gets split now.
5808 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5809 file descriptors might be low, and that setting 1000 would always raise it.
5810 It turns out that in some environments, the limit is already over 1000 and
5811 that lowering it causes trouble. So now Exim takes care not to decrease it.
5813 40. When delivering a message, the value of $return_path is set to $sender_
5814 address at the start of routing (routers may change the value). By an
5815 oversight, this default was not being set up when an address was tested by
5816 -bt or -bv, which affected the outcome if any router or filter referred to
5819 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5820 log only when "message" or "log_message" are set. However, if one of the
5821 conditions was an address verification, or a call to a nested ACL, the
5822 messages generated by the underlying test were being passed through. This
5823 no longer happens. The underlying message is available in $acl_verify_
5824 message for both "message" and "log_message" expansions, so it can be
5825 passed through if needed.
5827 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5828 new expansion $bh_ to give the encoded byte string without charset
5829 translation. Translation happens only if iconv() is available; HAVE_ICONV
5830 indicates this at build time. HEADERS_CHARSET gives the charset to
5831 translate to; headers_charset can change it in the configuration, and
5832 "headers charset" can change it in an individual filter file.
5834 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5835 that creates RFC 2047 encoded "words" labels them as that charset instead
5836 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5837 expansion operator; (ii) when Exim creates a From: line for a local
5838 message; (iii) when a header line is rewritten to include a "phrase" part.
5840 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5841 buggy, causing it to skip the first lines of messages whose message ID
5842 ended in 'D'. This would not have bitten before Exim release 4.14, because
5843 message IDs were unlikely to end in 'D' before then. The effect was to have
5844 incorrect size information for certain domains.
5846 45. #include "config.h" was missing at the start of the crypt16.c module. This
5847 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5850 46. If there was a timeout during a "random" callout check, Exim treated it as
5851 a failure of the random address, and carried on sending RSET and the real
5852 address. If the delay was just some slowness somewhere, the response to the
5853 original RCPT would be taken as a response to RSET and so on, causing
5854 mayhem of various kinds.
5856 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5857 when I implemented it. It didn't allow for the fact that some option values
5858 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5859 the right test for positive values.
5861 48. Domain names in DNS records are case-independent. Exim always looks them up
5862 in lower case. Some resolvers return domain names in exactly the case they
5863 appear in the zone file, that is, they may contain uppercase letters. Not
5864 all resolvers do this - some return always lower case. Exim was treating a
5865 change of case by a resolver as a change of domain, similar to a widening
5866 of a domain abbreviation. This triggered its re-routing code and so it was
5867 trying to route what was effectively the same domain again. This normally
5868 caused routing to fail (because the router wouldn't handle the domain
5869 twice). Now Exim checks for this case specially, and just changes the
5870 casing of the domain that it ultimately uses when it transmits the message
5873 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5876 50. If a filter generated a file delivery with a non-absolute name (possible if
5877 no home directory exists for the router), the forbid_file option was not
5880 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5881 the existing equality matching.
5883 52. Exim was using ints instead of ino_t variables in some places where it was
5884 dealing with inode numbers.
5886 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5887 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5888 called TMPDIR, and if it finds it is different, it changes its value.
5890 54. The smtp_printf() function is now made available to local_scan() so
5891 additional output lines can be written before returning. There is also an
5892 smtp_fflush() function to enable the detection of a dropped connection.
5893 The variables smtp_input and smtp_batched_input are exported to
5896 55. Changed the default runtime configuration: the message "Unknown user"
5897 has been removed from the ACL, and instead placed on the localuser router,
5898 using the cannot_route_message feature. This means that any verification
5899 failures that generate their own messages won't get overridden. Similarly,
5900 the "Unrouteable address" message that was in the ACL for unverifiable
5901 relay addresses has also been removed.
5903 56. Added hosts_avoid_esmtp to the smtp transport.
5905 57. The exicyclog script was not checking for the esoteric option
5906 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5907 will work only if exicyclog is run under the appropriate euid.
5909 58. Following a discussion on the list, the rules by which Exim recognises line
5910 endings on incoming messages have been changed. The -dropcr and drop_cr
5911 options are now no-ops, retained only for backwards compatibility. The
5912 following line terminators are recognized: LF CRLF CR. However, special
5913 processing applies to CR:
5915 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5916 nor a local message in the state where . is a terminator.
5918 (ii) If a bare CR is encountered in a header line, an extra space is added
5919 after the line terminator so as not to end the header. The reasoning
5920 behind this is that bare CRs in header lines are most likely either
5921 to be mistakes, or people trying to play silly games.
5923 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5924 was being incorrectly given as 18 bytes larger than it should have been.
5925 This is a VOB (very old bug).
5927 60. This may never have affected anything current, but just in case it has:
5928 When the local host is found other than at the start of a list of hosts,
5929 the local host, those with the same MX, and any that follow, are discarded.
5930 When the list in question was part of a longer list of hosts, the following
5931 hosts (not currently being processed) were also being discarded. This no
5932 longer happens. I'm not sure if this situation could ever has previously
5935 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5938 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5939 header. This is something that is recommended in a new Internet Draft, and
5940 is something that is documented as being done by Sendmail. There are two
5941 possible values. For messages generated by the autoreply transport, Exim
5944 Auto-Submitted: auto-replied
5946 whereas for all other generated messages (e.g. bounces) it adds
5948 Auto-Submitted: auto-generated
5950 63. The "personal" condition in filters now includes a test for the
5951 Auto-Submitted: header. If it contains the string "auto-" the message it
5952 not considered personal.
5954 64. Added rcpt_include_affixes as a generic transport option.
5956 65. Added queue_only_override (default true).
5958 66. Added the syslog_duplication option.
5960 67. If what should have been the first header line of a message consisted of
5961 a space followed by a colon, Exim was mis-interpreting it as a header line.
5962 It isn't of course - it is syntactically invalid and should therefore be
5963 treated as the start of the message body. The misbehaviour could have
5964 caused a number of strange effects, including loss of data in subsequent
5965 header lines, and spool format errors.
5967 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5968 client host had authenticated. This control can now be exercised by an ACL
5969 for more flexibility.
5971 69. By default, callouts do not happen when testing with -bh. There is now a
5972 variant, -bhc, which does actually run the callout code, including
5973 consulting and updating the callout cache.
5975 70. Added support for saslauthd authentication, courtesy of Alexander
5978 71. If statvfs() failed on the spool or log directories while checking their
5979 size for availability, Exim confusingly gave the error "space shortage".
5980 Furthermore, in debugging mode it crashed with a floating point exception.
5981 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5982 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5983 serious problem, Exim now writes to the main and panic logs when this
5984 happens, with details of the failure. It then refuses to accept the
5985 incoming message, giving the message "spool directory problem" or "log
5986 directory problem" with a 421 code for SMTP messages.
5988 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5989 0, 1, and 2 exist, because some OS complain for execs without them (see
5990 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5991 descriptors. However, the code omitted to check that the open succeeded,
5992 causing mysterious errors if for some reason the permissions on /dev/null
5993 got screwed. Now Exim writes a message to the main and panic logs, and
5994 bombs out if it can't open /dev/null.
5996 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5997 interact so that it is all more flexible. It is supposed to remain
5998 backwards compatible. Also added extra_local_interfaces.
6000 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6001 to bomb out with an assertion failure - to the client this appears as a
6002 connection drop. This problem occurs in the part of the code that was taken
6003 from the Samba project. Fortunately, the assertion is in a very simple
6004 function, so I have fixed this by reproducing the function inline in the
6005 one place where it is called, and arranging for authentication to fail
6006 instead of killing the process with assert().
6008 75. The SPA client code was not working when the server requested OEM rather
6009 than Unicode encoding.
6011 76. Added code to make require_files with a specific uid setting more usable in
6012 the case where statting the file as root fails - usually a non-root-mounted
6013 NFS file system. When this happens and the failure is EACCES, Exim now
6014 forks a subprocess and does the per-uid checking as the relevant uid.
6016 77. Added process_log_path.
6018 78. If log_file_path was not explicitly set, a setting of check_log_space or
6019 check_log_inodes was ignored.
6021 79. If a space check for the spool or log partitions fails, the incident is now
6022 logged. Of course, in the latter case the data may get lost...
6024 80. Added the %p formatting code to string_format() so that it can be used to
6025 print addresses in debug_print(). Adjusted all the address printing in the
6026 debugging in store.c to use %p rather than %d.
6028 81. There was a concern that a line of code in smtp_in.c could overflow a
6029 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6030 initially expressed, the concern was not well-founded, because trailing
6031 spaces are removed early. However, if the trailing spaces were followed by
6032 a NULL, they did not get removed, so the overflow was possible. Two fixes
6035 (a) I re-wrote the offending code in a cleaner fashion.
6036 (b) If an incoming SMTP command contains a NULL character, it is rejected
6039 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6040 runs initgroups(), so that if the Exim user is in any additional groups,
6041 they will be used during message reception.
6047 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob