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.
129 JH/25 Expand the coverage of the delivery $host and $host_address to
130 client authenticators run in verify callout. Bug 1476.
132 JH/26 Port service names are now accepted for tls_on_connect_ports, to
133 align with daemon_smtp_ports. Bug 72.
135 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
136 support and error reporting did not work properly.
142 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
144 PP/02 Make -n do something, by making it not do something.
145 When combined with -bP, the name of an option is not output.
147 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
150 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
151 $sender_host_name and config options to manage this, and basic check
154 PP/05 DSCP support for outbound connections and control modifier for inbound.
156 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
157 (Only plugin which currently uses this is kerberos4, which nobody should
158 be using, but we should make it available and other future plugins might
159 conceivably use it, even though it would break NAT; stuff *should* be
160 using channel bindings instead).
162 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
163 name; added for Sendmail compatibility; requires admin caller.
164 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
165 just ignore it); requires trusted caller.
166 Also parse but ignore: -Ac -Am -X<logfile>
169 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
171 TL/02 Add +smtp_confirmation as a default logging option.
173 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
174 Patch by Magnus Holmgren from 2007-02-20.
176 TL/04 Bugzilla 1281 - Spec typo.
177 Bugzilla 1283 - Spec typo.
178 Bugzilla 1290 - Spec grammar fixes.
180 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
182 TL/06 Add Experimental DMARC support using libopendmarc libraries.
184 TL/07 Fix an out of order global option causing a segfault. Reported to dev
185 mailing list by by Dmitry Isaikin.
187 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
189 JH/02 Support "G" suffix to numbers in ${if comparisons.
191 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
193 NM/01 Bugzilla 1197 - Spec typo
194 Bugzilla 1196 - Spec examples corrections
196 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
198 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
199 gnutls_enable_pkcs11, but renamed to more accurately indicate its
202 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
203 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
205 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
206 "acl {{name}{arg}...}", and optional args on acl condition
209 JH/05 Permit multiple router/transport headers_add/remove lines.
211 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
213 JH/07 Avoid using a waiting database for a single-message-only transport.
214 Performance patch from Paul Fisher. Bugzilla 1262.
216 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
219 JH/09 Add $headers_added variable, with content from use of ACL modifier
220 add_header (but not yet added to the message). Bugzilla 199.
222 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
223 Pulled from Bugzilla 817 by Wolfgang Breyha.
225 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
227 (nb: this is the same fix as in Exim 4.80.1)
229 JH/11 Add A= logging on delivery lines, and a client_set_id option on
232 JH/12 Add optional authenticated_sender logging to A= and a log_selector
235 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
237 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
238 advertise SMTP AUTH mechanism to us, instead of a generic
239 protocol violation error. Also, make Exim more robust to bad
240 data from the Dovecot auth socket.
242 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
244 When a queue runner is handling a message, Exim first routes the
245 recipient addresses, during which it prunes them based on the retry
246 hints database. After that it attempts to deliver the message to
247 any remaining recipients. It then updates the hints database using
250 So if a recipient address works intermittently, it can get repeatedly
251 deferred at routing time. The retry hints record remains fresh so the
252 address never reaches the final cutoff time.
254 This is a fairly common occurrence when a user is bumping up against
255 their storage quota. Exim had some logic in its local delivery code
256 to deal with this. However it did not apply to per-recipient defers
257 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
259 This change adds a proper retry rule check during routing so that the
260 final cutoff time is checked against the message's age. We only do
261 this check if there is an address retry record and there is not a
262 domain retry record; this implies that previous attempts to handle
263 the address had the retry_use_local_parts option turned on. We use
264 this as an approximation for the destination being like a local
265 delivery, as in LMTP.
267 I suspect this new check makes the old local delivery cutoff check
268 redundant, but I have not verified this so I left the code in place.
270 TF/02 Correct gecos expansion when From: is a prefix of the username.
272 Test 0254 submits a message to Exim with the header
276 When I ran the test suite under the user fanf2, Exim expanded
277 the header to contain my full name, whereas it should have added
278 a Resent-Sender: header. It erroneously treats any prefix of the
279 username as equal to the username.
281 This change corrects that bug.
283 GF/01 DCC debug and logging tidyup
284 Error conditions log to paniclog rather than rejectlog.
285 Debug lines prefixed by "DCC: " to remove any ambiguity.
287 TF/03 Avoid unnecessary rebuilds of lookup-related code.
289 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
290 Bug spotted by Jeremy Harris; was flawed since initial commit.
291 Would have resulted in OCSP responses post-SNI triggering an Exim
292 NULL dereference and crash.
294 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
296 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
297 Bug detection, analysis and fix by Samuel Thibault.
298 Bugzilla 1331, Debian bug #698092.
300 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
302 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
303 Server implementation by Todd Lyons, client by JH.
304 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
305 config variable "prdr_enable" controls whether the server
306 advertises the facility. If the client requests PRDR a new
307 acl_data_smtp_prdr ACL is called once for each recipient, after
308 the body content is received and before the acl_smtp_data ACL.
309 The client is controlled by bolth of: a hosts_try_prdr option
310 on the smtp transport, and the server advertisement.
311 Default client logging of deliveries and rejections involving
312 PRDR are flagged with the string "PRDR".
314 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
315 fclose(). Diagnosis by Todd Lyons.
317 PP/17 Update configure.default to handle IPv6 localhost better.
318 Patch by Alain Williams (plus minor tweaks).
321 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
322 This is now consistent with GnuTLS, and is now documented: the
323 previous undocumented portable approach to treating the option as
324 unset was to force an expansion failure. That still works, and
325 an empty string is now equivalent.
327 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
328 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
329 not performing validation itself.
331 PP/20 Added force_command boolean option to pipe transport.
332 Patch from Nick Koston, of cPanel Inc.
334 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
337 TF/04 Added udpsend ACL modifer and hexquote expansion operator
339 PP/21 Fix eximon continuous updating with timestamped log-files.
340 Broken in a format-string cleanup in 4.80, missed when I repaired the
341 other false fix of the same issue.
342 Report and fix from Heiko Schlichting.
345 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
346 Report from Prashanth Katuri.
348 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
349 It's SecureTransport, so affects any MacOS clients which use the
350 system-integrated TLS libraries, including email clients.
352 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
353 using a MIME ACL for non-SMTP local injection.
354 Report and assistance in diagnosis by Warren Baker.
356 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
358 JH/16 Fix comparisons for 64b. Bugzilla 1385.
360 TL/09 Add expansion variable $authenticated_fail_id to keep track of
361 last id that failed so it may be referenced in subsequent ACL's.
363 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
366 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
367 ldap library initialization, allowing self-signed CA's to be
368 used. Also properly sets require_cert option later in code by
369 using NULL (global ldap config) instead of ldap handle (per
370 session). Bug diagnosis and testing by alxgomz.
372 TL/12 Enhanced documentation in the ratelimit.pl script provided in
373 the src/util/ subdirectory.
375 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
376 renamed to Transport Post Delivery Action by Jeremy Harris, as
379 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
380 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
381 redis_servers = needs to be configured which will be used by the redis
382 lookup. Patch from Warren Baker, of The Packet Hub.
384 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
386 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
387 hostname or reverse DNS when processing a host list. Used suggestions
388 from multiple comments on this bug.
390 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
392 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
393 Missed a few lines, added it to make the runtest require no keyboard
396 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
397 contains upper case chars. Make router use caseful_local_part.
399 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
400 support when GnuTLS has been built with p11-kit.
406 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
408 This, or similar/improved, will also be change PP/11 of 4.82.
414 PP/01 Handle short writes when writing local log-files.
415 In practice, only affects FreeBSD (8 onwards).
416 Bugzilla 1053, with thanks to Dmitry Isaikin.
418 NM/01 Bugzilla 949 - Documentation tweak
420 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
423 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
425 PP/02 Implemented gsasl authenticator.
427 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
429 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
430 `pkg-config foo` for cflags/libs.
432 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
433 with rest of GSASL and with heimdal_gssapi.
435 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
436 `pkg-config foo` for cflags/libs for the TLS implementation.
438 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
439 properties get this fed in as external SSF. A number of robustness
440 and debugging improvements to the cyrus_sasl authenticator.
442 PP/08 cyrus_sasl server now expands the server_realm option.
444 PP/09 Bugzilla 1214 - Log authentication information in reject log.
445 Patch by Jeremy Harris.
447 PP/10 Added dbmjz lookup type.
449 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
451 PP/12 MAIL args handles TAB as well as SP, for better interop with
452 non-compliant senders.
453 Analysis and variant patch by Todd Lyons.
455 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
456 Bug report from Lars Müller <lars@samba.org> (via SUSE),
457 Patch from Dirk Mueller <dmueller@suse.com>
459 PP/13 tls_peerdn now print-escaped for spool files.
460 Observed some $tls_peerdn in wild which contained \n, which resulted
461 in spool file corruption.
463 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
464 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
465 or write after TLS renegotiation, which otherwise led to messages
468 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
469 as a tracking header (ie: a signed header comes before the signature).
470 Patch from Wolfgang Breyha.
472 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
473 comma-sep list; embedded commas doubled.
475 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
477 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
479 Report and patch from Dmitry Banschikov.
481 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
482 Removed SSL_clear() after SSL_new() which led to protocol negotiation
483 failures. We appear to now support TLS1.1+ with Exim.
485 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
486 lets Exim select keys and certificates based upon TLS SNI from client.
487 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
488 before an outbound SMTP session. New log_selector, +tls_sni.
490 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
491 NULL dereference. Report and patch from Alun Jones.
493 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
494 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
495 Not seeing resolver debug output on NetBSD, but suspect this is a
496 resolver implementation change.
498 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
499 Left warnings. Added "eximon gdb" invocation mode.
501 PP/21 Defaulting "accept_8bitmime" to true, not false.
503 PP/22 Added -bw for inetd wait mode support.
505 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
506 locate the relevant includes and libraries. Made this the default.
508 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
509 Bugzilla 1246, report and most of solution from Tomasz Kusy.
511 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
512 This may cause build issues on older platforms.
514 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
515 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
516 gnutls_require_mac & gnutls_require_protocols (no longer supported).
517 Added SNI support via GnuTLS too.
518 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
520 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
522 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
523 Applied second patch from Janne, implementing suggestion to default
524 multiple-strings-in-record handling to match SPF spec.
526 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
528 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
529 read-only, out of scope).
530 Patch from Wolfgang Breyha, report from Stuart Northfield.
532 PP/29 Fix three issues highlighted by clang analyser static analysis.
533 Only crash-plausible issue would require the Cambridge-specific
534 iplookup router and a misconfiguration.
535 Report from Marcin Mirosław.
537 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
539 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
540 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
541 As part of this, removing so much warning spew let me fix some minor
542 real issues in debug logging.
544 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
545 assignment on my part. Fixed.
547 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
548 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
549 Janne Snabb (who went above and beyond: thank you).
551 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
552 string otherwise requires a connection and a bunch more work and it's
553 relatively easy to get wrong. Should also expose TLS library linkage
556 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
557 64-bit ${eval} (JH/03).
559 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
560 GNU libc to support some of the 64-bit stuff, should not lead to
561 conflicts. Defined before os.h is pulled in, so if a given platform
562 needs to override this, it can.
564 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
565 protection layer was required, which is not implemented.
566 Bugzilla 1254, patch from Wolfgang Breyha.
568 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
569 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
570 tls_dhparam take prime identifiers. Also unbreak combination of
571 OpenSSL+DH_params+TLSSNI.
573 PP/39 Disable SSLv2 by default in OpenSSL support.
579 PP/01 Solaris build fix for Oracle's LDAP libraries.
580 Bugzilla 1109, patch from Stephen Usher.
582 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
584 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
587 TF/02 Fix a couple more cases where we did not log the error message
588 when unlink() failed. See also change 4.74-TF/03.
590 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
591 lock up or crash if it happened to be inside a call to libc when it
592 got a SIGUSR1 from exiwhat.
594 The SIGUSR1 handler appends the current process status to the process
595 log which is later printed by exiwhat. It used to use the general
596 purpose logging code to do this, but several functions it calls are
597 not safe for signals.
599 The new output code in the SIGUSR1 handler is specific to the process
600 log, and simple enough that it's easy to inspect for signal safety.
601 Removing some special cases also simplifies the general logging code.
602 Removing the spurious timestamps from the process log simplifies
605 TF/04 Improved ratelimit ACL condition.
607 The /noupdate option has been deprecated in favour of /readonly which
608 has clearer semantics. The /leaky, /strict, and /readonly update modes
609 are mutually exclusive. The update mode is no longer included in the
610 database key; it just determines when the database is updated. (This
611 means that when you upgrde Exim will forget old rate measurements.)
613 Exim now checks that the per_* options are used with an update mode that
614 makes sense for the current ACL. For example, when Exim is processing a
615 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
616 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
617 must specify per_mail/readonly. If you omit the update mode it defaults to
618 /leaky where that makes sense (as before) or /readonly where required.
620 The /noupdate option is now undocumented but still supported for
621 backwards compatibility. It is equivalent to /readonly except that in
622 ACLs where /readonly is required you may specify /leaky/noupdate or
623 /strict/noupdate which are treated the same as /readonly.
625 A useful new feature is the /count= option. This is a generalization
626 of the per_byte option, so that you can measure the throughput of other
627 aggregate values. For example, the per_byte option is now equivalent
628 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
630 The per_rcpt option has been generalized using the /count= mechanism
631 (though it's more complicated than the per_byte equivalence). When it is
632 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
633 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
634 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
635 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
636 Note that using per_rcpt with a non-readonly update mode in more than
637 one ACL will cause the recipients to be double-counted. (The per_mail
638 and per_byte options don't have this problem.)
640 The handling of very low rates has changed slightly. If the computed rate
641 is less than the event's count (usually one) then this event is the first
642 after a long gap. In this case the rate is set to the same as this event's
643 count, so that the first message of a spam run is counted properly.
645 The major new feature is a mechanism for counting the rate of unique
646 events. The new per_addr option counts the number of different
647 recipients that someone has sent messages to in the last time period. It
648 behaves like per_rcpt if all the recipient addresses are different, but
649 duplicate recipient addresses do not increase the measured rate. Like
650 the /count= option this is a general mechanism, so the per_addr option
651 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
652 example, measure the rate that a client uses different sender addresses
653 with the options per_mail/unique=$sender_address. There are further
654 details in the main documentation.
656 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
658 TF/06 Removed a few PCRE remnants.
660 TF/07 Automatically extract Exim's version number from tags in the git
661 repository when doing development or release builds.
663 PP/02 Raise smtp_cmd_buffer_size to 16kB.
664 Bugzilla 879. Patch from Paul Fisher.
666 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
667 Heavily based on revision 40f9a89a from Simon Arlott's tree.
670 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
672 PP/05 Variable $av_failed, true if the AV scanner deferred.
673 Bugzilla 1078. Patch from John Horne.
675 PP/06 Stop make process more reliably on build failure.
676 Bugzilla 1087. Patch from Heiko Schlittermann.
678 PP/07 Make maildir_use_size_file an _expandable_ boolean.
679 Bugzilla 1089. Patch from Heiko Schlittermann.
681 PP/08 Handle ${run} returning more data than OS pipe buffer size.
682 Bugzilla 1131. Patch from Holger Weiß.
684 PP/09 Handle IPv6 addresses with SPF.
685 Bugzilla 860. Patch from Wolfgang Breyha.
687 PP/10 GnuTLS: support TLS 1.2 & 1.1.
689 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
692 PP/11 match_* no longer expand right-hand-side by default.
693 New compile-time build option, EXPAND_LISTMATCH_RHS.
694 New expansion conditions, "inlist", "inlisti".
696 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
698 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
700 PP/14 fix log_write() format string regression from TF/03.
701 Bugzilla 1152. Patch from Dmitry Isaikin.
707 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
709 PP/02 Harmonised TLS library version reporting; only show if debugging.
710 Layout now matches that introduced for other libraries in 4.74 PP/03.
712 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
714 PP/04 New "dns_use_edns0" global option.
716 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
719 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
720 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
722 TK/01 Updated PolarSSL code to 0.14.2.
723 Bugzilla 1097. Patch from Andreas Metzler.
725 PP/07 Catch divide-by-zero in ${eval:...}.
728 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
731 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
732 format-string attack -- SECURITY: remote arbitrary code execution.
734 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
735 time unintentionally subject to list matching rules, letting the header
736 cause arbitrary Exim lookups (of items which can occur in lists, *not*
737 arbitrary string expansion). This allowed for information disclosure.
739 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
740 INT_MIN/-1 -- value coerced to INT_MAX.
746 NM/01 Workround for PCRE version dependency in version reporting
749 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
750 This fixes portability to compilers other than gcc, notably
751 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
753 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
754 makefiles for portability to HP-UX and POSIX correctness.
756 PP/01 Permit LOOKUP_foo enabling on the make command-line.
757 Also via indented variable definition in the Makefile.
758 (Debugging by Oliver Heesakkers).
760 PP/02 Restore caching of spamd results with expanded spamd_address.
761 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
763 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
764 Improves build reliability. Fix from: Frank Elsner
766 NM/02 Fix wide character breakage in the rfc2047 coding
767 Fixes bug 1064. Patch from Andrey N. Oktyabrski
769 NM/03 Allow underscore in dnslist lookups
770 Fixes bug 1026. Patch from Graeme Fowler
772 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
773 Code patches from Adam Ciarcinski of NetBSD.
775 NM/04 Fixed exiqgrep to cope with mailq missing size issue
778 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
779 is logged, to avoid truncation. Patch from John Horne.
781 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
782 Patch from Jakob Hirsch.
784 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
785 SQL string expansion failure details.
786 Patch from Andrey Oktyabrski.
788 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
789 Patch from Simon Arlott.
791 PP/09 New lookups functionality failed to compile on old gcc which rejects
792 extern declarations in function scope.
793 Patch from Oliver Fleischmann
795 PP/10 Use sig_atomic_t for flags set from signal handlers.
796 Check getgroups() return and improve debugging.
797 Fixed developed for diagnosis in bug 927 (which turned out to be
800 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
801 Patch from Mark Zealey.
803 PP/12 Bugzilla 1056: Improved spamd server selection.
804 Patch from Mark Zealey.
806 PP/13 Bugzilla 1086: Deal with maildir quota file races.
807 Based on patch from Heiko Schlittermann.
809 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
810 Patch from Uwe Doering, sign-off by Michael Haardt.
812 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
813 variable declaration deep within a block. Bug and patch from
816 PP/15 lookups-Makefile IRIX compatibilty coercion.
818 PP/16 Make DISABLE_DKIM build knob functional.
820 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
821 Patch by Simon Arlott
823 TF/03 Fix valgrind.h portability to C89 compilers that do not support
824 variable argument macros. Our copy now differs from upstream.
830 TF/01 Failure to get a lock on a hints database can have serious
831 consequences so log it to the panic log.
833 TF/02 Log LMTP confirmation messages in the same way as SMTP,
834 controlled using the smtp_confirmation log selector.
836 TF/03 Include the error message when we fail to unlink a spool file.
838 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
839 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
840 for maintaining out-of-tree patches for some time.
842 PP/01 Bugzilla 139: Documentation and portability issues.
843 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
844 Handle per-OS dynamic-module compilation flags.
846 PP/02 Let /dev/null have normal permissions.
847 The 4.73 fixes were a little too stringent and complained about the
848 permissions on /dev/null. Exempt it from some checks.
849 Reported by Andreas M. Kirchwitz.
851 PP/03 Report version information for many libraries, including
852 Exim version information for dynamically loaded libraries. Created
853 version.h, now support a version extension string for distributors
854 who patch heavily. Dynamic module ABI change.
856 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
857 privilege escalation vulnerability whereby the Exim run-time user
858 can cause root to append content of the attacker's choosing to
861 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
864 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
865 If dropping privileges for untrusted macros, we disabled normal logging
866 on the basis that it would fail; for the Exim run-time user, this is not
867 the case, and it resulted in successful deliveries going unlogged.
868 Fixed. Reported by Andreas Metzler.
874 PP/01 Date: & Message-Id: revert to normally being appended to a message,
875 only prepend for the Resent-* case. Fixes regression introduced in
876 Exim 4.70 by NM/22 for Bugzilla 607.
878 PP/02 Include check_rfc2047_length in configure.default because we're seeing
879 increasing numbers of administrators be bitten by this.
881 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
883 PP/03 Bugzilla 994: added openssl_options main configuration option.
885 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
887 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
889 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
891 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
892 without a peer certificate, leading to a segfault because of an
893 assumption that peers always have certificates. Be a little more
894 paranoid. Problem reported by Martin Tscholak.
896 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
897 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
898 NB: ClamAV planning to remove STREAM in "middle of 2010".
899 CL also introduces -bmalware, various -d+acl logging additions and
900 more caution in buffer sizes.
902 PP/09 Implemented reverse_ip expansion operator.
904 PP/10 Bugzilla 937: provide a "debug" ACL control.
906 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
908 PP/12 Bugzilla 973: Implement --version.
910 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
912 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
914 PP/15 Bugzilla 816: support multiple condition rules on Routers.
916 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
917 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
918 ignore trailing whitespace.
920 JJ/02 prevent non-panic DKIM error from being sent to paniclog
922 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
925 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
926 Notification from Dr Andrew Aitchison.
928 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
929 ExtendedDetectionInfo response format.
930 Notification from John Horne.
932 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
935 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
936 XSL and documented dependency on system catalogs, with examples of how
939 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
942 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
943 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
944 configuration file which is writeable by the Exim user or group.
946 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
947 of configuration files to cover files specified with the -C option if
948 they are going to be used with root privileges, not just the default
951 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
952 option (effectively making it always true).
954 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
955 files to be used while preserving root privileges.
957 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
958 that rogue child processes cannot use them.
960 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
961 run-time user, instead of root.
963 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
964 Exim run-time user without dropping privileges.
966 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
967 result string, instead of calling string_vformat() twice with the same
970 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
971 for other users. Others should always drop root privileges if they use
972 -C on the command line, even for a whitelisted configure file.
974 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
976 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
982 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
983 $data_path, and $header_path variables; fixed documentation bugs and
986 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
987 exipick to access non-standard spools, including the "frozen" queue
990 NM/01 Bugzilla 965: Support mysql stored procedures.
991 Patch from Alain Williams
993 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
995 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
996 Patch from Andreas Metzler
998 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
999 Patch from Kirill Miazine
1001 NM/05 Bugzilla 671: Added umask to procmail example.
1003 JJ/03 installed exipick 20100323.0, fixing doc bug
1005 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1006 directory. Notification and patch from Dan Rosenberg.
1008 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1010 TK/02 Improve log output when DKIM signing operation fails.
1012 MH/01 Treat the transport option dkim_domain as a colon separated
1013 list, not as a single string, and sign the message with each element,
1014 omitting multiple occurences of the same signer.
1016 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1017 Bugzilla 985, 986. Patch by Simon Arlott
1019 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1020 Patch by Simon Arlott
1022 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1023 MBX locking. Notification from Dan Rosenberg.
1029 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1031 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1033 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1035 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1037 NM/04 Bugzilla 915: Flag broken perl installation during build.
1043 TK/01 Added patch by Johannes Berg that expands the main option
1044 "spamd_address" if it starts with a dollar sign.
1046 TK/02 Write list of recipients to X-Envelope-Sender header when building
1047 the mbox-format spool file for content scanning (suggested by Jakob
1050 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1051 (http://www.dcc-servers.net/) support via dccifd. Activated by
1052 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1054 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1055 by Mark Daniel Reidel <mr@df.eu>.
1057 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1058 When building exim an external PCRE library is now needed -
1059 PCRE is a system library on the majority of modern systems.
1060 See entry on PCRE_LIBS in EDITME file.
1062 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1063 conversation. Added nologin parameter to request.
1064 Patch contributed by Kirill Miazine.
1066 TF/01 Do not log submission mode rewrites if they do not change the address.
1068 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1070 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1071 log files in place. Contributed by Roberto Lima.
1073 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1075 TF/03 Bugzilla 615: When checking the local_parts router precondition
1076 after a local_part_suffix or local_part_prefix option, Exim now
1077 does not use the address's named list lookup cache, since this
1078 contains cached lookups for the whole local part.
1080 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1081 Robert Millan. Documentation is in experimental-spec.txt.
1083 TF/04 Bugzilla 668: Fix parallel build (make -j).
1085 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1087 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1088 Patch provided by Jan Srzednicki.
1090 TF/05 Leading white space used to be stripped from $spam_report which
1091 wrecked the formatting. Now it is preserved.
1093 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1094 that they are available at delivery time.
1096 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1098 TF/08 TLS error reporting now respects the incoming_interface and
1099 incoming_port log selectors.
1101 TF/09 Produce a more useful error message if an SMTP transport's hosts
1102 setting expands to an empty string.
1104 NM/06 Bugzilla 744: EXPN did not work under TLS.
1105 Patch provided by Phil Pennock.
1107 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1108 Patch provided by Richard Godbee.
1110 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1111 acl_smtp_notquit, added index entry.
1113 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1114 Patch provided by Eugene Bujak.
1116 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1117 accept(). Patch provided by Maxim Dounin.
1119 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1120 Patch provided by Phil Pennock.
1122 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1124 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1125 Patch provided by Brad "anomie" Jorsch.
1127 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1128 Patch provided by Dean Brooks.
1130 TK/05 Add native DKIM support (does not depend on external libraries).
1132 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1133 Patch provided by Graeme Fowler.
1135 NM/16 Bugzilla 851: Documentation example syntax fix.
1137 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1139 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1142 NM/19 Bugzilla 745: TLS version reporting.
1143 Patch provided by Phil Pennock.
1145 NM/20 Bugzilla 167: bool: condition support.
1146 Patch provided by Phil Pennock.
1148 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1149 clients. Patch provided by Phil Pennock.
1151 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1152 Patch provided by Brad "anomie" Jorsch.
1154 NM/23 Bugzilla 687: Fix misparses in eximstats.
1155 Patch provided by Heiko Schlittermann.
1157 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1158 Patch provided by Heiko Schlittermann.
1160 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1161 plus update to original patch.
1163 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1165 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1166 Patch provided by David Brownlee.
1168 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1170 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1172 NM/30 Bugzilla 888: TLS documentation bugfixes.
1174 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1176 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1177 Unlike the original bugzilla I have changed all shell scripts in src tree.
1179 NM/33 Bugzilla 898: Transport filter timeout fix.
1180 Patch by Todd Rinaldo.
1182 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1183 Patch by Serge Demonchaux.
1185 NM/35 Bugzilla 39: Base64 decode bug fixes.
1186 Patch by Jakob Hirsch.
1188 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1190 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1192 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1194 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1200 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1201 ALT-N's libdkim that I have put here:
1202 http://duncanthrax.net/exim-experimental/
1204 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1205 were called 'true' and it seems that C99 defines that as a reserved
1206 keyword to be used with 'bool' variable types. That means you could
1207 not include C99-style headers which use bools without triggering
1208 build errors in sieve.c.
1210 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1211 as mailq or other aliases. Changed the --help handling significantly
1212 to do whats expected. exim_usage() emits usage/help information.
1214 SC/01 Added the -bylocaldomain option to eximstats.
1216 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1218 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1220 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1226 PH/01 Another patch from the Sieve maintainer.
1228 PH/02 When an IPv6 address is converted to a string for single-key lookup
1229 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1230 dots are used instead of colons so that keys in lsearch files need not
1231 contain colons. This was done some time before quoting was made available
1232 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1233 (notated using the quote facility) so as to distinguish them from IPv4
1234 keys. This meant that lookups for IP addresses in host lists did not work
1235 for iplsearch lookups.
1237 This has been fixed by arranging for IPv6 addresses to be expressed with
1238 colons if the lookup type is iplsearch. This is not incompatible, because
1239 previously such lookups could never work.
1241 The situation is now rather anomolous, since one *can* have colons in
1242 ordinary lsearch keys. However, making the change in all cases is
1243 incompatible and would probably break a number of configurations.
1245 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1248 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1249 conversion specification without a maximum field width, thereby enabling
1250 a rogue spamd server to cause a buffer overflow. While nobody in their
1251 right mind would setup Exim to query an untrusted spamd server, an
1252 attacker that gains access to a server running spamd could potentially
1253 exploit this vulnerability to run arbitrary code as the Exim user.
1255 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1256 $primary_hostname instead of what libspf2 thinks the hosts name is.
1258 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1259 a directory entry by the name of the lookup key. Previously, if a
1260 symlink pointed to a non-existing file or a file in a directory that
1261 Exim lacked permissions to read, a lookup for a key matching that
1262 symlink would fail. Now it is enough that a matching directory entry
1263 exists, symlink or not. (Bugzilla 503.)
1265 PH/03 The body_linecount and body_zerocount variables are now exported in the
1268 PH/04 Added the $dnslist_matched variable.
1270 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1271 This means they are set thereafter only if the connection becomes
1274 PH/06 Added the client_condition to authenticators so that some can be skipped
1275 by clients under certain conditions.
1277 PH/07 The error message for a badly-placed control=no_multiline_responses left
1278 "_responses" off the end of the name.
1280 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1282 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1283 (without spaces) instead of just copying the configuration text.
1285 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1287 PH/11 Added $max_received_linelength.
1289 PH/12 Added +ignore_defer and +include_defer to host lists.
1291 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1292 way in which PCRE > 7.0 is built.
1294 PH/14 Implemented queue_only_load_latch.
1296 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1297 MAIL command. The effect was to mangle the value on 64-bit systems.
1299 PH/16 Another patch from the Sieve maintainer.
1301 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1303 PH/18 If a system quota error occurred while trying to create the file for
1304 a maildir delivery, the message "Mailbox is full" was not appended to the
1305 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1306 only to a quota excession during the actual writing of the file.
1308 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1309 characters?) which causes problems in log lines. The DN values are now
1310 passed through string_printing() before being added to log lines.
1312 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1313 and InterBase are left for another time.)
1315 PH/21 Added message_body_newlines option.
1317 PH/22 Guard against possible overflow in moan_check_errorcopy().
1319 PH/23 POSIX allows open() to be a macro; guard against that.
1321 PH/24 If the recipient of an error message contained an @ in the local part
1322 (suitably quoted, of course), incorrect values were put in $domain and
1323 $local_part during the evaluation of errors_copy.
1329 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1330 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1333 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1334 issue a MAIL command.
1336 PH/02 In an ACL statement such as
1338 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1340 if a client was not listed at all, or was listed with a value other than
1341 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1342 the condition was not true (as it should be), so access was not denied.
1343 The bug was that the ! inversion was incorrectly passed on to the second
1344 item. This has been fixed.
1346 PH/03 Added additional dnslists conditions == and =& which are different from
1347 = and & when the dns lookup returns more than one IP address.
1349 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1350 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1352 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1353 FSYNC, which compiles an option called disable_fsync that allows for
1354 bypassing fsync(). The documentation is heavily laced with warnings.
1356 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1358 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1359 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1360 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1361 including adding "make clean"; (3) Added -fPIC when compiling the test
1362 dynamically loaded module, to get rid of a warning.
1364 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1365 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1366 = 0s. The bug is otherwise harmless.
1368 PH/07 There was a bug in the dovecot authenticator such that the value of
1369 $auth1 could be overwritten, and so not correctly preserved, after a
1370 successful authentication. This usually meant that the value preserved by
1371 the server_setid option was incorrect.
1373 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1375 PH/09 Installed PCRE release 7.0.
1377 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1378 run for batched SMTP input. It is now run at the start of every message
1379 in the batch. While fixing this I discovered that the process information
1380 (output by running exiwhat) was not always getting set for -bs and -bS
1381 input. This is fixed, and it now also says "batched" for BSMTP.
1383 PH/11 Added control=no_pipelining.
1385 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1386 patch, slightly modified), and move the expansion of helo_data till after
1387 the connection is made in the smtp transport (so it can use these
1390 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1392 PH/14 Added log_selector = +pid.
1394 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1396 PH/16 Add ${if forany and ${if forall.
1398 PH/17 Added dsn_from option to vary the From: line in DSNs.
1400 PH/18 Flush SMTP output before performing a callout, unless control =
1401 no_callout_flush is set.
1403 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1404 was true (the default) a successful delivery failed to delete the retry
1405 item, thus causing premature timeout of the address. The bug is now
1408 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1410 PH/21 Long custom messages for fakedefer and fakereject are now split up
1411 into multiline reponses in the same way that messages for "deny" and
1412 other ACL rejections are.
1414 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1415 with slight modification.
1417 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1418 draft, changing the syntax and factoring some duplicate code".
1420 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1421 for deliveries of the second and subsequent messages over the same SMTP
1424 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1425 ${reduce, with only minor "tidies".
1427 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1429 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1430 expansion side effects.
1432 PH/27 When a message times out after an over-quota error from an Exim-imposed
1433 quota, the bounce message says "mailbox is full". This message was not
1434 being given when it was a system quota that was exceeded. It now should
1437 MH/03 Made $recipients available in local_scan(). local_scan() already has
1438 better access to the recipient list through recipients_list[], but
1439 $recipients can be useful in postmaster-provided expansion strings.
1441 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1442 in the case of a MAIL command with additional options following the
1443 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1444 were accidentally chopped off.
1446 PH/29 SMTP synchronization checks are implemented when a command is read -
1447 there is a check that no more input is waiting when there shouldn't be
1448 any. However, for some commands, a delay in an ACL can mean that it is
1449 some time before the response is written. In this time, more input might
1450 arrive, invalidly. So now there are extra checks after an ACL has run for
1451 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1452 pipelining has not been advertised.
1454 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1456 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1457 returned in the SMTP response when smtp_return_error_details was false.
1458 This has been fixed.
1460 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1461 instead of the C I/O that was originally supplied, because problems were
1462 reported on Solaris.
1464 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1465 Exim which did not show up earlier: it was assuming that a call to
1466 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1467 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1468 was a macro that became an assignment, so it seemed to work. This has
1469 changed to a proper function call with a void return, hence the compile
1470 error. Exim's code has been fixed.
1472 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1475 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1477 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1479 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1480 because it turns out to be too big or there is a timeout) there is no
1481 "Completed" line in the log. When some messages of this type were
1482 selected by exigrep, they were listed as "not completed". Others were
1483 picked up by some special patterns. I have improved the selection
1484 criteria to be more general.
1486 PH/38 The host_find_failed option in the manualroute router can now be set
1487 to "ignore", to completely ignore a host whose IP address cannot be
1488 found. If all hosts are ignored, the behaviour is controlled by the new
1489 host_all_ignored option.
1491 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1492 homing or because of multiple MX records with /mx) generated more than
1493 one IP address, and the following item turned out to be the local host,
1494 all the secondary addresses of the first item were incorrectly removed
1495 from the list, along with the local host and any following hosts (which
1496 is what is supposed to happen).
1498 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1499 whoever called Exim into the -H file. In the case of the daemon it was
1500 behaving confusingly. When first started, it used values for whoever
1501 started the daemon, but after a SIGHUP it used the Exim user (because it
1502 calls itself on a restart). I have changed the code so that it now always
1505 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1506 message are rejected with the same error (e.g. no authentication or bad
1507 sender address), and a DATA command is nevertheless sent (as can happen
1508 with PIPELINING or a stupid MUA), the error message that was given to the
1509 RCPT commands is included in the rejection of the DATA command. This is
1510 intended to be helpful for MUAs that show only the final error to their
1513 PH/42 Another patch from the Sieve maintainer.
1515 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1516 Eximstats - Fixed some broken HTML links and added missing column headers
1518 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1519 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1521 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1523 PH/43 Yet another patch from the Sieve maintainer.
1525 PH/44 I found a way to check for a TCP/IP connection going away before sending
1526 the response to the final '.' that terminates a message, but only in the
1527 case where the client has not sent further data following the '.'
1528 (unfortunately, this is allowed). However, in many cases there won't be
1529 any further data because there won't be any more messages to send. A call
1530 to select() can be used: if it shows that the input is "ready", there is
1531 either input waiting, or the socket has been closed. An attempt to read
1532 the next input character can distinguish the two cases. Previously, Exim
1533 would have sent an OK response which the client would never have see.
1534 This could lead to message repetition. This fix should cure that, at
1535 least in a lot of common cases.
1537 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1538 advertised in response to EHLO.
1544 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1545 fixed by 4.65/MH/01 (is this a record?) are fixed:
1547 (i) An empty string was always treated as zero by the numeric comparison
1548 operators. This behaviour has been restored.
1550 (ii) It is documented that the numeric comparison operators always treat
1551 their arguments as decimal numbers. This was broken in that numbers
1552 starting with 0 were being interpreted as octal.
1554 While fixing these problems I realized that there was another issue that
1555 hadn't been noticed. Values of message_size_limit (both the global option
1556 and the transport option) were treated as octal if they started with 0.
1557 The documentation was vague. These values are now always treated as
1558 decimal, and I will make that clear in the documentation.
1564 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1565 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1568 MH/01 Don't check that the operands of numeric comparison operators are
1569 integers when their expansion is in "skipping" mode (fixes bug
1570 introduced by 4.64-PH/07).
1572 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1573 child addresses, Exim now panics and dies. Previously, because the count
1574 is held in a short int, deliveries were likely to be lost. As such a
1575 large number of recipients for a single message is ridiculous
1576 (performance will be very, very poor), I have chosen to impose a limit
1577 rather than extend the field.
1583 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1584 leftover -K file (the existence of which was triggered by #402).
1585 While we were at it, introduced process PID as part of the -K
1586 filename. This should rule out race conditions when creating
1589 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1590 processing considerably. Previous code took too long for large mails,
1591 triggering a timeout which in turn triggers #401.
1593 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1594 in the DK code in transports.c. sendfile() is not really portable,
1595 hence the _LINUX specificness.
1597 TF/01 In the add_headers option to the mail command in an Exim filter,
1598 there was a bug that Exim would claim a syntax error in any
1599 header after the first one which had an odd number of characters
1602 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1603 callout verification, Exim cached a "reject" for the entire domain. This
1604 is correct for most verifications, but it is not correct for a recipient
1605 verification with use_sender or use_postmaster set, because in that case
1606 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1607 case of MAIL FROM:<> rejection from other early rejections (e.g.
1608 rejection of HELO). When verifying a recipient using a non-null MAIL
1609 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1610 Whatever the result of the callout, the value of the domain cache is
1611 left unchanged (for any other kind of callout, getting as far as trying
1612 RCPT means that the domain itself is ok).
1614 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1617 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1618 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1619 session, the daemon ignores ECONNECT errors and logs others; it now
1620 ignores EPIPE as well.
1622 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1623 (quoted-printable decoding).
1625 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1626 later the small subsequent patch to fix an introduced bug.
1628 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1630 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1632 PH/08 An error is now given if message_size_limit is specified negative.
1634 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1635 to be given (somewhat) arbitrary names.
1637 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1640 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1641 miscellaneous code fixes
1643 PH/10 Added the log_reject_target ACL modifier to specify where to log
1646 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1647 hostname. This is wrong, because it relates to the incoming message (and
1648 probably the interface on which it is arriving) and not to the outgoing
1649 callout (which could be using a different interface). This has been
1650 changed to use the value of the helo_data option from the smtp transport
1651 instead - this is what is used when a message is actually being sent. If
1652 there is no remote transport (possible with a router that sets up host
1653 addresses), $smtp_active_hostname is used.
1655 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1656 tweaks were necessary in order to get it to work (see also 21 below):
1657 (a) The code assumed that strncpy() returns a negative number on buffer
1658 overflow, which isn't the case. Replaced with Exim's string_format()
1660 (b) There were several signed/unsigned issues. I just did the minimum
1661 hacking in of casts. There is scope for a larger refactoring.
1662 (c) The code used strcasecmp() which is not a standard C function.
1663 Replaced with Exim's strcmpic() function.
1664 (d) The code set only $1; it now sets $auth1 as well.
1665 (e) A simple test gave the error "authentication client didn't specify
1666 service in request". It would seem that Dovecot has changed its
1667 interface. Fortunately there's a specification; I followed it and
1668 changed what the client sends and it appears to be working now.
1670 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1673 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1674 address A is aliased to B and C, where B exists and C does not. Without
1675 -v the output is "A verified" because verification stops after a
1676 successful redirection if more than one address is generated. However,
1677 with -v the child addresses are also verified. Exim was outputting "A
1678 failed to verify" and then showing the successful verification for C,
1679 with its parentage. It now outputs "B failed to verify", showing B's
1680 parentage before showing the successful verification of C.
1682 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1683 look up a TXT record in a specific list after matching in a combined
1686 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1687 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1688 they consult the DNS. I had assumed they would set it the way they
1689 wanted; and indeed my experiments on Linux seem to show that in some
1690 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1691 To be on the safe side, however, I have now made the interface to
1692 host_find_byname() similar to host_find_bydns(), with an argument
1693 containing the DNS resolver options. The host_find_byname() function now
1694 sets these options at its start, just as host_find_bydns() does. The smtp
1695 transport options dns_qualify_single and dns_search_parents are passed to
1696 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1697 of host_find_byname() use the default settings of RES_DEFNAMES
1698 (qualify_single) but not RES_DNSRCH (search_parents).
1700 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1701 spool_read_header() do less string testing, by means of a preliminary
1702 switch on the second character of optional "-foo" lines. (This is
1703 overdue, caused by the large number of possibilities that now exist.
1704 Originally there were few.) While I was there, I also converted the
1705 str(n)cmp tests so they don't re-test the leading "-" and the first
1706 character, in the hope this might squeeze out yet more improvement.
1708 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1709 flag allowing group syntax was set by the header_syntax check but not
1710 turned off, possible causing trouble later; (2) The flag was not being
1711 set at all for the header_verify test, causing "group"-style headers to
1712 be rejected. I have now set it in this case, and also caused header_
1713 verify to ignore an empty address taken from a group. While doing this, I
1714 came across some other cases where the code for allowing group syntax
1715 while scanning a header line wasn't quite right (mostly, not resetting
1716 the flag correctly in the right place). These bugs could have caused
1717 trouble for malformed header lines. I hope it is now all correct.
1719 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1720 with the "reply" argument non-NULL. The code, however (which originally
1721 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1722 but it didn't always do it. This confused somebody who was copying the
1723 code for some other use. I have removed all the tests.
1725 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1726 feature that was used to support insecure browsers during the U.S. crypto
1727 embargo. It requires special client support, and Exim is probably the
1728 only MTA that supported it -- and would never use it because real RSA is
1729 always available. This code has been removed, because it had the bad
1730 effect of slowing Exim down by computing (never used) parameters for the
1731 RSA_EXPORT functionality.
1733 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1734 authenticator to fail if there's a tab character in the incoming data
1735 (there should never be unless someone is messing about, as it's supposed
1736 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1737 if the connection is using TLS or if the remote IP is the same as the
1738 local IP, and the "valid-client-cert option" if a client certificate has
1741 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1742 authenticators. This can be used for authorization after authentication
1743 succeeds. (In the case of plaintext, it servers for both authentication
1746 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1747 if any retry times were supplied.
1749 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1750 connection, where there is no client IP address to check. In this
1751 situation, the verify now always succeeds.
1753 PH/25 Applied John Jetmore's -Mset patch.
1755 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1757 PH/27 In a string expansion for a processed (not raw) header when multiple
1758 headers of the same name were present, leading whitespace was being
1759 removed from all of them, but trailing whitespace was being removed only
1760 from the last one. Now trailing whitespace is removed from each header
1761 before concatenation. Completely empty headers in a concatenation (as
1762 before) are ignored.
1764 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1765 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1767 PH/29 [Removed. This was a change that I later backed out, and forgot to
1768 correct the ChangeLog entry (that I had efficiently created) before
1769 committing the later change.]
1771 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1772 address errors (4xx response to RCPT) over the same connection as other
1773 messages routed to the same hosts. Such deliveries are always "forced",
1774 so retry times are not inspected. This resulted in far too many retries
1775 for the affected addresses. The effect occurred only when there were more
1776 hosts than the hosts_max_try setting in the smtp transport when it had
1777 the 4xx errors. Those hosts that it had tried were not added to the list
1778 of hosts for which the message was waiting, so if all were tried, there
1779 was no problem. Two fixes have been applied:
1781 (i) If there are any address or message errors in an SMTP delivery, none
1782 of the hosts (tried or untried) are now added to the list of hosts
1783 for which the message is waiting, so the message should not be a
1784 candidate for sending over the same connection that was used for a
1785 successful delivery of some other message. This seems entirely
1786 reasonable: after all the message is NOT "waiting for some host".
1787 This is so "obvious" that I'm not sure why it wasn't done
1788 previously. Hope I haven't missed anything, but it can't do any
1789 harm, as the worst effect is to miss an optimization.
1791 (ii) If, despite (i), such a delivery is accidentally attempted, the
1792 routing retry time is respected, so at least it doesn't keep
1793 hammering the server.
1795 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1796 in ${readsocket because some servers need this prod.
1798 PH/32 Added some extra debug output when updating a wait-xxx database.
1800 PH/33 The hint "could be header name not terminated by colon", which has been
1801 given for certain expansion errors for a long time, was not being given
1802 for the ${if def:h_colon_omitted{... case.
1804 PH/34 The spec says: "With one important exception, whenever a domain list is
1805 being scanned, $domain contains the subject domain." There was at least
1806 one case where this was not true.
1808 PH/35 The error "getsockname() failed: connection reset by peer" was being
1809 written to the panic log as well as the main log, but it isn't really
1810 panic-worthy as it just means the connection died rather early on. I have
1811 removed the panic log writing for the ECONNRESET error when getsockname()
1814 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1815 runs only) independently of the message's sender address. This meant
1816 that, if the 4xx error was in fact related to the sender, a different
1817 message to the same recipient with a different sender could confuse
1818 things. In particualar, this can happen when sending to a greylisting
1819 server, but other circumstances could also provoke similar problems.
1820 I have changed the default so that the retry time for these errors is now
1821 based a combination of the sender and recipient addresses. This change
1822 can be overridden by setting address_retry_include_sender=false in the
1825 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1826 remote server are returned as part of bounce messages. This was not
1827 happening for LMTP over a pipe (the lmtp transport), but now it is the
1828 same for both kinds of LMTP.
1830 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1831 in header lines that were in fact CNAMEs. This is no longer the case.
1833 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1834 and queue runs started by the daemon processed all messages. This has
1835 been fixed so that -R and -S can now usefully be given with -q<time>.
1837 PH/40 Import PCRE release 6.7 (fixes some bugs).
1839 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1841 PH/42 Give an error if -q is specified more than once.
1843 PH/43 Renamed the variables $interface_address and $interface_port as
1844 $received_ip_address and $received_port, to make it clear that these
1845 values apply to message reception, and not to the outgoing interface when
1846 a message is delivered. (The old names remain recognized, of course.)
1848 PH/44 There was no timeout on the connect() call when using a Unix domain
1849 socket in the ${readsocket expansion. There now is.
1851 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1852 be meaningful with "accept".
1854 SC/01 Eximstats V1.43
1855 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1857 SC/02 Eximstats V1.44
1858 Use a glob alias rather than an array ref in the generated
1859 parser. This improves both readability and performance.
1861 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1862 Collect SpamAssassin and rejection statistics.
1863 Don't display local sender or destination tables unless
1864 there is data to show.
1865 Added average volumes into the top table text output.
1867 SC/04 Eximstats V1.46
1868 Collect data on the number of addresses (recipients)
1869 as well as the number of messages.
1871 SC/05 Eximstats V1.47
1872 Added 'Message too big' to the list of mail rejection
1873 reasons (thanks to Marco Gaiarin).
1875 SC/06 Eximstats V1.48
1876 Mainlog lines which have GMT offsets and are too short to
1877 have a flag are now skipped.
1879 SC/07 Eximstats V1.49 (Alain Williams)
1880 Added the -emptyok flag.
1882 SC/08 Eximstats V1.50
1883 Fixes for obtaining the IP address from reject messages.
1885 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1886 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1887 whitesspace changes from 4.64-PH/27
1889 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1892 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1893 are found, allow negative numbers in numeric criteria)
1895 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1897 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1898 to match changes made in 4.64-PH/43
1900 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1902 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1903 "domain" is actually an IP address, and force a failure. This locks out
1904 those revolvers/nameservers that support "A-for-A" lookups, in
1905 contravention of the specifications.
1907 PH/48 When a host name was looked up from an IP address, and the subsequent
1908 forward lookup of the name timed out, the host name was left in
1909 $sender_host_name, contrary to the specification.
1911 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1912 restricted to single-key lookups, Exim was not diagnosing an error if
1913 * or *@ was used with a query-style lookup.
1915 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1917 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1918 long ago, but noone interested enough thought of it. Let's just say that
1919 the "1.1" means that there are some new functions that weren't there at
1920 some point in the past.
1922 PH/51 Error processing for expansion failure of helo_data from an smtp
1923 transport during callout processing was broken.
1925 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1926 tested/used via the -bh/-bhc/-bs options.
1928 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1929 bug, fixed in subsequent PCRE releases).
1931 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1932 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1934 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1940 SC/01 Use a glob alias rather than an array ref in eximstats generated
1941 parser. This improves both readability and performance.
1943 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1944 Don't display local sender or destination tables in eximstats unless
1945 there is data to show.
1946 Added average volumes into the eximstats top table text output.
1948 SC/03 Collect data on the number of addresses (recipients) as well
1949 as the number of messages in eximstats.
1951 TF/01 Correct an error in the documentation for the redirect router. Exim
1952 does (usually) call initgroups() when daemonizing.
1954 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1955 with consistent privilege compared to when running as a daemon.
1957 TF/03 Note in the spec that $authenticated_id is not set for local
1958 submissions from trusted users.
1960 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1961 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1963 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1964 by adding some example configuration directives to the default
1965 configuration file. A little bit of work is required to uncomment the
1966 directives and define how usernames and passwords are checked, but
1967 there is now a framework to start from.
1969 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1970 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1971 without this. I don't know how relevant this is to other LDAP libraries.
1973 PH/02 Add the verb name to the "unknown ACL verb" error.
1975 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1977 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1979 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1980 directory not expanded when it should be if an expanded home directory
1981 was set for the address (which is overridden by the transport).
1983 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1986 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1987 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1988 because it is too late at that time, and has no effect.
1990 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1991 security issue with \' (bugzilla #107). I could not use the
1992 PQescapeStringConn() function, because it needs a PGconn value as one of
1995 PH/08 When testing addresses using -bt, indicate those final addresses that
1996 are duplicates that would not cause an additional delivery. At least one
1997 person was confused, thinking that -bt output corresponded to deliveries.
1998 (Suppressing duplicates isn't a good idea as you lose the information
1999 about possibly different redirections that led to the duplicates.)
2001 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2002 systems where poll() doesn't work, in particular OS X.
2004 PH/10 Added more information to debugging output for retry time not reached.
2006 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2007 operations in malware.c.
2009 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2012 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2013 syslog_duplication set false, log lines that would normally be written
2014 both the the main log and to the reject log were not written to syslog at
2017 PH/14 In the default configuration, change the use of "message" in ACL warn
2018 statements to "add_header".
2020 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2021 not followed by a command (e.g. "seen endif").
2023 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2024 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2027 PH/17 Added extra conditions to the default value of delay_warning_condition
2031 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2032 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2033 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2036 The Auto-Submitted: and various List- headers are standardised, whereas I
2037 don't think Precedence: ever was.
2039 PH/18 Refactored debugging code in route_finduser() to show more information,
2040 in particular, the error code if getpwnam() issues one.
2042 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2043 This is apparently needed in addition to the PH/07 change above to avoid
2044 any possible encoding problems.
2046 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2047 but not after initializing Perl.
2049 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2050 output them only if debugging. By default they are written stderr,
2051 apparently, which is not desirable.
2053 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2056 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2059 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2061 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2062 authenticated or an ident call has been made. Suppress the default
2063 values for $authenticated_id and $authenticated_sender (but permit -oMai
2064 and -oMas) when testing with -bh.
2066 PH/24 Re-jigged the order of the tests in the default configuration so that the
2067 tests for valid domains and recipients precede the DNS black list and CSA
2068 tests, on the grounds that those ones are more expensive.
2070 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2071 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2072 This bug exists in every version of Exim that I still have, right back to
2075 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2076 However, an attempt to turn on case-sensitivity in a regex key by
2077 including (?-i) didn't work because the subject string was already
2078 lowercased, and the effects were non-intuitive. It turns out that a
2079 one-line patch can be used to allow (?-i) to work as expected.
2085 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2086 other effects) broke the use of negated acl sub-conditions.
2088 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2091 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2092 "Deny" causes Exim to reject the incoming connection with a 554 error.
2093 Unfortunately, if there is a major crisis, such as a disk failure,
2094 tcp-wrappers gives "deny", whereas what one would like would be some
2095 kind of temporary error. A kludge has been added to help with this.
2096 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2097 554 error is used if errno is still zero or contains ENOENT (which occurs
2098 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2101 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2103 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2104 errors. Otherwise a message that provokes a temporary error (when other
2105 messages do not) can cause a whole host to time out.
2107 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2108 addresses were routed directly to files or pipes from a redirect router.
2109 File deliveries just didn't batch; pipe deliveries might have suffered
2112 PH/06 A failure to get a lock for a hints database would erroneously always say
2113 "Failed to get write lock", even when it was really a read lock.
2115 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2116 of 0600. This has been changed to use the value of the lockfile_mode
2117 option (which defaults to 0600).
2119 PH/08 Applied small patch from the Sieve maintainer.
2121 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2122 folder from quota calculations, a direct delivery into this folder messed
2123 up the contents of the maildirsize file. This was because the regex was
2124 used only to exclude .Trash (or whatever) when the size of the mailbox
2125 was calculated. There was no check that a delivery was happening into an
2126 excluded directory. This bug has been fixed by ignoring all quota
2127 processing for deliveries into excluded directories.
2129 PH/10 Added the maildirfolder_create_regex option to appendfile.
2135 PH/01 The code for finding all the local interface addresses on a FreeBSD
2136 system running IPv6 was broken. This may well have applied to all BSD
2137 systems, as well as to others that have similar system calls. The broken
2138 code found IPv4 interfaces correctly, but gave incorrect values for the
2139 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2140 that it would not match correctly against @[] and not recognize the IPv6
2143 PH/02 The ipliteral router was not recognizing addresses of the form user@
2144 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2146 PH/03 Added disable_ipv6.
2148 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2149 decoded form, because it is most often used to construct To: headers
2150 lines in autoreplies, and the decoded form may well be syntactically
2151 invalid. However, $reply_address has leading white space removed, and all
2152 newlines turned into spaces so that the autoreply transport does not
2155 PH/05 If group was specified without a user on a router, and no group or user
2156 was specified on a transport, the group from the router was ignored.
2158 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2159 for visible compile-time settings that can be used to change these
2160 numbers, for those that want even more. Backwards compatibility with old
2161 spool files has been maintained. However, going back to a previous Exim
2162 release will lost any variables that are in spool files.
2164 PH/07 Two small changes when running in the test harness: increase delay when
2165 passing a TCP/IP connection to a new process, in case the original
2166 process has to generate a bounce, and remove special handling of
2167 127.0.0.2 (sic), which is no longer necessary.
2169 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2170 be the same on different OS.
2172 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2175 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2176 whether --show-vars was specified or not
2178 JJ/02 exipick: Added support for new ACL variable spool format introduced
2181 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2182 syntactically invalid From: or Reply-to: line, and a filter used this to
2183 generate an autoreply, and therefore failed to obtain an address for the
2184 autoreply, Exim could try to deliver to a non-existent relative file
2185 name, causing unrelated and misleading errors. What now happens is that
2186 it logs this as a hard delivery error, but does not attempt to create a
2189 PH/11 The exinext utility has a -C option for testing purposes, but although
2190 the given file was scanned by exinext itself; it wasn't being passed on
2191 when Exim was called.
2193 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2194 an end-of-file indication when reading a command response.
2196 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2197 compiled. In many other places in Exim, IPv6 addresses are always
2198 recognized, so I have changed this. It also means that IPv4 domain
2199 literals of the form [IPV4:n.n.n.n] are now always recognized.
2201 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2202 used if the router is not running as root, for example, when verifying at
2203 ACL time, or when using -bh. The debugging output from this situation was
2204 non-existent - all you got was a failure to exec. I have made two
2207 (a) Failures to set uid/gid, the current directory, or a process leader
2208 in a subprocess such as that created by queryprogram now generate
2209 suitable debugging ouput when -d is set.
2211 (b) The queryprogram router detects when it is not running as root,
2212 outputs suitable debugging information if -d is set, and then runs
2213 the subprocess without attempting to change uid/gid.
2215 PH/15 Minor change to Makefile for building test_host (undocumented testing
2218 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2219 additional section of a DNS packet that returns MX or SRV records.
2220 Instead, it always explicitly searches for A/AAAA records. This avoids
2221 major problems that occur when a DNS server includes only records of one
2222 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2223 fixed another bug: if SRV records were looked up and the corresponding
2224 address records were *not* found in the additional section, the port
2225 values from the SRV records were lost.
2227 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2228 using the correct key (the original address) when searching the retry
2229 rules in order to find which one to use for generating the retry hint.
2231 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2232 adding the default one. Similarly, if it contains a Reply-To: header, the
2233 errors_reply_to option, if set, is not used.
2235 PH/19 When calculating a retry time, Exim used to measure the "time since
2236 failure" by looking at the "first failed" field in the retry record. Now
2237 it does not use this if it is later than than the arrival time of the
2238 message. Instead it uses the arrival time. This makes for better
2239 behaviour in cases where some deliveries succeed, thus re-setting the
2240 "first failed" field. An example is a quota failure for a huge message
2241 when small messages continue to be delivered. Without this change, the
2242 "time since failure" will always be short, possible causing more frequent
2243 delivery attempts for the huge message than are intended.
2244 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2246 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2247 $1, $2, $3) because the numerical variables can be reset during some
2248 expansion items (e.g. "match"), thereby losing the authentication data.
2250 PH/21 Make -bV show the size of off_t variables so that the test suite can
2251 decide whether to run tests for quotas > 2G.
2253 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2254 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2255 filecount value is greater than 2G or if a quota value is greater than 2G
2256 on a system where the size of off_t is not greater than 4, a panic error
2259 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2260 never match. The debug and -bh output now contains an explicit error
2261 message indicating a malformed IPv4 address or mask.
2263 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2264 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2265 PH/23 above applies.
2267 PH/25 Do not write to syslog when running in the test harness. The only
2268 occasion when this arises is a failure to open the main or panic logs
2269 (for which there is an explicit test).
2271 PH/26 Added the /no_tell option to "control=freeze".
2273 PH/27 If a host name lookup failed very early in a connection, for example, if
2274 the IP address matched host_lookup and the reverse lookup yielded a name
2275 that did not have a forward lookup, an error message of the form "no IP
2276 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2277 could be logged. Now it outputs the IP address instead of "NULL".
2279 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2280 allows the sender and the authenticated sender to be set when
2281 submitting a message from within Exim. Since child_open_exim() is
2282 documented for local_scan(), the new function should be too.
2284 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2285 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2286 results in an empty string is now treated as unset.
2288 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2290 PH/31 Added sender_verify_fail logging option.
2292 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2293 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2294 tidied the source and removed it altogether.
2296 PH/33 When a queue run was abandoned because the load average was too high, a
2297 log line was always written; now it is written only if the queue_run log
2298 selector is set. In addition, the log line for abandonment now contains
2299 information about the queue run such as the pid. This is always present
2300 in "start" and "stop" lines but was omitted from the "abandon" line.
2302 PH/34 Omit spaces between a header name and the colon in the error message that
2303 is given when verify = headers_syntax fails (if there are lots of them,
2304 the message gets confusing).
2306 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2307 names, as there are now some PTR records that contain slashes. This check
2308 is only to protect against broken name servers that fall over on strange
2309 characters, so the fact that it applies to all lookups doesn't matter.
2311 PH/36 Now that the new test suite is complete, we can remove some of the
2312 special code in Exim that was needed for the old test suite. For example,
2313 sorting DNS records because real resolvers return them in an arbitrary
2314 order. The new test suite's fake resolver always returns records in the
2317 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2318 bounces) except when queue_only is set, to avoid logging races between
2319 the different processes.
2321 PH/38 Panic-die if .include specifies a non-absolute path.
2323 PH/39 A tweak to the "H" retry rule from its user.
2325 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2326 a label. They prevented compilation on older perls.
2328 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2329 a warning to be raised on newish perls.
2331 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2332 on queue. Changes to match documented behaviour of showing count of
2333 messages matching specified criteria.
2335 PH/40 Changed the default ident timeout from 30s to 5s.
2337 PH/41 Added support for the use of login_cap features, on those BSD systems
2338 that have them, for controlling the resources used by pipe deliveries.
2340 PH/42 The content-scanning code uses fopen() to create files in which to put
2341 message data. Previously it was not paying any attention to the mode of
2342 the files. Exim runs with umask(0) because the rest of the code creates
2343 files with open(), and sets the required mode explicitly. Thus, these
2344 files were ending up world-writeable. This was not a big issue, because,
2345 being within the spool directory, they were not world-accessible. I have
2346 created a function called modefopen, which takes an additional mode
2347 argument. It sets umask(777), creates the file, chmods it to the required
2348 mode, then resets the umask. All the relevant calls to fopen() in the
2349 content scanning code have been changed to use this function.
2351 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2352 to 24 hours. This avoids potential overflow problems when processing G
2353 and H retry rules. I suspect nobody ever tinkers with this value.
2355 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2357 PH/45 When the plaintext authenticator is running as a client, the server's
2358 challenges are checked to ensure they are valid base64 strings. By
2359 default, the authentication attempt is cancelled if an invalid string is
2360 received. Setting client_ignore_invalid_base64 true ignores these errors.
2361 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2362 they are received. Thus, the responses can be made to depend on the
2363 challenges. If an invalid string is ignored, an empty string is placed in
2366 PH/46 Messages that are created by the autoreply transport now contains a
2367 References: header, in accordance with RFCs 2822 and 3834.
2369 PH/47 Added authenticated_sender_force to the smtp transport.
2371 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2373 PH/49 Installed latest patch from the Sieve maintainer.
2375 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2376 was also set, the appendfile transport was unnecessarily scanning a
2377 directory of message files (e.g. for maildir delivery) to find the count
2378 of files (along with the size), even though it did not need this
2379 information. It now does the scan only if it needs to find either the
2380 size of the count of files.
2382 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2384 PH/52 Two bugs concerned with error handling when the smtp transport is
2387 (i) Exim was not creating retry information for temporary errors given
2388 for individual recipients after the DATA command when the smtp transport
2389 was used in LMTP mode. This meant that they could be retried too
2390 frequently, and not timed out correctly.
2392 (ii) Exim was setting the flag that allows error details to be returned
2393 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2394 recipients that were returned after the DATA command.
2396 PH/53 This is related to PH/52, but is more general: for any failing address,
2397 when detailed error information was permitted to be returned to the
2398 sender, but the error was temporary, then after the final timeout, only
2399 "retry timeout exceeded" was returned. Now it returns the full error as
2400 well as "retry timeout exceeded".
2402 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2403 do this, and (what is worse) MTAs that accept it.
2405 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2406 will now be deprecated.
2408 PH/56 New os.c-cygwin from the Cygwin maintainer.
2410 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2411 formats (previously only available in exim formats via -bpr, -bpru,
2412 and -bpra. Now also available in native and exiqgrep formats)
2414 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2415 with very large, slow to parse queues
2417 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2419 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2421 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2422 responses to authentication challenges, though it was showing the
2423 challenges; (ii) I've removed the CR characters from the debug output for
2426 PH/58 Allow for the insertion of a newline as well as a space when a string
2427 is turned into more than one encoded-word during RFC 2047 encoding. The
2428 Sieve code now uses this.
2430 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2431 data_4xx, lost_connection, tls_required.
2433 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2434 message was being sent as an SMTP response.
2436 PH/61 Add -l and -k options to exicyclog.
2438 PH/62 When verifying, if an address was redirected to one new address, so that
2439 verification continued, and the new address failed or deferred after
2440 having set something in $address_data, the value of $address_data was not
2441 passed back to the ACL. This was different to the case when no
2442 redirection occurred. The value is now passed back in both cases.
2444 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2445 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2446 use login_cap.h, so on its own it isn't the distinguishing feature. The
2447 new name refers directly to the setclassresources() function.
2449 PH/65 Added configuration files for NetBSD3.
2451 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2453 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2454 is preferred over IPv4.
2456 PH/68 The bounce_return_message and bounce_return_body options were not being
2457 honoured for bounces generated during the reception of non-SMTP messages.
2458 In particular, this applied to messages rejected by the ACL. This bug has
2459 been fixed. However, if bounce_return_message is true and bounce_return_
2460 body is false, the headers that are returned for a non-SMTP message
2461 include only those that have been read before the error was detected.
2462 (In the case of an ACL rejection, they have all been read.)
2464 PH/69 The HTML version of the specification is now built in a directory called
2465 spec_html instead of spec.html, because the latter looks like a path with
2466 a MIME-type, and this confuses some software.
2468 PH/70 Catch two compiler warnings in sieve.c.
2470 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2471 function verify_get_ident() calls ip_connect() to connect a socket, but
2472 if the "connect()" function timed out, ip_connect() used to close the
2473 socket. However, verify_get_ident() also closes the socket later, and in
2474 between Exim writes to the log, which may get opened at this point. When
2475 the socket was closed in ip_connect(), the log could get the same file
2476 descriptor number as the socket. This naturally causes chaos. The fix is
2477 not to close the socket in ip_connect(); the socket should be closed by
2478 the function that creates it. There was only one place in the code where
2479 this was missing, in the iplookup router, which I don't think anybody now
2480 uses, but I've fixed it anyway.
2482 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2483 well as to direct DNS lookups. Otherwise the handling of names in host
2484 lists is inconsistent and therefore confusing.
2490 PH/01 Two changes to the default runtime configuration:
2492 (1) Move the checks for relay_from_hosts and authenticated clients from
2493 after to before the (commented out) DNS black list checks.
2495 (2) Add control=submission to the relay_from_hosts and authenticated
2496 clients checks, on the grounds that messages accepted by these
2497 statements are most likely to be submissions.
2499 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2501 (1) Generate an error if the third argument for the ${prvs expansion is
2504 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2507 (3) Reset the variables that are obtained from the first argument of
2508 ${prvscheck and used in the second argument before leaving the code,
2509 because their memory is reclaimed, so using them afterwards may do
2512 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2513 one (it's much easier than Tom thought :-).
2515 (5) Because of (4), we can now allow for the use of $prvscheck_result
2516 inside the third argument.
2518 PH/03 For some reason, the default setting of PATH when running a command from
2519 a pipe transport was just "/usr/bin". I have changed it to
2522 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2523 anything to be listed in the output from -bV.
2525 PH/05 When a filter generated an autoreply, the entire To: header line was
2526 quoted in the delivery log line, like this:
2528 => >A.N.Other <ano@some.domain> <original@ddress> ...
2530 This has been changed so that it extracts the operative address. There
2531 may be more than one such address. If so, they are comma-separated, like
2534 => >ano@some.domain,ona@other.domain <original@ddress> ...
2536 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2537 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2538 not being looked up in the rDNS to get a host name, Exim was showing the
2539 IP address twice in Received: lines, even though the IP addresses were
2540 identical. For example:
2542 Received: from [1.2.3.4] (helo=[1.2.3.4])
2544 However, if the real host name was known, it was omitting the HELO data
2545 if it matched the actual IP address. This has been tidied up so that it
2546 doesn't show the same IP address twice.
2548 PH/07 When both +timestamp and +memory debugging was on, the value given by
2549 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2550 called by the debug printing, thereby overwriting the timestamp buffer.
2551 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2553 PH/08 When the original message was included in an autoreply transport, it
2554 always said "this is a copy of the message, including all the headers",
2555 even if body_only or headers_only was set. It now gives an appropriate
2558 PH/09 Applied a patch from the Sieve maintainer which:
2560 o fixes some comments
2561 o adds the (disabled) notify extension core
2562 o adds some debug output for the result of if/elsif tests
2563 o points to the current vacation draft in the documentation
2564 and documents the missing references header update
2568 o fixes a bug in processing the envelope test (when testing
2569 multiple envelope elements, the last element determinted the
2572 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2573 Electronic Mail") by including:
2575 Auto-submitted: auto-generated
2577 in the messages that it generates (bounce messages and others, such as
2578 warnings). In the case of bounce messages for non-SMTP mesages, there was
2579 also a typo: it was using "Auto_submitted" (underscore instead of
2580 hyphen). Since every message generated by Exim is necessarily in response
2581 to another message, thes have all been changed to:
2583 Auto-Submitted: auto-replied
2585 in accordance with these statements in the RFC:
2587 The auto-replied keyword:
2589 - SHOULD be used on messages sent in direct response to another
2590 message by an automatic process,
2592 - MUST NOT be used on manually-generated messages,
2594 - MAY be used on Delivery Status Notifications (DSNs) and Message
2595 Disposition Notifications (MDNs),
2597 - MUST NOT be used on messages generated by automatic or periodic
2598 processes, except for messages which are automatic responses to
2601 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2602 to the default Received: header definition.
2604 PH/12 Added log selector acl_warn_skipped (default on).
2606 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2607 variables because (a) they are in the wrong storage pool and (b) even if
2608 they were copied, it wouldn't work properly because of the caching.
2610 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2611 checking when decoding. Apparently there are clients that generate
2612 overlong encoded strings. Why am I not surprised?
2614 PH/15 If the first argument of "${if match_address" was not empty, but did not
2615 contain an "@" character, Exim crashed. Now it writes a panic log message
2616 and treats the condition as false.
2618 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2620 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2621 extension "envelope-auth". The code is finished and in agreement with
2622 other implementations, but there is no documentation so far and in fact,
2623 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2624 not changing the active code.
2626 Print executed "if" and "elsif" statements when debugging is used. This
2627 helps a great deal to understand what a filter does.
2629 Document more things not specified clearly in RFC3028. I had all this
2630 sorted out, when out of a sudden new issues came to my mind. Oops."
2632 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2635 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2636 canonical form (as documented). However, after a host name lookup from
2637 the IP address, check_host() was doing a simple string comparison with
2638 addresses acquired from the DNS when checking that the found name did
2639 have the original IP as one of its addresses. Since any found IPv6
2640 addresses are likely to be in abbreviated form, the comparison could
2641 fail. Luckily, there already exists a function for doing the comparison
2642 by converting both addresses to binary, so now that is used instead of
2643 the text comparison.
2645 PH/20 There was another similar case to PH/19, when a complete host name was
2646 given in a host list; looking up its IP address could give an abbreviated
2647 form, whereas the current host's name might or might not be abbreviated.
2648 The same fix has been applied.
2654 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2655 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2658 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2659 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2661 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2663 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2664 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2665 Also, for "Auto-Submitted", treat anything other than "no" as
2666 non-personal, in accordance with RFC 3834. (Previously it treated
2667 anything starting "auto-" as non-personal.)
2669 TF/01 The control=submission/name=... option had a problem with syntax
2670 errors if the name included a slash character. The /name= option
2671 now slurps the rest of the string, so it can include any characters
2672 but it must come last in the list of options (after /sender_retain
2675 PH/05 Some modifications to the interface to the fake nameserver for the new
2683 TK/01 Added the "success_on_redirect" address verification option. See
2684 NewStuff for rationale and an example.
2686 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2688 PH/02 Patch to exigrep to allow it to work on syslog lines.
2690 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2691 fread() to skip over the body file's header line, because in Cygwin the
2692 header line is locked and is inaccessible.
2694 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2695 co-exist for some time) to make it clear that it is the Exim ID that is
2696 referenced, not the Message-ID: header line.
2698 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2699 string_format() function, because snprintf() does not exist on all
2702 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2703 ${stat: expansion item.
2705 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2706 protocol synchronization error", to keep the pedants happy.
2708 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2709 well as for IRIX systems, when gcc is being used. See the host.c source
2712 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2714 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2717 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2718 transport and to the smtp transport in LMTP mode.
2720 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2722 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2723 run a filter in a subprocess. This could lead to confusion in subsequent
2724 lookups in the parent process. There should also be a search_tidyup() at
2725 the end of the subprocess.
2727 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2728 only if the host matched helo_try_verify_hosts, which caused the
2729 verification to occur when the EHLO/HELO command was issued. The ACL just
2730 tested the remembered result. Now, if a previous verification attempt has
2731 not happened, "verify = helo" does it there and then.
2733 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2735 TK/03 Fix log output including CR from clamd.
2737 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2738 exist provoked a memory error which could cause a segfault.
2740 PH/15 Installed PCRE 6.2
2742 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2744 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2745 of the problem. Specifically, suggested +O2 rather than +O1 for the
2748 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2750 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2751 record, the port was not correctly logged when the outgoing_port log
2752 selector was set (it logged the transort's default port).
2754 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2755 fallback_hosts, and "hosts" in the smtp transport.
2757 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2758 host errors such as "Connection refused".
2760 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2761 authentication with radiusclient 0.4.9:
2763 - Error returned from rc_read_config was caught wrongly
2764 - Username/password not passed on to radius server due to wrong length.
2766 The presumption is that some radiusclient API changes for 4.51/PH/17
2767 were not taken care of correctly. The code is still untested by me (my
2768 Linux distribution still has 0.3.2 of radiusclient), but it was
2769 contributed by a Radius user.
2771 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2772 expanding the "port" option of the smtp transport.
2774 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2775 while reading a MIME header. Thanks to Tom Hughes for a patch.
2777 PH/24 Include config.h inside local_scan.h so that configuration settings are
2780 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2781 that in an ACL for RCPT (for example), you can examine exactly what was
2784 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2785 commands, but it was not correctly comparing the address with the actual
2786 client host address. Thus, it would show the EHLO address in Received:
2787 header lines when this was not necessary.
2789 PH/27 Added the % operator to ${eval:}.
2791 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2792 it should be ignoring failures (because with -C, for example, it has lost
2793 privilege). It wasn't ignoring creation failures other than "already
2796 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2799 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2800 because an input line was too long, either on its own, or by virtue of
2801 too many continuations, the temporary file was not being removed, and the
2802 return code was incorrect.
2804 PH/31 Missing "BOOL" in function definition in filtertest.c.
2806 PH/32 Applied Sieve patches from the maintainer.
2808 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2810 PH/33 Added "verify = not_blind".
2812 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2813 Local/Makefile (with some defaults set). These are used in built scripts
2814 such as exicyclog, but they have never been used in the exim_install
2815 script (though there are many overriding facilities there). I have
2816 arranged that the exim_install script now takes note of these two
2819 PH/35 Installed configuration files for Dragonfly.
2821 PH/36 When a locally submitted message by a trusted user did not contain a
2822 From: header, and the sender address was obtained from -f or from an SMTP
2823 MAIL command, and the trusted user did not use -F to supply a sender
2824 name, $originator_name was incorrectly used when constructing a From:
2825 header. Furthermore, $originator_name was used for submission mode
2826 messages from external hosts without From: headers in a similar way,
2827 which is clearly wrong.
2829 PH/37 Added control=suppress_local_fixups.
2831 PH/38 When log_selector = +received_sender was set, and the addition of the
2832 sender made the log line's construction buffer exactly full, or one byte
2833 less than full, an overflow happened when the terminating "\n" was
2836 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2837 when the result of a list match is failure because a DNS lookup failed.
2839 PH/40 RM_COMMAND is now used in the building process.
2841 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2842 the "build-* directories that it finds.
2844 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2845 address in a domain literal was a prefix of an interface address.
2847 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2848 when verifying a sender address, unless rewrite_headers is false.
2850 PH/44 Wrote a long comment about why errors_to addresses are verified as
2851 recipients, not senders.
2853 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2854 the ratelimit ACL was added.
2856 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2858 PH/46 Added extra information about PostgreSQL errors to the error string.
2860 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2861 avoiding the need to install special zones in a real server. This is
2862 backwards compatible; if it can't find the fake resolver, it drops back.
2863 Thus, both old and new test suites can be run.
2865 TF/02 Added util/ratelimit.pl
2867 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2868 clock is set back in time.
2870 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2871 Candler <B.Candler@pobox.com>.
2873 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2874 OK for addresses that are the result of redirections.
2876 PH/48 A number of further additions for the benefit of the new test suite,
2877 including a fake gethostbyname() that interfaces to the fake DNS resolver
2880 TF/06 The fix for widen_domains has also been applied to qualify_single and
2881 search_parents which are the other dnslookup options that can cause
2884 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2887 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2889 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2890 service" if it encountered an MX record with an empty target hostname.
2891 The message is now "an MX or SRV record indicated no SMTP service".
2893 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2894 if the DNS of the sending site is misconfigured. This is quite a
2895 common situation. This change restores the behaviour of treating a
2896 helo verification defer as a failure.
2898 PH/51 If self=fail was set on a router, the bounce message did not include the
2899 actual error message.
2905 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2907 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2908 command itself ended in error, the underlying message about the transport
2909 filter timeout was being overwritten with the pipe command error. Now the
2910 underlying error message should be appended to the second error message.
2912 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2914 PH/02 Reduce the amount of output that "make" produces by default. Full output
2915 can still be requested.
2917 PH/03 The warning log line about a condition test deferring for a "warn" verb
2918 was being output only once per connection, rather than after each
2919 occurrence (because it was using the same function as for successful
2920 "warn" verbs). This seems wrong, so I have changed it.
2922 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2923 it should not have, which might have caused a crash in the right
2924 circumstances, but probably never did.
2926 PH/04 Installed a modified version of Tony Finch's patch to make submission
2927 mode fix the return path as well as the Sender: header line, and to
2928 add a /name= option so that you can make the user's friendly name appear
2931 TF/03 Added the control = fakedefer ACL modifier.
2933 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2934 Mark Lowes for thorough testing.
2936 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2938 TK/03 Merged latest SRS patch from Miles Wilton.
2940 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2941 with the definition in sysexits.h (which is #included earlier).
2942 Fortunately, Exim does not actually use EX_OK. The code used to try to
2943 preserve the sysexits.h value, by assumimg that macro definitions were
2944 scanned for macro replacements. I have been disabused of this notion,
2945 so now the code just undefines EX_OK before #including unistd.h.
2947 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2948 in the smtp transport. When a block could not be written in a single
2949 write() function, the timeout was being re-applied to each part-write.
2950 This seems wrong - if the receiver was accepting one byte at a time it
2951 would take for ever. The timeout is now adjusted when this happens. It
2952 doesn't have to be particularly precise.
2954 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2955 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2957 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2958 without a domain if the check to <postmaster@domain> fails.
2960 SC/01 Eximstats: added -xls and the ability to specify output files
2961 (patch written by Frank Heydlauf).
2963 SC/02 Eximstats: use FileHandles for outputing results.
2965 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2967 SC/04 Eximstats: fixed display of large numbers with -nvr option
2969 SC/05 Eximstats: fixed merging of reports with empty tables.
2971 SC/06 Eximstats: added the -include_original_destination flag
2973 SC/07 Eximstats: removed tabs and trailing whitespace.
2975 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2977 TK/06 MBOX spool code: Add real "From " MBOX separator line
2978 so the .eml file is really in mbox format (even though
2979 most programs do not really care). Patch from Alex Miller.
2981 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2982 The latter is generated from $received_to and is only set if the
2983 message has one envelope recipient. SA can use these headers,
2984 obviously out-of-the-box. Patch from Alex Miller.
2986 PH/08 The ${def test on a variable was returning false if the variable's
2987 value was "0", contrary to what the specification has always said!
2988 The result should be true unless the variable is empty.
2990 PH/09 The syntax error of a character other than { following "${if
2991 def:variable_name" (after optional whitespace) was not being diagnosed.
2992 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2993 accidental colon was present, for example, could give incorrect results.
2995 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2996 result is used (not including appendfile, where other changes are about
2999 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3000 This involved changing a lot of size variables from int to off_t. It
3001 should work with maildirs and everything.
3003 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3004 spamd dying while we are connected to it.
3006 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3009 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3010 function for BDB 4.3.
3012 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3014 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3015 the sources for PCRE have been re-arranged and more files are now
3018 PH/15 The code I had for printing potentially long long variables in PH/11
3019 above was not the best (it lost precision). The length of off_t variables
3020 is now inspected at build time, and an appropriate printing format (%ld
3021 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3022 to be "long long int" or "long int". This is needed for the internal
3023 formatting function string_vformat().
3025 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3026 the configuration file to be ":syslog", then the script "guesses" where
3027 the logs files are, rather than using the compiled in default. In our
3028 case the guess is not the same as the compiled default, so the script
3029 suddenly stopped working when I started to use syslog. The patch checks
3030 to see if log_file_path is "". If so, it attempts to read it from exim
3031 with no configuration file to get the compiled in version, before it
3032 falls back to the previous guessing code."
3034 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3035 implementing BATV in an Exim configuration. See NewStuff for the gory
3038 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3039 Makefile that are specific to HP-UX.
3041 PH/18 If the "use_postmaster" option was set for a recipient callout together
3042 with the "random" option, the postmaster address was used as the MAIL
3043 FROM address for the random test, but not for the subsequent recipient
3044 test. It is now used for both.
3046 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3047 patch removes a few documentation additions to RFC 3028, because the
3048 latest draft now contains them. It adds the new en;ascii-case comparator
3049 and a new error check for 8bit text in MIME parts. Comparator and
3050 require names are now matched exactly. I enabled the subaddress
3051 extension, but it is not well tested yet (read: it works for me)."
3053 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3054 rework some of the code of TK/09 above to avoid the hardwired use of
3055 "%lld" and "long long". Replaced the call to snprintf() with a call to
3058 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3059 records point to non-existent hosts", "retry timeout exceeded", and
3060 "retry time not reached for any host after a long failure period".
3062 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3063 experimental DomainKeys support:
3065 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3066 (2) On an error such as an illegally used "control", the wrong name for
3067 the control was given.
3069 These problems did NOT occur unless DomainKeys support was compiled.
3071 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3073 PH/24 Added ${if match_ip condition.
3075 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3076 databases so that it will be absolutely obvious if a crash occurs in the
3077 DB library. This is a regular occurrence (often caused by mis-matched
3080 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3081 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3082 picked up on a user's system that detects such things. There doesn't seem
3083 to be a gcc warning option for this - only an attribute that has to be
3084 put on the function's prototype. It seems that in Fedora Core 4 they have
3085 set this on a number of new functions. No doubt there will be more in due
3088 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3089 specify a transport. However, if an address that was verified by such a
3090 router was the subject of a callout, Exim crashed because it tried to
3091 read the rcpt_include_affixes from the non-existent transport. Now it
3092 just assumes that the setting of that option is false. This bug was
3093 introduced by 4.51/PH/31.
3095 PH/28 Changed -d+all to exclude +memory, because that information is very
3096 rarely of interest, but it makes the output a lot bigger. People tend to
3097 do -d+all out of habit.
3099 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3100 code in os-type was giving problems when libc.so lives in lib64, like on
3103 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3104 aren't the modern standard, and it seems that some systems' include files
3105 don't always have them. Exim was already checking for some of the newer
3106 ones like T_AAAA, and defining it itself. I've added checks for all the
3107 record types that Exim uses.
3109 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3110 not automatically generating a new one, as it is supposed to. This
3111 prevented TLS from working. If the file did exist, but contained invalid
3112 data, a new version was generated, as expected. It was only the case of a
3113 non-existent file that was broken.
3115 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3116 with a change in libdomainkeys > 0.64.
3118 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3119 from DNS. If the selector record carries the flag, it now has
3120 precedence over the domain-wide flag.
3122 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3124 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3125 the use of a port name that isn't defined in /etc/services) occurred, the
3126 message was deferred as in a normal delivery, and thus remained on the
3127 spool, instead of being failed because of the mua_wrapper setting. This
3128 is now fixed, and I tidied up some of the mua_wrapper messages at the
3131 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3132 the messages in a hash of arrays rather than using individual hashes.
3133 This is a bit cleaner and results in dramatic memory savings, albeit
3134 at a slight CPU cost.
3136 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3137 as requested by Marc Sherman.
3139 SC/10 Eximstats: added histograms for user specified patterns as requested
3142 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3144 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3145 fopen() in the content-scanning modules that did not already have it.
3151 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3152 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3154 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3156 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3158 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3159 to test Sieve filters that use "vacation".
3161 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3162 that changes the way the GnuTLS parameters are stored in the cache file.
3163 The new format can be generated externally. For backward compatibility,
3164 if the data in the cache doesn't make sense, Exim assumes it has read an
3165 old-format file, and it generates new data and writes a new file. This
3166 means that you can't go back to an older release without removing the
3169 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3170 work if there are any delivery delays because "one_time" forces the
3171 parent to be marked "delivered", so its unseen clone is never tried
3172 again. For this reason, Exim now forbids the simultaneous setting of
3175 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3176 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3177 entry if you want to know the details. The fix, however, neglected to
3178 consider the case when local delivery batching is involved. The test for
3179 "previously delivered" was not happening when checking to see if an
3180 address could be batched with a previous (undelivered) one; under
3181 certain circumstances this could lead to multiple deliveries to the same
3184 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3185 in its include files, and this causes problems building Exim.
3187 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3188 header_syntax) but Exim was just ignoring anything given after a slash.
3189 In particular, this caused confusion with an attempt to use "verify =
3190 reverse_host_lookup/defer_ok". An error is now given when options are
3191 supplied for verify items that do not have them. (Maybe reverse_host_
3192 lookup should have a defer_ok option, but that's a different point.)
3194 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3195 defined by RFC 821) to 2048, because there were problems with some AUTH
3196 commands, and RFC 1869 says the size should be increased for extended
3197 SMTP commands that take arguments.
3199 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3202 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3203 "unknown" error; now it says that the functionality isn't in the binary.
3205 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3206 an address' error message when a string expansion fails (syntax or
3207 whatever). Otherwise the password may appear in the log. Following change
3208 PH/42 below, there is no longer a chance of it appearing in a bounce
3211 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3213 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3214 of its addresses was ever tried. (Bugzilla bug #2.)
3216 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3217 the result incorrectly in the debug output. (It correctly added a newline
3218 to what was transported.)
3220 TF/01 Added $received_time.
3222 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3223 commented out examples of how to interface to a virus scanner and to
3224 SpamAssassin. Also added commented examples of av_scanner and
3225 spamd_address settings.
3227 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3228 and controls are allowed in which ACLs. There were a couple of minor
3229 errors. Some of the entries in the conditions table (which is a table of
3230 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3231 negation of where the condition IS allowed.
3233 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3235 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3236 header file does not have a version number, so I've had to invent a new
3237 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3238 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3239 radiusclient), but it was contributed by a Radius user.
3241 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3242 files or directories, for OpenSSL.
3244 PH/19 When an Exim process that is running as root has to create an Exim log
3245 file, it does so in a subprocess that runs as exim:exim so as to get the
3246 ownership right at creation (otherwise, other Exim processes might see
3247 the file with the wrong ownership). There was no test for failure of this
3248 fork() call, which would lead to the process getting stuck as it waited
3249 for a non-existent subprocess. Forks do occasionally fail when resources
3250 run out. I reviewed all the other calls to fork(); they all seem to check
3253 PH/20 When checking for unexpected SMTP input at connect time (before writing
3254 the banner), Exim was not dealing correctly with a non-positive return
3255 from the read() function. If the client had disconnected by this time,
3256 the result was a log entry for a synchronization error with an empty
3257 string after "input=" when read() returned zero. If read() returned -1
3258 (an event I could not check), uninitialized data bytes were printed.
3259 There were reports of junk text (parts of files, etc) appearing after
3262 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3264 PH/22 Added support for macro redefinition, and (re)definition in between
3265 driver and ACL definitions.
3267 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3268 forgetting to use the resulting value; it was using the unexpanded value.
3270 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3271 hadn't been configured. The fix is from Juergen Kreileder, who
3272 understands it better than I do:
3274 "Here's what I see happening with three configured cyrus_sasl
3275 authenticators configured (plain, login, cram-md5):
3277 On startup auth_cyrus_sasl_init() gets called for each of these.
3278 This means three calls to sasl_listmech() without a specified mech_list.
3279 => SASL tests which mechs of all available mechs actually work
3280 => three warnings about OTP not working
3281 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3283 With the patch, sasl_listmech() also gets called three times. But now
3284 SASL's mech_list option is set to the server_mech specified in the the
3285 authenticator. Or in other words, the answer from sasl_listmech()
3286 gets limited to just the mech you're testing for (which is different
3288 => the return list contains just 'plain' or 'login', 'cram-md5' or
3289 nothing depending on the value of ob->server_mech.
3291 I've just tested the patch: Authentication still works fine,
3292 unavailable mechs specified in the exim configuration are still
3293 caught, and the auth.log warnings about OTP are gone."
3295 PH/25 When debugging is enabled, the contents of the command line are added
3296 to the debugging output, even when log_selector=+arguments is not
3299 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3300 answer is "GNU", and only if the return is "GNU/something" is the answer
3303 PH/27 $acl_verify_message is now set immediately after the failure of a
3304 verification in an ACL, and so is available in subsequent modifiers. In
3305 particular, the message can be preserved by coding like this:
3307 warn !verify = sender
3308 set acl_m0 = $acl_verify_message
3310 Previously, $acl_verify_message was set only while expanding "message"
3311 and "log_message" when a very denied access.
3313 PH/28 Modified OS/os.c-Linux with
3315 -#ifndef OS_LOAD_AVERAGE
3316 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3318 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3319 nomenclature these days.)
3321 PH/29 Installed patch from the Sieve maintainer that adds the options
3322 sieve_useraddress and sieve_subaddress to the redirect router.
3324 PH/30 In these circumstances:
3325 . Two addresses routed to the same list of hosts;
3326 . First host does not offer TLS;
3327 . First host accepts first address;
3328 . First host gives temporary error to second address;
3329 . Second host offers TLS and a TLS session is established;
3330 . Second host accepts second address.
3331 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3332 and peerdn, if requested) that were in fact used only for the second
3335 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3336 attention to any local part prefix or suffix that was matched by the
3337 router that accepted the address. It now behaves in the same way as it
3338 does for delivery: the affixes are removed from the local part unless
3339 rcpt_include_affixes is set on the transport.
3341 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3342 timeout during the DATA phase of an incoming message.
3344 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3345 applied a patch sanctioned by the Sieve maintainer.
3347 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3348 the uid or gid is negative. A case of a negative gid caused this to be
3349 noticed. The fix allows for either to be negative.
3351 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3352 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3355 PH/36 The change PH/12 above was broken. Fixed it.
3357 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3358 the grounds that routing the same address twice would always produce the
3359 same answer. This might have been true once, but it is certainly no
3360 longer true now. Routing a child address may depend on the previous
3361 routing that produced that child. Some complicated redirection strategies
3362 went wrong when messages had multiple recipients, and made Exim's
3363 behaviour dependent on the order in which the addresses were given.
3365 I have moved the duplicate checking until after the routing is complete.
3366 Exim scans the addresses that are assigned to local and remote
3367 transports, and removes any duplicates. This means that more work will be
3368 done, as duplicates will always all be routed, but duplicates are
3369 presumably rare, so I don't expect this is of any significance.
3371 For deliveries to pipes, files, and autoreplies, the duplicate checking
3372 still happens during the routing process, since they are not going to be
3375 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3376 It corrects a timeout issue with spamd. This is Ian's comment: "The
3377 background is that sometimes spamd either never reads data from a
3378 connection it has accepted, or it never writes response data. The exiscan
3379 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3380 blindly assumes that writes won't block so it may never time out."
3382 PH/39 Allow G after quota size as well as K and M.
3384 PH/40 The value set for $authenticated_id in an authenticator may not contain
3385 binary zeroes or newlines because the value is written to log lines and
3386 to spool files. There was no check on this. Now the value is run through
3387 the string_printing() function so that such characters are converted to
3388 printable escape sequences.
3390 PH/41 $message_linecount is a new variable that contains the total number of
3391 lines in the message. Compare $body_linecount, which is the count for the
3394 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3395 bounce and delay warning messages, except in certain special cases, which
3398 (a) An SMTP error message from a remote host;
3399 (b) A message specified in a :fail: redirection;
3400 (c) A message specified in a "fail" command in a system filter;
3401 (d) A message specified in a FAIL return from the queryprogram router;
3402 (e) A message specified by the cannot_route_message router option.
3404 In these cases only, Exim does include the error details in bounce and
3405 warning messages. There are also a few cases where bland messages such
3406 as "unrouteable address" or "local delivery error" are given.
3408 PH/43 $value is now also set for the "else" part of a ${run expansion.
3410 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3411 being worked on, but at least Exim now implements the latest version to
3414 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3415 process to complete was treated as a delivery failure, a timeout while
3416 writing the message to the pipe was logged, but erroneously treated as a
3417 successful delivery. Such timeouts include transport filter timeouts. For
3418 consistency with the overall process timeout, these timeouts are now
3419 treated as errors, giving rise to delivery failures by default. However,
3420 there is now a new Boolean option for the pipe transport called
3421 timeout_defer, which, if set TRUE, converts the failures into defers for
3422 both kinds of timeout. A transport filter timeout is now identified in
3425 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3426 systems where "make" and "gmake" are different, calling "gmake" at top
3427 level broke things. I've arranged for the value of $(MAKE) to be passed
3428 from the Makefile to this script so that it can call the same version of
3432 A note about Exim versions 4.44 and 4.50
3433 ----------------------------------------
3435 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3436 changes of various kinds. As a consequence, a big documentation update was
3437 needed. This delayed the release for rather longer than seemed good, especially
3438 in the light of a couple of (minor) security issues. Therefore, the changes
3439 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3440 release. So 4.44 and 4.50 are in effect two different branches that both start
3443 I have left the 4.50 change log unchanged; it contains all the changes since
3444 4.43. The change log for 4.44 is below; many of its items are identical to
3445 those for 4.50. This seems to be the most sensible way to preserve the
3446 historical information.
3452 1. Minor wording change to the doc/README.SIEVE file.
3454 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3455 computation of the current number of files was incorrect.
3457 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3458 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3461 4. Give more explanation in the error message when the command for a transport
3462 filter fails to execute.
3464 5. There are several places where Exim runs a non-Exim command in a
3465 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3466 was being done only for the command run by the queryprogram router. It is
3467 now done for all such subprocesses. The other cases are: ${run, transport
3468 filters, and the commands run by the lmtp and pipe transports.
3470 6. Added CONFIGURE_GROUP build-time option.
3472 7. Some older OS have a limit of 256 on the maximum number of file
3473 descriptors. Exim was using setrlimit() to set 1000 as a large value
3474 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3475 systems. I've change it so that if it can't get 1000, it tries for 256.
3477 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3478 was an oversight, and furthermore, ever since the addition of extra
3479 controls (e.g. 4.43/32), the checks on when to allow different forms of
3480 "control" were broken. There should now be diagnostics for all cases when a
3481 control that does not make sense is encountered.
3483 9. Added the /retain_sender option to "control=submission".
3485 10. $recipients is now available in the predata ACL (oversight).
3487 11. Tidy the search cache before the fork to do a delivery from a message
3488 received from the command line. Otherwise the child will trigger a lookup
3489 failure and thereby defer the delivery if it tries to use (for example) a
3490 cached ldap connection that the parent has called unbind on.
3492 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3493 of $address_data from the recipient verification was clobbered by the
3494 sender verification.
3496 13. The value of address_data from a sender verification is now available in
3497 $sender_address_data in subsequent conditions in the ACL statement.
3499 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3501 15. Added a new option "connect=<time>" to callout options, to set a different
3504 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3505 was its contents. (It was OK if the option was not defined at all.)
3507 17. A "Completed" log line is now written for messages that are removed from
3508 the spool by the -Mrm option.
3510 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3511 information about exactly what failed.
3513 19. Added -dd to debug only the daemon process.
3515 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3516 handles timeouts, both on the server side and network timeouts. Renamed the
3517 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3519 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3520 It is now set to "smtps".
3522 22. $host_address is now set to the target address during the checking of
3523 ignore_target_hosts.
3525 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3526 being passed; this would have caused $sender_host_name to have been used if
3527 matching the list had actually called for a host name (not very likely,
3528 since this list is usually IP addresses). A host name is now passed as
3531 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3532 code that specifies a non-restarting handler (typically sigaction() in
3533 modern systems) in an attempt to fix a rare and obscure crash bug.
3535 25. Narrowed the window for a race in the daemon that could cause it to ignore
3536 SIGCHLD signals. This is not a major problem, because they are used only to
3537 wake it up if nothing else does.
3539 26. A malformed maildirsize file could cause Exim to calculate negative values
3540 for the mailbox size or file count. Odd effects could occur as a result.
3541 The maildirsize information is now recalculated if the size or filecount
3544 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3545 support for a long time. Removed HAVE_SYS_VFS_H.
3547 28. Installed the latest version of exipick from John Jetmore.
3549 29. In an address list, if the pattern was not a regular expression, an empty
3550 subject address (from a bounce message) matched only if the pattern was an
3551 empty string. Non-empty patterns were not even tested. This was the wrong
3552 because it is perfectly reasonable to use an empty address as part of a
3553 database query. An empty address is now tested by patterns that are
3554 lookups. However, all the other forms of pattern expect the subject to
3555 contain a local part and a domain, and therefore, for them, an empty
3556 address still always fails if the pattern is not itself empty.
3558 30. Exim went into a mad DNS loop when attempting to do a callout where the
3559 host was specified on an smtp transport, and looking it up yielded more
3560 than one IP address.
3562 31. Re-factored the code for checking spool and log partition space into a
3563 function that finds that data and another that does the check. The former
3564 is then used to implement four new variables: $spool_space, $log_space,
3565 $spool_inodes, and $log_inodes.
3567 32. The RFC2047 encoding function was originally intended for short strings
3568 such as real names; it was not keeping to the 75-character limit for
3569 encoded words that the RFC imposes. It now respects the limit, and
3570 generates multiple encoded words if necessary. To be on the safe side, I
3571 have increased the buffer size for the ${rfc2047: expansion operator from
3574 33. It is now permitted to omit both strings after an "if" condition; if the
3575 condition is true, the result is "true". As before, when the second string
3576 is omitted, a false condition yields an empty string. This makes it less
3577 cumbersome to write custom ACL and router conditions.
3579 34. Failure to deliver a bounce message always caused it to be frozen, even if
3580 there was an errors_to setting on the router. The errors_to setting is now
3583 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3584 canonical form (fully expanded) before being placed in
3585 $sender_host_address.
3587 36. The table in the code that translates DNS record types into text (T_A to
3588 "A" for instance) was missing entries for NS and CNAME. It is just possible
3589 that this could have caused confusion if both these types were looked up
3590 for the same domain, because the text type is used as part of Exim's
3591 per-process caching. But the chance of anyone hitting this buglet seems
3594 37. The dnsdb lookup has been extended in a number of ways.
3596 (1) There is a new type, "zns", which walks up the domain tree until it
3597 finds some nameserver records. It should be used with care.
3599 (2) There is a new type, "mxh", which is like "mx" except that it returns
3600 just the host names, not the priorities.
3602 (3) It is now possible to give a list of domains (or IP addresses) to be
3603 looked up. The behaviour when one of the lookups defers can be
3604 controlled by a keyword.
3606 (4) It is now possible to specify the separator character for use when
3607 multiple records are returned.
3609 38. The dnslists ACL condition has been extended: it is now possible to supply
3610 a list of IP addresses and/or domains to be looked up in a particular DNS
3613 39. Added log_selector=+queue_time_overall.
3615 40. When running the queue in the test harness, wait just a tad after forking a
3616 delivery process, to get repeatability of debugging output.
3618 41. Include certificate and key file names in error message when GnuTLS fails
3619 to set them up, because the GnuTLS error message doesn't include the name
3620 of the failing file when there is a problem reading it.
3622 42. Allow both -bf and -bF in the same test run.
3624 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3626 44. The "Exiscan patch" is now merged into the mainline Exim source.
3628 45. Sometimes the final signoff response after QUIT could fail to get
3629 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3630 < 0 before doing a fflush(). This bug looks as though it goes back to the
3631 introduction of TLS in release 3.20, but "sometimes" must have been rare
3632 because the tests only now provoked it.
3634 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3635 (this can affect the format of dates).
3637 47. exim_tidydb, when checking for the continued existence of a message for
3638 which it has found a message-specific retry record, was not finding
3639 messages that were in split spool directories. Consequently, it was
3640 deleting retry records that should have stayed in existence.
3642 48. Steve fixed some bugs in eximstats.
3644 49. The SPA authentication driver was not abandoning authentication and moving
3645 on to the next authenticator when an expansion was forced to fail,
3646 contradicting the general specification for all authenticators. Instead it
3647 was generating a temporary error. It now behaves as specified.
3649 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3650 (the order specifies the preference for clients). The order is now AES256,
3651 AES128, 3DES, ARCFOUR128.
3653 51. Small patch to Sieve code - explicitly set From: when generating an
3656 52. Exim crashed if a remote delivery caused a very long error message to be
3657 recorded - for instance if somebody sent an entire SpamAssassin report back
3658 as a large number of 550 error lines. This bug was coincidentally fixed by
3659 increasing the size of one of Exim's internal buffers (big_buffer) that
3660 happened as part of the Exiscan merge. However, to be on the safe side, I
3661 have made the code more robust (and fixed the comments that describe what
3664 53. Now that there can be additional text after "Completed" in log lines (if
3665 the queue_time_overall log selector is set), a one-byte patch to exigrep
3666 was needed to allow it to recognize "Completed" as not the last thing in
3669 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3670 patch that reportedly fixes this has been added. I am not expert enough to
3671 create a test for it. This is what the patch creator wrote:
3673 "I found a little strange behaviour of ldap code when working with
3674 Windows 2003 AD Domain, where users was placed in more than one
3675 Organization Units. When I tried to give exim partial DN, the exit code
3676 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3677 But simultaneously result of request was absolutely normal ldap result,
3678 so I produce this patch..."
3680 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3681 REFERENCE, so I have modified the code to exclude the patch when that macro
3684 55. Some experimental protocols are using DNS PTR records for new purposes. The
3685 keys for these records are domain names, not reversed IP addresses. The
3686 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3687 leaves it alone. Component reversal etc. now happens only for IP addresses.
3690 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3692 57. Double the size of the debug message buffer (to 2048) so that more of very
3693 long debug lines gets shown.
3695 58. The exicyclog utility now does better if the number of log files to keep
3696 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3698 59. Two changes related to the smtp_active_hostname option:
3700 (1) $smtp_active_hostname is now available as a variable.
3701 (2) The default for smtp_banner uses $smtp_active_hostname instead
3702 of $primary_hostname.
3704 60. The host_aton() function is supposed to be passed a string that is known
3705 to be a valid IP address. However, in the case of IPv6 addresses, it was
3706 not checking this. This is a hostage to fortune. Exim now panics and dies
3707 if the condition is not met. A case was found where this could be provoked
3708 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3709 components; fortuitously, this particular loophole had already been fixed
3710 by change 4.50/55 above.
3712 If there are any other similar loopholes, the new check in host_aton()
3713 itself should stop them being exploited. The report I received stated that
3714 data on the command line could provoke the exploit when Exim was running as
3715 exim, but did not say which command line option was involved. All I could
3716 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3717 running as the user.
3720 61. There was a buffer overflow vulnerability in the SPA authentication code
3721 (which came originally from the Samba project). I have added a test to the
3722 spa_base64_to_bits() function which I hope fixes it.
3725 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3726 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3728 63. The daemon start-up calls getloadavg() while still root for those OS that
3729 need the first call to be done as root, but it missed one case: when
3730 deliver_queue_load_max is set with deliver_drop_privilege. This is
3731 necessary for the benefit of the queue runner, because there is no re-exec
3732 when deliver_drop_privilege is set.
3734 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3735 This has been fixed.
3737 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3738 was in use, was not putting the data itself into the right store pool;
3739 consequently, it could be overwritten for a subsequent message in the same
3740 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3743 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3745 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3746 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3747 the function were treating the return as a boolean value, which happened to
3748 work because 0=false and not-0=true, but is not correct code.
3750 68. The host_aton() function was not handling scoped IPv6 addresses (those
3751 with, for example, "%eth0" on the end) correctly.
3753 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3754 negated items (that is, ~something) in unsigned ints. Some compilers
3755 apparently mutter when there is no cast.
3757 70. If an address verification called from an ACL failed, and did not produce a
3758 user-specific message (i.e. there was only a "system" message), nothing was
3759 put in $acl_verify_message. In this situation, it now puts the system
3762 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3763 session; change 4.31/43 added the unwanted input to the log line - except
3764 that it did not do this in the start of session case. It now does.
3766 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3767 This is wrong and can cause the other end to generate a synchronization
3768 error if it is another Exim or anything else that does the synchronization
3769 check. A QUIT command is no longer sent after a timeout.
3771 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3772 during host lookups.
3774 74. The defer_ok option of callout verification was not working if it was used
3775 when verifying addresses in header lines, that is, for this case:
3777 verify = header_sender/callout=defer_ok
3779 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3780 those file descriptors could be used for SMTP connections. If anything
3781 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3782 could be sent to the SMTP client, causing chaos. The daemon now opens
3783 stdin, stdout, and stderr to /dev/null when it puts itself into the
3786 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3787 log by default. The user can override this with suitable Perl magic.
3789 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3790 the log message when discard triggers, was not working for the DATA ACL or
3791 for the non-SMTP ACL.
3793 78. Error message wording change in sieve.c.
3795 79. If smtp_accept_max_per_host was set, the number of connections could be
3796 restricted to fewer than expected, because the daemon was trying to set up
3797 a new connection before checking whether the processes handling previous
3798 connections had finished. The check for completed processes is now done
3799 earlier. On busy systems, this bug wouldn't be noticed because something
3800 else would have woken the daemon, and it would have reaped the completed
3803 80. If a message was submitted locally by a user whose login name contained one
3804 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3805 It caused a spool format error. I have fixed the spool reading code. A
3806 related problem was that the "from" clause in the Received: line became
3807 illegal because of the space(s). It is now covered by ${quote_local_part.
3809 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3812 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3814 83. Patch from Sieve maintainer to fix unterminated string problem in
3815 "vacation" handling.
3817 84. Some minor changes to the Linux configuration files to help with other
3818 OS variants using glibc.
3820 85. One more patch for Sieve to update vacation handling to latest spec.
3823 ----------------------------------------------------
3824 See the note above about the 4.44 and 4.50 releases.
3825 ----------------------------------------------------
3831 1. Change 4.43/35 introduced a bug that caused file counts to be
3832 incorrectly computed when quota_filecount was set in an appendfile
3835 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3836 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3839 3. Give more explanation in the error message when the command for a transport
3840 filter fails to execute.
3842 4. There are several places where Exim runs a non-Exim command in a
3843 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3844 was being done only for the command run by the queryprogram router. It is
3845 now done for all such subprocesses. The other cases are: ${run, transport
3846 filters, and the commands run by the lmtp and pipe transports.
3848 5. Some older OS have a limit of 256 on the maximum number of file
3849 descriptors. Exim was using setrlimit() to set 1000 as a large value
3850 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3851 systems. I've change it so that if it can't get 1000, it tries for 256.
3853 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3854 was an oversight, and furthermore, ever since the addition of extra
3855 controls (e.g. 4.43/32), the checks on when to allow different forms of
3856 "control" were broken. There should now be diagnostics for all cases when a
3857 control that does not make sense is encountered.
3859 7. $recipients is now available in the predata ACL (oversight).
3861 8. Tidy the search cache before the fork to do a delivery from a message
3862 received from the command line. Otherwise the child will trigger a lookup
3863 failure and thereby defer the delivery if it tries to use (for example) a
3864 cached ldap connection that the parent has called unbind on.
3866 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3867 of $address_data from the recipient verification was clobbered by the
3868 sender verification.
3870 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3871 was its contents. (It was OK if the option was not defined at all.)
3873 11. A "Completed" log line is now written for messages that are removed from
3874 the spool by the -Mrm option.
3876 12. $host_address is now set to the target address during the checking of
3877 ignore_target_hosts.
3879 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3880 being passed; this would have caused $sender_host_name to have been used if
3881 matching the list had actually called for a host name (not very likely,
3882 since this list is usually IP addresses). A host name is now passed as
3885 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3886 code that specifies a non-restarting handler (typically sigaction() in
3887 modern systems) in an attempt to fix a rare and obscure crash bug.
3889 15. Narrowed the window for a race in the daemon that could cause it to ignore
3890 SIGCHLD signals. This is not a major problem, because they are used only to
3891 wake it up if nothing else does.
3893 16. A malformed maildirsize file could cause Exim to calculate negative values
3894 for the mailbox size or file count. Odd effects could occur as a result.
3895 The maildirsize information is now recalculated if the size or filecount
3898 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3899 support for a long time. Removed HAVE_SYS_VFS_H.
3901 18. Updated exipick to current release from John Jetmore.
3903 19. Allow an empty sender to be matched against a lookup in an address list.
3904 Previously the only cases considered were a regular expression, or an
3907 20. Exim went into a mad DNS lookup loop when doing a callout where the
3908 host was specified on the transport, if the DNS lookup yielded more than
3911 21. The RFC2047 encoding function was originally intended for short strings
3912 such as real names; it was not keeping to the 75-character limit for
3913 encoded words that the RFC imposes. It now respects the limit, and
3914 generates multiple encoded words if necessary. To be on the safe side, I
3915 have increased the buffer size for the ${rfc2047: expansion operator from
3918 22. Failure to deliver a bounce message always caused it to be frozen, even if
3919 there was an errors_to setting on the router. The errors_to setting is now
3922 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3923 canonical form (fully expanded) before being placed in
3924 $sender_host_address.
3926 24. Updated eximstats to version 1.33
3928 25. Include certificate and key file names in error message when GnuTLS fails
3929 to set them up, because the GnuTLS error message doesn't include the name
3930 of the failing file when there is a problem reading it.
3932 26. Expand error message when OpenSSL has problems setting up cert/key files.
3935 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3936 (this can affect the format of dates).
3938 28. exim_tidydb, when checking for the continued existence of a message for
3939 which it has found a message-specific retry record, was not finding
3940 messages that were in split spool directories. Consequently, it was
3941 deleting retry records that should have stayed in existence.
3943 29. eximstats updated to version 1.35
3944 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3945 1.35 - bugfix such that pie charts by volume are generated correctly
3947 30. The SPA authentication driver was not abandoning authentication and moving
3948 on to the next authenticator when an expansion was forced to fail,
3949 contradicting the general specification for all authenticators. Instead it
3950 was generating a temporary error. It now behaves as specified.
3952 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3953 (the order specifies the preference for clients). The order is now AES256,
3954 AES128, 3DES, ARCFOUR128.
3956 31. Small patch to Sieve code - explicitly set From: when generating an
3959 32. Exim crashed if a remote delivery caused a very long error message to be
3960 recorded - for instance if somebody sent an entire SpamAssassin report back
3961 as a large number of 550 error lines. This bug was coincidentally fixed by
3962 increasing the size of one of Exim's internal buffers (big_buffer) that
3963 happened as part of the Exiscan merge. However, to be on the safe side, I
3964 have made the code more robust (and fixed the comments that describe what
3967 33. Some experimental protocols are using DNS PTR records for new purposes. The
3968 keys for these records are domain names, not reversed IP addresses. The
3969 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3970 leaves it alone. Component reversal etc. now happens only for IP addresses.
3973 34. The host_aton() function is supposed to be passed a string that is known
3974 to be a valid IP address. However, in the case of IPv6 addresses, it was
3975 not checking this. This is a hostage to fortune. Exim now panics and dies
3976 if the condition is not met. A case was found where this could be provoked
3977 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3978 components; fortuitously, this particular loophole had already been fixed
3979 by change 4.50/55 or 4.44/33 above.
3981 If there are any other similar loopholes, the new check in host_aton()
3982 itself should stop them being exploited. The report I received stated that
3983 data on the command line could provoke the exploit when Exim was running as
3984 exim, but did not say which command line option was involved. All I could
3985 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3986 running as the user.
3989 35. There was a buffer overflow vulnerability in the SPA authentication code
3990 (which came originally from the Samba project). I have added a test to the
3991 spa_base64_to_bits() function which I hope fixes it.
3994 36. The daemon start-up calls getloadavg() while still root for those OS that
3995 need the first call to be done as root, but it missed one case: when
3996 deliver_queue_load_max is set with deliver_drop_privilege. This is
3997 necessary for the benefit of the queue runner, because there is no re-exec
3998 when deliver_drop_privilege is set.
4000 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4001 was in use, was not putting the data itself into the right store pool;
4002 consequently, it could be overwritten for a subsequent message in the same
4003 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4006 38. Sometimes the final signoff response after QUIT could fail to get
4007 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4008 < 0 before doing a fflush(). This bug looks as though it goes back to the
4009 introduction of TLS in release 3.20, but "sometimes" must have been rare
4010 because the tests only now provoked it.
4016 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4017 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4018 Now it returns an integer. A number of calls were still expecting a T/F
4019 return. Fortuitously, in all cases, the tests worked in OK situations,
4020 which is the norm. However, things would have gone wrong on any write
4021 failures on the smtp file descriptor. This function is used when sending
4022 messages over SMTP and also when doing verify callouts.
4024 2. When Exim is called to do synchronous delivery of a locally submitted
4025 message (the -odf or -odi options), it no longer closes stderr before doing
4028 3. Implemented the mua_wrapper option.
4030 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4032 5. Implemented the functions header_remove(), header_testname(),
4033 header_add_at_position(), and receive_remove_recipient(), and exported them
4036 6. If an ACL "warn" statement specified the addition of headers, Exim already
4037 inserted X-ACL-Warn: at the start if there was no header name. However, it
4038 was not making this test for the second and subsequent header lines if
4039 there were newlines in the string. This meant that an invalid header could
4040 be inserted if Exim was badly configured.
4042 7. Allow an ACL "warn" statement to add header lines at the start or after all
4043 the Received: headers, as well as at the end.
4045 8. Added the rcpt_4xx retry error code.
4047 9. Added postmaster_mailfrom=xxx to callout verification option.
4049 10. Added mailfrom=xxxx to the callout verification option, for verify=
4052 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4053 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4055 12. Inserted some casts to stop certain compilers warning when using pointer
4056 differences as field lengths or precisions in printf-type calls (mostly
4057 affecting debugging statements).
4059 13. Added optional readline() support for -be (dynamically loaded).
4061 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4062 same clock tick as a message's arrival, so that its received time was the
4063 same as the "first fail" time on the retry record, and that message
4064 remained on the queue past the ultimate address timeout, every queue runner
4065 would try a delivery (because it was past the ultimate address timeout) but
4066 after another failure, the ultimate address timeout, which should have then
4067 bounced the address, did not kick in. This was a "< instead of <=" error;
4068 in most cases the first failure would have been in the next clock tick
4069 after the received time, and all would be well.
4071 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4072 being recognized when the domain list was tested by the match_domain
4073 condition in an expansion string.
4075 16. Added the ${str2b64: operator.
4077 17. Exim was always calling setrlimit() to set a large limit for the number of
4078 processes, without checking whether the existing limit was already
4079 adequate. (It did check for the limit on file descriptors.) Furthermore,
4080 errors from getrlimit() and setrlimit() were being ignored. Now they are
4081 logged to the main and panic logs, but Exim does carry on, to try to do its
4082 job under whatever limits there are.
4084 18. Imported PCRE 5.0.
4086 19. Trivial typo in log message " temporarily refused connection" (the leading
4089 20. If the log selector return_path_on_delivery was set and an address was
4090 redirected to /dev/null, the delivery process crashed because it assumed
4091 that a return path would always be set for a "successful" delivery. In this
4092 case, the whole delivery is bypassed as an optimization, and therefore no
4095 21. Internal re-arrangement: the function for sending a challenge and reading
4096 a response while authentication was assuming a zero-terminated challenge
4097 string. It's now changed to take a pointer and a length, to allow for
4098 binary data in such strings.
4100 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4102 23. Exim was not respecting finduser_retries when seeking the login of the
4103 uid under which it was called; it was always trying 10 times. (The default
4104 setting of finduser_retries is zero.) Also, it was sleeping after the final
4105 failure, which is pointless.
4107 24. Implemented tls_on_connect_ports.
4109 25. Implemented acl_smtp_predata.
4111 26. If the domain in control=submission is set empty, Exim assumes that the
4112 authenticated id is a complete email address when it generates From: or
4113 Sender: header lines.
4115 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4116 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4117 chown and chgrp in /bin and hostname in /usr/bin.
4119 28. Exim was keeping the "process log" file open after each use, just as it
4120 does for the main log. This opens the possibility of it remaining open for
4121 long periods when the USR1 signal hits a daemon. Occasional processlog
4122 errors were reported, that could have been caused by this. Anyway, it seems
4123 much more sensible not to leave this file open at all, so that is what now
4126 29. The long-running daemon process does not normally write to the log once it
4127 has entered its main loop, and it closes the log before doing so. This is
4128 so that log files can straightforwardly be renamed and moved. However,
4129 there are a couple of unusual error situations where the daemon does write
4130 log entries, and I had neglected to close the log afterwards.
4132 30. The text of an SMTP error response that was received during a remote
4133 delivery was being truncated at 512 bytes. This is too short for some of
4134 the long messages that one sometimes sees. I've increased the limit to
4137 31. It is now possible to make retry rules that apply only when a message has a
4138 specific sender, in particular, an empty sender.
4140 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4141 it possible to be selective about when SMTP synchronization is enforced.
4143 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4145 32. Implemented hosts_connection_nolog.
4147 33. Added an ACL for QUIT.
4149 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4152 35. Added mailbox_size and mailbox_filecount to appendfile.
4154 36. Added control = no_multiline_responses to ACLs.
4156 37. There was a bug in the logic of the code that waits for the clock to tick
4157 in the case where the clock went backwards by a substantial amount such
4158 that the microsecond fraction of "now" was more than the microsecond
4159 fraction of "then" (but the whole seconds number was less).
4161 38. Added support for the libradius Radius client library this is found on
4162 FreeBSD (previously only the radiusclient library was supported).
4168 1. When certain lookups returned multiple values in the form name=value, the
4169 quoting of the values was not always being done properly. Specifically:
4170 (a) If the value started with a double quote, but contained no whitespace,
4172 (b) If the value contained whitespace other than a space character (i.e.
4173 tabs or newlines or carriage returns) it was not quoted.
4174 This fix has been applied to the mysql and pgsql lookups by writing a
4175 separate quoting function and calling it from the lookup code. The fix
4176 should probably also be applied to nisplus, ibase and oracle lookups, but
4177 since I cannot test any of those, I have not disturbed their existing code.
4179 2. A hit in the callout cache for a specific address caused a log line with no
4180 reason for rejecting RCPT. Now it says "Previous (cached) callout
4181 verification failure".
4183 3. There was an off-by-one bug in the queryprogram router. An over-long
4184 return line was truncated at 256 instead of 255 characters, thereby
4185 overflowing its buffer with the terminating zero. As well as fixing this, I
4186 have increased the buffer size to 1024 (and made a note to document this).
4188 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4189 when Exim is waiting for an SMTP response from a remote server, Exim
4190 restarts its select() call on the socket, thereby resetting its timeout.
4191 This is not a problem when such interrupts are rare. Somebody set up a cron
4192 job to run exiwhat every 2 minutes, which is less than the normal select()
4193 timeout (5 or 10 minutes). This meant that the select() timeout never
4194 kicked in because it was always reset. I have fixed this by comparing the
4195 time when an interrupt arrives with the time at the start of the first call
4196 to select(). If more time than the timeout has elapsed, the interrupt is
4197 treated as a timeout.
4199 5. Some internal re-factoring in preparation for the addition of Sieve
4200 extensions (by MH). In particular, the "personal" test is moved to a
4201 separate function, and given an option for scanning Cc: and Bcc: (which is
4202 not set for Exim filters).
4204 6. When Exim created an email address using the login of the caller as the
4205 local part (e.g. when creating a From: or Sender: header line), it was not
4206 quoting the local part when it contained special characters such as @.
4208 7. Installed new OpenBSD configuration files.
4210 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4211 try to make them clearer.
4213 9. Callout options, other than the timeout value, were being ignored when
4214 verifying sender addresses in header lines. For example, when using
4216 verify = header_sender/callout=no_cache
4218 the cache was (incorrectly) being used.
4220 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4221 only the Cygwin environment.
4223 11. When return_path_on_delivery was set as a log selector, if different remote
4224 addresses in the same message used different return paths and parallel
4225 remote delivery occurred, the wrong values would sometimes be logged.
4226 (Whenever a remote delivery process finished, the return path value from
4227 the most recently started remote delivery process was logged.)
4229 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4230 lines when AUTH and/or TLS are in use. This is the "received protocol"
4231 field. Exim used to use "asmtp" for authenticated SMTP, without any
4232 indication (in the protocol name) for TLS use. Now it follows the RFC and
4233 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4234 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4235 names appear in log lines as well as in Received: header lines.
4237 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4238 extensions, and comparison tests, and to fix some bugs.
4240 14. Changes to the "personal" filter test:
4242 (1) The test was buggy in that it was just doing the equivalent of
4243 "contains" tests on header lines. For example, if a user's address was
4244 anne@some.where, the "personal" test would incorrectly be true for
4246 To: susanne@some.where
4248 This test is now done by extracting each address from the header in turn,
4249 and checking the entire address. Other tests that are part of "personal"
4250 are now done using regular expressions (for example, to check local parts
4251 of addresses in From: header lines).
4253 (2) The list of non-personal local parts in From: addresses has been
4254 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4255 taken from the Sieve specification recommendations.
4257 (3) If the message contains any header line starting with "List-" it is
4258 treated as non-personal.
4260 (4) The test for "circular" in the Subject: header line has been removed
4261 because it now seems ill-conceived.
4263 15. Minor typos in src/EDITME comments corrected.
4265 16. Installed latest exipick from John Jetmore.
4267 17. If headers_add on a router specified a text string that was too long for
4268 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4269 of string_sprintf() is now avoided.
4271 18. $message_body_size was not set (it was always zero) when running the DATA
4272 ACL and the local_scan() function.
4274 19. For the "mail" command in an Exim filter, no default was being set for
4275 the once_repeat time, causing a random time value to be used if "once" was
4276 specified. (If the value happened to be <= 0, no repeat happened.) The
4277 default is now 0s, meaning "never repeat". The "vacation" command was OK
4278 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4279 (I found it when inspecting the code).
4281 20. There is now an overall timeout for performing a callout verification. It
4282 defaults to 4 times the callout timeout, which applies to individual SMTP
4283 commands during the callout. The overall timeout applies when there is more
4284 than one host that can be tried. The timeout is checked before trying the
4285 next host. This prevents very long delays if there are a large number of
4286 hosts and all are timing out (e.g. when the network connections are timing
4287 out). The value of the overall timeout can be changed by specifying an
4288 additional sub-option for "callout", called "maxwait". For example:
4290 verify = sender/callout=5s,maxwait=20s
4292 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4293 to the end before writing, but this should make it even safer).
4295 22. Exim was forgetting that it had advertised PIPELINING for the second and
4296 subsequent messages on an SMTP connection. It was also not resetting its
4297 memory on STARTTLS and an internal HELO.
4299 23. When Exim logs an SMTP synchronization error within a session, it now
4300 records whether PIPELINING has been advertised or not.
4302 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4303 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4304 rather than long int.
4306 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4308 26. Added the never_mail option to autoreply.
4314 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4315 crash if the getsockname() call failed; this can happen if a connection is
4316 closed very soon after it is established. The problem was simply in the
4317 order in which certain operations were done, causing Exim to try to write
4318 to the SMTP stream before it had set up the file descriptor. The bug has
4319 been fixed by making things happen in the correct order.
4325 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4326 before the connection was closed, thus losing the rejection response.
4328 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4329 some early Solaris releases, but causes trouble in current releases where
4330 socklen_t is defined.
4332 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4335 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4338 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4339 that releasing the top of it at the end releases what was used for sub-
4340 expansions (unless the block got too big). However, discard this block if
4341 the first thing is a variable or header, so that we can use its block when
4342 it is dynamic (useful for very large $message_headers, for example).
4344 6. Lookups now cache *every* query, not just the most recent. A new, separate
4345 store pool is used for this. It can be recovered when all lookup caches are
4346 flushed. Lookups now release memory at the end of their result strings.
4347 This has involved some general refactoring of the lookup sources.
4349 7. Some code has been added to the store_xxx() functions to reduce the amount
4350 of flapping under certain conditions.
4352 8. log_incoming_interface used to affect only the <= reception log lines. Now
4353 it causes the local interface and port to be added to several more SMTP log
4354 lines, for example "SMTP connection from", and rejection lines.
4356 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4358 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4360 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4361 could be overwritten at the end of the current message (or the start of a
4362 new message if it was set in a HELO ACL). The value is now preserved for
4363 the duration of the SMTP connection.
4365 12. If a transport had a headers_rewrite setting, and a matching header line
4366 contained an unqualified address, that address was qualified, even if it
4367 did not match any rewriting rules. The underlying bug was that the values
4368 of the flags that permit the existence of unqualified sender and recipient
4369 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4370 non-local messages, and by -bnq for local messages) were not being
4371 preserved with the message after it was received.
4373 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4374 "next input=" as part of the text comprising the host identity instead of
4375 the correct text. The code was using the same buffer for two different
4376 strings. However, depending on which order the printing function evaluated
4377 its arguments, the bug did not always show up. Under Linux, for example, my
4378 test suite worked just fine.
4380 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4381 doesn't work with some older versions of Perl. It has been changed to "my",
4382 which in any case is probably the better facility to use.
4384 15. A really picky compiler found some instances of statements for creating
4385 error messages that either had too many or two few arguments for the format
4388 16. The size of the buffer for calls to the DNS resolver has been increased
4389 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4390 for addresses that have a lot of PTR records. This alleviates a problem; it
4391 does not fully solve it.
4393 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4394 buffer now truncates the list and logs the incident, which is the same
4395 action as happens when Exim is looking up a host name and its aliases.
4396 Previously in this situation something unpredictable would happen;
4397 sometimes it was "internal error: store_reset failed".
4399 18. If a server dropped the connection unexpectedly when an Exim client was
4400 using GnuTLS and trying to read a response, the client delivery process
4401 crashed while trying to generate an error log message.
4403 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4404 string, for example:
4406 warn message = H1: something\nH2: something
4408 the text was added as a single header line from Exim's point of view
4409 though it ended up OK in the delivered message. However, searching for the
4410 second and subsequent header lines using $h_h2: did not work. This has been
4411 fixed. Similarly, if a system filter added multiple headers in this way,
4412 the routers could not see them.
4414 20. Expanded the error message when iplsearch is called with an invalid key to
4415 suggest using net-iplsearch in a host list.
4417 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4418 ACLs are no longer actually imposed (and a message to that effect is
4421 22. If a "gecos" field in a passwd entry contained escaped characters, in
4422 particular, if it contained a \" sequence, Exim got it wrong when building
4423 a From: or a Sender: header from that name. A second bug also caused
4424 incorrect handling when an unquoted " was present following a character
4425 that needed quoting.
4427 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4428 was not being matched caselessly.
4430 24. Arranged for all hyphens in the exim.8 source to be escaped with
4433 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4434 properly. Recipient callout cache records were still being keyed to include
4435 the sender, even when use_sender was set false. This led to far more
4436 callouts that were necessary. The sender is no longer included in the key
4437 when use_sender is false.
4439 26. Added "control = submission" modifier to ACLs.
4441 27. Added the ${base62d: operator to decode base 62 numbers.
4443 28. dnsdb lookups can now access SRV records.
4445 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4446 the configuration file.
4448 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4449 (-v) mode. This makes the output for a verbose queue run more intelligible.
4451 31. Added a use_postmaster feature to recipient callouts.
4453 32. Added the $body_zerocount variable, containing the number of binary zero
4454 bytes in the message body.
4456 33. The time of last modification of the "new" subdirectory is now used as the
4457 "mailbox time last read" when there is a quota error for a maildir
4460 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4462 35. Added +ignore_unknown as a special item in host lists.
4464 36. Code for decoding IPv6 addresses in host lists is now included, even if
4465 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4466 address was recognized as an IP address, but was then not correctly decoded
4467 into binary, causing unexpected and incorrect effects when compared with
4474 1. Very minor rewording of debugging text in manualroute to say "list of
4475 hosts" instead of "hostlist".
4477 2. If verify=header_syntax was set, and a header line with an unqualified
4478 address (no domain) and a large number of spaces between the end of the
4479 name and the colon was received, the reception process suffered a buffer
4480 overflow, and (when I tested it) crashed. This was caused by some obsolete
4481 code that should have been removed. The fix is to remove it!
4483 3. When running in the test harness, delay a bit after writing a bounce
4484 message to get a bit more predictability in the log output.
4486 4. Added a call to search_tidyup() just before forking a reception process. In
4487 theory, someone could use a lookup in the expansion of smtp_accept_max_
4488 per_host which, without the tidyup, could leave open a database connection.
4490 5. Added the variables $recipient_data and $sender_data which get set from a
4491 lookup success in an ACL "recipients" or "senders" condition, or a router
4492 "senders" option, similar to $domain_data and $local_part_data.
4494 6. Moved the writing of debug_print from before to after the "senders" test
4497 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4498 problems for message scanning, either using a data ACL, or using
4499 local_scan() because the Received: header was not generated till after they
4500 were called (in order to set the time as the time of reception completion).
4501 I have revised the way this works. The header is now generated after the
4502 body is received, but before the ACL or local_scan() are called. After they
4503 are run, the timestamp in the header is updated.
4509 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4510 before starting a queue runner without re-exec. This happened only when
4511 deliver_drop_privilege was set or when the Exim user was set to root. The
4512 effect of the bug was that timeouts during subsequent deliveries caused
4513 crashes instead of being properly handled. The handler is now left at its
4514 default (and expected) setting.
4516 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4517 message, again when deliver_drop_privilege is set or Exim is run as root.
4518 The bug described in (1) was not present in this case, but the tidying up
4519 of the other signals was missing. I have made the two cases consistent.
4521 3. The ignore_target_hosts setting on a manualroute router was being ignored
4522 for hosts that were looked up using the /MX notation.
4524 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4527 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4528 operated on the sender address. After changing the $sender_address to <>
4529 for the sender address verify, Exim was re-instated it as the original
4530 (before rewriting) address, but remembering that it had rewritten it, so it
4531 wasn't rewriting it again. This bug also had the effect of breaking the
4532 sender address verification caching when the sender address was rewritten.
4534 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4535 This has been changed so that if the ip literal address matches
4536 ignore_target_hosts, the router declines.
4538 7. Added expansion conditions match_domain, match_address, and match_local_
4539 part (NOT match_host).
4541 8. The placeholder for the Received: header didn't have a length field set.
4543 9. Added code to Exim itself and to exim_lock to test for a specific race
4544 condition that could lead to file corruption when using MBX delivery. The
4545 issue is with the lockfile that is created in /tmp. If this file is removed
4546 after a process has opened it but before that process has acquired a lock,
4547 there is the potential for a second process to recreate the file and also
4548 acquire a lock. This could lead to two Exim processes writing to the file
4549 at the same time. The added code performs the same test as UW imapd; it
4550 checks after acquiring the lock that its file descriptor still refers to
4551 the same named file.
4553 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4554 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4557 11. Added the smtp_active_hostname option. If used, this will typically be made
4558 to depend on the incoming interface address. Because $interface_address is
4559 not set up until the daemon has forked a reception process, error responses
4560 that can happen earlier (such as "too many connections") no longer contain
4563 12. If an expansion in a condition on a "warn" statement fails because a lookup
4564 defers, the "warn" statement is abandoned, and the next ACL statement is
4565 processed. Previously this caused the whole ACL to be aborted.
4567 13. Added the iplsearch lookup type.
4569 14. Added ident_timeout as a log selector.
4571 15. Added tls_certificate_verified as a log selector.
4573 16. Added a global option tls_require_ciphers (compare the smtp transport
4574 option of the same name). This controls incoming TLS connections.
4576 17. I finally figured out how to make tls_require_ciphers do a similar thing
4577 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4578 before starting the TLS session.
4580 18. Tabs are now shown as \t in -bP output.
4582 19. If the log selector return_path_on_delivery was set, Exim crashed when
4583 bouncing a message because it had too many Received: header lines.
4585 20. If two routers both had headers_remove settings, and the first one included
4586 a superfluous trailing colon, the final name in the first list and the
4587 first name in the second list were incorrectly joined into one item (with a
4588 colon in the middle).
4594 1. Added -C and -D options to the exinext utility, mainly to make it easier
4595 to include in the automated testing, but these could be helpful when
4596 multiple configurations are in use.
4598 2. The exinext utility was not formatting the output nicely when there was
4599 an alternate port involved in the retry record key, nor when there was a
4600 message id as well (for retries that were specific to a specific message
4601 and a specific host). It was also confused by IPv6 addresses, because of
4602 the additional colons they contain. I have fixed the IPv4 problem, and
4603 patched it up to do a reasonable job for IPv6.
4605 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4606 delivery, the log line now contains "pipelined" if PIPELINING was used.
4608 4. An SMTP transport process used to panic and die if the bind() call to set
4609 an explicit outgoing interface failed. This has been changed; it is now
4610 treated in the same way as a connect() failure.
4612 5. A reference to $sender_host_name in the part of a conditional expansion
4613 that was being skipped was still causing a DNS lookup. This no longer
4616 6. The def: expansion condition was not recognizing references to header lines
4617 that used bh_ and bheader_.
4619 7. Added the _cache feature to named lists.
4621 8. The code for checking quota_filecount in the appendfile transport was
4622 allowing one more file than it should have been.
4624 9. For compatibility with Sendmail, the command line option
4632 and sets the incoming protocol and host name (for trusted callers). The
4633 host name and its colon can be omitted when only the protocol is to be set.
4634 Note the Exim already has two private options, -pd and -ps, that refer to
4635 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4636 "s", but I don't think that's a major issue.
4638 10. A number of refactoring changes to the code, none of which should affect
4641 (a) The number of logging options was getting close to filling up the
4642 32-bit word that was used as a bit map. I have split them into two classes:
4643 those that are passed in the argument to log_write(), and those that are
4644 only ever tested independently outside of that function. These are now in
4645 separate 32-bit words, so there is plenty of room for expansion again.
4646 There is no change in the user interface or the logging behaviour.
4648 (b) When building, for example, log lines, the code previously used a
4649 macro that called string_cat() twice, in order to add two strings. This is
4650 not really sufficiently general. Furthermore, there was one instance where
4651 it was actually wrong because one of the argument was used twice, and in
4652 one call a function was used. (As it happened, calling the function twice
4653 did not affect the overall behaviour.) The macro has been replaced by a
4654 function that can join an arbitrary number of extra strings onto a growing
4657 (c) The code for expansion conditions now uses a table and a binary chop
4658 instead of a serial search (which was left over from when there were very
4659 few conditions). Also, it now recognizes conditions like "pam" even when
4660 the relevant support is not compiled in: a suitably worded error message is
4661 given if an attempt is made to use such a condition.
4663 11. Added ${time_interval:xxxxx}.
4665 12. A bug was causing one of the ddress fields not to be passed back correctly
4666 from remote delivery subprocesses. The field in question was not being
4667 subsequently used, so this caused to problems in practice.
4669 13. Added new log selectors queue_time and deliver_time.
4671 14. Might have fixed a bug in maildirsizefile handling that threw up
4672 "unexpected character" debug warnings, and recalculated the data
4673 unnecessarily. In any case, I expanded the warning message to give more
4676 15. Added the message "Restricted characters in address" to the statements in
4677 the default ACL that block characters like @ and % in local parts.
4679 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4680 Three changes have been made:
4682 (a) There was a serious bug; a negative response to MAIL caused the whole
4683 recipient domain to be cached as invalid, thereby blocking all messages
4684 to all local parts at the same domain, from all senders. This bug has
4685 been fixed. The domain is no longer cached after a negative response to
4686 MAIL if the sender used is not empty.
4688 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4691 (c) A new callout option, "use_sender" has been added for people who want
4692 the modified behaviour.
4698 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4701 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4702 indeed breaks things for older releases.
4704 3. Added additional logging to the case where there is a problem reading data
4705 from a filter that is running in a subprocess using a pipe, in order to
4706 try to track down a specific problem.
4708 4. Testing facility fudge: when running in the test harness and attempting
4709 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4710 getting "No route to host". Convert this to a timeout.
4712 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4715 6. Some OS don't have socklen_t but use size_t instead. This affects the
4716 fifth argument of getsockopt() amongst other things. This is now
4717 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4718 can be set for individual OS. I have set it for SunOS5, OSF1, and
4719 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4720 some earlier ones do not.
4722 7. Change 4.30/15 was not doing the test caselessly.
4724 8. The standard form for an IPv6 address literal was being rejected by address
4725 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4726 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4727 this, as well as the form without the "IPv6" on the front (but only when
4728 address literals are enabled, of course).
4730 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4732 10. Exim crashed if a message with an empty sender address specified by -f
4733 encountered a router with an errors_to setting. This could be provoked only
4734 by a command such as
4738 where an empty string was supplied; "<>" did not hit this bug.
4740 11. Installed PCRE release 4.5.
4742 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4743 remained set. It is now erased.
4745 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4746 times from message ids (which are base 36 rather than the normal 62).
4748 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4749 were being counted as actual protocol errors, and logged if the log
4750 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4751 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4752 and DATA following a set of rejected RCPTs do not count as protocol errors.
4753 In other words, Exim assumes they were pipelined, though this may not
4754 actually be the case. Of course, in all cases the client gets an
4755 appropriate error code.
4757 15. If a lookup fails in an ACL condition, a message about the failure may
4758 be available; it is used if testing the ACL cannot continue, because most
4759 such messages specify what the cause of the deferral is. However, some
4760 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4761 that caused an old message to be retained and used if a later statement
4762 caused a defer, replacing the real cause of the deferral.
4764 16. If an IP address had so many PTR records that the DNS lookup buffer
4765 was not large enough to hold them, Exim could crash while trying to process
4766 the truncated data. It now detects and logs this case.
4768 17. Further to 4.21/58, another change has been made: if (and only if) the
4769 first line of a message (the first header line) ends with CRLF, a bare LF
4770 in a subsequent header line has a space inserted after it, so as not to
4771 terminate the header.
4773 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4774 unnecessarily, used atoi() instead of strtol(), and didn't check the
4775 termination when getting file sizes from file names by regex.
4777 19. Completely re-implemented the support for maildirsize files, in the light
4778 of a number of problems with the previous contributed implementation
4779 (4.30/29). In particular:
4781 . If the quota is zero, the maildirsize file is maintained, but no quota is
4784 . If the maildir directory does not exist, it is created before any attempt
4785 to write a maildirsize file.
4787 . The quota value in the file is just a cache; if the quota is changed in
4788 the transport, the new value overrides.
4790 . A regular expression is available for excluding directories from the
4793 20. The autoreply transport checks the characters in options that define the
4794 message's headers; it allows continued headers, but it was checking with
4795 isspace() after an embedded newline instead of explicitly looking for a
4798 21. If all the "regular" hosts to which an address was routed had passed their
4799 expiry times, and had not reached their retry times, the address was
4800 bounced, even if fallback hosts were defined. Now Exim should go on to try
4803 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4804 equivalent code in the SMTP transport. Some hosts send humungous responses
4805 to HELO/EHLO, more than 1024 it seems.
4807 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4808 gives compiler warnings in some environments. I've now done it "properly",
4811 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4812 (because of problems with the built-in one) was declared to return uschar *
4813 instead of char *, causing compiler failure.
4815 25. Fixed a file descriptor leak when processing alias/forward files.
4817 26. Fixed a minor format string issue in dbfn.c.
4819 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4821 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4822 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4825 29. When a sender address is verified, it is cached, to save repeating the test
4826 when there is more than one recipient in a message. However, when the
4827 verification involves a callout, it is possible for different callout
4828 options to be set for different recipients. It is too complicated to keep
4829 track of this in the cache, so now Exim always runs a verification when a
4830 callout is required, relying on the callout cache for the optimization.
4831 The overhead is duplication of the address routing, but this should not be
4834 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4835 to be verified with callout=postmaster, and the main callout worked but the
4836 postmaster check failed, the verification correctly failed. However, if a
4837 subsequent RCPT command asked for sender verification *without* the
4838 postmaster check, incorrect caching caused this verification also to fail,
4841 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4842 it was not caching the DNS options (qualify_single, search_parents) that
4843 were used when the lookup failed. A subsequent lookup with different
4844 options therefore always gave the same answer, though there were cases
4845 where it should not have. (Example: a "domains = !$mx_any" option on a
4846 dnslookup router: the "domains" option is always processed without any
4847 widening, but the router might have qualify_single set.) Now Exim uses the
4848 cached value only when the same options are set.
4850 32. Added John Jetmore's "exipick" utility to the distribution.
4852 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4853 than a timeout (e.g. a certificate is required, and is not provided), an
4854 Exim server now closes the connection immediately. Previously it waited for
4855 the client to close - but if the client is SSL, it seems that they each
4856 wait for each other, leading to a delay before one of them times out.
4858 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4859 maintained 0.8.x compatibility because I don't think many are using it, and
4860 it is clearly obsolete.
4862 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4865 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4866 client certificate was expired. A simple patch fixes this, though I don't
4867 understand the full logic of why the verify callback is called multiple
4870 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4871 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4872 which causes problems with some clients (such as the Certicom SSL Plus
4873 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4874 disables the coutermeasure allowing Eudora to connect."
4876 38. Exim was not checking that a write() to a log file succeeded. This could
4877 lead to Bad Things if a log got too big, in particular if it hit a file
4878 size limit. Exim now panics and dies if it cannot write to a log file, just
4879 as it does if it cannot open a log file.
4881 39. Modified OS/Makefile-Linux so that it now contains
4883 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4885 The two -D definitions ensure that Exim is compiled with large file
4886 support, which makes it possible to handle log files that are bigger than
4889 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4890 instance) a domain was checked against a named list that involved a lookup,
4891 causing $domain_data to be set, then another domain was checked against the
4892 same list, then the first domain was re-checked, the value of $domain_data
4893 after the final check could be wrong. In particular, if the second check
4894 failed, it could be set empty. This bug probably also applied to
4897 41. The strip_trailing_dot option was not being applied to the address given
4898 with the -f command-line option.
4900 42. The code for reading a message's header from the spool was incrementing
4901 $received_count, but never initializing it. This meant that the value was
4902 incorrect (doubled) while delivering a message in the same process in which
4903 it was received. In the most common configuration of Exim, this never
4904 happens - a fresh exec is done - but it can happen when
4905 deliver_drop_privilege is set.
4907 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4908 it now includes up to 150 characters of the unexpected data in the log
4911 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4912 and building data strings. The size of both of these buffers was 10 000
4913 bytes - far larger than anybody would *ever* want, thought I. Needless to
4914 say, somebody hit the limit. I have increased the maximum line length to
4915 20 000 and the maximum data length of concatenated lines to 100 000. I have
4916 also fixed two bugs, because there was no checking on these buffers. Tsk,
4917 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4918 buffer is too small.
4920 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4921 lsearch lookups. Now it does.
4923 46. When parsing a route_list item in a manualroute router, a fixed-length
4924 buffer was used for the list of hosts. I made this 1024 bytes long,
4925 thinking that nobody would ever have a list of hosts that long. Wrong.
4926 Somebody had a whole pile of complicated expansion conditions, and the
4927 string was silently truncated, leading to an expansion error. It turns out
4928 that it is easier to change to an unlimited length (owing to other changes
4929 that have happened since this code was originally written) than to build
4930 structure for giving a limitation error. The length of the item that
4931 expands into the list of hosts is now unlimited.
4933 47. The lsearch lookup could not handle data where the length of text line was
4934 more than 4095 characters. Such lines were truncated, leading to shortened
4935 data being returned. It should now handle lines of any length.
4937 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4938 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4941 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4942 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4943 no longer happens in comment lines. A list of replacements is now placed
4944 at the head of all of the source files, except those whose only change is
4945 to replace PERL_COMMAND in the very first #! line.
4947 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4948 messages on the queue, with a bottom-up merge sort, using code contributed
4949 by Michael Haardt. This should make operations like -bp somewhat faster on
4950 large queues. It won't affect queue runners, except when queue_run_in_order
4953 51. Installed eximstats 1.31 in the distribution.
4955 52. Added support for SRV lookups to the dnslookup router.
4957 53. If an ACL referred to $message_body or $message_body_end, the value was not
4958 reset for any messages that followed in the same SMTP session.
4960 54. The store-handling optimization for building very long strings was not
4961 differentiating between the different store pools. I don't think this
4962 actually made any difference in practice, but I've tidied it.
4964 55. While running the routers to verify a sender address, $sender_address
4965 was still set to the sender address. This is wrong, because when routing to
4966 send a bounce to the sender, it would be empty. Therefore, I have changed
4967 it so that, while verifying a sender address, $sender_address is set to <>.
4968 (There is no change to what happens when verifying a recipient address.)
4970 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4971 target A or AAAA records (if not already returned) without resetting the
4972 qualify_single or search_parents options of the DNS resolver. These are
4973 inappropriate in this case because the targets of MX and SRV records must
4974 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4975 target that, when qualified, matched something in the local domain. These
4976 two options are now turned off when doing these lookups.
4978 57. It seems that at least some releases of Reiserfs (which does not have the
4979 concept of a fixed number of inodes) returns zero and not -1 for the
4980 number of available inodes. This interacted badly with check_spool_inodes,
4981 which assumed that -1 was the "no such thing" setting. What I have done is
4982 to check that the total number of inodes is greater than zero before doing
4983 the test of how many are available.
4985 58. When a "warn" ACL statement has a log_message modifier, the message is
4986 remembered, and not repeated. This is to avoid a lot of repetition when a
4987 message has many recipients that cause the same warning to be written.
4988 Howewer, Exim was preserving the list of already written lines for an
4989 entire SMTP session, which doesn't seem right. The memory is now reset if a
4990 new message is started.
4992 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4993 debugging output unless log_rewrite was also set.
4995 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4996 of (int)(handle) when we know that handle contains (void *)(-1).
4998 61. The Exim daemon panic-logs an error return when it closes the incoming
4999 connection. However "connection reset by peer" seems to be common, and
5000 isn't really an error worthy of noting specially, so that particular error
5003 62. When Exim is trying to find all the local interfaces, it used to panic and
5004 die if the ioctl to get the interface flags failed. However, it seems that
5005 on at least one OS (Solaris 9) it is possible to have an interface that is
5006 included in the list of interfaces, but for which you get a failure error
5007 for this call. This happens when the interface is not "plumbed" into a
5008 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5009 failure of the "get flags" call assumes that the interface is down.
5011 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5012 makes life easier for people who are doing arithmetic on fields extracted
5013 from dates, where you often get leading zeros that should not be
5014 interpreted as octal.
5016 64. Added qualify_domain to the redirect router, to override the global
5019 65. If a pathologically long header line contained very many addresses (the
5020 report of this problem mentioned 10 000) and each of them was rewritten,
5021 Exim could use up a very large amount of memory. (It kept on making new
5022 copies of the header line as it rewrote, and never released the old ones.)
5023 At the expense of a bit more processing, the header rewriting function has
5024 been changed so that it no longer eats memory in this way.
5026 66. The generation of the Received: header has been moved from the time that a
5027 message starts to be received, to the time that it finishes. The timestamp
5028 in the Received: header should now be very close to that of the <= log
5029 line. There are two side-effects of this change:
5031 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5032 logged header lines no longer include the local Received: line, because
5033 it has not yet been created. The same applies to a copy of the message
5034 that is returned to a non-SMTP sender when a message is rejected.
5036 (b) When a filter file is tested using -bf, no additional Received: header
5037 is added to the test message. After some thought, I decided that this
5040 This change does not affect the value of $received_for. It is still set
5041 after address rewriting, but before local_scan() is called.
5043 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5045 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5046 gave an unhelpful panic error message, and a defer error. I have managed to
5047 change this behaviour so that it now rejects any supplied certificate,
5048 which seems right, as the list of acceptable certificates is empty.
5050 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5051 gave an unhelpful defer error. I have not managed to make this reject any
5052 supplied certificates, but the error message it gives is "no certificate
5053 supplied", which is not helpful.
5055 70. exigrep's output now also includes lines that are not associated with any
5056 message, but which match the given pattern. Implemented by a patch from
5057 Martin Sluka, which also tidied up the Perl a bit.
5059 71. Recipient callout verification, like sender verification, was using <> in
5060 the MAIL FROM command. This isn't really the right thing, since the actual
5061 sender may affect whether the remote host accepts the recipient or not. I
5062 have changed it to use the actual sender in the callout; this means that
5063 the cache record is now keyed on a recipient/sender pair, not just the
5064 recipient address. There doesn't seem to be a real danger of callout loops,
5065 since a callout by the remote host to check the sender would use <>.
5066 [SEE ABOVE: changed after hitting problems.]
5068 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5069 temporary errors. However, in the case of such a code being given after
5070 the end of a data transmission (i.e. after ".") Exim was failing to write
5071 a retry record for the message. (Yes, there was some broken host that was
5072 actually sending 8xx at this point.)
5074 73. An unknown lookup type in a host list could cause Exim to panic-die when
5075 the list was checked. (An example that provoked this was putting <; in the
5076 middle of a list instead of at the start.) If this happened during a DATA
5077 ACL check, a -D file could be left lying around. This kind of configuration
5078 error no longer causes Exim to die; instead it causes a defer errror. The
5079 incident is still logged to the main and panic logs.
5081 74. Buglet left over from Exim 3 conversion. The message "too many messages
5082 in one connection" was written to the rejectlog but not the mainlog, except
5083 when address rewriting (yes!) was being logged.
5085 75. Added write_rejectlog option.
5087 76. When a system filter was run not as root (that is, when system_filter_user
5088 was set), the values of the $n variables were not being returned to the
5089 main process; thus, they were not subsequently available in the $sn
5092 77. Added +return_path_on_delivery log selector.
5094 78. A connection timeout was being treated differently from recipients deferred
5095 when testing hosts_max_try with a message that was older than the host's
5096 retry timeout. (The host should not be counted, thus allowing all hosts to
5097 be tried at least once before bouncing.) This may have been the cause of an
5098 occasionally reported bug whereby a message would remain on the queue
5099 longer than the retry timeout, but would be bounced if a delivery was
5100 forced. I say "may" because I never totally pinned down the problem;
5101 setting up timeout/retry tests is difficult. See also the next item.
5103 79. The ultimate address timeout was not being applied to errors that involved
5104 a combination of host plus message (for example, a timeout on a MAIL
5105 command). When an address resolved to a number of possible hosts, and they
5106 were not all tried for each delivery (e.g. because of hosts_max_try), a
5107 message could remain on the queue longer than the retry timeout.
5109 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5112 81. Fixed an obscure SMTP outgoing bug which required at least the following
5113 conditions: (a) there was another message waiting for the same server;
5114 (b) the server returned 5xx to all RCPT commands in the first message so
5115 that the message was not completed; (c) the server dropped the connection
5116 or gave a negative response to the RSET that Exim sends to abort the
5117 transaction. The observed case was a dropped connection after DATA that had
5118 been sent in pipelining mode. That is, the server had advertised PIPELINING
5119 but was not implementing it correctly. The effect of the bug was incorrect
5120 behaviour, such as trying another host, and this could lead to a crash.
5126 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5127 and daemon.c were passed as pointers to ints; they should have been
5128 pointers to socklen_t variables (which are typically unsigned ints).
5130 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5133 3. Fixed a really odd bug that affected only the testing scheme; patching a
5134 certain fixed string in the binary changed the value of another string that
5135 happened to be identical to the end of the original first string.
5137 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5138 name", it returns that address as the IP address. On some operating
5139 systems (e.g. Solaris), it also passes back the IP address string as the
5140 "host name". However, on others (e.g. Linux), it passes back an empty
5141 string. Exim wasn't checking for this, and was changing the host name to an
5142 empty string, assuming it had been canonicized.
5144 5. Although rare, it is permitted to have more than one PTR record for a given
5145 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5146 all the names associated with an address, because they do in Solaris.
5147 However, it seems that they do not in Linux for data that comes from the
5148 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5149 I found this out when I moved to a new Linux workstation and tried to run
5150 the Exim test suite.
5152 To get round this problem I have changed the code so that it now does its
5153 own call to the DNS to look up PTR records when searching for a host name.
5154 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5155 addresses that are only in /etc/hosts are still found.
5157 This behaviour is, however, controlled by an option called host_lookup_
5158 order, which defaults to "bydns:byaddr". If people want to use the other
5159 order, or indeed, just use one or the other means of lookup, they can
5160 specify it in this variable.
5162 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5163 some operating systems, this comes back from gethostbyaddr() as an empty
5164 string, and this is what Exim used to test for. However, it seems that in
5165 other systems, "." is yielded. Exim now tests for this case too.
5167 7. The values of check_spool_space and check_log_space are now held internally
5168 as a number of kilobytes instead of an absolute number of bytes. If a
5169 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5170 kilobyte. This means that much larger values can be stored.
5172 8. Exim monitor: an attempt to get the action menu when not actually pointing
5173 at a message produces an empty menu entitled "No message selected". This
5174 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5175 no entries in it ("Shell widget menu has zero width and/or height"). So I
5176 have added a single, blank menu entry in this case.
5178 9. Added ${quote_local_part.
5180 10. MIME decoding is now applied to the contents of Subject: header lines when
5183 11. Now that a reference to $sender_host_address automatically causes a reverse
5184 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5185 host lookup before query-style lookups in lists that might use this
5186 variable. This has therefore been abolished, and the "net-" prefix is no
5187 longer necessary for query-style lookups.
5189 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5190 have been a typo for LFLAGS, so it has been changed.
5192 13. The install script calls Exim with "-C /dev/null" in order to find the
5193 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5194 to be output. Howeve, since Exim outputs its version number before the
5195 error, it didn't break the script. It just looked ugly. I fixed this by
5196 always allowing "-C /dev/null" if the caller is root.
5198 14. Ignore overlarge ACL variable number when reading spool file - insurance
5199 against a later release with more variables having written the file.
5201 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5202 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5203 this, as well as the form without the "IPv6" on the front.
5205 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5206 OS/Makefile-Darwin file.
5208 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5209 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5211 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5212 variable. However, some OS use ints for inodes. I've added cast to long int
5213 to get rid of the compiler warning.
5215 19. I had forgotten to lock out "/../" in configuration file names when
5216 ALT_CONFIG_PREFIX was set.
5218 20. Routers used for verification do not need to specify transports. However,
5219 if such a router generated a host list, and callout was configured, Exim
5220 crashed, because it could not find a port number from the (non-existent)
5221 transport. It now assumes port 25 in this circumstance.
5223 21. Added the -t option to exigrep.
5225 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5226 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5227 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5228 them is set, LOOKUP_LSEARCH is forced.
5230 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5232 24. Added sender and host information to the "rejected by local_scan()" log
5233 line; previously there was no indication of these.
5235 25. Added .include_if_exists.
5237 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5238 Linux. It turns out that not all file systems support this. Apparently some
5239 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5240 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5241 not supported on the file descriptor, is now ignored when Exim is trying to
5242 sync a directory. This applies only to Linux.
5244 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5246 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5247 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5250 29. Added support for maildirsize files from supplied patch (modified a bit).
5252 30. The use of :fail: followed by an empty string could lead Exim to respond to
5253 sender verification failures with (e.g.):
5255 550 Verification failed for <xxx>
5256 550 Sender verify failed
5258 where the first response line was missing the '-' that indicates it is not
5259 the final line of the response.
5261 31. The loop for finding the name of the user that called Exim had a hardwired
5262 limit of 10; it now uses the value of finduser_retries, which is used for
5263 all other user lookups.
5265 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5268 33. Exim was neglecting to zero errno before one call of strtol() when
5269 expanding a string and expecting an integer value. On some systems this
5270 resulted in spurious "integer overflow" errors. Also, it was casting the
5271 result into an int without checking.
5273 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5274 did not work. The code looks as if it has *never* worked, though it appears
5275 to have been documented since at least releast 1.62. I have made it work.
5277 35. The "timeout_DNS" error in retry rules, also documented since at least
5278 1.62, also never worked. As it isn't clear exactly what this means, and
5279 clearly it isn't a major issue, I have abolished the feature by treating it
5280 as "timeout", and writing a warning to the main and panic logs.
5282 36. The display of retry rules for -brt wasn't always showing the error code
5285 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5286 timeout_connect_A, timeout_connect_MX.
5288 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5289 to the empty sender.
5291 39. The daemon was not analysing the content of -oX till after it had closed
5292 stderr and disconnected from the controlling terminal. This meant that any
5293 syntax errors were only noted on the panic log, and the return code from
5294 the command was 0. By re-arranging the code a little, I've made the
5295 decoding happen first, so such errors now appear on stderr, and the return
5296 code is 1. However, the actual setting up of the sockets still happens in
5297 the disconnected process, so errors there are still only recorded on the
5300 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5301 connections (as happens on some IP stacks) was logged at start up time as
5302 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5303 it from "IPv6 and IPv4", which means that two separate sockets are being
5306 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5307 says whether AF_INET or AF_INET6 was passed as an argument.
5309 42. Exiwhat output was messed up when time zones were included in log
5312 43. Exiwhat now gives more information about the daemon's listening ports,
5313 and whether -tls-on-connect was used.
5315 44. The "port" option of the smtp transport is now expanded.
5317 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5318 silently ignored. Now an error message is written to the main and panic
5321 46. There's a new ACL modifier called "logwrite" which writes to a log file
5322 as soon as it is encountered.
5324 47. Added $local_user_uid and $local_user_gid at routing time.
5326 48. Exim crashed when trying to verify a sender address that was being
5329 49. Exim was recognizing only a space character after ".include". It now also
5330 recognizes a tab character.
5332 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5333 extracting the relevant information from the specification. The man page no
5334 longer contains scrambled data for the -d option, and I've added a section
5335 at the front about calling Exim under different names.
5337 51. Added "extra_headers" argument to the "mail" command in filter files.
5339 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5342 53. Installed eximstats 1.29.
5344 54. Added transport_filter_timeout as a generic transport option.
5346 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5347 Cc: header lines. This was required by RFC 822, but it not required by RFC
5350 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5351 incoming messages that did not have them. Now it does so only if the
5352 message originates locally, that is, if there is no associated remote host
5353 address. When Resent- header lines are present, this applies to the Resent-
5354 lines rather than the non-Resent- lines.
5356 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5357 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5359 58. Messages for configuration errors now include the name of the main
5360 configuration file - useful now that there may be more than one file in a
5361 list (.included file names were always shown).
5363 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5364 for those rare installations that do not start the daemon as root or run it
5365 setuid root. I've cut out the call to initgroups() if the daemon is not
5368 60. The Exim user and group can now be bound into the binary as text strings
5369 that are looked up at the start of Exim's processing.
5371 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5373 62. Added $mailstore_basename variable.
5375 63. Installed patch to sieve.c from Michael Haardt.
5377 64. When Exim failed to open the panic log after failing to open the main log,
5378 the original message it was trying to log was written to stderr and debug
5379 output, but if they were not available (the usual case in production), it
5380 was lost. Now it is written to syslog before the two lines that record the
5381 failures to open the logs.
5383 65. Users' Exim filters run in subprocesses under the user's uid. It is
5384 possible for a "deliver" command or an alias in a "personal" command to
5385 provoke an address rewrite. If logging of address rewriting is configured,
5386 this fails because the process is not running as root or exim. There may be
5387 a better way of dealing with this, but for the moment (because 4.30 needs
5388 to be released), I have disabled address rewrite logging when running a
5389 filter in a non-root, non-exim process.
5395 1. The buildconfig auxiliary program wasn't quoting the value set for
5396 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5397 not defined. This bug was masked in 4.22 by the effect that was fixed in
5400 2. Some messages that were rejected after a message id was allocated were
5401 shown as "incomplete" by exigrep. It no longer does this for messages that
5402 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5404 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5405 have allow_domain_literals set, the ID did not get logged in the <= line.
5406 Domain literals are now always recognized in Message-ID: header lines.
5408 4. The first argument for a ${extract expansion item is the key name or field
5409 number. Leading and trailing spaces in this item were not being ignored,
5410 causing some misleading effects.
5412 5. When deliver_drop_privilege was set, single queue runner processes started
5413 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5414 same command in the process it spins off) were not dropping privilege.
5416 6. When the daemon running as "exim" started a queue runner, it always
5417 re-executed Exim in the spun-off process. This is a waste of effort when
5418 deliver_drop_privilege is set. The new process now just calls the
5419 queue-runner function directly.
5425 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5428 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5429 ignored. Though the use of -oP was forcing the writing of a pid file, it
5430 was always written to the default place.
5432 3. If the message "no IP address found for host xxxx" is generated during
5433 incoming verification, it is now followed by identification of the incoming
5434 connection (so you can more easily find what provoked it).
5436 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5438 5. Added some features to "harden" Exim a bit more against certain attacks:
5440 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5441 be put in Local/Makefile. This is like the never_users runtime option,
5442 but it cannot be overridden. The default setting is "root".
5444 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5445 prefix string with which any file named in a -C command line option
5448 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5449 is retained for -C and -D only if the caller of Exim is root. Without
5450 it, the exim user may also use -C and -D and retain privilege.
5452 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5453 command line option is disabled.
5455 6. Macro names set by the -D option must start with an upper case letter, just
5456 like macro names defined in the configuration file.
5458 7. Added "dereference=" facility to LDAP.
5460 8. Two instances of the typo "uknown" in the source files are fixed.
5462 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5463 the Configure-Makefile script screwed up while processing it.
5465 10. Incorporated PCRE 4.4.
5467 11. The SMTP synchronization check was not operating right at the start of an
5468 SMTP session. For example, it could not catch a HELO sent before the client
5469 waited for the greeting. There is now a check for outstanding input at the
5470 point when the greeting is written. Because of the duplex, asynchronous
5471 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5472 way, but not yet received, when the check is performed.
5474 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5475 on TCP/IP sockets, because this apparently causes some broken clients to
5478 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5479 unchanged) from the Cygwin maintainer.
5481 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5482 was supported instead of testing for mbx. Effectively a typo.
5484 15. The spa authenticator server code was not checking that the input it
5485 received was valid base64.
5487 16. The debug output line for the "set" modifier in ACLs was not showing the
5488 name of the variable that was being set.
5490 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5492 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5493 lookup on its own. Something else was needed to trigger the lookup. For
5494 example, a match in host_lookup or the need for a host name in a host list.
5495 Now, if $sender_host_name is referenced and the host name has not yet been
5496 looked up, a lookup is performed. If the lookup fails, the variable remains
5497 empty, and $host_lookup_failed is set to "1".
5499 19. Added "eqi" as a case-independent comparison operator.
5501 20. The saslauthd authentication condition could segfault if neither service
5502 nor realm was specified.
5504 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5505 error message that was logged was misleading, and incoming SMTP
5506 connections were dropped. The message is now more accurate, and temporary
5507 errors are given to SMTP connections.
5509 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5510 (except RSET) with a 421 error, until QUIT is received. However, it was
5511 failing to send a response to QUIT.
5513 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5514 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5515 ACL is now run after the helo_try_verify_hosts code.
5517 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5518 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5519 case-independent, so other case variants are also recognized). Apparently
5520 some systems use these upper case variants.
5522 25. If more than two messages were waiting for the same host, and a transport
5523 filter was specified for the transport, Exim sent two messages over the
5524 same TCP/IP connection, and then failed with "socket operation on non-
5525 socket" when it tried to send the third.
5527 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5529 27. The extern definition of crypt16() in expand.c was not being excluded when
5530 the OS had its own crypt16() function.
5532 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5533 as a preferred synonym for return_size_limit, both as an option and as an
5536 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5538 30. Changed the default configuration ACL to relax the local part checking rule
5539 for addresses that are not in any local domains. For these addresses,
5540 slashes and pipe symbols are allowed within local parts, but the sequence
5541 /../ is explicitly forbidden.
5543 31. SPA server authentication was not clearing the challenge buffer before
5546 32. log_message in a "warn" ACL statement was writing to the reject log as
5547 well as to the main log, which contradicts the documentation and doesn't
5548 seem right (because no rejection is happening). So I have stopped it.
5550 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5551 However, I am unable to do any testing of this.
5553 34. Fixed an infelicity in the appendfile transport. When checking directories
5554 for a mailbox, to see if any needed to be created, it was accidentally
5555 using path names with one or more superfluous leading slashes; tracing
5556 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5558 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5559 contents are added to the log line that is written for every discarded
5560 recipient. (Previously a log_message setting was ignored.)
5562 36. The ${quote: operator now quotes the string if it is empty.
5564 37. The install script runs exim in order to find its version number. If for
5565 some reason other than non-existence or emptiness, which it checks, it
5566 could not run './exim', it was installing it with an empty version number,
5567 i.e. as "exim-". This error state is now caught, and the installation is
5570 38. An argument was missing from the function that creates an error message
5571 when Exim fails to connect to the socket for saslauthd authentication.
5572 This could cause Exim to crash, or give a corrupted message.
5574 39. Added isip, isip4, and isip6 to ${if conditions.
5576 40. The ACL variables $acl_xx are now saved with the message, and can be
5577 accessed later in routers, transports, and filters.
5579 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5580 strings in the file are not string-expanded.
5582 42. If a MAIL command specified a SIZE value that was too large to fit into an
5583 int variable, the check against message_size_limit failed. Such values are
5584 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5585 day this will have to be increased, but I don't think I want to be around
5586 when emails are that large.
5593 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5594 iconv() is not standard in FreeBSD.
5596 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5597 IPv6 enabled. The observed symptom was a segmentation fault on return from
5598 the function os_common_find_running_interfaces() in src/os.c.
5600 3. In the check_special_case() function in daemon.c I had used "errno" as an
5601 argument name, which causes warnings on some systems. This was basically a
5602 typo, since it was named "eno" in the comments!
5604 4. The code that waits for the clock to tick (at a resolution of some fraction
5605 of a second) so as to ensure message-id uniqueness was always waiting for
5606 at least one whole tick, when it could have waited for less. [This is
5607 almost certainly not relevant at current processor speeds, where it is
5608 unlikely to ever wait at all. But we try to future-proof.]
5610 5. The function that sleeps for a time interval that includes fractions of a
5611 second contained a race. It did not block SIGALRM between setting the
5612 timer, and suspending (a couple of lines later). If the interval was short
5613 and the sigsuspend() was delayed until after it had expired, the suspension
5614 never ended. On busy systems this could lead to processes getting stuck for
5617 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5618 process, before it forks any delivery processes. The open lookup caching
5619 mechanism meant that the open file or database connection was passed into
5620 the delivery process. The problem was that delivery processes always tidy
5621 up cached lookup data. This could cause a problem for the next delivery
5622 process started by the queue runner, because the external queue runner
5623 process does not know about the closure. So the next delivery process
5624 still has data in the lookup cache. In the case of a file lookup, there was
5625 no problem because closing a file descriptor in a subprocess doesn't affect
5626 the parent. However, if the lookup was caching a connection to a database,
5627 the connection was closed, and the second delivery process was likely to
5628 see errors such as "PGSQL: query failed: server closed the connection
5629 unexpectedly". The problem has been fixed by closing all cached lookups
5630 in a queue runner before running a delivery process.
5632 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5633 seem to have the "const" qualifier which it has on other OS. I've
5636 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5637 *of the same type* (client or server) with the same public name that an
5638 error should be diagnosed.
5640 9. When Exim looked up a host name for an IP address, but failed to find the
5641 original IP address when looking up the host name (a safety check), it
5642 output the message "<ip address> does not match any IP for NULL", which was
5643 confusing, to say the least. The bug was that the host name should have
5644 appeared instead of "NULL".
5646 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5647 user that is built into the binary, and the -C or -D options is used, root
5648 privilege is dropped before the configuration file is read. In addition,
5649 logging is switched to stderr instead of the normal log files. If the
5650 configuration then re-defines the Exim user, the unprivileged environment
5651 is probably not what is expected, so Exim logs a panic warning message (but
5654 However, if deliver_drop_privilege is set, the unprivileged state may well
5655 be exactly what is intended, so the warning has been cut out in that case,
5656 and Exim is allowed to try to write to its normal log files.
5662 1. smtp_return_error_details was not giving details for temporary sender
5663 or receiver verification errors.
5665 2. Diagnose a configuration error if two authenticators have the same public
5668 3. Exim used not to create the message log file for a message until the first
5669 delivery attempt. This could be confusing when incoming messages were held
5670 for policy or load reasons. The message log file is now created at the time
5671 the message is received, and an initial "Received" line is written to it.
5673 4. The automatically generated man page for command line options had a minor
5674 bug that caused no ill effects; however, a more serious problem was that
5675 the procedure for building the man page automatically didn't always
5676 operate. Consequently, release 4.20 contains an out-of-date version. This
5677 shouldn't happen again.
5679 5. When building Exim with embedded Perl support, the script that builds the
5680 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5681 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5683 6. The freeze_tell option was not being used for messages that were frozen on
5684 arrival, either by an ACL or by local_scan().
5686 7. Added the smtp_incomplete_transaction log selector.
5688 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5689 was accepting AUTH without a new EHLO.
5691 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5692 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5695 10. Logging of TCP/IP connections (when configured) now happens in the main
5696 daemon process instead of the child process, so that the TCP/IP connection
5697 count is more accurate (but it can never be perfect).
5699 11. The use of "drop" in a nested ACL was not being handled correctly in the
5700 outer ACL. Now, if condition failure induced by the nested "drop" causes
5701 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5702 or "require"), the connection is dropped.
5704 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5705 that yield "discard" can now be used with an "accept" or a "discard" verb,
5706 but an error is generated for any others (because I can't see a useful way
5707 to define what should happen).
5709 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5710 are now processed in the same way as lines in the Exim configuration file.
5711 In particular, continuation lines are supported.
5713 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5715 15. Added -ti meaning -t -i.
5717 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5718 domains, and warn by logging if others are found.
5720 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5721 returned from GIFCONF when Exim is trying to find the list of interfaces on
5722 a host. The code in os.c has been modified to copy each ifreq to an aligned
5723 structure in all cases.
5725 Also, in some cases, the returned ifreq's were being copied to a 'struct
5726 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5727 means the last couple of bytes of an IPv6 address could be chopped if the
5728 ifreq contained only a normal sockaddr (14 bytes storage).
5730 18. Named domain lists were not supported in the hosts_treat_as_local option.
5731 An entry such as +xxxx was not recognized, and was treated as a literal
5734 19. Ensure that header lines added by a DATA ACL are included in the reject log
5735 if the ACL subsequently rejects the message.
5737 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5738 MD5 (which is deprecated).
5740 21. When testing a filter file using -bf, Exim was writing a message when it
5741 took the sender from a "From " line in the message, but it was not doing so
5742 when it took $return_path from a Return-Path: header line. It now does.
5744 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5745 with a valid header line field name (a series of printing characters
5746 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5748 23. Changed "disc" in the source to "disk" to conform to the documentation and
5749 the book and for uniformity.
5751 24. Ignore Sendmail's -Ooption=value command line item.
5753 25. When execve() failed while trying to run a command in a pipe transport,
5754 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5755 could be confused with a return value of 69 from the command itself. This
5756 has been changed to 127, the value the shell returns if it is asked to run
5757 a non-existent command. The wording for the related log line suggests a
5758 non-existent command as the problem.
5760 26. If received_header_text expands to an empty string, do not add a Received:
5761 header line to the message. (Well, it adds a token one on the spool, but
5762 marks it "old" so that it doesn't get used or transmitted.)
5764 27. Installed eximstats 1.28 (addition of -nt option).
5766 28. There was no check for failure on the call to getsockname() in the daemon
5767 code. This can fail if there is a shortage of resources on the system, with
5768 ENOMEM, for example. A temporary error is now given on failure.
5770 29. Contrary to the C standard, it seems that in some environments, the
5771 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5772 program. Exim now does this explicitly; it affects the formatting of
5773 timestamps using strftime().
5775 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5776 complaints. I've now initialized all the variables, to avoid this.
5778 32. Header lines added by a system filter were not being "seen" during
5779 transport-time rewrites.
5781 33. The info_callback() function passed to OpenSSL is set up with type void
5782 (*)(SSL *, int, int), as described somewhere. However, when calling the
5783 function (actually a macro) that sets it up, the type void(*)() is
5784 expected. I've put in a cast to prevent warnings from picky compilers.
5786 34. If a DNS black list lookup found a CNAME record, but there were no A
5787 records associated with the domain it pointed at, Exim crashed.
5789 35. If a DNS black list lookup returned more than one A record, Exim ignored
5790 all but the first. It now scans all returned addresses if a particular IP
5791 value is being sought. In this situation, the contents of the
5792 $dnslist_value variable are a list of all the addresses, separated by a
5795 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5796 to accept a host name and all its aliases if the forward lookup for any of
5797 them yielded the IP address of the incoming connection. Now it accepts only
5798 those names whose forward lookup yields the correct IP address. Any other
5799 names are discarded. This closes a loophole whereby a rogue DNS
5800 administrator could create reverse DNS records to break through a
5801 wildcarded host restriction in an ACL.
5803 37. If a user filter or a system filter that ran in a subprocess used any of
5804 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5805 the wrong values were passed to the pipe command ($thisaddress had the
5806 value of $0, $0 had the value of $1, etc). This bug was introduced by
5807 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5809 38. Improved the line breaking for long SMTP error messages from ACLs.
5810 Previously, if there was no break point between 40 and 75 characters, Exim
5811 left the rest of the message alone. Two changes have been made: (a) I've
5812 reduced the minimum length to 35 characters; (b) if it can't find a break
5813 point between 35 and 75 characters, it looks ahead and uses the first one
5814 that it finds. This may give the occasional overlong line, but at least the
5815 remaining text gets split now.
5817 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5818 file descriptors might be low, and that setting 1000 would always raise it.
5819 It turns out that in some environments, the limit is already over 1000 and
5820 that lowering it causes trouble. So now Exim takes care not to decrease it.
5822 40. When delivering a message, the value of $return_path is set to $sender_
5823 address at the start of routing (routers may change the value). By an
5824 oversight, this default was not being set up when an address was tested by
5825 -bt or -bv, which affected the outcome if any router or filter referred to
5828 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5829 log only when "message" or "log_message" are set. However, if one of the
5830 conditions was an address verification, or a call to a nested ACL, the
5831 messages generated by the underlying test were being passed through. This
5832 no longer happens. The underlying message is available in $acl_verify_
5833 message for both "message" and "log_message" expansions, so it can be
5834 passed through if needed.
5836 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5837 new expansion $bh_ to give the encoded byte string without charset
5838 translation. Translation happens only if iconv() is available; HAVE_ICONV
5839 indicates this at build time. HEADERS_CHARSET gives the charset to
5840 translate to; headers_charset can change it in the configuration, and
5841 "headers charset" can change it in an individual filter file.
5843 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5844 that creates RFC 2047 encoded "words" labels them as that charset instead
5845 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5846 expansion operator; (ii) when Exim creates a From: line for a local
5847 message; (iii) when a header line is rewritten to include a "phrase" part.
5849 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5850 buggy, causing it to skip the first lines of messages whose message ID
5851 ended in 'D'. This would not have bitten before Exim release 4.14, because
5852 message IDs were unlikely to end in 'D' before then. The effect was to have
5853 incorrect size information for certain domains.
5855 45. #include "config.h" was missing at the start of the crypt16.c module. This
5856 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5859 46. If there was a timeout during a "random" callout check, Exim treated it as
5860 a failure of the random address, and carried on sending RSET and the real
5861 address. If the delay was just some slowness somewhere, the response to the
5862 original RCPT would be taken as a response to RSET and so on, causing
5863 mayhem of various kinds.
5865 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5866 when I implemented it. It didn't allow for the fact that some option values
5867 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5868 the right test for positive values.
5870 48. Domain names in DNS records are case-independent. Exim always looks them up
5871 in lower case. Some resolvers return domain names in exactly the case they
5872 appear in the zone file, that is, they may contain uppercase letters. Not
5873 all resolvers do this - some return always lower case. Exim was treating a
5874 change of case by a resolver as a change of domain, similar to a widening
5875 of a domain abbreviation. This triggered its re-routing code and so it was
5876 trying to route what was effectively the same domain again. This normally
5877 caused routing to fail (because the router wouldn't handle the domain
5878 twice). Now Exim checks for this case specially, and just changes the
5879 casing of the domain that it ultimately uses when it transmits the message
5882 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5885 50. If a filter generated a file delivery with a non-absolute name (possible if
5886 no home directory exists for the router), the forbid_file option was not
5889 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5890 the existing equality matching.
5892 52. Exim was using ints instead of ino_t variables in some places where it was
5893 dealing with inode numbers.
5895 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5896 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5897 called TMPDIR, and if it finds it is different, it changes its value.
5899 54. The smtp_printf() function is now made available to local_scan() so
5900 additional output lines can be written before returning. There is also an
5901 smtp_fflush() function to enable the detection of a dropped connection.
5902 The variables smtp_input and smtp_batched_input are exported to
5905 55. Changed the default runtime configuration: the message "Unknown user"
5906 has been removed from the ACL, and instead placed on the localuser router,
5907 using the cannot_route_message feature. This means that any verification
5908 failures that generate their own messages won't get overridden. Similarly,
5909 the "Unrouteable address" message that was in the ACL for unverifiable
5910 relay addresses has also been removed.
5912 56. Added hosts_avoid_esmtp to the smtp transport.
5914 57. The exicyclog script was not checking for the esoteric option
5915 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5916 will work only if exicyclog is run under the appropriate euid.
5918 58. Following a discussion on the list, the rules by which Exim recognises line
5919 endings on incoming messages have been changed. The -dropcr and drop_cr
5920 options are now no-ops, retained only for backwards compatibility. The
5921 following line terminators are recognized: LF CRLF CR. However, special
5922 processing applies to CR:
5924 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5925 nor a local message in the state where . is a terminator.
5927 (ii) If a bare CR is encountered in a header line, an extra space is added
5928 after the line terminator so as not to end the header. The reasoning
5929 behind this is that bare CRs in header lines are most likely either
5930 to be mistakes, or people trying to play silly games.
5932 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5933 was being incorrectly given as 18 bytes larger than it should have been.
5934 This is a VOB (very old bug).
5936 60. This may never have affected anything current, but just in case it has:
5937 When the local host is found other than at the start of a list of hosts,
5938 the local host, those with the same MX, and any that follow, are discarded.
5939 When the list in question was part of a longer list of hosts, the following
5940 hosts (not currently being processed) were also being discarded. This no
5941 longer happens. I'm not sure if this situation could ever has previously
5944 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5947 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5948 header. This is something that is recommended in a new Internet Draft, and
5949 is something that is documented as being done by Sendmail. There are two
5950 possible values. For messages generated by the autoreply transport, Exim
5953 Auto-Submitted: auto-replied
5955 whereas for all other generated messages (e.g. bounces) it adds
5957 Auto-Submitted: auto-generated
5959 63. The "personal" condition in filters now includes a test for the
5960 Auto-Submitted: header. If it contains the string "auto-" the message it
5961 not considered personal.
5963 64. Added rcpt_include_affixes as a generic transport option.
5965 65. Added queue_only_override (default true).
5967 66. Added the syslog_duplication option.
5969 67. If what should have been the first header line of a message consisted of
5970 a space followed by a colon, Exim was mis-interpreting it as a header line.
5971 It isn't of course - it is syntactically invalid and should therefore be
5972 treated as the start of the message body. The misbehaviour could have
5973 caused a number of strange effects, including loss of data in subsequent
5974 header lines, and spool format errors.
5976 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5977 client host had authenticated. This control can now be exercised by an ACL
5978 for more flexibility.
5980 69. By default, callouts do not happen when testing with -bh. There is now a
5981 variant, -bhc, which does actually run the callout code, including
5982 consulting and updating the callout cache.
5984 70. Added support for saslauthd authentication, courtesy of Alexander
5987 71. If statvfs() failed on the spool or log directories while checking their
5988 size for availability, Exim confusingly gave the error "space shortage".
5989 Furthermore, in debugging mode it crashed with a floating point exception.
5990 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5991 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5992 serious problem, Exim now writes to the main and panic logs when this
5993 happens, with details of the failure. It then refuses to accept the
5994 incoming message, giving the message "spool directory problem" or "log
5995 directory problem" with a 421 code for SMTP messages.
5997 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5998 0, 1, and 2 exist, because some OS complain for execs without them (see
5999 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6000 descriptors. However, the code omitted to check that the open succeeded,
6001 causing mysterious errors if for some reason the permissions on /dev/null
6002 got screwed. Now Exim writes a message to the main and panic logs, and
6003 bombs out if it can't open /dev/null.
6005 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6006 interact so that it is all more flexible. It is supposed to remain
6007 backwards compatible. Also added extra_local_interfaces.
6009 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6010 to bomb out with an assertion failure - to the client this appears as a
6011 connection drop. This problem occurs in the part of the code that was taken
6012 from the Samba project. Fortunately, the assertion is in a very simple
6013 function, so I have fixed this by reproducing the function inline in the
6014 one place where it is called, and arranging for authentication to fail
6015 instead of killing the process with assert().
6017 75. The SPA client code was not working when the server requested OEM rather
6018 than Unicode encoding.
6020 76. Added code to make require_files with a specific uid setting more usable in
6021 the case where statting the file as root fails - usually a non-root-mounted
6022 NFS file system. When this happens and the failure is EACCES, Exim now
6023 forks a subprocess and does the per-uid checking as the relevant uid.
6025 77. Added process_log_path.
6027 78. If log_file_path was not explicitly set, a setting of check_log_space or
6028 check_log_inodes was ignored.
6030 79. If a space check for the spool or log partitions fails, the incident is now
6031 logged. Of course, in the latter case the data may get lost...
6033 80. Added the %p formatting code to string_format() so that it can be used to
6034 print addresses in debug_print(). Adjusted all the address printing in the
6035 debugging in store.c to use %p rather than %d.
6037 81. There was a concern that a line of code in smtp_in.c could overflow a
6038 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6039 initially expressed, the concern was not well-founded, because trailing
6040 spaces are removed early. However, if the trailing spaces were followed by
6041 a NULL, they did not get removed, so the overflow was possible. Two fixes
6044 (a) I re-wrote the offending code in a cleaner fashion.
6045 (b) If an incoming SMTP command contains a NULL character, it is rejected
6048 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6049 runs initgroups(), so that if the Exim user is in any additional groups,
6050 they will be used during message reception.
6056 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob