1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
15 TL/02 The BSD's have an arc4random API. One of the functions to induce
16 adding randomness was arc4random_stir(), but it has been removed in
17 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
18 function when detected.
20 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
21 cause callback expansion.
23 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
24 syntax errors in an expansion can be treated as a string instead of
25 logging or causing an error, due to the internal use of bool_lax
26 instead of bool when processing it.
28 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
29 server certificates when making smtp deliveries.
31 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
36 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
37 checkers that were complaining about end of non-void function with no
40 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
41 This was a regression intruduced in 4.83 by another bugfix.
43 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
45 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
46 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
52 TF/01 Correctly close the server side of TLS when forking for delivery.
54 When a message was received over SMTP with TLS, Exim failed to clear up
55 the incoming connection properly after forking off the child process to
56 deliver the message. In some situations the subsequent outgoing
57 delivery connection happened to have the same fd number as the incoming
58 connection previously had. Exim would try to use TLS and fail, logging
59 a "Bad file descriptor" error.
61 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
62 utilities have not been installed.
64 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
65 temporary space as the ACL may create new global variables.
67 TL/01 LDAP support uses per connection or global context settings, depending
68 upon the detected version of the libraries at build time.
70 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
71 to extract and use the src ip:port in logging and expansions as if it
72 were a direct connection from the outside internet. PPv2 support was
73 updated based on HAProxy spec change in May 2014.
75 JH/02 Add ${listextract {number}{list}{success}{fail}}.
77 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
78 Properly escape header and check for NULL return.
80 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
83 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
85 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
86 characters in header names, implemented as a verify condition.
87 Contributed by Michael Fischer v. Mollard.
89 TL/05 Rename SPF condition results err_perm and err_temp to standardized
90 results permerror and temperror. Previous values are deprecated but
91 still accepted. In a future release, err_perm and err_temp will be
92 completely removed, which will be a backward incompatibility if the
93 ACL tests for either of these two old results. Patch contributed by
94 user bes-internal on the mailing list.
96 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
98 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
99 selectors, in both main and reject logs.
101 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
104 JH/07 Add malware type "sock" for talking to simple daemon.
106 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
108 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
109 routers/transports under cutthrough routing.
111 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
112 numbers. Touch up "bool" conditional to keep the same definition.
114 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
116 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
118 JH/12 Expand items in router/transport headers_add or headers_remove lists
119 individually rather than the list as a whole. Bug 1452.
121 Required for reasonable handling of multiple headers_ options when
122 they may be empty; requires that headers_remove items with embedded
123 colons must have them doubled (or the list-separator changed).
125 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
126 view the policy declared in the DMARC record. Currently, $dmarc_status
127 is a combined value of both the record presence and the result of the
130 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
132 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
133 dnslookup router and the smtp transport (applying to the forward
136 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
137 of ldap servers used for a specific lookup. Patch provided by Heiko
140 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
141 New variable $lookup_dnssec_authenticated for observability.
143 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
144 Patch submitted by Lars Timman.
146 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
148 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
149 Requires trusted mode and valid format message id, aborts otherwise.
150 Patch contributed by Heiko Schlichting.
152 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
153 certextract with support for various fields. Bug 1358.
155 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
156 is requested by default, modifiable by smtp transport option
159 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
160 operate on certificate variables to give certificate fingerprints
161 Also new ${sha256:cert_variable}.
163 JH/23 The PRDR feature is moved from being Experimental into the mainline.
165 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
166 Christian Aistleitner.
168 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
170 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
171 file. Patch from Wolfgang Breyha.
173 JH/25 Expand the coverage of the delivery $host and $host_address to
174 client authenticators run in verify callout. Bug 1476.
176 JH/26 Port service names are now accepted for tls_on_connect_ports, to
177 align with daemon_smtp_ports. Bug 72.
179 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
180 support and error reporting did not work properly.
182 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
183 and is readable. Patch from Andrew Colin Kissa.
185 TL/14 Enhance documentation of ${run expansion and how it parses the
186 commandline after expansion, particularly in the case when an
187 unquoted variable expansion results in an empty value.
189 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
191 PP/02 Fix internal collision of T_APL on systems which support RFC3123
192 by renaming away from it. Addresses GH issue 15, reported by
195 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
197 TL/15 SECURITY: prevent double expansion in math comparison functions
198 (can expand unsanitized data). Not remotely exploitable.
205 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
207 PP/02 Make -n do something, by making it not do something.
208 When combined with -bP, the name of an option is not output.
210 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
213 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
214 $sender_host_name and config options to manage this, and basic check
217 PP/05 DSCP support for outbound connections and control modifier for inbound.
219 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
220 (Only plugin which currently uses this is kerberos4, which nobody should
221 be using, but we should make it available and other future plugins might
222 conceivably use it, even though it would break NAT; stuff *should* be
223 using channel bindings instead).
225 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
226 name; added for Sendmail compatibility; requires admin caller.
227 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
228 just ignore it); requires trusted caller.
229 Also parse but ignore: -Ac -Am -X<logfile>
232 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
234 TL/02 Add +smtp_confirmation as a default logging option.
236 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
237 Patch by Magnus Holmgren from 2007-02-20.
239 TL/04 Bugzilla 1281 - Spec typo.
240 Bugzilla 1283 - Spec typo.
241 Bugzilla 1290 - Spec grammar fixes.
243 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
245 TL/06 Add Experimental DMARC support using libopendmarc libraries.
247 TL/07 Fix an out of order global option causing a segfault. Reported to dev
248 mailing list by by Dmitry Isaikin.
250 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
252 JH/02 Support "G" suffix to numbers in ${if comparisons.
254 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
256 NM/01 Bugzilla 1197 - Spec typo
257 Bugzilla 1196 - Spec examples corrections
259 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
261 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
262 gnutls_enable_pkcs11, but renamed to more accurately indicate its
265 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
266 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
268 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
269 "acl {{name}{arg}...}", and optional args on acl condition
272 JH/05 Permit multiple router/transport headers_add/remove lines.
274 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
276 JH/07 Avoid using a waiting database for a single-message-only transport.
277 Performance patch from Paul Fisher. Bugzilla 1262.
279 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
282 JH/09 Add $headers_added variable, with content from use of ACL modifier
283 add_header (but not yet added to the message). Bugzilla 199.
285 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
286 Pulled from Bugzilla 817 by Wolfgang Breyha.
288 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
290 (nb: this is the same fix as in Exim 4.80.1)
292 JH/11 Add A= logging on delivery lines, and a client_set_id option on
295 JH/12 Add optional authenticated_sender logging to A= and a log_selector
298 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
300 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
301 advertise SMTP AUTH mechanism to us, instead of a generic
302 protocol violation error. Also, make Exim more robust to bad
303 data from the Dovecot auth socket.
305 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
307 When a queue runner is handling a message, Exim first routes the
308 recipient addresses, during which it prunes them based on the retry
309 hints database. After that it attempts to deliver the message to
310 any remaining recipients. It then updates the hints database using
313 So if a recipient address works intermittently, it can get repeatedly
314 deferred at routing time. The retry hints record remains fresh so the
315 address never reaches the final cutoff time.
317 This is a fairly common occurrence when a user is bumping up against
318 their storage quota. Exim had some logic in its local delivery code
319 to deal with this. However it did not apply to per-recipient defers
320 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
322 This change adds a proper retry rule check during routing so that the
323 final cutoff time is checked against the message's age. We only do
324 this check if there is an address retry record and there is not a
325 domain retry record; this implies that previous attempts to handle
326 the address had the retry_use_local_parts option turned on. We use
327 this as an approximation for the destination being like a local
328 delivery, as in LMTP.
330 I suspect this new check makes the old local delivery cutoff check
331 redundant, but I have not verified this so I left the code in place.
333 TF/02 Correct gecos expansion when From: is a prefix of the username.
335 Test 0254 submits a message to Exim with the header
339 When I ran the test suite under the user fanf2, Exim expanded
340 the header to contain my full name, whereas it should have added
341 a Resent-Sender: header. It erroneously treats any prefix of the
342 username as equal to the username.
344 This change corrects that bug.
346 GF/01 DCC debug and logging tidyup
347 Error conditions log to paniclog rather than rejectlog.
348 Debug lines prefixed by "DCC: " to remove any ambiguity.
350 TF/03 Avoid unnecessary rebuilds of lookup-related code.
352 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
353 Bug spotted by Jeremy Harris; was flawed since initial commit.
354 Would have resulted in OCSP responses post-SNI triggering an Exim
355 NULL dereference and crash.
357 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
359 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
360 Bug detection, analysis and fix by Samuel Thibault.
361 Bugzilla 1331, Debian bug #698092.
363 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
365 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
366 Server implementation by Todd Lyons, client by JH.
367 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
368 config variable "prdr_enable" controls whether the server
369 advertises the facility. If the client requests PRDR a new
370 acl_data_smtp_prdr ACL is called once for each recipient, after
371 the body content is received and before the acl_smtp_data ACL.
372 The client is controlled by bolth of: a hosts_try_prdr option
373 on the smtp transport, and the server advertisement.
374 Default client logging of deliveries and rejections involving
375 PRDR are flagged with the string "PRDR".
377 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
378 fclose(). Diagnosis by Todd Lyons.
380 PP/17 Update configure.default to handle IPv6 localhost better.
381 Patch by Alain Williams (plus minor tweaks).
384 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
385 This is now consistent with GnuTLS, and is now documented: the
386 previous undocumented portable approach to treating the option as
387 unset was to force an expansion failure. That still works, and
388 an empty string is now equivalent.
390 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
391 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
392 not performing validation itself.
394 PP/20 Added force_command boolean option to pipe transport.
395 Patch from Nick Koston, of cPanel Inc.
397 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
400 TF/04 Added udpsend ACL modifer and hexquote expansion operator
402 PP/21 Fix eximon continuous updating with timestamped log-files.
403 Broken in a format-string cleanup in 4.80, missed when I repaired the
404 other false fix of the same issue.
405 Report and fix from Heiko Schlichting.
408 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
409 Report from Prashanth Katuri.
411 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
412 It's SecureTransport, so affects any MacOS clients which use the
413 system-integrated TLS libraries, including email clients.
415 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
416 using a MIME ACL for non-SMTP local injection.
417 Report and assistance in diagnosis by Warren Baker.
419 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
421 JH/16 Fix comparisons for 64b. Bugzilla 1385.
423 TL/09 Add expansion variable $authenticated_fail_id to keep track of
424 last id that failed so it may be referenced in subsequent ACL's.
426 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
429 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
430 ldap library initialization, allowing self-signed CA's to be
431 used. Also properly sets require_cert option later in code by
432 using NULL (global ldap config) instead of ldap handle (per
433 session). Bug diagnosis and testing by alxgomz.
435 TL/12 Enhanced documentation in the ratelimit.pl script provided in
436 the src/util/ subdirectory.
438 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
439 renamed to Transport Post Delivery Action by Jeremy Harris, as
442 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
443 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
444 redis_servers = needs to be configured which will be used by the redis
445 lookup. Patch from Warren Baker, of The Packet Hub.
447 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
449 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
450 hostname or reverse DNS when processing a host list. Used suggestions
451 from multiple comments on this bug.
453 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
455 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
456 Missed a few lines, added it to make the runtest require no keyboard
459 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
460 contains upper case chars. Make router use caseful_local_part.
462 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
463 support when GnuTLS has been built with p11-kit.
469 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
471 This, or similar/improved, will also be change PP/11 of 4.82.
477 PP/01 Handle short writes when writing local log-files.
478 In practice, only affects FreeBSD (8 onwards).
479 Bugzilla 1053, with thanks to Dmitry Isaikin.
481 NM/01 Bugzilla 949 - Documentation tweak
483 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
486 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
488 PP/02 Implemented gsasl authenticator.
490 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
492 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
493 `pkg-config foo` for cflags/libs.
495 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
496 with rest of GSASL and with heimdal_gssapi.
498 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
499 `pkg-config foo` for cflags/libs for the TLS implementation.
501 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
502 properties get this fed in as external SSF. A number of robustness
503 and debugging improvements to the cyrus_sasl authenticator.
505 PP/08 cyrus_sasl server now expands the server_realm option.
507 PP/09 Bugzilla 1214 - Log authentication information in reject log.
508 Patch by Jeremy Harris.
510 PP/10 Added dbmjz lookup type.
512 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
514 PP/12 MAIL args handles TAB as well as SP, for better interop with
515 non-compliant senders.
516 Analysis and variant patch by Todd Lyons.
518 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
519 Bug report from Lars Müller <lars@samba.org> (via SUSE),
520 Patch from Dirk Mueller <dmueller@suse.com>
522 PP/13 tls_peerdn now print-escaped for spool files.
523 Observed some $tls_peerdn in wild which contained \n, which resulted
524 in spool file corruption.
526 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
527 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
528 or write after TLS renegotiation, which otherwise led to messages
531 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
532 as a tracking header (ie: a signed header comes before the signature).
533 Patch from Wolfgang Breyha.
535 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
536 comma-sep list; embedded commas doubled.
538 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
540 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
542 Report and patch from Dmitry Banschikov.
544 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
545 Removed SSL_clear() after SSL_new() which led to protocol negotiation
546 failures. We appear to now support TLS1.1+ with Exim.
548 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
549 lets Exim select keys and certificates based upon TLS SNI from client.
550 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
551 before an outbound SMTP session. New log_selector, +tls_sni.
553 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
554 NULL dereference. Report and patch from Alun Jones.
556 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
557 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
558 Not seeing resolver debug output on NetBSD, but suspect this is a
559 resolver implementation change.
561 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
562 Left warnings. Added "eximon gdb" invocation mode.
564 PP/21 Defaulting "accept_8bitmime" to true, not false.
566 PP/22 Added -bw for inetd wait mode support.
568 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
569 locate the relevant includes and libraries. Made this the default.
571 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
572 Bugzilla 1246, report and most of solution from Tomasz Kusy.
574 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
575 This may cause build issues on older platforms.
577 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
578 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
579 gnutls_require_mac & gnutls_require_protocols (no longer supported).
580 Added SNI support via GnuTLS too.
581 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
583 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
585 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
586 Applied second patch from Janne, implementing suggestion to default
587 multiple-strings-in-record handling to match SPF spec.
589 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
591 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
592 read-only, out of scope).
593 Patch from Wolfgang Breyha, report from Stuart Northfield.
595 PP/29 Fix three issues highlighted by clang analyser static analysis.
596 Only crash-plausible issue would require the Cambridge-specific
597 iplookup router and a misconfiguration.
598 Report from Marcin Mirosław.
600 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
602 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
603 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
604 As part of this, removing so much warning spew let me fix some minor
605 real issues in debug logging.
607 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
608 assignment on my part. Fixed.
610 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
611 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
612 Janne Snabb (who went above and beyond: thank you).
614 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
615 string otherwise requires a connection and a bunch more work and it's
616 relatively easy to get wrong. Should also expose TLS library linkage
619 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
620 64-bit ${eval} (JH/03).
622 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
623 GNU libc to support some of the 64-bit stuff, should not lead to
624 conflicts. Defined before os.h is pulled in, so if a given platform
625 needs to override this, it can.
627 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
628 protection layer was required, which is not implemented.
629 Bugzilla 1254, patch from Wolfgang Breyha.
631 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
632 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
633 tls_dhparam take prime identifiers. Also unbreak combination of
634 OpenSSL+DH_params+TLSSNI.
636 PP/39 Disable SSLv2 by default in OpenSSL support.
642 PP/01 Solaris build fix for Oracle's LDAP libraries.
643 Bugzilla 1109, patch from Stephen Usher.
645 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
647 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
650 TF/02 Fix a couple more cases where we did not log the error message
651 when unlink() failed. See also change 4.74-TF/03.
653 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
654 lock up or crash if it happened to be inside a call to libc when it
655 got a SIGUSR1 from exiwhat.
657 The SIGUSR1 handler appends the current process status to the process
658 log which is later printed by exiwhat. It used to use the general
659 purpose logging code to do this, but several functions it calls are
660 not safe for signals.
662 The new output code in the SIGUSR1 handler is specific to the process
663 log, and simple enough that it's easy to inspect for signal safety.
664 Removing some special cases also simplifies the general logging code.
665 Removing the spurious timestamps from the process log simplifies
668 TF/04 Improved ratelimit ACL condition.
670 The /noupdate option has been deprecated in favour of /readonly which
671 has clearer semantics. The /leaky, /strict, and /readonly update modes
672 are mutually exclusive. The update mode is no longer included in the
673 database key; it just determines when the database is updated. (This
674 means that when you upgrde Exim will forget old rate measurements.)
676 Exim now checks that the per_* options are used with an update mode that
677 makes sense for the current ACL. For example, when Exim is processing a
678 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
679 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
680 must specify per_mail/readonly. If you omit the update mode it defaults to
681 /leaky where that makes sense (as before) or /readonly where required.
683 The /noupdate option is now undocumented but still supported for
684 backwards compatibility. It is equivalent to /readonly except that in
685 ACLs where /readonly is required you may specify /leaky/noupdate or
686 /strict/noupdate which are treated the same as /readonly.
688 A useful new feature is the /count= option. This is a generalization
689 of the per_byte option, so that you can measure the throughput of other
690 aggregate values. For example, the per_byte option is now equivalent
691 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
693 The per_rcpt option has been generalized using the /count= mechanism
694 (though it's more complicated than the per_byte equivalence). When it is
695 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
696 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
697 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
698 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
699 Note that using per_rcpt with a non-readonly update mode in more than
700 one ACL will cause the recipients to be double-counted. (The per_mail
701 and per_byte options don't have this problem.)
703 The handling of very low rates has changed slightly. If the computed rate
704 is less than the event's count (usually one) then this event is the first
705 after a long gap. In this case the rate is set to the same as this event's
706 count, so that the first message of a spam run is counted properly.
708 The major new feature is a mechanism for counting the rate of unique
709 events. The new per_addr option counts the number of different
710 recipients that someone has sent messages to in the last time period. It
711 behaves like per_rcpt if all the recipient addresses are different, but
712 duplicate recipient addresses do not increase the measured rate. Like
713 the /count= option this is a general mechanism, so the per_addr option
714 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
715 example, measure the rate that a client uses different sender addresses
716 with the options per_mail/unique=$sender_address. There are further
717 details in the main documentation.
719 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
721 TF/06 Removed a few PCRE remnants.
723 TF/07 Automatically extract Exim's version number from tags in the git
724 repository when doing development or release builds.
726 PP/02 Raise smtp_cmd_buffer_size to 16kB.
727 Bugzilla 879. Patch from Paul Fisher.
729 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
730 Heavily based on revision 40f9a89a from Simon Arlott's tree.
733 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
735 PP/05 Variable $av_failed, true if the AV scanner deferred.
736 Bugzilla 1078. Patch from John Horne.
738 PP/06 Stop make process more reliably on build failure.
739 Bugzilla 1087. Patch from Heiko Schlittermann.
741 PP/07 Make maildir_use_size_file an _expandable_ boolean.
742 Bugzilla 1089. Patch from Heiko Schlittermann.
744 PP/08 Handle ${run} returning more data than OS pipe buffer size.
745 Bugzilla 1131. Patch from Holger Weiß.
747 PP/09 Handle IPv6 addresses with SPF.
748 Bugzilla 860. Patch from Wolfgang Breyha.
750 PP/10 GnuTLS: support TLS 1.2 & 1.1.
752 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
755 PP/11 match_* no longer expand right-hand-side by default.
756 New compile-time build option, EXPAND_LISTMATCH_RHS.
757 New expansion conditions, "inlist", "inlisti".
759 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
761 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
763 PP/14 fix log_write() format string regression from TF/03.
764 Bugzilla 1152. Patch from Dmitry Isaikin.
770 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
772 PP/02 Harmonised TLS library version reporting; only show if debugging.
773 Layout now matches that introduced for other libraries in 4.74 PP/03.
775 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
777 PP/04 New "dns_use_edns0" global option.
779 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
782 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
783 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
785 TK/01 Updated PolarSSL code to 0.14.2.
786 Bugzilla 1097. Patch from Andreas Metzler.
788 PP/07 Catch divide-by-zero in ${eval:...}.
791 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
794 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
795 format-string attack -- SECURITY: remote arbitrary code execution.
797 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
798 time unintentionally subject to list matching rules, letting the header
799 cause arbitrary Exim lookups (of items which can occur in lists, *not*
800 arbitrary string expansion). This allowed for information disclosure.
802 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
803 INT_MIN/-1 -- value coerced to INT_MAX.
809 NM/01 Workround for PCRE version dependency in version reporting
812 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
813 This fixes portability to compilers other than gcc, notably
814 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
816 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
817 makefiles for portability to HP-UX and POSIX correctness.
819 PP/01 Permit LOOKUP_foo enabling on the make command-line.
820 Also via indented variable definition in the Makefile.
821 (Debugging by Oliver Heesakkers).
823 PP/02 Restore caching of spamd results with expanded spamd_address.
824 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
826 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
827 Improves build reliability. Fix from: Frank Elsner
829 NM/02 Fix wide character breakage in the rfc2047 coding
830 Fixes bug 1064. Patch from Andrey N. Oktyabrski
832 NM/03 Allow underscore in dnslist lookups
833 Fixes bug 1026. Patch from Graeme Fowler
835 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
836 Code patches from Adam Ciarcinski of NetBSD.
838 NM/04 Fixed exiqgrep to cope with mailq missing size issue
841 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
842 is logged, to avoid truncation. Patch from John Horne.
844 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
845 Patch from Jakob Hirsch.
847 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
848 SQL string expansion failure details.
849 Patch from Andrey Oktyabrski.
851 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
852 Patch from Simon Arlott.
854 PP/09 New lookups functionality failed to compile on old gcc which rejects
855 extern declarations in function scope.
856 Patch from Oliver Fleischmann
858 PP/10 Use sig_atomic_t for flags set from signal handlers.
859 Check getgroups() return and improve debugging.
860 Fixed developed for diagnosis in bug 927 (which turned out to be
863 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
864 Patch from Mark Zealey.
866 PP/12 Bugzilla 1056: Improved spamd server selection.
867 Patch from Mark Zealey.
869 PP/13 Bugzilla 1086: Deal with maildir quota file races.
870 Based on patch from Heiko Schlittermann.
872 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
873 Patch from Uwe Doering, sign-off by Michael Haardt.
875 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
876 variable declaration deep within a block. Bug and patch from
879 PP/15 lookups-Makefile IRIX compatibilty coercion.
881 PP/16 Make DISABLE_DKIM build knob functional.
883 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
884 Patch by Simon Arlott
886 TF/03 Fix valgrind.h portability to C89 compilers that do not support
887 variable argument macros. Our copy now differs from upstream.
893 TF/01 Failure to get a lock on a hints database can have serious
894 consequences so log it to the panic log.
896 TF/02 Log LMTP confirmation messages in the same way as SMTP,
897 controlled using the smtp_confirmation log selector.
899 TF/03 Include the error message when we fail to unlink a spool file.
901 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
902 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
903 for maintaining out-of-tree patches for some time.
905 PP/01 Bugzilla 139: Documentation and portability issues.
906 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
907 Handle per-OS dynamic-module compilation flags.
909 PP/02 Let /dev/null have normal permissions.
910 The 4.73 fixes were a little too stringent and complained about the
911 permissions on /dev/null. Exempt it from some checks.
912 Reported by Andreas M. Kirchwitz.
914 PP/03 Report version information for many libraries, including
915 Exim version information for dynamically loaded libraries. Created
916 version.h, now support a version extension string for distributors
917 who patch heavily. Dynamic module ABI change.
919 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
920 privilege escalation vulnerability whereby the Exim run-time user
921 can cause root to append content of the attacker's choosing to
924 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
927 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
928 If dropping privileges for untrusted macros, we disabled normal logging
929 on the basis that it would fail; for the Exim run-time user, this is not
930 the case, and it resulted in successful deliveries going unlogged.
931 Fixed. Reported by Andreas Metzler.
937 PP/01 Date: & Message-Id: revert to normally being appended to a message,
938 only prepend for the Resent-* case. Fixes regression introduced in
939 Exim 4.70 by NM/22 for Bugzilla 607.
941 PP/02 Include check_rfc2047_length in configure.default because we're seeing
942 increasing numbers of administrators be bitten by this.
944 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
946 PP/03 Bugzilla 994: added openssl_options main configuration option.
948 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
950 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
952 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
954 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
955 without a peer certificate, leading to a segfault because of an
956 assumption that peers always have certificates. Be a little more
957 paranoid. Problem reported by Martin Tscholak.
959 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
960 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
961 NB: ClamAV planning to remove STREAM in "middle of 2010".
962 CL also introduces -bmalware, various -d+acl logging additions and
963 more caution in buffer sizes.
965 PP/09 Implemented reverse_ip expansion operator.
967 PP/10 Bugzilla 937: provide a "debug" ACL control.
969 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
971 PP/12 Bugzilla 973: Implement --version.
973 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
975 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
977 PP/15 Bugzilla 816: support multiple condition rules on Routers.
979 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
980 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
981 ignore trailing whitespace.
983 JJ/02 prevent non-panic DKIM error from being sent to paniclog
985 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
988 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
989 Notification from Dr Andrew Aitchison.
991 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
992 ExtendedDetectionInfo response format.
993 Notification from John Horne.
995 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
998 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
999 XSL and documented dependency on system catalogs, with examples of how
1002 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1005 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1006 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1007 configuration file which is writeable by the Exim user or group.
1009 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1010 of configuration files to cover files specified with the -C option if
1011 they are going to be used with root privileges, not just the default
1014 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1015 option (effectively making it always true).
1017 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1018 files to be used while preserving root privileges.
1020 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1021 that rogue child processes cannot use them.
1023 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1024 run-time user, instead of root.
1026 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1027 Exim run-time user without dropping privileges.
1029 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1030 result string, instead of calling string_vformat() twice with the same
1033 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1034 for other users. Others should always drop root privileges if they use
1035 -C on the command line, even for a whitelisted configure file.
1037 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1039 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1045 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1046 $data_path, and $header_path variables; fixed documentation bugs and
1049 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1050 exipick to access non-standard spools, including the "frozen" queue
1053 NM/01 Bugzilla 965: Support mysql stored procedures.
1054 Patch from Alain Williams
1056 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1058 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1059 Patch from Andreas Metzler
1061 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1062 Patch from Kirill Miazine
1064 NM/05 Bugzilla 671: Added umask to procmail example.
1066 JJ/03 installed exipick 20100323.0, fixing doc bug
1068 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1069 directory. Notification and patch from Dan Rosenberg.
1071 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1073 TK/02 Improve log output when DKIM signing operation fails.
1075 MH/01 Treat the transport option dkim_domain as a colon separated
1076 list, not as a single string, and sign the message with each element,
1077 omitting multiple occurences of the same signer.
1079 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1080 Bugzilla 985, 986. Patch by Simon Arlott
1082 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1083 Patch by Simon Arlott
1085 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1086 MBX locking. Notification from Dan Rosenberg.
1092 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1094 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1096 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1098 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1100 NM/04 Bugzilla 915: Flag broken perl installation during build.
1106 TK/01 Added patch by Johannes Berg that expands the main option
1107 "spamd_address" if it starts with a dollar sign.
1109 TK/02 Write list of recipients to X-Envelope-Sender header when building
1110 the mbox-format spool file for content scanning (suggested by Jakob
1113 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1114 (http://www.dcc-servers.net/) support via dccifd. Activated by
1115 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1117 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1118 by Mark Daniel Reidel <mr@df.eu>.
1120 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1121 When building exim an external PCRE library is now needed -
1122 PCRE is a system library on the majority of modern systems.
1123 See entry on PCRE_LIBS in EDITME file.
1125 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1126 conversation. Added nologin parameter to request.
1127 Patch contributed by Kirill Miazine.
1129 TF/01 Do not log submission mode rewrites if they do not change the address.
1131 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1133 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1134 log files in place. Contributed by Roberto Lima.
1136 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1138 TF/03 Bugzilla 615: When checking the local_parts router precondition
1139 after a local_part_suffix or local_part_prefix option, Exim now
1140 does not use the address's named list lookup cache, since this
1141 contains cached lookups for the whole local part.
1143 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1144 Robert Millan. Documentation is in experimental-spec.txt.
1146 TF/04 Bugzilla 668: Fix parallel build (make -j).
1148 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1150 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1151 Patch provided by Jan Srzednicki.
1153 TF/05 Leading white space used to be stripped from $spam_report which
1154 wrecked the formatting. Now it is preserved.
1156 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1157 that they are available at delivery time.
1159 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1161 TF/08 TLS error reporting now respects the incoming_interface and
1162 incoming_port log selectors.
1164 TF/09 Produce a more useful error message if an SMTP transport's hosts
1165 setting expands to an empty string.
1167 NM/06 Bugzilla 744: EXPN did not work under TLS.
1168 Patch provided by Phil Pennock.
1170 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1171 Patch provided by Richard Godbee.
1173 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1174 acl_smtp_notquit, added index entry.
1176 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1177 Patch provided by Eugene Bujak.
1179 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1180 accept(). Patch provided by Maxim Dounin.
1182 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1183 Patch provided by Phil Pennock.
1185 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1187 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1188 Patch provided by Brad "anomie" Jorsch.
1190 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1191 Patch provided by Dean Brooks.
1193 TK/05 Add native DKIM support (does not depend on external libraries).
1195 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1196 Patch provided by Graeme Fowler.
1198 NM/16 Bugzilla 851: Documentation example syntax fix.
1200 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1202 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1205 NM/19 Bugzilla 745: TLS version reporting.
1206 Patch provided by Phil Pennock.
1208 NM/20 Bugzilla 167: bool: condition support.
1209 Patch provided by Phil Pennock.
1211 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1212 clients. Patch provided by Phil Pennock.
1214 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1215 Patch provided by Brad "anomie" Jorsch.
1217 NM/23 Bugzilla 687: Fix misparses in eximstats.
1218 Patch provided by Heiko Schlittermann.
1220 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1221 Patch provided by Heiko Schlittermann.
1223 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1224 plus update to original patch.
1226 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1228 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1229 Patch provided by David Brownlee.
1231 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1233 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1235 NM/30 Bugzilla 888: TLS documentation bugfixes.
1237 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1239 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1240 Unlike the original bugzilla I have changed all shell scripts in src tree.
1242 NM/33 Bugzilla 898: Transport filter timeout fix.
1243 Patch by Todd Rinaldo.
1245 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1246 Patch by Serge Demonchaux.
1248 NM/35 Bugzilla 39: Base64 decode bug fixes.
1249 Patch by Jakob Hirsch.
1251 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1253 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1255 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1257 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1263 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1264 ALT-N's libdkim that I have put here:
1265 http://duncanthrax.net/exim-experimental/
1267 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1268 were called 'true' and it seems that C99 defines that as a reserved
1269 keyword to be used with 'bool' variable types. That means you could
1270 not include C99-style headers which use bools without triggering
1271 build errors in sieve.c.
1273 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1274 as mailq or other aliases. Changed the --help handling significantly
1275 to do whats expected. exim_usage() emits usage/help information.
1277 SC/01 Added the -bylocaldomain option to eximstats.
1279 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1281 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1283 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1289 PH/01 Another patch from the Sieve maintainer.
1291 PH/02 When an IPv6 address is converted to a string for single-key lookup
1292 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1293 dots are used instead of colons so that keys in lsearch files need not
1294 contain colons. This was done some time before quoting was made available
1295 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1296 (notated using the quote facility) so as to distinguish them from IPv4
1297 keys. This meant that lookups for IP addresses in host lists did not work
1298 for iplsearch lookups.
1300 This has been fixed by arranging for IPv6 addresses to be expressed with
1301 colons if the lookup type is iplsearch. This is not incompatible, because
1302 previously such lookups could never work.
1304 The situation is now rather anomolous, since one *can* have colons in
1305 ordinary lsearch keys. However, making the change in all cases is
1306 incompatible and would probably break a number of configurations.
1308 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1311 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1312 conversion specification without a maximum field width, thereby enabling
1313 a rogue spamd server to cause a buffer overflow. While nobody in their
1314 right mind would setup Exim to query an untrusted spamd server, an
1315 attacker that gains access to a server running spamd could potentially
1316 exploit this vulnerability to run arbitrary code as the Exim user.
1318 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1319 $primary_hostname instead of what libspf2 thinks the hosts name is.
1321 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1322 a directory entry by the name of the lookup key. Previously, if a
1323 symlink pointed to a non-existing file or a file in a directory that
1324 Exim lacked permissions to read, a lookup for a key matching that
1325 symlink would fail. Now it is enough that a matching directory entry
1326 exists, symlink or not. (Bugzilla 503.)
1328 PH/03 The body_linecount and body_zerocount variables are now exported in the
1331 PH/04 Added the $dnslist_matched variable.
1333 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1334 This means they are set thereafter only if the connection becomes
1337 PH/06 Added the client_condition to authenticators so that some can be skipped
1338 by clients under certain conditions.
1340 PH/07 The error message for a badly-placed control=no_multiline_responses left
1341 "_responses" off the end of the name.
1343 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1345 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1346 (without spaces) instead of just copying the configuration text.
1348 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1350 PH/11 Added $max_received_linelength.
1352 PH/12 Added +ignore_defer and +include_defer to host lists.
1354 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1355 way in which PCRE > 7.0 is built.
1357 PH/14 Implemented queue_only_load_latch.
1359 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1360 MAIL command. The effect was to mangle the value on 64-bit systems.
1362 PH/16 Another patch from the Sieve maintainer.
1364 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1366 PH/18 If a system quota error occurred while trying to create the file for
1367 a maildir delivery, the message "Mailbox is full" was not appended to the
1368 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1369 only to a quota excession during the actual writing of the file.
1371 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1372 characters?) which causes problems in log lines. The DN values are now
1373 passed through string_printing() before being added to log lines.
1375 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1376 and InterBase are left for another time.)
1378 PH/21 Added message_body_newlines option.
1380 PH/22 Guard against possible overflow in moan_check_errorcopy().
1382 PH/23 POSIX allows open() to be a macro; guard against that.
1384 PH/24 If the recipient of an error message contained an @ in the local part
1385 (suitably quoted, of course), incorrect values were put in $domain and
1386 $local_part during the evaluation of errors_copy.
1392 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1393 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1396 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1397 issue a MAIL command.
1399 PH/02 In an ACL statement such as
1401 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1403 if a client was not listed at all, or was listed with a value other than
1404 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1405 the condition was not true (as it should be), so access was not denied.
1406 The bug was that the ! inversion was incorrectly passed on to the second
1407 item. This has been fixed.
1409 PH/03 Added additional dnslists conditions == and =& which are different from
1410 = and & when the dns lookup returns more than one IP address.
1412 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1413 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1415 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1416 FSYNC, which compiles an option called disable_fsync that allows for
1417 bypassing fsync(). The documentation is heavily laced with warnings.
1419 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1421 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1422 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1423 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1424 including adding "make clean"; (3) Added -fPIC when compiling the test
1425 dynamically loaded module, to get rid of a warning.
1427 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1428 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1429 = 0s. The bug is otherwise harmless.
1431 PH/07 There was a bug in the dovecot authenticator such that the value of
1432 $auth1 could be overwritten, and so not correctly preserved, after a
1433 successful authentication. This usually meant that the value preserved by
1434 the server_setid option was incorrect.
1436 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1438 PH/09 Installed PCRE release 7.0.
1440 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1441 run for batched SMTP input. It is now run at the start of every message
1442 in the batch. While fixing this I discovered that the process information
1443 (output by running exiwhat) was not always getting set for -bs and -bS
1444 input. This is fixed, and it now also says "batched" for BSMTP.
1446 PH/11 Added control=no_pipelining.
1448 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1449 patch, slightly modified), and move the expansion of helo_data till after
1450 the connection is made in the smtp transport (so it can use these
1453 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1455 PH/14 Added log_selector = +pid.
1457 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1459 PH/16 Add ${if forany and ${if forall.
1461 PH/17 Added dsn_from option to vary the From: line in DSNs.
1463 PH/18 Flush SMTP output before performing a callout, unless control =
1464 no_callout_flush is set.
1466 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1467 was true (the default) a successful delivery failed to delete the retry
1468 item, thus causing premature timeout of the address. The bug is now
1471 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1473 PH/21 Long custom messages for fakedefer and fakereject are now split up
1474 into multiline reponses in the same way that messages for "deny" and
1475 other ACL rejections are.
1477 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1478 with slight modification.
1480 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1481 draft, changing the syntax and factoring some duplicate code".
1483 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1484 for deliveries of the second and subsequent messages over the same SMTP
1487 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1488 ${reduce, with only minor "tidies".
1490 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1492 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1493 expansion side effects.
1495 PH/27 When a message times out after an over-quota error from an Exim-imposed
1496 quota, the bounce message says "mailbox is full". This message was not
1497 being given when it was a system quota that was exceeded. It now should
1500 MH/03 Made $recipients available in local_scan(). local_scan() already has
1501 better access to the recipient list through recipients_list[], but
1502 $recipients can be useful in postmaster-provided expansion strings.
1504 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1505 in the case of a MAIL command with additional options following the
1506 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1507 were accidentally chopped off.
1509 PH/29 SMTP synchronization checks are implemented when a command is read -
1510 there is a check that no more input is waiting when there shouldn't be
1511 any. However, for some commands, a delay in an ACL can mean that it is
1512 some time before the response is written. In this time, more input might
1513 arrive, invalidly. So now there are extra checks after an ACL has run for
1514 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1515 pipelining has not been advertised.
1517 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1519 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1520 returned in the SMTP response when smtp_return_error_details was false.
1521 This has been fixed.
1523 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1524 instead of the C I/O that was originally supplied, because problems were
1525 reported on Solaris.
1527 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1528 Exim which did not show up earlier: it was assuming that a call to
1529 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1530 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1531 was a macro that became an assignment, so it seemed to work. This has
1532 changed to a proper function call with a void return, hence the compile
1533 error. Exim's code has been fixed.
1535 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1538 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1540 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1542 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1543 because it turns out to be too big or there is a timeout) there is no
1544 "Completed" line in the log. When some messages of this type were
1545 selected by exigrep, they were listed as "not completed". Others were
1546 picked up by some special patterns. I have improved the selection
1547 criteria to be more general.
1549 PH/38 The host_find_failed option in the manualroute router can now be set
1550 to "ignore", to completely ignore a host whose IP address cannot be
1551 found. If all hosts are ignored, the behaviour is controlled by the new
1552 host_all_ignored option.
1554 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1555 homing or because of multiple MX records with /mx) generated more than
1556 one IP address, and the following item turned out to be the local host,
1557 all the secondary addresses of the first item were incorrectly removed
1558 from the list, along with the local host and any following hosts (which
1559 is what is supposed to happen).
1561 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1562 whoever called Exim into the -H file. In the case of the daemon it was
1563 behaving confusingly. When first started, it used values for whoever
1564 started the daemon, but after a SIGHUP it used the Exim user (because it
1565 calls itself on a restart). I have changed the code so that it now always
1568 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1569 message are rejected with the same error (e.g. no authentication or bad
1570 sender address), and a DATA command is nevertheless sent (as can happen
1571 with PIPELINING or a stupid MUA), the error message that was given to the
1572 RCPT commands is included in the rejection of the DATA command. This is
1573 intended to be helpful for MUAs that show only the final error to their
1576 PH/42 Another patch from the Sieve maintainer.
1578 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1579 Eximstats - Fixed some broken HTML links and added missing column headers
1581 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1582 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1584 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1586 PH/43 Yet another patch from the Sieve maintainer.
1588 PH/44 I found a way to check for a TCP/IP connection going away before sending
1589 the response to the final '.' that terminates a message, but only in the
1590 case where the client has not sent further data following the '.'
1591 (unfortunately, this is allowed). However, in many cases there won't be
1592 any further data because there won't be any more messages to send. A call
1593 to select() can be used: if it shows that the input is "ready", there is
1594 either input waiting, or the socket has been closed. An attempt to read
1595 the next input character can distinguish the two cases. Previously, Exim
1596 would have sent an OK response which the client would never have see.
1597 This could lead to message repetition. This fix should cure that, at
1598 least in a lot of common cases.
1600 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1601 advertised in response to EHLO.
1607 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1608 fixed by 4.65/MH/01 (is this a record?) are fixed:
1610 (i) An empty string was always treated as zero by the numeric comparison
1611 operators. This behaviour has been restored.
1613 (ii) It is documented that the numeric comparison operators always treat
1614 their arguments as decimal numbers. This was broken in that numbers
1615 starting with 0 were being interpreted as octal.
1617 While fixing these problems I realized that there was another issue that
1618 hadn't been noticed. Values of message_size_limit (both the global option
1619 and the transport option) were treated as octal if they started with 0.
1620 The documentation was vague. These values are now always treated as
1621 decimal, and I will make that clear in the documentation.
1627 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1628 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1631 MH/01 Don't check that the operands of numeric comparison operators are
1632 integers when their expansion is in "skipping" mode (fixes bug
1633 introduced by 4.64-PH/07).
1635 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1636 child addresses, Exim now panics and dies. Previously, because the count
1637 is held in a short int, deliveries were likely to be lost. As such a
1638 large number of recipients for a single message is ridiculous
1639 (performance will be very, very poor), I have chosen to impose a limit
1640 rather than extend the field.
1646 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1647 leftover -K file (the existence of which was triggered by #402).
1648 While we were at it, introduced process PID as part of the -K
1649 filename. This should rule out race conditions when creating
1652 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1653 processing considerably. Previous code took too long for large mails,
1654 triggering a timeout which in turn triggers #401.
1656 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1657 in the DK code in transports.c. sendfile() is not really portable,
1658 hence the _LINUX specificness.
1660 TF/01 In the add_headers option to the mail command in an Exim filter,
1661 there was a bug that Exim would claim a syntax error in any
1662 header after the first one which had an odd number of characters
1665 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1666 callout verification, Exim cached a "reject" for the entire domain. This
1667 is correct for most verifications, but it is not correct for a recipient
1668 verification with use_sender or use_postmaster set, because in that case
1669 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1670 case of MAIL FROM:<> rejection from other early rejections (e.g.
1671 rejection of HELO). When verifying a recipient using a non-null MAIL
1672 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1673 Whatever the result of the callout, the value of the domain cache is
1674 left unchanged (for any other kind of callout, getting as far as trying
1675 RCPT means that the domain itself is ok).
1677 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1680 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1681 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1682 session, the daemon ignores ECONNECT errors and logs others; it now
1683 ignores EPIPE as well.
1685 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1686 (quoted-printable decoding).
1688 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1689 later the small subsequent patch to fix an introduced bug.
1691 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1693 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1695 PH/08 An error is now given if message_size_limit is specified negative.
1697 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1698 to be given (somewhat) arbitrary names.
1700 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1703 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1704 miscellaneous code fixes
1706 PH/10 Added the log_reject_target ACL modifier to specify where to log
1709 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1710 hostname. This is wrong, because it relates to the incoming message (and
1711 probably the interface on which it is arriving) and not to the outgoing
1712 callout (which could be using a different interface). This has been
1713 changed to use the value of the helo_data option from the smtp transport
1714 instead - this is what is used when a message is actually being sent. If
1715 there is no remote transport (possible with a router that sets up host
1716 addresses), $smtp_active_hostname is used.
1718 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1719 tweaks were necessary in order to get it to work (see also 21 below):
1720 (a) The code assumed that strncpy() returns a negative number on buffer
1721 overflow, which isn't the case. Replaced with Exim's string_format()
1723 (b) There were several signed/unsigned issues. I just did the minimum
1724 hacking in of casts. There is scope for a larger refactoring.
1725 (c) The code used strcasecmp() which is not a standard C function.
1726 Replaced with Exim's strcmpic() function.
1727 (d) The code set only $1; it now sets $auth1 as well.
1728 (e) A simple test gave the error "authentication client didn't specify
1729 service in request". It would seem that Dovecot has changed its
1730 interface. Fortunately there's a specification; I followed it and
1731 changed what the client sends and it appears to be working now.
1733 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1736 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1737 address A is aliased to B and C, where B exists and C does not. Without
1738 -v the output is "A verified" because verification stops after a
1739 successful redirection if more than one address is generated. However,
1740 with -v the child addresses are also verified. Exim was outputting "A
1741 failed to verify" and then showing the successful verification for C,
1742 with its parentage. It now outputs "B failed to verify", showing B's
1743 parentage before showing the successful verification of C.
1745 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1746 look up a TXT record in a specific list after matching in a combined
1749 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1750 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1751 they consult the DNS. I had assumed they would set it the way they
1752 wanted; and indeed my experiments on Linux seem to show that in some
1753 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1754 To be on the safe side, however, I have now made the interface to
1755 host_find_byname() similar to host_find_bydns(), with an argument
1756 containing the DNS resolver options. The host_find_byname() function now
1757 sets these options at its start, just as host_find_bydns() does. The smtp
1758 transport options dns_qualify_single and dns_search_parents are passed to
1759 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1760 of host_find_byname() use the default settings of RES_DEFNAMES
1761 (qualify_single) but not RES_DNSRCH (search_parents).
1763 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1764 spool_read_header() do less string testing, by means of a preliminary
1765 switch on the second character of optional "-foo" lines. (This is
1766 overdue, caused by the large number of possibilities that now exist.
1767 Originally there were few.) While I was there, I also converted the
1768 str(n)cmp tests so they don't re-test the leading "-" and the first
1769 character, in the hope this might squeeze out yet more improvement.
1771 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1772 flag allowing group syntax was set by the header_syntax check but not
1773 turned off, possible causing trouble later; (2) The flag was not being
1774 set at all for the header_verify test, causing "group"-style headers to
1775 be rejected. I have now set it in this case, and also caused header_
1776 verify to ignore an empty address taken from a group. While doing this, I
1777 came across some other cases where the code for allowing group syntax
1778 while scanning a header line wasn't quite right (mostly, not resetting
1779 the flag correctly in the right place). These bugs could have caused
1780 trouble for malformed header lines. I hope it is now all correct.
1782 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1783 with the "reply" argument non-NULL. The code, however (which originally
1784 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1785 but it didn't always do it. This confused somebody who was copying the
1786 code for some other use. I have removed all the tests.
1788 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1789 feature that was used to support insecure browsers during the U.S. crypto
1790 embargo. It requires special client support, and Exim is probably the
1791 only MTA that supported it -- and would never use it because real RSA is
1792 always available. This code has been removed, because it had the bad
1793 effect of slowing Exim down by computing (never used) parameters for the
1794 RSA_EXPORT functionality.
1796 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1797 authenticator to fail if there's a tab character in the incoming data
1798 (there should never be unless someone is messing about, as it's supposed
1799 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1800 if the connection is using TLS or if the remote IP is the same as the
1801 local IP, and the "valid-client-cert option" if a client certificate has
1804 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1805 authenticators. This can be used for authorization after authentication
1806 succeeds. (In the case of plaintext, it servers for both authentication
1809 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1810 if any retry times were supplied.
1812 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1813 connection, where there is no client IP address to check. In this
1814 situation, the verify now always succeeds.
1816 PH/25 Applied John Jetmore's -Mset patch.
1818 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1820 PH/27 In a string expansion for a processed (not raw) header when multiple
1821 headers of the same name were present, leading whitespace was being
1822 removed from all of them, but trailing whitespace was being removed only
1823 from the last one. Now trailing whitespace is removed from each header
1824 before concatenation. Completely empty headers in a concatenation (as
1825 before) are ignored.
1827 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1828 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1830 PH/29 [Removed. This was a change that I later backed out, and forgot to
1831 correct the ChangeLog entry (that I had efficiently created) before
1832 committing the later change.]
1834 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1835 address errors (4xx response to RCPT) over the same connection as other
1836 messages routed to the same hosts. Such deliveries are always "forced",
1837 so retry times are not inspected. This resulted in far too many retries
1838 for the affected addresses. The effect occurred only when there were more
1839 hosts than the hosts_max_try setting in the smtp transport when it had
1840 the 4xx errors. Those hosts that it had tried were not added to the list
1841 of hosts for which the message was waiting, so if all were tried, there
1842 was no problem. Two fixes have been applied:
1844 (i) If there are any address or message errors in an SMTP delivery, none
1845 of the hosts (tried or untried) are now added to the list of hosts
1846 for which the message is waiting, so the message should not be a
1847 candidate for sending over the same connection that was used for a
1848 successful delivery of some other message. This seems entirely
1849 reasonable: after all the message is NOT "waiting for some host".
1850 This is so "obvious" that I'm not sure why it wasn't done
1851 previously. Hope I haven't missed anything, but it can't do any
1852 harm, as the worst effect is to miss an optimization.
1854 (ii) If, despite (i), such a delivery is accidentally attempted, the
1855 routing retry time is respected, so at least it doesn't keep
1856 hammering the server.
1858 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1859 in ${readsocket because some servers need this prod.
1861 PH/32 Added some extra debug output when updating a wait-xxx database.
1863 PH/33 The hint "could be header name not terminated by colon", which has been
1864 given for certain expansion errors for a long time, was not being given
1865 for the ${if def:h_colon_omitted{... case.
1867 PH/34 The spec says: "With one important exception, whenever a domain list is
1868 being scanned, $domain contains the subject domain." There was at least
1869 one case where this was not true.
1871 PH/35 The error "getsockname() failed: connection reset by peer" was being
1872 written to the panic log as well as the main log, but it isn't really
1873 panic-worthy as it just means the connection died rather early on. I have
1874 removed the panic log writing for the ECONNRESET error when getsockname()
1877 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1878 runs only) independently of the message's sender address. This meant
1879 that, if the 4xx error was in fact related to the sender, a different
1880 message to the same recipient with a different sender could confuse
1881 things. In particualar, this can happen when sending to a greylisting
1882 server, but other circumstances could also provoke similar problems.
1883 I have changed the default so that the retry time for these errors is now
1884 based a combination of the sender and recipient addresses. This change
1885 can be overridden by setting address_retry_include_sender=false in the
1888 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1889 remote server are returned as part of bounce messages. This was not
1890 happening for LMTP over a pipe (the lmtp transport), but now it is the
1891 same for both kinds of LMTP.
1893 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1894 in header lines that were in fact CNAMEs. This is no longer the case.
1896 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1897 and queue runs started by the daemon processed all messages. This has
1898 been fixed so that -R and -S can now usefully be given with -q<time>.
1900 PH/40 Import PCRE release 6.7 (fixes some bugs).
1902 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1904 PH/42 Give an error if -q is specified more than once.
1906 PH/43 Renamed the variables $interface_address and $interface_port as
1907 $received_ip_address and $received_port, to make it clear that these
1908 values apply to message reception, and not to the outgoing interface when
1909 a message is delivered. (The old names remain recognized, of course.)
1911 PH/44 There was no timeout on the connect() call when using a Unix domain
1912 socket in the ${readsocket expansion. There now is.
1914 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1915 be meaningful with "accept".
1917 SC/01 Eximstats V1.43
1918 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1920 SC/02 Eximstats V1.44
1921 Use a glob alias rather than an array ref in the generated
1922 parser. This improves both readability and performance.
1924 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1925 Collect SpamAssassin and rejection statistics.
1926 Don't display local sender or destination tables unless
1927 there is data to show.
1928 Added average volumes into the top table text output.
1930 SC/04 Eximstats V1.46
1931 Collect data on the number of addresses (recipients)
1932 as well as the number of messages.
1934 SC/05 Eximstats V1.47
1935 Added 'Message too big' to the list of mail rejection
1936 reasons (thanks to Marco Gaiarin).
1938 SC/06 Eximstats V1.48
1939 Mainlog lines which have GMT offsets and are too short to
1940 have a flag are now skipped.
1942 SC/07 Eximstats V1.49 (Alain Williams)
1943 Added the -emptyok flag.
1945 SC/08 Eximstats V1.50
1946 Fixes for obtaining the IP address from reject messages.
1948 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1949 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1950 whitesspace changes from 4.64-PH/27
1952 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1955 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1956 are found, allow negative numbers in numeric criteria)
1958 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1960 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1961 to match changes made in 4.64-PH/43
1963 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1965 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1966 "domain" is actually an IP address, and force a failure. This locks out
1967 those revolvers/nameservers that support "A-for-A" lookups, in
1968 contravention of the specifications.
1970 PH/48 When a host name was looked up from an IP address, and the subsequent
1971 forward lookup of the name timed out, the host name was left in
1972 $sender_host_name, contrary to the specification.
1974 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1975 restricted to single-key lookups, Exim was not diagnosing an error if
1976 * or *@ was used with a query-style lookup.
1978 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1980 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1981 long ago, but noone interested enough thought of it. Let's just say that
1982 the "1.1" means that there are some new functions that weren't there at
1983 some point in the past.
1985 PH/51 Error processing for expansion failure of helo_data from an smtp
1986 transport during callout processing was broken.
1988 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1989 tested/used via the -bh/-bhc/-bs options.
1991 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1992 bug, fixed in subsequent PCRE releases).
1994 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1995 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1997 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2003 SC/01 Use a glob alias rather than an array ref in eximstats generated
2004 parser. This improves both readability and performance.
2006 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2007 Don't display local sender or destination tables in eximstats unless
2008 there is data to show.
2009 Added average volumes into the eximstats top table text output.
2011 SC/03 Collect data on the number of addresses (recipients) as well
2012 as the number of messages in eximstats.
2014 TF/01 Correct an error in the documentation for the redirect router. Exim
2015 does (usually) call initgroups() when daemonizing.
2017 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2018 with consistent privilege compared to when running as a daemon.
2020 TF/03 Note in the spec that $authenticated_id is not set for local
2021 submissions from trusted users.
2023 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2024 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2026 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2027 by adding some example configuration directives to the default
2028 configuration file. A little bit of work is required to uncomment the
2029 directives and define how usernames and passwords are checked, but
2030 there is now a framework to start from.
2032 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2033 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2034 without this. I don't know how relevant this is to other LDAP libraries.
2036 PH/02 Add the verb name to the "unknown ACL verb" error.
2038 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2040 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2042 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2043 directory not expanded when it should be if an expanded home directory
2044 was set for the address (which is overridden by the transport).
2046 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2049 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2050 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2051 because it is too late at that time, and has no effect.
2053 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2054 security issue with \' (bugzilla #107). I could not use the
2055 PQescapeStringConn() function, because it needs a PGconn value as one of
2058 PH/08 When testing addresses using -bt, indicate those final addresses that
2059 are duplicates that would not cause an additional delivery. At least one
2060 person was confused, thinking that -bt output corresponded to deliveries.
2061 (Suppressing duplicates isn't a good idea as you lose the information
2062 about possibly different redirections that led to the duplicates.)
2064 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2065 systems where poll() doesn't work, in particular OS X.
2067 PH/10 Added more information to debugging output for retry time not reached.
2069 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2070 operations in malware.c.
2072 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2075 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2076 syslog_duplication set false, log lines that would normally be written
2077 both the the main log and to the reject log were not written to syslog at
2080 PH/14 In the default configuration, change the use of "message" in ACL warn
2081 statements to "add_header".
2083 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2084 not followed by a command (e.g. "seen endif").
2086 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2087 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2090 PH/17 Added extra conditions to the default value of delay_warning_condition
2094 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2095 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2096 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2099 The Auto-Submitted: and various List- headers are standardised, whereas I
2100 don't think Precedence: ever was.
2102 PH/18 Refactored debugging code in route_finduser() to show more information,
2103 in particular, the error code if getpwnam() issues one.
2105 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2106 This is apparently needed in addition to the PH/07 change above to avoid
2107 any possible encoding problems.
2109 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2110 but not after initializing Perl.
2112 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2113 output them only if debugging. By default they are written stderr,
2114 apparently, which is not desirable.
2116 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2119 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2122 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2124 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2125 authenticated or an ident call has been made. Suppress the default
2126 values for $authenticated_id and $authenticated_sender (but permit -oMai
2127 and -oMas) when testing with -bh.
2129 PH/24 Re-jigged the order of the tests in the default configuration so that the
2130 tests for valid domains and recipients precede the DNS black list and CSA
2131 tests, on the grounds that those ones are more expensive.
2133 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2134 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2135 This bug exists in every version of Exim that I still have, right back to
2138 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2139 However, an attempt to turn on case-sensitivity in a regex key by
2140 including (?-i) didn't work because the subject string was already
2141 lowercased, and the effects were non-intuitive. It turns out that a
2142 one-line patch can be used to allow (?-i) to work as expected.
2148 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2149 other effects) broke the use of negated acl sub-conditions.
2151 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2154 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2155 "Deny" causes Exim to reject the incoming connection with a 554 error.
2156 Unfortunately, if there is a major crisis, such as a disk failure,
2157 tcp-wrappers gives "deny", whereas what one would like would be some
2158 kind of temporary error. A kludge has been added to help with this.
2159 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2160 554 error is used if errno is still zero or contains ENOENT (which occurs
2161 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2164 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2166 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2167 errors. Otherwise a message that provokes a temporary error (when other
2168 messages do not) can cause a whole host to time out.
2170 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2171 addresses were routed directly to files or pipes from a redirect router.
2172 File deliveries just didn't batch; pipe deliveries might have suffered
2175 PH/06 A failure to get a lock for a hints database would erroneously always say
2176 "Failed to get write lock", even when it was really a read lock.
2178 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2179 of 0600. This has been changed to use the value of the lockfile_mode
2180 option (which defaults to 0600).
2182 PH/08 Applied small patch from the Sieve maintainer.
2184 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2185 folder from quota calculations, a direct delivery into this folder messed
2186 up the contents of the maildirsize file. This was because the regex was
2187 used only to exclude .Trash (or whatever) when the size of the mailbox
2188 was calculated. There was no check that a delivery was happening into an
2189 excluded directory. This bug has been fixed by ignoring all quota
2190 processing for deliveries into excluded directories.
2192 PH/10 Added the maildirfolder_create_regex option to appendfile.
2198 PH/01 The code for finding all the local interface addresses on a FreeBSD
2199 system running IPv6 was broken. This may well have applied to all BSD
2200 systems, as well as to others that have similar system calls. The broken
2201 code found IPv4 interfaces correctly, but gave incorrect values for the
2202 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2203 that it would not match correctly against @[] and not recognize the IPv6
2206 PH/02 The ipliteral router was not recognizing addresses of the form user@
2207 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2209 PH/03 Added disable_ipv6.
2211 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2212 decoded form, because it is most often used to construct To: headers
2213 lines in autoreplies, and the decoded form may well be syntactically
2214 invalid. However, $reply_address has leading white space removed, and all
2215 newlines turned into spaces so that the autoreply transport does not
2218 PH/05 If group was specified without a user on a router, and no group or user
2219 was specified on a transport, the group from the router was ignored.
2221 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2222 for visible compile-time settings that can be used to change these
2223 numbers, for those that want even more. Backwards compatibility with old
2224 spool files has been maintained. However, going back to a previous Exim
2225 release will lost any variables that are in spool files.
2227 PH/07 Two small changes when running in the test harness: increase delay when
2228 passing a TCP/IP connection to a new process, in case the original
2229 process has to generate a bounce, and remove special handling of
2230 127.0.0.2 (sic), which is no longer necessary.
2232 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2233 be the same on different OS.
2235 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2238 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2239 whether --show-vars was specified or not
2241 JJ/02 exipick: Added support for new ACL variable spool format introduced
2244 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2245 syntactically invalid From: or Reply-to: line, and a filter used this to
2246 generate an autoreply, and therefore failed to obtain an address for the
2247 autoreply, Exim could try to deliver to a non-existent relative file
2248 name, causing unrelated and misleading errors. What now happens is that
2249 it logs this as a hard delivery error, but does not attempt to create a
2252 PH/11 The exinext utility has a -C option for testing purposes, but although
2253 the given file was scanned by exinext itself; it wasn't being passed on
2254 when Exim was called.
2256 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2257 an end-of-file indication when reading a command response.
2259 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2260 compiled. In many other places in Exim, IPv6 addresses are always
2261 recognized, so I have changed this. It also means that IPv4 domain
2262 literals of the form [IPV4:n.n.n.n] are now always recognized.
2264 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2265 used if the router is not running as root, for example, when verifying at
2266 ACL time, or when using -bh. The debugging output from this situation was
2267 non-existent - all you got was a failure to exec. I have made two
2270 (a) Failures to set uid/gid, the current directory, or a process leader
2271 in a subprocess such as that created by queryprogram now generate
2272 suitable debugging ouput when -d is set.
2274 (b) The queryprogram router detects when it is not running as root,
2275 outputs suitable debugging information if -d is set, and then runs
2276 the subprocess without attempting to change uid/gid.
2278 PH/15 Minor change to Makefile for building test_host (undocumented testing
2281 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2282 additional section of a DNS packet that returns MX or SRV records.
2283 Instead, it always explicitly searches for A/AAAA records. This avoids
2284 major problems that occur when a DNS server includes only records of one
2285 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2286 fixed another bug: if SRV records were looked up and the corresponding
2287 address records were *not* found in the additional section, the port
2288 values from the SRV records were lost.
2290 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2291 using the correct key (the original address) when searching the retry
2292 rules in order to find which one to use for generating the retry hint.
2294 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2295 adding the default one. Similarly, if it contains a Reply-To: header, the
2296 errors_reply_to option, if set, is not used.
2298 PH/19 When calculating a retry time, Exim used to measure the "time since
2299 failure" by looking at the "first failed" field in the retry record. Now
2300 it does not use this if it is later than than the arrival time of the
2301 message. Instead it uses the arrival time. This makes for better
2302 behaviour in cases where some deliveries succeed, thus re-setting the
2303 "first failed" field. An example is a quota failure for a huge message
2304 when small messages continue to be delivered. Without this change, the
2305 "time since failure" will always be short, possible causing more frequent
2306 delivery attempts for the huge message than are intended.
2307 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2309 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2310 $1, $2, $3) because the numerical variables can be reset during some
2311 expansion items (e.g. "match"), thereby losing the authentication data.
2313 PH/21 Make -bV show the size of off_t variables so that the test suite can
2314 decide whether to run tests for quotas > 2G.
2316 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2317 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2318 filecount value is greater than 2G or if a quota value is greater than 2G
2319 on a system where the size of off_t is not greater than 4, a panic error
2322 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2323 never match. The debug and -bh output now contains an explicit error
2324 message indicating a malformed IPv4 address or mask.
2326 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2327 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2328 PH/23 above applies.
2330 PH/25 Do not write to syslog when running in the test harness. The only
2331 occasion when this arises is a failure to open the main or panic logs
2332 (for which there is an explicit test).
2334 PH/26 Added the /no_tell option to "control=freeze".
2336 PH/27 If a host name lookup failed very early in a connection, for example, if
2337 the IP address matched host_lookup and the reverse lookup yielded a name
2338 that did not have a forward lookup, an error message of the form "no IP
2339 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2340 could be logged. Now it outputs the IP address instead of "NULL".
2342 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2343 allows the sender and the authenticated sender to be set when
2344 submitting a message from within Exim. Since child_open_exim() is
2345 documented for local_scan(), the new function should be too.
2347 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2348 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2349 results in an empty string is now treated as unset.
2351 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2353 PH/31 Added sender_verify_fail logging option.
2355 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2356 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2357 tidied the source and removed it altogether.
2359 PH/33 When a queue run was abandoned because the load average was too high, a
2360 log line was always written; now it is written only if the queue_run log
2361 selector is set. In addition, the log line for abandonment now contains
2362 information about the queue run such as the pid. This is always present
2363 in "start" and "stop" lines but was omitted from the "abandon" line.
2365 PH/34 Omit spaces between a header name and the colon in the error message that
2366 is given when verify = headers_syntax fails (if there are lots of them,
2367 the message gets confusing).
2369 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2370 names, as there are now some PTR records that contain slashes. This check
2371 is only to protect against broken name servers that fall over on strange
2372 characters, so the fact that it applies to all lookups doesn't matter.
2374 PH/36 Now that the new test suite is complete, we can remove some of the
2375 special code in Exim that was needed for the old test suite. For example,
2376 sorting DNS records because real resolvers return them in an arbitrary
2377 order. The new test suite's fake resolver always returns records in the
2380 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2381 bounces) except when queue_only is set, to avoid logging races between
2382 the different processes.
2384 PH/38 Panic-die if .include specifies a non-absolute path.
2386 PH/39 A tweak to the "H" retry rule from its user.
2388 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2389 a label. They prevented compilation on older perls.
2391 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2392 a warning to be raised on newish perls.
2394 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2395 on queue. Changes to match documented behaviour of showing count of
2396 messages matching specified criteria.
2398 PH/40 Changed the default ident timeout from 30s to 5s.
2400 PH/41 Added support for the use of login_cap features, on those BSD systems
2401 that have them, for controlling the resources used by pipe deliveries.
2403 PH/42 The content-scanning code uses fopen() to create files in which to put
2404 message data. Previously it was not paying any attention to the mode of
2405 the files. Exim runs with umask(0) because the rest of the code creates
2406 files with open(), and sets the required mode explicitly. Thus, these
2407 files were ending up world-writeable. This was not a big issue, because,
2408 being within the spool directory, they were not world-accessible. I have
2409 created a function called modefopen, which takes an additional mode
2410 argument. It sets umask(777), creates the file, chmods it to the required
2411 mode, then resets the umask. All the relevant calls to fopen() in the
2412 content scanning code have been changed to use this function.
2414 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2415 to 24 hours. This avoids potential overflow problems when processing G
2416 and H retry rules. I suspect nobody ever tinkers with this value.
2418 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2420 PH/45 When the plaintext authenticator is running as a client, the server's
2421 challenges are checked to ensure they are valid base64 strings. By
2422 default, the authentication attempt is cancelled if an invalid string is
2423 received. Setting client_ignore_invalid_base64 true ignores these errors.
2424 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2425 they are received. Thus, the responses can be made to depend on the
2426 challenges. If an invalid string is ignored, an empty string is placed in
2429 PH/46 Messages that are created by the autoreply transport now contains a
2430 References: header, in accordance with RFCs 2822 and 3834.
2432 PH/47 Added authenticated_sender_force to the smtp transport.
2434 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2436 PH/49 Installed latest patch from the Sieve maintainer.
2438 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2439 was also set, the appendfile transport was unnecessarily scanning a
2440 directory of message files (e.g. for maildir delivery) to find the count
2441 of files (along with the size), even though it did not need this
2442 information. It now does the scan only if it needs to find either the
2443 size of the count of files.
2445 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2447 PH/52 Two bugs concerned with error handling when the smtp transport is
2450 (i) Exim was not creating retry information for temporary errors given
2451 for individual recipients after the DATA command when the smtp transport
2452 was used in LMTP mode. This meant that they could be retried too
2453 frequently, and not timed out correctly.
2455 (ii) Exim was setting the flag that allows error details to be returned
2456 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2457 recipients that were returned after the DATA command.
2459 PH/53 This is related to PH/52, but is more general: for any failing address,
2460 when detailed error information was permitted to be returned to the
2461 sender, but the error was temporary, then after the final timeout, only
2462 "retry timeout exceeded" was returned. Now it returns the full error as
2463 well as "retry timeout exceeded".
2465 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2466 do this, and (what is worse) MTAs that accept it.
2468 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2469 will now be deprecated.
2471 PH/56 New os.c-cygwin from the Cygwin maintainer.
2473 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2474 formats (previously only available in exim formats via -bpr, -bpru,
2475 and -bpra. Now also available in native and exiqgrep formats)
2477 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2478 with very large, slow to parse queues
2480 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2482 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2484 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2485 responses to authentication challenges, though it was showing the
2486 challenges; (ii) I've removed the CR characters from the debug output for
2489 PH/58 Allow for the insertion of a newline as well as a space when a string
2490 is turned into more than one encoded-word during RFC 2047 encoding. The
2491 Sieve code now uses this.
2493 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2494 data_4xx, lost_connection, tls_required.
2496 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2497 message was being sent as an SMTP response.
2499 PH/61 Add -l and -k options to exicyclog.
2501 PH/62 When verifying, if an address was redirected to one new address, so that
2502 verification continued, and the new address failed or deferred after
2503 having set something in $address_data, the value of $address_data was not
2504 passed back to the ACL. This was different to the case when no
2505 redirection occurred. The value is now passed back in both cases.
2507 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2508 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2509 use login_cap.h, so on its own it isn't the distinguishing feature. The
2510 new name refers directly to the setclassresources() function.
2512 PH/65 Added configuration files for NetBSD3.
2514 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2516 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2517 is preferred over IPv4.
2519 PH/68 The bounce_return_message and bounce_return_body options were not being
2520 honoured for bounces generated during the reception of non-SMTP messages.
2521 In particular, this applied to messages rejected by the ACL. This bug has
2522 been fixed. However, if bounce_return_message is true and bounce_return_
2523 body is false, the headers that are returned for a non-SMTP message
2524 include only those that have been read before the error was detected.
2525 (In the case of an ACL rejection, they have all been read.)
2527 PH/69 The HTML version of the specification is now built in a directory called
2528 spec_html instead of spec.html, because the latter looks like a path with
2529 a MIME-type, and this confuses some software.
2531 PH/70 Catch two compiler warnings in sieve.c.
2533 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2534 function verify_get_ident() calls ip_connect() to connect a socket, but
2535 if the "connect()" function timed out, ip_connect() used to close the
2536 socket. However, verify_get_ident() also closes the socket later, and in
2537 between Exim writes to the log, which may get opened at this point. When
2538 the socket was closed in ip_connect(), the log could get the same file
2539 descriptor number as the socket. This naturally causes chaos. The fix is
2540 not to close the socket in ip_connect(); the socket should be closed by
2541 the function that creates it. There was only one place in the code where
2542 this was missing, in the iplookup router, which I don't think anybody now
2543 uses, but I've fixed it anyway.
2545 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2546 well as to direct DNS lookups. Otherwise the handling of names in host
2547 lists is inconsistent and therefore confusing.
2553 PH/01 Two changes to the default runtime configuration:
2555 (1) Move the checks for relay_from_hosts and authenticated clients from
2556 after to before the (commented out) DNS black list checks.
2558 (2) Add control=submission to the relay_from_hosts and authenticated
2559 clients checks, on the grounds that messages accepted by these
2560 statements are most likely to be submissions.
2562 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2564 (1) Generate an error if the third argument for the ${prvs expansion is
2567 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2570 (3) Reset the variables that are obtained from the first argument of
2571 ${prvscheck and used in the second argument before leaving the code,
2572 because their memory is reclaimed, so using them afterwards may do
2575 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2576 one (it's much easier than Tom thought :-).
2578 (5) Because of (4), we can now allow for the use of $prvscheck_result
2579 inside the third argument.
2581 PH/03 For some reason, the default setting of PATH when running a command from
2582 a pipe transport was just "/usr/bin". I have changed it to
2585 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2586 anything to be listed in the output from -bV.
2588 PH/05 When a filter generated an autoreply, the entire To: header line was
2589 quoted in the delivery log line, like this:
2591 => >A.N.Other <ano@some.domain> <original@ddress> ...
2593 This has been changed so that it extracts the operative address. There
2594 may be more than one such address. If so, they are comma-separated, like
2597 => >ano@some.domain,ona@other.domain <original@ddress> ...
2599 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2600 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2601 not being looked up in the rDNS to get a host name, Exim was showing the
2602 IP address twice in Received: lines, even though the IP addresses were
2603 identical. For example:
2605 Received: from [1.2.3.4] (helo=[1.2.3.4])
2607 However, if the real host name was known, it was omitting the HELO data
2608 if it matched the actual IP address. This has been tidied up so that it
2609 doesn't show the same IP address twice.
2611 PH/07 When both +timestamp and +memory debugging was on, the value given by
2612 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2613 called by the debug printing, thereby overwriting the timestamp buffer.
2614 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2616 PH/08 When the original message was included in an autoreply transport, it
2617 always said "this is a copy of the message, including all the headers",
2618 even if body_only or headers_only was set. It now gives an appropriate
2621 PH/09 Applied a patch from the Sieve maintainer which:
2623 o fixes some comments
2624 o adds the (disabled) notify extension core
2625 o adds some debug output for the result of if/elsif tests
2626 o points to the current vacation draft in the documentation
2627 and documents the missing references header update
2631 o fixes a bug in processing the envelope test (when testing
2632 multiple envelope elements, the last element determinted the
2635 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2636 Electronic Mail") by including:
2638 Auto-submitted: auto-generated
2640 in the messages that it generates (bounce messages and others, such as
2641 warnings). In the case of bounce messages for non-SMTP mesages, there was
2642 also a typo: it was using "Auto_submitted" (underscore instead of
2643 hyphen). Since every message generated by Exim is necessarily in response
2644 to another message, thes have all been changed to:
2646 Auto-Submitted: auto-replied
2648 in accordance with these statements in the RFC:
2650 The auto-replied keyword:
2652 - SHOULD be used on messages sent in direct response to another
2653 message by an automatic process,
2655 - MUST NOT be used on manually-generated messages,
2657 - MAY be used on Delivery Status Notifications (DSNs) and Message
2658 Disposition Notifications (MDNs),
2660 - MUST NOT be used on messages generated by automatic or periodic
2661 processes, except for messages which are automatic responses to
2664 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2665 to the default Received: header definition.
2667 PH/12 Added log selector acl_warn_skipped (default on).
2669 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2670 variables because (a) they are in the wrong storage pool and (b) even if
2671 they were copied, it wouldn't work properly because of the caching.
2673 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2674 checking when decoding. Apparently there are clients that generate
2675 overlong encoded strings. Why am I not surprised?
2677 PH/15 If the first argument of "${if match_address" was not empty, but did not
2678 contain an "@" character, Exim crashed. Now it writes a panic log message
2679 and treats the condition as false.
2681 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2683 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2684 extension "envelope-auth". The code is finished and in agreement with
2685 other implementations, but there is no documentation so far and in fact,
2686 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2687 not changing the active code.
2689 Print executed "if" and "elsif" statements when debugging is used. This
2690 helps a great deal to understand what a filter does.
2692 Document more things not specified clearly in RFC3028. I had all this
2693 sorted out, when out of a sudden new issues came to my mind. Oops."
2695 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2698 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2699 canonical form (as documented). However, after a host name lookup from
2700 the IP address, check_host() was doing a simple string comparison with
2701 addresses acquired from the DNS when checking that the found name did
2702 have the original IP as one of its addresses. Since any found IPv6
2703 addresses are likely to be in abbreviated form, the comparison could
2704 fail. Luckily, there already exists a function for doing the comparison
2705 by converting both addresses to binary, so now that is used instead of
2706 the text comparison.
2708 PH/20 There was another similar case to PH/19, when a complete host name was
2709 given in a host list; looking up its IP address could give an abbreviated
2710 form, whereas the current host's name might or might not be abbreviated.
2711 The same fix has been applied.
2717 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2718 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2721 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2722 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2724 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2726 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2727 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2728 Also, for "Auto-Submitted", treat anything other than "no" as
2729 non-personal, in accordance with RFC 3834. (Previously it treated
2730 anything starting "auto-" as non-personal.)
2732 TF/01 The control=submission/name=... option had a problem with syntax
2733 errors if the name included a slash character. The /name= option
2734 now slurps the rest of the string, so it can include any characters
2735 but it must come last in the list of options (after /sender_retain
2738 PH/05 Some modifications to the interface to the fake nameserver for the new
2746 TK/01 Added the "success_on_redirect" address verification option. See
2747 NewStuff for rationale and an example.
2749 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2751 PH/02 Patch to exigrep to allow it to work on syslog lines.
2753 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2754 fread() to skip over the body file's header line, because in Cygwin the
2755 header line is locked and is inaccessible.
2757 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2758 co-exist for some time) to make it clear that it is the Exim ID that is
2759 referenced, not the Message-ID: header line.
2761 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2762 string_format() function, because snprintf() does not exist on all
2765 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2766 ${stat: expansion item.
2768 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2769 protocol synchronization error", to keep the pedants happy.
2771 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2772 well as for IRIX systems, when gcc is being used. See the host.c source
2775 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2777 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2780 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2781 transport and to the smtp transport in LMTP mode.
2783 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2785 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2786 run a filter in a subprocess. This could lead to confusion in subsequent
2787 lookups in the parent process. There should also be a search_tidyup() at
2788 the end of the subprocess.
2790 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2791 only if the host matched helo_try_verify_hosts, which caused the
2792 verification to occur when the EHLO/HELO command was issued. The ACL just
2793 tested the remembered result. Now, if a previous verification attempt has
2794 not happened, "verify = helo" does it there and then.
2796 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2798 TK/03 Fix log output including CR from clamd.
2800 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2801 exist provoked a memory error which could cause a segfault.
2803 PH/15 Installed PCRE 6.2
2805 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2807 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2808 of the problem. Specifically, suggested +O2 rather than +O1 for the
2811 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2813 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2814 record, the port was not correctly logged when the outgoing_port log
2815 selector was set (it logged the transort's default port).
2817 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2818 fallback_hosts, and "hosts" in the smtp transport.
2820 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2821 host errors such as "Connection refused".
2823 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2824 authentication with radiusclient 0.4.9:
2826 - Error returned from rc_read_config was caught wrongly
2827 - Username/password not passed on to radius server due to wrong length.
2829 The presumption is that some radiusclient API changes for 4.51/PH/17
2830 were not taken care of correctly. The code is still untested by me (my
2831 Linux distribution still has 0.3.2 of radiusclient), but it was
2832 contributed by a Radius user.
2834 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2835 expanding the "port" option of the smtp transport.
2837 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2838 while reading a MIME header. Thanks to Tom Hughes for a patch.
2840 PH/24 Include config.h inside local_scan.h so that configuration settings are
2843 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2844 that in an ACL for RCPT (for example), you can examine exactly what was
2847 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2848 commands, but it was not correctly comparing the address with the actual
2849 client host address. Thus, it would show the EHLO address in Received:
2850 header lines when this was not necessary.
2852 PH/27 Added the % operator to ${eval:}.
2854 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2855 it should be ignoring failures (because with -C, for example, it has lost
2856 privilege). It wasn't ignoring creation failures other than "already
2859 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2862 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2863 because an input line was too long, either on its own, or by virtue of
2864 too many continuations, the temporary file was not being removed, and the
2865 return code was incorrect.
2867 PH/31 Missing "BOOL" in function definition in filtertest.c.
2869 PH/32 Applied Sieve patches from the maintainer.
2871 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2873 PH/33 Added "verify = not_blind".
2875 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2876 Local/Makefile (with some defaults set). These are used in built scripts
2877 such as exicyclog, but they have never been used in the exim_install
2878 script (though there are many overriding facilities there). I have
2879 arranged that the exim_install script now takes note of these two
2882 PH/35 Installed configuration files for Dragonfly.
2884 PH/36 When a locally submitted message by a trusted user did not contain a
2885 From: header, and the sender address was obtained from -f or from an SMTP
2886 MAIL command, and the trusted user did not use -F to supply a sender
2887 name, $originator_name was incorrectly used when constructing a From:
2888 header. Furthermore, $originator_name was used for submission mode
2889 messages from external hosts without From: headers in a similar way,
2890 which is clearly wrong.
2892 PH/37 Added control=suppress_local_fixups.
2894 PH/38 When log_selector = +received_sender was set, and the addition of the
2895 sender made the log line's construction buffer exactly full, or one byte
2896 less than full, an overflow happened when the terminating "\n" was
2899 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2900 when the result of a list match is failure because a DNS lookup failed.
2902 PH/40 RM_COMMAND is now used in the building process.
2904 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2905 the "build-* directories that it finds.
2907 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2908 address in a domain literal was a prefix of an interface address.
2910 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2911 when verifying a sender address, unless rewrite_headers is false.
2913 PH/44 Wrote a long comment about why errors_to addresses are verified as
2914 recipients, not senders.
2916 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2917 the ratelimit ACL was added.
2919 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2921 PH/46 Added extra information about PostgreSQL errors to the error string.
2923 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2924 avoiding the need to install special zones in a real server. This is
2925 backwards compatible; if it can't find the fake resolver, it drops back.
2926 Thus, both old and new test suites can be run.
2928 TF/02 Added util/ratelimit.pl
2930 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2931 clock is set back in time.
2933 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2934 Candler <B.Candler@pobox.com>.
2936 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2937 OK for addresses that are the result of redirections.
2939 PH/48 A number of further additions for the benefit of the new test suite,
2940 including a fake gethostbyname() that interfaces to the fake DNS resolver
2943 TF/06 The fix for widen_domains has also been applied to qualify_single and
2944 search_parents which are the other dnslookup options that can cause
2947 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2950 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2952 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2953 service" if it encountered an MX record with an empty target hostname.
2954 The message is now "an MX or SRV record indicated no SMTP service".
2956 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2957 if the DNS of the sending site is misconfigured. This is quite a
2958 common situation. This change restores the behaviour of treating a
2959 helo verification defer as a failure.
2961 PH/51 If self=fail was set on a router, the bounce message did not include the
2962 actual error message.
2968 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2970 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2971 command itself ended in error, the underlying message about the transport
2972 filter timeout was being overwritten with the pipe command error. Now the
2973 underlying error message should be appended to the second error message.
2975 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2977 PH/02 Reduce the amount of output that "make" produces by default. Full output
2978 can still be requested.
2980 PH/03 The warning log line about a condition test deferring for a "warn" verb
2981 was being output only once per connection, rather than after each
2982 occurrence (because it was using the same function as for successful
2983 "warn" verbs). This seems wrong, so I have changed it.
2985 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2986 it should not have, which might have caused a crash in the right
2987 circumstances, but probably never did.
2989 PH/04 Installed a modified version of Tony Finch's patch to make submission
2990 mode fix the return path as well as the Sender: header line, and to
2991 add a /name= option so that you can make the user's friendly name appear
2994 TF/03 Added the control = fakedefer ACL modifier.
2996 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2997 Mark Lowes for thorough testing.
2999 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3001 TK/03 Merged latest SRS patch from Miles Wilton.
3003 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3004 with the definition in sysexits.h (which is #included earlier).
3005 Fortunately, Exim does not actually use EX_OK. The code used to try to
3006 preserve the sysexits.h value, by assumimg that macro definitions were
3007 scanned for macro replacements. I have been disabused of this notion,
3008 so now the code just undefines EX_OK before #including unistd.h.
3010 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3011 in the smtp transport. When a block could not be written in a single
3012 write() function, the timeout was being re-applied to each part-write.
3013 This seems wrong - if the receiver was accepting one byte at a time it
3014 would take for ever. The timeout is now adjusted when this happens. It
3015 doesn't have to be particularly precise.
3017 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3018 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3020 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3021 without a domain if the check to <postmaster@domain> fails.
3023 SC/01 Eximstats: added -xls and the ability to specify output files
3024 (patch written by Frank Heydlauf).
3026 SC/02 Eximstats: use FileHandles for outputing results.
3028 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3030 SC/04 Eximstats: fixed display of large numbers with -nvr option
3032 SC/05 Eximstats: fixed merging of reports with empty tables.
3034 SC/06 Eximstats: added the -include_original_destination flag
3036 SC/07 Eximstats: removed tabs and trailing whitespace.
3038 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3040 TK/06 MBOX spool code: Add real "From " MBOX separator line
3041 so the .eml file is really in mbox format (even though
3042 most programs do not really care). Patch from Alex Miller.
3044 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3045 The latter is generated from $received_to and is only set if the
3046 message has one envelope recipient. SA can use these headers,
3047 obviously out-of-the-box. Patch from Alex Miller.
3049 PH/08 The ${def test on a variable was returning false if the variable's
3050 value was "0", contrary to what the specification has always said!
3051 The result should be true unless the variable is empty.
3053 PH/09 The syntax error of a character other than { following "${if
3054 def:variable_name" (after optional whitespace) was not being diagnosed.
3055 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3056 accidental colon was present, for example, could give incorrect results.
3058 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3059 result is used (not including appendfile, where other changes are about
3062 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3063 This involved changing a lot of size variables from int to off_t. It
3064 should work with maildirs and everything.
3066 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3067 spamd dying while we are connected to it.
3069 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3072 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3073 function for BDB 4.3.
3075 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3077 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3078 the sources for PCRE have been re-arranged and more files are now
3081 PH/15 The code I had for printing potentially long long variables in PH/11
3082 above was not the best (it lost precision). The length of off_t variables
3083 is now inspected at build time, and an appropriate printing format (%ld
3084 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3085 to be "long long int" or "long int". This is needed for the internal
3086 formatting function string_vformat().
3088 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3089 the configuration file to be ":syslog", then the script "guesses" where
3090 the logs files are, rather than using the compiled in default. In our
3091 case the guess is not the same as the compiled default, so the script
3092 suddenly stopped working when I started to use syslog. The patch checks
3093 to see if log_file_path is "". If so, it attempts to read it from exim
3094 with no configuration file to get the compiled in version, before it
3095 falls back to the previous guessing code."
3097 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3098 implementing BATV in an Exim configuration. See NewStuff for the gory
3101 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3102 Makefile that are specific to HP-UX.
3104 PH/18 If the "use_postmaster" option was set for a recipient callout together
3105 with the "random" option, the postmaster address was used as the MAIL
3106 FROM address for the random test, but not for the subsequent recipient
3107 test. It is now used for both.
3109 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3110 patch removes a few documentation additions to RFC 3028, because the
3111 latest draft now contains them. It adds the new en;ascii-case comparator
3112 and a new error check for 8bit text in MIME parts. Comparator and
3113 require names are now matched exactly. I enabled the subaddress
3114 extension, but it is not well tested yet (read: it works for me)."
3116 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3117 rework some of the code of TK/09 above to avoid the hardwired use of
3118 "%lld" and "long long". Replaced the call to snprintf() with a call to
3121 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3122 records point to non-existent hosts", "retry timeout exceeded", and
3123 "retry time not reached for any host after a long failure period".
3125 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3126 experimental DomainKeys support:
3128 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3129 (2) On an error such as an illegally used "control", the wrong name for
3130 the control was given.
3132 These problems did NOT occur unless DomainKeys support was compiled.
3134 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3136 PH/24 Added ${if match_ip condition.
3138 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3139 databases so that it will be absolutely obvious if a crash occurs in the
3140 DB library. This is a regular occurrence (often caused by mis-matched
3143 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3144 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3145 picked up on a user's system that detects such things. There doesn't seem
3146 to be a gcc warning option for this - only an attribute that has to be
3147 put on the function's prototype. It seems that in Fedora Core 4 they have
3148 set this on a number of new functions. No doubt there will be more in due
3151 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3152 specify a transport. However, if an address that was verified by such a
3153 router was the subject of a callout, Exim crashed because it tried to
3154 read the rcpt_include_affixes from the non-existent transport. Now it
3155 just assumes that the setting of that option is false. This bug was
3156 introduced by 4.51/PH/31.
3158 PH/28 Changed -d+all to exclude +memory, because that information is very
3159 rarely of interest, but it makes the output a lot bigger. People tend to
3160 do -d+all out of habit.
3162 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3163 code in os-type was giving problems when libc.so lives in lib64, like on
3166 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3167 aren't the modern standard, and it seems that some systems' include files
3168 don't always have them. Exim was already checking for some of the newer
3169 ones like T_AAAA, and defining it itself. I've added checks for all the
3170 record types that Exim uses.
3172 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3173 not automatically generating a new one, as it is supposed to. This
3174 prevented TLS from working. If the file did exist, but contained invalid
3175 data, a new version was generated, as expected. It was only the case of a
3176 non-existent file that was broken.
3178 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3179 with a change in libdomainkeys > 0.64.
3181 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3182 from DNS. If the selector record carries the flag, it now has
3183 precedence over the domain-wide flag.
3185 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3187 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3188 the use of a port name that isn't defined in /etc/services) occurred, the
3189 message was deferred as in a normal delivery, and thus remained on the
3190 spool, instead of being failed because of the mua_wrapper setting. This
3191 is now fixed, and I tidied up some of the mua_wrapper messages at the
3194 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3195 the messages in a hash of arrays rather than using individual hashes.
3196 This is a bit cleaner and results in dramatic memory savings, albeit
3197 at a slight CPU cost.
3199 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3200 as requested by Marc Sherman.
3202 SC/10 Eximstats: added histograms for user specified patterns as requested
3205 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3207 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3208 fopen() in the content-scanning modules that did not already have it.
3214 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3215 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3217 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3219 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3221 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3222 to test Sieve filters that use "vacation".
3224 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3225 that changes the way the GnuTLS parameters are stored in the cache file.
3226 The new format can be generated externally. For backward compatibility,
3227 if the data in the cache doesn't make sense, Exim assumes it has read an
3228 old-format file, and it generates new data and writes a new file. This
3229 means that you can't go back to an older release without removing the
3232 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3233 work if there are any delivery delays because "one_time" forces the
3234 parent to be marked "delivered", so its unseen clone is never tried
3235 again. For this reason, Exim now forbids the simultaneous setting of
3238 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3239 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3240 entry if you want to know the details. The fix, however, neglected to
3241 consider the case when local delivery batching is involved. The test for
3242 "previously delivered" was not happening when checking to see if an
3243 address could be batched with a previous (undelivered) one; under
3244 certain circumstances this could lead to multiple deliveries to the same
3247 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3248 in its include files, and this causes problems building Exim.
3250 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3251 header_syntax) but Exim was just ignoring anything given after a slash.
3252 In particular, this caused confusion with an attempt to use "verify =
3253 reverse_host_lookup/defer_ok". An error is now given when options are
3254 supplied for verify items that do not have them. (Maybe reverse_host_
3255 lookup should have a defer_ok option, but that's a different point.)
3257 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3258 defined by RFC 821) to 2048, because there were problems with some AUTH
3259 commands, and RFC 1869 says the size should be increased for extended
3260 SMTP commands that take arguments.
3262 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3265 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3266 "unknown" error; now it says that the functionality isn't in the binary.
3268 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3269 an address' error message when a string expansion fails (syntax or
3270 whatever). Otherwise the password may appear in the log. Following change
3271 PH/42 below, there is no longer a chance of it appearing in a bounce
3274 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3276 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3277 of its addresses was ever tried. (Bugzilla bug #2.)
3279 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3280 the result incorrectly in the debug output. (It correctly added a newline
3281 to what was transported.)
3283 TF/01 Added $received_time.
3285 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3286 commented out examples of how to interface to a virus scanner and to
3287 SpamAssassin. Also added commented examples of av_scanner and
3288 spamd_address settings.
3290 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3291 and controls are allowed in which ACLs. There were a couple of minor
3292 errors. Some of the entries in the conditions table (which is a table of
3293 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3294 negation of where the condition IS allowed.
3296 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3298 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3299 header file does not have a version number, so I've had to invent a new
3300 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3301 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3302 radiusclient), but it was contributed by a Radius user.
3304 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3305 files or directories, for OpenSSL.
3307 PH/19 When an Exim process that is running as root has to create an Exim log
3308 file, it does so in a subprocess that runs as exim:exim so as to get the
3309 ownership right at creation (otherwise, other Exim processes might see
3310 the file with the wrong ownership). There was no test for failure of this
3311 fork() call, which would lead to the process getting stuck as it waited
3312 for a non-existent subprocess. Forks do occasionally fail when resources
3313 run out. I reviewed all the other calls to fork(); they all seem to check
3316 PH/20 When checking for unexpected SMTP input at connect time (before writing
3317 the banner), Exim was not dealing correctly with a non-positive return
3318 from the read() function. If the client had disconnected by this time,
3319 the result was a log entry for a synchronization error with an empty
3320 string after "input=" when read() returned zero. If read() returned -1
3321 (an event I could not check), uninitialized data bytes were printed.
3322 There were reports of junk text (parts of files, etc) appearing after
3325 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3327 PH/22 Added support for macro redefinition, and (re)definition in between
3328 driver and ACL definitions.
3330 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3331 forgetting to use the resulting value; it was using the unexpanded value.
3333 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3334 hadn't been configured. The fix is from Juergen Kreileder, who
3335 understands it better than I do:
3337 "Here's what I see happening with three configured cyrus_sasl
3338 authenticators configured (plain, login, cram-md5):
3340 On startup auth_cyrus_sasl_init() gets called for each of these.
3341 This means three calls to sasl_listmech() without a specified mech_list.
3342 => SASL tests which mechs of all available mechs actually work
3343 => three warnings about OTP not working
3344 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3346 With the patch, sasl_listmech() also gets called three times. But now
3347 SASL's mech_list option is set to the server_mech specified in the the
3348 authenticator. Or in other words, the answer from sasl_listmech()
3349 gets limited to just the mech you're testing for (which is different
3351 => the return list contains just 'plain' or 'login', 'cram-md5' or
3352 nothing depending on the value of ob->server_mech.
3354 I've just tested the patch: Authentication still works fine,
3355 unavailable mechs specified in the exim configuration are still
3356 caught, and the auth.log warnings about OTP are gone."
3358 PH/25 When debugging is enabled, the contents of the command line are added
3359 to the debugging output, even when log_selector=+arguments is not
3362 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3363 answer is "GNU", and only if the return is "GNU/something" is the answer
3366 PH/27 $acl_verify_message is now set immediately after the failure of a
3367 verification in an ACL, and so is available in subsequent modifiers. In
3368 particular, the message can be preserved by coding like this:
3370 warn !verify = sender
3371 set acl_m0 = $acl_verify_message
3373 Previously, $acl_verify_message was set only while expanding "message"
3374 and "log_message" when a very denied access.
3376 PH/28 Modified OS/os.c-Linux with
3378 -#ifndef OS_LOAD_AVERAGE
3379 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3381 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3382 nomenclature these days.)
3384 PH/29 Installed patch from the Sieve maintainer that adds the options
3385 sieve_useraddress and sieve_subaddress to the redirect router.
3387 PH/30 In these circumstances:
3388 . Two addresses routed to the same list of hosts;
3389 . First host does not offer TLS;
3390 . First host accepts first address;
3391 . First host gives temporary error to second address;
3392 . Second host offers TLS and a TLS session is established;
3393 . Second host accepts second address.
3394 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3395 and peerdn, if requested) that were in fact used only for the second
3398 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3399 attention to any local part prefix or suffix that was matched by the
3400 router that accepted the address. It now behaves in the same way as it
3401 does for delivery: the affixes are removed from the local part unless
3402 rcpt_include_affixes is set on the transport.
3404 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3405 timeout during the DATA phase of an incoming message.
3407 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3408 applied a patch sanctioned by the Sieve maintainer.
3410 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3411 the uid or gid is negative. A case of a negative gid caused this to be
3412 noticed. The fix allows for either to be negative.
3414 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3415 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3418 PH/36 The change PH/12 above was broken. Fixed it.
3420 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3421 the grounds that routing the same address twice would always produce the
3422 same answer. This might have been true once, but it is certainly no
3423 longer true now. Routing a child address may depend on the previous
3424 routing that produced that child. Some complicated redirection strategies
3425 went wrong when messages had multiple recipients, and made Exim's
3426 behaviour dependent on the order in which the addresses were given.
3428 I have moved the duplicate checking until after the routing is complete.
3429 Exim scans the addresses that are assigned to local and remote
3430 transports, and removes any duplicates. This means that more work will be
3431 done, as duplicates will always all be routed, but duplicates are
3432 presumably rare, so I don't expect this is of any significance.
3434 For deliveries to pipes, files, and autoreplies, the duplicate checking
3435 still happens during the routing process, since they are not going to be
3438 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3439 It corrects a timeout issue with spamd. This is Ian's comment: "The
3440 background is that sometimes spamd either never reads data from a
3441 connection it has accepted, or it never writes response data. The exiscan
3442 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3443 blindly assumes that writes won't block so it may never time out."
3445 PH/39 Allow G after quota size as well as K and M.
3447 PH/40 The value set for $authenticated_id in an authenticator may not contain
3448 binary zeroes or newlines because the value is written to log lines and
3449 to spool files. There was no check on this. Now the value is run through
3450 the string_printing() function so that such characters are converted to
3451 printable escape sequences.
3453 PH/41 $message_linecount is a new variable that contains the total number of
3454 lines in the message. Compare $body_linecount, which is the count for the
3457 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3458 bounce and delay warning messages, except in certain special cases, which
3461 (a) An SMTP error message from a remote host;
3462 (b) A message specified in a :fail: redirection;
3463 (c) A message specified in a "fail" command in a system filter;
3464 (d) A message specified in a FAIL return from the queryprogram router;
3465 (e) A message specified by the cannot_route_message router option.
3467 In these cases only, Exim does include the error details in bounce and
3468 warning messages. There are also a few cases where bland messages such
3469 as "unrouteable address" or "local delivery error" are given.
3471 PH/43 $value is now also set for the "else" part of a ${run expansion.
3473 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3474 being worked on, but at least Exim now implements the latest version to
3477 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3478 process to complete was treated as a delivery failure, a timeout while
3479 writing the message to the pipe was logged, but erroneously treated as a
3480 successful delivery. Such timeouts include transport filter timeouts. For
3481 consistency with the overall process timeout, these timeouts are now
3482 treated as errors, giving rise to delivery failures by default. However,
3483 there is now a new Boolean option for the pipe transport called
3484 timeout_defer, which, if set TRUE, converts the failures into defers for
3485 both kinds of timeout. A transport filter timeout is now identified in
3488 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3489 systems where "make" and "gmake" are different, calling "gmake" at top
3490 level broke things. I've arranged for the value of $(MAKE) to be passed
3491 from the Makefile to this script so that it can call the same version of
3495 A note about Exim versions 4.44 and 4.50
3496 ----------------------------------------
3498 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3499 changes of various kinds. As a consequence, a big documentation update was
3500 needed. This delayed the release for rather longer than seemed good, especially
3501 in the light of a couple of (minor) security issues. Therefore, the changes
3502 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3503 release. So 4.44 and 4.50 are in effect two different branches that both start
3506 I have left the 4.50 change log unchanged; it contains all the changes since
3507 4.43. The change log for 4.44 is below; many of its items are identical to
3508 those for 4.50. This seems to be the most sensible way to preserve the
3509 historical information.
3515 1. Minor wording change to the doc/README.SIEVE file.
3517 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3518 computation of the current number of files was incorrect.
3520 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3521 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3524 4. Give more explanation in the error message when the command for a transport
3525 filter fails to execute.
3527 5. There are several places where Exim runs a non-Exim command in a
3528 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3529 was being done only for the command run by the queryprogram router. It is
3530 now done for all such subprocesses. The other cases are: ${run, transport
3531 filters, and the commands run by the lmtp and pipe transports.
3533 6. Added CONFIGURE_GROUP build-time option.
3535 7. Some older OS have a limit of 256 on the maximum number of file
3536 descriptors. Exim was using setrlimit() to set 1000 as a large value
3537 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3538 systems. I've change it so that if it can't get 1000, it tries for 256.
3540 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3541 was an oversight, and furthermore, ever since the addition of extra
3542 controls (e.g. 4.43/32), the checks on when to allow different forms of
3543 "control" were broken. There should now be diagnostics for all cases when a
3544 control that does not make sense is encountered.
3546 9. Added the /retain_sender option to "control=submission".
3548 10. $recipients is now available in the predata ACL (oversight).
3550 11. Tidy the search cache before the fork to do a delivery from a message
3551 received from the command line. Otherwise the child will trigger a lookup
3552 failure and thereby defer the delivery if it tries to use (for example) a
3553 cached ldap connection that the parent has called unbind on.
3555 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3556 of $address_data from the recipient verification was clobbered by the
3557 sender verification.
3559 13. The value of address_data from a sender verification is now available in
3560 $sender_address_data in subsequent conditions in the ACL statement.
3562 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3564 15. Added a new option "connect=<time>" to callout options, to set a different
3567 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3568 was its contents. (It was OK if the option was not defined at all.)
3570 17. A "Completed" log line is now written for messages that are removed from
3571 the spool by the -Mrm option.
3573 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3574 information about exactly what failed.
3576 19. Added -dd to debug only the daemon process.
3578 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3579 handles timeouts, both on the server side and network timeouts. Renamed the
3580 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3582 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3583 It is now set to "smtps".
3585 22. $host_address is now set to the target address during the checking of
3586 ignore_target_hosts.
3588 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3589 being passed; this would have caused $sender_host_name to have been used if
3590 matching the list had actually called for a host name (not very likely,
3591 since this list is usually IP addresses). A host name is now passed as
3594 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3595 code that specifies a non-restarting handler (typically sigaction() in
3596 modern systems) in an attempt to fix a rare and obscure crash bug.
3598 25. Narrowed the window for a race in the daemon that could cause it to ignore
3599 SIGCHLD signals. This is not a major problem, because they are used only to
3600 wake it up if nothing else does.
3602 26. A malformed maildirsize file could cause Exim to calculate negative values
3603 for the mailbox size or file count. Odd effects could occur as a result.
3604 The maildirsize information is now recalculated if the size or filecount
3607 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3608 support for a long time. Removed HAVE_SYS_VFS_H.
3610 28. Installed the latest version of exipick from John Jetmore.
3612 29. In an address list, if the pattern was not a regular expression, an empty
3613 subject address (from a bounce message) matched only if the pattern was an
3614 empty string. Non-empty patterns were not even tested. This was the wrong
3615 because it is perfectly reasonable to use an empty address as part of a
3616 database query. An empty address is now tested by patterns that are
3617 lookups. However, all the other forms of pattern expect the subject to
3618 contain a local part and a domain, and therefore, for them, an empty
3619 address still always fails if the pattern is not itself empty.
3621 30. Exim went into a mad DNS loop when attempting to do a callout where the
3622 host was specified on an smtp transport, and looking it up yielded more
3623 than one IP address.
3625 31. Re-factored the code for checking spool and log partition space into a
3626 function that finds that data and another that does the check. The former
3627 is then used to implement four new variables: $spool_space, $log_space,
3628 $spool_inodes, and $log_inodes.
3630 32. The RFC2047 encoding function was originally intended for short strings
3631 such as real names; it was not keeping to the 75-character limit for
3632 encoded words that the RFC imposes. It now respects the limit, and
3633 generates multiple encoded words if necessary. To be on the safe side, I
3634 have increased the buffer size for the ${rfc2047: expansion operator from
3637 33. It is now permitted to omit both strings after an "if" condition; if the
3638 condition is true, the result is "true". As before, when the second string
3639 is omitted, a false condition yields an empty string. This makes it less
3640 cumbersome to write custom ACL and router conditions.
3642 34. Failure to deliver a bounce message always caused it to be frozen, even if
3643 there was an errors_to setting on the router. The errors_to setting is now
3646 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3647 canonical form (fully expanded) before being placed in
3648 $sender_host_address.
3650 36. The table in the code that translates DNS record types into text (T_A to
3651 "A" for instance) was missing entries for NS and CNAME. It is just possible
3652 that this could have caused confusion if both these types were looked up
3653 for the same domain, because the text type is used as part of Exim's
3654 per-process caching. But the chance of anyone hitting this buglet seems
3657 37. The dnsdb lookup has been extended in a number of ways.
3659 (1) There is a new type, "zns", which walks up the domain tree until it
3660 finds some nameserver records. It should be used with care.
3662 (2) There is a new type, "mxh", which is like "mx" except that it returns
3663 just the host names, not the priorities.
3665 (3) It is now possible to give a list of domains (or IP addresses) to be
3666 looked up. The behaviour when one of the lookups defers can be
3667 controlled by a keyword.
3669 (4) It is now possible to specify the separator character for use when
3670 multiple records are returned.
3672 38. The dnslists ACL condition has been extended: it is now possible to supply
3673 a list of IP addresses and/or domains to be looked up in a particular DNS
3676 39. Added log_selector=+queue_time_overall.
3678 40. When running the queue in the test harness, wait just a tad after forking a
3679 delivery process, to get repeatability of debugging output.
3681 41. Include certificate and key file names in error message when GnuTLS fails
3682 to set them up, because the GnuTLS error message doesn't include the name
3683 of the failing file when there is a problem reading it.
3685 42. Allow both -bf and -bF in the same test run.
3687 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3689 44. The "Exiscan patch" is now merged into the mainline Exim source.
3691 45. Sometimes the final signoff response after QUIT could fail to get
3692 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3693 < 0 before doing a fflush(). This bug looks as though it goes back to the
3694 introduction of TLS in release 3.20, but "sometimes" must have been rare
3695 because the tests only now provoked it.
3697 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3698 (this can affect the format of dates).
3700 47. exim_tidydb, when checking for the continued existence of a message for
3701 which it has found a message-specific retry record, was not finding
3702 messages that were in split spool directories. Consequently, it was
3703 deleting retry records that should have stayed in existence.
3705 48. Steve fixed some bugs in eximstats.
3707 49. The SPA authentication driver was not abandoning authentication and moving
3708 on to the next authenticator when an expansion was forced to fail,
3709 contradicting the general specification for all authenticators. Instead it
3710 was generating a temporary error. It now behaves as specified.
3712 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3713 (the order specifies the preference for clients). The order is now AES256,
3714 AES128, 3DES, ARCFOUR128.
3716 51. Small patch to Sieve code - explicitly set From: when generating an
3719 52. Exim crashed if a remote delivery caused a very long error message to be
3720 recorded - for instance if somebody sent an entire SpamAssassin report back
3721 as a large number of 550 error lines. This bug was coincidentally fixed by
3722 increasing the size of one of Exim's internal buffers (big_buffer) that
3723 happened as part of the Exiscan merge. However, to be on the safe side, I
3724 have made the code more robust (and fixed the comments that describe what
3727 53. Now that there can be additional text after "Completed" in log lines (if
3728 the queue_time_overall log selector is set), a one-byte patch to exigrep
3729 was needed to allow it to recognize "Completed" as not the last thing in
3732 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3733 patch that reportedly fixes this has been added. I am not expert enough to
3734 create a test for it. This is what the patch creator wrote:
3736 "I found a little strange behaviour of ldap code when working with
3737 Windows 2003 AD Domain, where users was placed in more than one
3738 Organization Units. When I tried to give exim partial DN, the exit code
3739 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3740 But simultaneously result of request was absolutely normal ldap result,
3741 so I produce this patch..."
3743 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3744 REFERENCE, so I have modified the code to exclude the patch when that macro
3747 55. Some experimental protocols are using DNS PTR records for new purposes. The
3748 keys for these records are domain names, not reversed IP addresses. The
3749 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3750 leaves it alone. Component reversal etc. now happens only for IP addresses.
3753 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3755 57. Double the size of the debug message buffer (to 2048) so that more of very
3756 long debug lines gets shown.
3758 58. The exicyclog utility now does better if the number of log files to keep
3759 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3761 59. Two changes related to the smtp_active_hostname option:
3763 (1) $smtp_active_hostname is now available as a variable.
3764 (2) The default for smtp_banner uses $smtp_active_hostname instead
3765 of $primary_hostname.
3767 60. The host_aton() function is supposed to be passed a string that is known
3768 to be a valid IP address. However, in the case of IPv6 addresses, it was
3769 not checking this. This is a hostage to fortune. Exim now panics and dies
3770 if the condition is not met. A case was found where this could be provoked
3771 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3772 components; fortuitously, this particular loophole had already been fixed
3773 by change 4.50/55 above.
3775 If there are any other similar loopholes, the new check in host_aton()
3776 itself should stop them being exploited. The report I received stated that
3777 data on the command line could provoke the exploit when Exim was running as
3778 exim, but did not say which command line option was involved. All I could
3779 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3780 running as the user.
3783 61. There was a buffer overflow vulnerability in the SPA authentication code
3784 (which came originally from the Samba project). I have added a test to the
3785 spa_base64_to_bits() function which I hope fixes it.
3788 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3789 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3791 63. The daemon start-up calls getloadavg() while still root for those OS that
3792 need the first call to be done as root, but it missed one case: when
3793 deliver_queue_load_max is set with deliver_drop_privilege. This is
3794 necessary for the benefit of the queue runner, because there is no re-exec
3795 when deliver_drop_privilege is set.
3797 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3798 This has been fixed.
3800 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3801 was in use, was not putting the data itself into the right store pool;
3802 consequently, it could be overwritten for a subsequent message in the same
3803 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3806 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3808 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3809 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3810 the function were treating the return as a boolean value, which happened to
3811 work because 0=false and not-0=true, but is not correct code.
3813 68. The host_aton() function was not handling scoped IPv6 addresses (those
3814 with, for example, "%eth0" on the end) correctly.
3816 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3817 negated items (that is, ~something) in unsigned ints. Some compilers
3818 apparently mutter when there is no cast.
3820 70. If an address verification called from an ACL failed, and did not produce a
3821 user-specific message (i.e. there was only a "system" message), nothing was
3822 put in $acl_verify_message. In this situation, it now puts the system
3825 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3826 session; change 4.31/43 added the unwanted input to the log line - except
3827 that it did not do this in the start of session case. It now does.
3829 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3830 This is wrong and can cause the other end to generate a synchronization
3831 error if it is another Exim or anything else that does the synchronization
3832 check. A QUIT command is no longer sent after a timeout.
3834 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3835 during host lookups.
3837 74. The defer_ok option of callout verification was not working if it was used
3838 when verifying addresses in header lines, that is, for this case:
3840 verify = header_sender/callout=defer_ok
3842 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3843 those file descriptors could be used for SMTP connections. If anything
3844 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3845 could be sent to the SMTP client, causing chaos. The daemon now opens
3846 stdin, stdout, and stderr to /dev/null when it puts itself into the
3849 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3850 log by default. The user can override this with suitable Perl magic.
3852 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3853 the log message when discard triggers, was not working for the DATA ACL or
3854 for the non-SMTP ACL.
3856 78. Error message wording change in sieve.c.
3858 79. If smtp_accept_max_per_host was set, the number of connections could be
3859 restricted to fewer than expected, because the daemon was trying to set up
3860 a new connection before checking whether the processes handling previous
3861 connections had finished. The check for completed processes is now done
3862 earlier. On busy systems, this bug wouldn't be noticed because something
3863 else would have woken the daemon, and it would have reaped the completed
3866 80. If a message was submitted locally by a user whose login name contained one
3867 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3868 It caused a spool format error. I have fixed the spool reading code. A
3869 related problem was that the "from" clause in the Received: line became
3870 illegal because of the space(s). It is now covered by ${quote_local_part.
3872 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3875 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3877 83. Patch from Sieve maintainer to fix unterminated string problem in
3878 "vacation" handling.
3880 84. Some minor changes to the Linux configuration files to help with other
3881 OS variants using glibc.
3883 85. One more patch for Sieve to update vacation handling to latest spec.
3886 ----------------------------------------------------
3887 See the note above about the 4.44 and 4.50 releases.
3888 ----------------------------------------------------
3894 1. Change 4.43/35 introduced a bug that caused file counts to be
3895 incorrectly computed when quota_filecount was set in an appendfile
3898 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3899 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3902 3. Give more explanation in the error message when the command for a transport
3903 filter fails to execute.
3905 4. There are several places where Exim runs a non-Exim command in a
3906 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3907 was being done only for the command run by the queryprogram router. It is
3908 now done for all such subprocesses. The other cases are: ${run, transport
3909 filters, and the commands run by the lmtp and pipe transports.
3911 5. Some older OS have a limit of 256 on the maximum number of file
3912 descriptors. Exim was using setrlimit() to set 1000 as a large value
3913 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3914 systems. I've change it so that if it can't get 1000, it tries for 256.
3916 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3917 was an oversight, and furthermore, ever since the addition of extra
3918 controls (e.g. 4.43/32), the checks on when to allow different forms of
3919 "control" were broken. There should now be diagnostics for all cases when a
3920 control that does not make sense is encountered.
3922 7. $recipients is now available in the predata ACL (oversight).
3924 8. Tidy the search cache before the fork to do a delivery from a message
3925 received from the command line. Otherwise the child will trigger a lookup
3926 failure and thereby defer the delivery if it tries to use (for example) a
3927 cached ldap connection that the parent has called unbind on.
3929 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3930 of $address_data from the recipient verification was clobbered by the
3931 sender verification.
3933 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3934 was its contents. (It was OK if the option was not defined at all.)
3936 11. A "Completed" log line is now written for messages that are removed from
3937 the spool by the -Mrm option.
3939 12. $host_address is now set to the target address during the checking of
3940 ignore_target_hosts.
3942 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3943 being passed; this would have caused $sender_host_name to have been used if
3944 matching the list had actually called for a host name (not very likely,
3945 since this list is usually IP addresses). A host name is now passed as
3948 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3949 code that specifies a non-restarting handler (typically sigaction() in
3950 modern systems) in an attempt to fix a rare and obscure crash bug.
3952 15. Narrowed the window for a race in the daemon that could cause it to ignore
3953 SIGCHLD signals. This is not a major problem, because they are used only to
3954 wake it up if nothing else does.
3956 16. A malformed maildirsize file could cause Exim to calculate negative values
3957 for the mailbox size or file count. Odd effects could occur as a result.
3958 The maildirsize information is now recalculated if the size or filecount
3961 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3962 support for a long time. Removed HAVE_SYS_VFS_H.
3964 18. Updated exipick to current release from John Jetmore.
3966 19. Allow an empty sender to be matched against a lookup in an address list.
3967 Previously the only cases considered were a regular expression, or an
3970 20. Exim went into a mad DNS lookup loop when doing a callout where the
3971 host was specified on the transport, if the DNS lookup yielded more than
3974 21. The RFC2047 encoding function was originally intended for short strings
3975 such as real names; it was not keeping to the 75-character limit for
3976 encoded words that the RFC imposes. It now respects the limit, and
3977 generates multiple encoded words if necessary. To be on the safe side, I
3978 have increased the buffer size for the ${rfc2047: expansion operator from
3981 22. Failure to deliver a bounce message always caused it to be frozen, even if
3982 there was an errors_to setting on the router. The errors_to setting is now
3985 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3986 canonical form (fully expanded) before being placed in
3987 $sender_host_address.
3989 24. Updated eximstats to version 1.33
3991 25. Include certificate and key file names in error message when GnuTLS fails
3992 to set them up, because the GnuTLS error message doesn't include the name
3993 of the failing file when there is a problem reading it.
3995 26. Expand error message when OpenSSL has problems setting up cert/key files.
3998 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3999 (this can affect the format of dates).
4001 28. exim_tidydb, when checking for the continued existence of a message for
4002 which it has found a message-specific retry record, was not finding
4003 messages that were in split spool directories. Consequently, it was
4004 deleting retry records that should have stayed in existence.
4006 29. eximstats updated to version 1.35
4007 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4008 1.35 - bugfix such that pie charts by volume are generated correctly
4010 30. The SPA authentication driver was not abandoning authentication and moving
4011 on to the next authenticator when an expansion was forced to fail,
4012 contradicting the general specification for all authenticators. Instead it
4013 was generating a temporary error. It now behaves as specified.
4015 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4016 (the order specifies the preference for clients). The order is now AES256,
4017 AES128, 3DES, ARCFOUR128.
4019 31. Small patch to Sieve code - explicitly set From: when generating an
4022 32. Exim crashed if a remote delivery caused a very long error message to be
4023 recorded - for instance if somebody sent an entire SpamAssassin report back
4024 as a large number of 550 error lines. This bug was coincidentally fixed by
4025 increasing the size of one of Exim's internal buffers (big_buffer) that
4026 happened as part of the Exiscan merge. However, to be on the safe side, I
4027 have made the code more robust (and fixed the comments that describe what
4030 33. Some experimental protocols are using DNS PTR records for new purposes. The
4031 keys for these records are domain names, not reversed IP addresses. The
4032 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4033 leaves it alone. Component reversal etc. now happens only for IP addresses.
4036 34. The host_aton() function is supposed to be passed a string that is known
4037 to be a valid IP address. However, in the case of IPv6 addresses, it was
4038 not checking this. This is a hostage to fortune. Exim now panics and dies
4039 if the condition is not met. A case was found where this could be provoked
4040 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4041 components; fortuitously, this particular loophole had already been fixed
4042 by change 4.50/55 or 4.44/33 above.
4044 If there are any other similar loopholes, the new check in host_aton()
4045 itself should stop them being exploited. The report I received stated that
4046 data on the command line could provoke the exploit when Exim was running as
4047 exim, but did not say which command line option was involved. All I could
4048 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4049 running as the user.
4052 35. There was a buffer overflow vulnerability in the SPA authentication code
4053 (which came originally from the Samba project). I have added a test to the
4054 spa_base64_to_bits() function which I hope fixes it.
4057 36. The daemon start-up calls getloadavg() while still root for those OS that
4058 need the first call to be done as root, but it missed one case: when
4059 deliver_queue_load_max is set with deliver_drop_privilege. This is
4060 necessary for the benefit of the queue runner, because there is no re-exec
4061 when deliver_drop_privilege is set.
4063 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4064 was in use, was not putting the data itself into the right store pool;
4065 consequently, it could be overwritten for a subsequent message in the same
4066 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4069 38. Sometimes the final signoff response after QUIT could fail to get
4070 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4071 < 0 before doing a fflush(). This bug looks as though it goes back to the
4072 introduction of TLS in release 3.20, but "sometimes" must have been rare
4073 because the tests only now provoked it.
4079 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4080 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4081 Now it returns an integer. A number of calls were still expecting a T/F
4082 return. Fortuitously, in all cases, the tests worked in OK situations,
4083 which is the norm. However, things would have gone wrong on any write
4084 failures on the smtp file descriptor. This function is used when sending
4085 messages over SMTP and also when doing verify callouts.
4087 2. When Exim is called to do synchronous delivery of a locally submitted
4088 message (the -odf or -odi options), it no longer closes stderr before doing
4091 3. Implemented the mua_wrapper option.
4093 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4095 5. Implemented the functions header_remove(), header_testname(),
4096 header_add_at_position(), and receive_remove_recipient(), and exported them
4099 6. If an ACL "warn" statement specified the addition of headers, Exim already
4100 inserted X-ACL-Warn: at the start if there was no header name. However, it
4101 was not making this test for the second and subsequent header lines if
4102 there were newlines in the string. This meant that an invalid header could
4103 be inserted if Exim was badly configured.
4105 7. Allow an ACL "warn" statement to add header lines at the start or after all
4106 the Received: headers, as well as at the end.
4108 8. Added the rcpt_4xx retry error code.
4110 9. Added postmaster_mailfrom=xxx to callout verification option.
4112 10. Added mailfrom=xxxx to the callout verification option, for verify=
4115 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4116 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4118 12. Inserted some casts to stop certain compilers warning when using pointer
4119 differences as field lengths or precisions in printf-type calls (mostly
4120 affecting debugging statements).
4122 13. Added optional readline() support for -be (dynamically loaded).
4124 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4125 same clock tick as a message's arrival, so that its received time was the
4126 same as the "first fail" time on the retry record, and that message
4127 remained on the queue past the ultimate address timeout, every queue runner
4128 would try a delivery (because it was past the ultimate address timeout) but
4129 after another failure, the ultimate address timeout, which should have then
4130 bounced the address, did not kick in. This was a "< instead of <=" error;
4131 in most cases the first failure would have been in the next clock tick
4132 after the received time, and all would be well.
4134 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4135 being recognized when the domain list was tested by the match_domain
4136 condition in an expansion string.
4138 16. Added the ${str2b64: operator.
4140 17. Exim was always calling setrlimit() to set a large limit for the number of
4141 processes, without checking whether the existing limit was already
4142 adequate. (It did check for the limit on file descriptors.) Furthermore,
4143 errors from getrlimit() and setrlimit() were being ignored. Now they are
4144 logged to the main and panic logs, but Exim does carry on, to try to do its
4145 job under whatever limits there are.
4147 18. Imported PCRE 5.0.
4149 19. Trivial typo in log message " temporarily refused connection" (the leading
4152 20. If the log selector return_path_on_delivery was set and an address was
4153 redirected to /dev/null, the delivery process crashed because it assumed
4154 that a return path would always be set for a "successful" delivery. In this
4155 case, the whole delivery is bypassed as an optimization, and therefore no
4158 21. Internal re-arrangement: the function for sending a challenge and reading
4159 a response while authentication was assuming a zero-terminated challenge
4160 string. It's now changed to take a pointer and a length, to allow for
4161 binary data in such strings.
4163 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4165 23. Exim was not respecting finduser_retries when seeking the login of the
4166 uid under which it was called; it was always trying 10 times. (The default
4167 setting of finduser_retries is zero.) Also, it was sleeping after the final
4168 failure, which is pointless.
4170 24. Implemented tls_on_connect_ports.
4172 25. Implemented acl_smtp_predata.
4174 26. If the domain in control=submission is set empty, Exim assumes that the
4175 authenticated id is a complete email address when it generates From: or
4176 Sender: header lines.
4178 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4179 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4180 chown and chgrp in /bin and hostname in /usr/bin.
4182 28. Exim was keeping the "process log" file open after each use, just as it
4183 does for the main log. This opens the possibility of it remaining open for
4184 long periods when the USR1 signal hits a daemon. Occasional processlog
4185 errors were reported, that could have been caused by this. Anyway, it seems
4186 much more sensible not to leave this file open at all, so that is what now
4189 29. The long-running daemon process does not normally write to the log once it
4190 has entered its main loop, and it closes the log before doing so. This is
4191 so that log files can straightforwardly be renamed and moved. However,
4192 there are a couple of unusual error situations where the daemon does write
4193 log entries, and I had neglected to close the log afterwards.
4195 30. The text of an SMTP error response that was received during a remote
4196 delivery was being truncated at 512 bytes. This is too short for some of
4197 the long messages that one sometimes sees. I've increased the limit to
4200 31. It is now possible to make retry rules that apply only when a message has a
4201 specific sender, in particular, an empty sender.
4203 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4204 it possible to be selective about when SMTP synchronization is enforced.
4206 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4208 32. Implemented hosts_connection_nolog.
4210 33. Added an ACL for QUIT.
4212 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4215 35. Added mailbox_size and mailbox_filecount to appendfile.
4217 36. Added control = no_multiline_responses to ACLs.
4219 37. There was a bug in the logic of the code that waits for the clock to tick
4220 in the case where the clock went backwards by a substantial amount such
4221 that the microsecond fraction of "now" was more than the microsecond
4222 fraction of "then" (but the whole seconds number was less).
4224 38. Added support for the libradius Radius client library this is found on
4225 FreeBSD (previously only the radiusclient library was supported).
4231 1. When certain lookups returned multiple values in the form name=value, the
4232 quoting of the values was not always being done properly. Specifically:
4233 (a) If the value started with a double quote, but contained no whitespace,
4235 (b) If the value contained whitespace other than a space character (i.e.
4236 tabs or newlines or carriage returns) it was not quoted.
4237 This fix has been applied to the mysql and pgsql lookups by writing a
4238 separate quoting function and calling it from the lookup code. The fix
4239 should probably also be applied to nisplus, ibase and oracle lookups, but
4240 since I cannot test any of those, I have not disturbed their existing code.
4242 2. A hit in the callout cache for a specific address caused a log line with no
4243 reason for rejecting RCPT. Now it says "Previous (cached) callout
4244 verification failure".
4246 3. There was an off-by-one bug in the queryprogram router. An over-long
4247 return line was truncated at 256 instead of 255 characters, thereby
4248 overflowing its buffer with the terminating zero. As well as fixing this, I
4249 have increased the buffer size to 1024 (and made a note to document this).
4251 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4252 when Exim is waiting for an SMTP response from a remote server, Exim
4253 restarts its select() call on the socket, thereby resetting its timeout.
4254 This is not a problem when such interrupts are rare. Somebody set up a cron
4255 job to run exiwhat every 2 minutes, which is less than the normal select()
4256 timeout (5 or 10 minutes). This meant that the select() timeout never
4257 kicked in because it was always reset. I have fixed this by comparing the
4258 time when an interrupt arrives with the time at the start of the first call
4259 to select(). If more time than the timeout has elapsed, the interrupt is
4260 treated as a timeout.
4262 5. Some internal re-factoring in preparation for the addition of Sieve
4263 extensions (by MH). In particular, the "personal" test is moved to a
4264 separate function, and given an option for scanning Cc: and Bcc: (which is
4265 not set for Exim filters).
4267 6. When Exim created an email address using the login of the caller as the
4268 local part (e.g. when creating a From: or Sender: header line), it was not
4269 quoting the local part when it contained special characters such as @.
4271 7. Installed new OpenBSD configuration files.
4273 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4274 try to make them clearer.
4276 9. Callout options, other than the timeout value, were being ignored when
4277 verifying sender addresses in header lines. For example, when using
4279 verify = header_sender/callout=no_cache
4281 the cache was (incorrectly) being used.
4283 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4284 only the Cygwin environment.
4286 11. When return_path_on_delivery was set as a log selector, if different remote
4287 addresses in the same message used different return paths and parallel
4288 remote delivery occurred, the wrong values would sometimes be logged.
4289 (Whenever a remote delivery process finished, the return path value from
4290 the most recently started remote delivery process was logged.)
4292 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4293 lines when AUTH and/or TLS are in use. This is the "received protocol"
4294 field. Exim used to use "asmtp" for authenticated SMTP, without any
4295 indication (in the protocol name) for TLS use. Now it follows the RFC and
4296 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4297 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4298 names appear in log lines as well as in Received: header lines.
4300 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4301 extensions, and comparison tests, and to fix some bugs.
4303 14. Changes to the "personal" filter test:
4305 (1) The test was buggy in that it was just doing the equivalent of
4306 "contains" tests on header lines. For example, if a user's address was
4307 anne@some.where, the "personal" test would incorrectly be true for
4309 To: susanne@some.where
4311 This test is now done by extracting each address from the header in turn,
4312 and checking the entire address. Other tests that are part of "personal"
4313 are now done using regular expressions (for example, to check local parts
4314 of addresses in From: header lines).
4316 (2) The list of non-personal local parts in From: addresses has been
4317 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4318 taken from the Sieve specification recommendations.
4320 (3) If the message contains any header line starting with "List-" it is
4321 treated as non-personal.
4323 (4) The test for "circular" in the Subject: header line has been removed
4324 because it now seems ill-conceived.
4326 15. Minor typos in src/EDITME comments corrected.
4328 16. Installed latest exipick from John Jetmore.
4330 17. If headers_add on a router specified a text string that was too long for
4331 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4332 of string_sprintf() is now avoided.
4334 18. $message_body_size was not set (it was always zero) when running the DATA
4335 ACL and the local_scan() function.
4337 19. For the "mail" command in an Exim filter, no default was being set for
4338 the once_repeat time, causing a random time value to be used if "once" was
4339 specified. (If the value happened to be <= 0, no repeat happened.) The
4340 default is now 0s, meaning "never repeat". The "vacation" command was OK
4341 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4342 (I found it when inspecting the code).
4344 20. There is now an overall timeout for performing a callout verification. It
4345 defaults to 4 times the callout timeout, which applies to individual SMTP
4346 commands during the callout. The overall timeout applies when there is more
4347 than one host that can be tried. The timeout is checked before trying the
4348 next host. This prevents very long delays if there are a large number of
4349 hosts and all are timing out (e.g. when the network connections are timing
4350 out). The value of the overall timeout can be changed by specifying an
4351 additional sub-option for "callout", called "maxwait". For example:
4353 verify = sender/callout=5s,maxwait=20s
4355 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4356 to the end before writing, but this should make it even safer).
4358 22. Exim was forgetting that it had advertised PIPELINING for the second and
4359 subsequent messages on an SMTP connection. It was also not resetting its
4360 memory on STARTTLS and an internal HELO.
4362 23. When Exim logs an SMTP synchronization error within a session, it now
4363 records whether PIPELINING has been advertised or not.
4365 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4366 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4367 rather than long int.
4369 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4371 26. Added the never_mail option to autoreply.
4377 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4378 crash if the getsockname() call failed; this can happen if a connection is
4379 closed very soon after it is established. The problem was simply in the
4380 order in which certain operations were done, causing Exim to try to write
4381 to the SMTP stream before it had set up the file descriptor. The bug has
4382 been fixed by making things happen in the correct order.
4388 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4389 before the connection was closed, thus losing the rejection response.
4391 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4392 some early Solaris releases, but causes trouble in current releases where
4393 socklen_t is defined.
4395 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4398 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4401 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4402 that releasing the top of it at the end releases what was used for sub-
4403 expansions (unless the block got too big). However, discard this block if
4404 the first thing is a variable or header, so that we can use its block when
4405 it is dynamic (useful for very large $message_headers, for example).
4407 6. Lookups now cache *every* query, not just the most recent. A new, separate
4408 store pool is used for this. It can be recovered when all lookup caches are
4409 flushed. Lookups now release memory at the end of their result strings.
4410 This has involved some general refactoring of the lookup sources.
4412 7. Some code has been added to the store_xxx() functions to reduce the amount
4413 of flapping under certain conditions.
4415 8. log_incoming_interface used to affect only the <= reception log lines. Now
4416 it causes the local interface and port to be added to several more SMTP log
4417 lines, for example "SMTP connection from", and rejection lines.
4419 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4421 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4423 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4424 could be overwritten at the end of the current message (or the start of a
4425 new message if it was set in a HELO ACL). The value is now preserved for
4426 the duration of the SMTP connection.
4428 12. If a transport had a headers_rewrite setting, and a matching header line
4429 contained an unqualified address, that address was qualified, even if it
4430 did not match any rewriting rules. The underlying bug was that the values
4431 of the flags that permit the existence of unqualified sender and recipient
4432 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4433 non-local messages, and by -bnq for local messages) were not being
4434 preserved with the message after it was received.
4436 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4437 "next input=" as part of the text comprising the host identity instead of
4438 the correct text. The code was using the same buffer for two different
4439 strings. However, depending on which order the printing function evaluated
4440 its arguments, the bug did not always show up. Under Linux, for example, my
4441 test suite worked just fine.
4443 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4444 doesn't work with some older versions of Perl. It has been changed to "my",
4445 which in any case is probably the better facility to use.
4447 15. A really picky compiler found some instances of statements for creating
4448 error messages that either had too many or two few arguments for the format
4451 16. The size of the buffer for calls to the DNS resolver has been increased
4452 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4453 for addresses that have a lot of PTR records. This alleviates a problem; it
4454 does not fully solve it.
4456 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4457 buffer now truncates the list and logs the incident, which is the same
4458 action as happens when Exim is looking up a host name and its aliases.
4459 Previously in this situation something unpredictable would happen;
4460 sometimes it was "internal error: store_reset failed".
4462 18. If a server dropped the connection unexpectedly when an Exim client was
4463 using GnuTLS and trying to read a response, the client delivery process
4464 crashed while trying to generate an error log message.
4466 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4467 string, for example:
4469 warn message = H1: something\nH2: something
4471 the text was added as a single header line from Exim's point of view
4472 though it ended up OK in the delivered message. However, searching for the
4473 second and subsequent header lines using $h_h2: did not work. This has been
4474 fixed. Similarly, if a system filter added multiple headers in this way,
4475 the routers could not see them.
4477 20. Expanded the error message when iplsearch is called with an invalid key to
4478 suggest using net-iplsearch in a host list.
4480 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4481 ACLs are no longer actually imposed (and a message to that effect is
4484 22. If a "gecos" field in a passwd entry contained escaped characters, in
4485 particular, if it contained a \" sequence, Exim got it wrong when building
4486 a From: or a Sender: header from that name. A second bug also caused
4487 incorrect handling when an unquoted " was present following a character
4488 that needed quoting.
4490 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4491 was not being matched caselessly.
4493 24. Arranged for all hyphens in the exim.8 source to be escaped with
4496 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4497 properly. Recipient callout cache records were still being keyed to include
4498 the sender, even when use_sender was set false. This led to far more
4499 callouts that were necessary. The sender is no longer included in the key
4500 when use_sender is false.
4502 26. Added "control = submission" modifier to ACLs.
4504 27. Added the ${base62d: operator to decode base 62 numbers.
4506 28. dnsdb lookups can now access SRV records.
4508 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4509 the configuration file.
4511 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4512 (-v) mode. This makes the output for a verbose queue run more intelligible.
4514 31. Added a use_postmaster feature to recipient callouts.
4516 32. Added the $body_zerocount variable, containing the number of binary zero
4517 bytes in the message body.
4519 33. The time of last modification of the "new" subdirectory is now used as the
4520 "mailbox time last read" when there is a quota error for a maildir
4523 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4525 35. Added +ignore_unknown as a special item in host lists.
4527 36. Code for decoding IPv6 addresses in host lists is now included, even if
4528 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4529 address was recognized as an IP address, but was then not correctly decoded
4530 into binary, causing unexpected and incorrect effects when compared with
4537 1. Very minor rewording of debugging text in manualroute to say "list of
4538 hosts" instead of "hostlist".
4540 2. If verify=header_syntax was set, and a header line with an unqualified
4541 address (no domain) and a large number of spaces between the end of the
4542 name and the colon was received, the reception process suffered a buffer
4543 overflow, and (when I tested it) crashed. This was caused by some obsolete
4544 code that should have been removed. The fix is to remove it!
4546 3. When running in the test harness, delay a bit after writing a bounce
4547 message to get a bit more predictability in the log output.
4549 4. Added a call to search_tidyup() just before forking a reception process. In
4550 theory, someone could use a lookup in the expansion of smtp_accept_max_
4551 per_host which, without the tidyup, could leave open a database connection.
4553 5. Added the variables $recipient_data and $sender_data which get set from a
4554 lookup success in an ACL "recipients" or "senders" condition, or a router
4555 "senders" option, similar to $domain_data and $local_part_data.
4557 6. Moved the writing of debug_print from before to after the "senders" test
4560 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4561 problems for message scanning, either using a data ACL, or using
4562 local_scan() because the Received: header was not generated till after they
4563 were called (in order to set the time as the time of reception completion).
4564 I have revised the way this works. The header is now generated after the
4565 body is received, but before the ACL or local_scan() are called. After they
4566 are run, the timestamp in the header is updated.
4572 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4573 before starting a queue runner without re-exec. This happened only when
4574 deliver_drop_privilege was set or when the Exim user was set to root. The
4575 effect of the bug was that timeouts during subsequent deliveries caused
4576 crashes instead of being properly handled. The handler is now left at its
4577 default (and expected) setting.
4579 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4580 message, again when deliver_drop_privilege is set or Exim is run as root.
4581 The bug described in (1) was not present in this case, but the tidying up
4582 of the other signals was missing. I have made the two cases consistent.
4584 3. The ignore_target_hosts setting on a manualroute router was being ignored
4585 for hosts that were looked up using the /MX notation.
4587 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4590 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4591 operated on the sender address. After changing the $sender_address to <>
4592 for the sender address verify, Exim was re-instated it as the original
4593 (before rewriting) address, but remembering that it had rewritten it, so it
4594 wasn't rewriting it again. This bug also had the effect of breaking the
4595 sender address verification caching when the sender address was rewritten.
4597 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4598 This has been changed so that if the ip literal address matches
4599 ignore_target_hosts, the router declines.
4601 7. Added expansion conditions match_domain, match_address, and match_local_
4602 part (NOT match_host).
4604 8. The placeholder for the Received: header didn't have a length field set.
4606 9. Added code to Exim itself and to exim_lock to test for a specific race
4607 condition that could lead to file corruption when using MBX delivery. The
4608 issue is with the lockfile that is created in /tmp. If this file is removed
4609 after a process has opened it but before that process has acquired a lock,
4610 there is the potential for a second process to recreate the file and also
4611 acquire a lock. This could lead to two Exim processes writing to the file
4612 at the same time. The added code performs the same test as UW imapd; it
4613 checks after acquiring the lock that its file descriptor still refers to
4614 the same named file.
4616 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4617 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4620 11. Added the smtp_active_hostname option. If used, this will typically be made
4621 to depend on the incoming interface address. Because $interface_address is
4622 not set up until the daemon has forked a reception process, error responses
4623 that can happen earlier (such as "too many connections") no longer contain
4626 12. If an expansion in a condition on a "warn" statement fails because a lookup
4627 defers, the "warn" statement is abandoned, and the next ACL statement is
4628 processed. Previously this caused the whole ACL to be aborted.
4630 13. Added the iplsearch lookup type.
4632 14. Added ident_timeout as a log selector.
4634 15. Added tls_certificate_verified as a log selector.
4636 16. Added a global option tls_require_ciphers (compare the smtp transport
4637 option of the same name). This controls incoming TLS connections.
4639 17. I finally figured out how to make tls_require_ciphers do a similar thing
4640 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4641 before starting the TLS session.
4643 18. Tabs are now shown as \t in -bP output.
4645 19. If the log selector return_path_on_delivery was set, Exim crashed when
4646 bouncing a message because it had too many Received: header lines.
4648 20. If two routers both had headers_remove settings, and the first one included
4649 a superfluous trailing colon, the final name in the first list and the
4650 first name in the second list were incorrectly joined into one item (with a
4651 colon in the middle).
4657 1. Added -C and -D options to the exinext utility, mainly to make it easier
4658 to include in the automated testing, but these could be helpful when
4659 multiple configurations are in use.
4661 2. The exinext utility was not formatting the output nicely when there was
4662 an alternate port involved in the retry record key, nor when there was a
4663 message id as well (for retries that were specific to a specific message
4664 and a specific host). It was also confused by IPv6 addresses, because of
4665 the additional colons they contain. I have fixed the IPv4 problem, and
4666 patched it up to do a reasonable job for IPv6.
4668 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4669 delivery, the log line now contains "pipelined" if PIPELINING was used.
4671 4. An SMTP transport process used to panic and die if the bind() call to set
4672 an explicit outgoing interface failed. This has been changed; it is now
4673 treated in the same way as a connect() failure.
4675 5. A reference to $sender_host_name in the part of a conditional expansion
4676 that was being skipped was still causing a DNS lookup. This no longer
4679 6. The def: expansion condition was not recognizing references to header lines
4680 that used bh_ and bheader_.
4682 7. Added the _cache feature to named lists.
4684 8. The code for checking quota_filecount in the appendfile transport was
4685 allowing one more file than it should have been.
4687 9. For compatibility with Sendmail, the command line option
4695 and sets the incoming protocol and host name (for trusted callers). The
4696 host name and its colon can be omitted when only the protocol is to be set.
4697 Note the Exim already has two private options, -pd and -ps, that refer to
4698 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4699 "s", but I don't think that's a major issue.
4701 10. A number of refactoring changes to the code, none of which should affect
4704 (a) The number of logging options was getting close to filling up the
4705 32-bit word that was used as a bit map. I have split them into two classes:
4706 those that are passed in the argument to log_write(), and those that are
4707 only ever tested independently outside of that function. These are now in
4708 separate 32-bit words, so there is plenty of room for expansion again.
4709 There is no change in the user interface or the logging behaviour.
4711 (b) When building, for example, log lines, the code previously used a
4712 macro that called string_cat() twice, in order to add two strings. This is
4713 not really sufficiently general. Furthermore, there was one instance where
4714 it was actually wrong because one of the argument was used twice, and in
4715 one call a function was used. (As it happened, calling the function twice
4716 did not affect the overall behaviour.) The macro has been replaced by a
4717 function that can join an arbitrary number of extra strings onto a growing
4720 (c) The code for expansion conditions now uses a table and a binary chop
4721 instead of a serial search (which was left over from when there were very
4722 few conditions). Also, it now recognizes conditions like "pam" even when
4723 the relevant support is not compiled in: a suitably worded error message is
4724 given if an attempt is made to use such a condition.
4726 11. Added ${time_interval:xxxxx}.
4728 12. A bug was causing one of the ddress fields not to be passed back correctly
4729 from remote delivery subprocesses. The field in question was not being
4730 subsequently used, so this caused to problems in practice.
4732 13. Added new log selectors queue_time and deliver_time.
4734 14. Might have fixed a bug in maildirsizefile handling that threw up
4735 "unexpected character" debug warnings, and recalculated the data
4736 unnecessarily. In any case, I expanded the warning message to give more
4739 15. Added the message "Restricted characters in address" to the statements in
4740 the default ACL that block characters like @ and % in local parts.
4742 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4743 Three changes have been made:
4745 (a) There was a serious bug; a negative response to MAIL caused the whole
4746 recipient domain to be cached as invalid, thereby blocking all messages
4747 to all local parts at the same domain, from all senders. This bug has
4748 been fixed. The domain is no longer cached after a negative response to
4749 MAIL if the sender used is not empty.
4751 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4754 (c) A new callout option, "use_sender" has been added for people who want
4755 the modified behaviour.
4761 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4764 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4765 indeed breaks things for older releases.
4767 3. Added additional logging to the case where there is a problem reading data
4768 from a filter that is running in a subprocess using a pipe, in order to
4769 try to track down a specific problem.
4771 4. Testing facility fudge: when running in the test harness and attempting
4772 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4773 getting "No route to host". Convert this to a timeout.
4775 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4778 6. Some OS don't have socklen_t but use size_t instead. This affects the
4779 fifth argument of getsockopt() amongst other things. This is now
4780 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4781 can be set for individual OS. I have set it for SunOS5, OSF1, and
4782 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4783 some earlier ones do not.
4785 7. Change 4.30/15 was not doing the test caselessly.
4787 8. The standard form for an IPv6 address literal was being rejected by address
4788 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4789 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4790 this, as well as the form without the "IPv6" on the front (but only when
4791 address literals are enabled, of course).
4793 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4795 10. Exim crashed if a message with an empty sender address specified by -f
4796 encountered a router with an errors_to setting. This could be provoked only
4797 by a command such as
4801 where an empty string was supplied; "<>" did not hit this bug.
4803 11. Installed PCRE release 4.5.
4805 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4806 remained set. It is now erased.
4808 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4809 times from message ids (which are base 36 rather than the normal 62).
4811 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4812 were being counted as actual protocol errors, and logged if the log
4813 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4814 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4815 and DATA following a set of rejected RCPTs do not count as protocol errors.
4816 In other words, Exim assumes they were pipelined, though this may not
4817 actually be the case. Of course, in all cases the client gets an
4818 appropriate error code.
4820 15. If a lookup fails in an ACL condition, a message about the failure may
4821 be available; it is used if testing the ACL cannot continue, because most
4822 such messages specify what the cause of the deferral is. However, some
4823 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4824 that caused an old message to be retained and used if a later statement
4825 caused a defer, replacing the real cause of the deferral.
4827 16. If an IP address had so many PTR records that the DNS lookup buffer
4828 was not large enough to hold them, Exim could crash while trying to process
4829 the truncated data. It now detects and logs this case.
4831 17. Further to 4.21/58, another change has been made: if (and only if) the
4832 first line of a message (the first header line) ends with CRLF, a bare LF
4833 in a subsequent header line has a space inserted after it, so as not to
4834 terminate the header.
4836 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4837 unnecessarily, used atoi() instead of strtol(), and didn't check the
4838 termination when getting file sizes from file names by regex.
4840 19. Completely re-implemented the support for maildirsize files, in the light
4841 of a number of problems with the previous contributed implementation
4842 (4.30/29). In particular:
4844 . If the quota is zero, the maildirsize file is maintained, but no quota is
4847 . If the maildir directory does not exist, it is created before any attempt
4848 to write a maildirsize file.
4850 . The quota value in the file is just a cache; if the quota is changed in
4851 the transport, the new value overrides.
4853 . A regular expression is available for excluding directories from the
4856 20. The autoreply transport checks the characters in options that define the
4857 message's headers; it allows continued headers, but it was checking with
4858 isspace() after an embedded newline instead of explicitly looking for a
4861 21. If all the "regular" hosts to which an address was routed had passed their
4862 expiry times, and had not reached their retry times, the address was
4863 bounced, even if fallback hosts were defined. Now Exim should go on to try
4866 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4867 equivalent code in the SMTP transport. Some hosts send humungous responses
4868 to HELO/EHLO, more than 1024 it seems.
4870 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4871 gives compiler warnings in some environments. I've now done it "properly",
4874 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4875 (because of problems with the built-in one) was declared to return uschar *
4876 instead of char *, causing compiler failure.
4878 25. Fixed a file descriptor leak when processing alias/forward files.
4880 26. Fixed a minor format string issue in dbfn.c.
4882 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4884 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4885 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4888 29. When a sender address is verified, it is cached, to save repeating the test
4889 when there is more than one recipient in a message. However, when the
4890 verification involves a callout, it is possible for different callout
4891 options to be set for different recipients. It is too complicated to keep
4892 track of this in the cache, so now Exim always runs a verification when a
4893 callout is required, relying on the callout cache for the optimization.
4894 The overhead is duplication of the address routing, but this should not be
4897 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4898 to be verified with callout=postmaster, and the main callout worked but the
4899 postmaster check failed, the verification correctly failed. However, if a
4900 subsequent RCPT command asked for sender verification *without* the
4901 postmaster check, incorrect caching caused this verification also to fail,
4904 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4905 it was not caching the DNS options (qualify_single, search_parents) that
4906 were used when the lookup failed. A subsequent lookup with different
4907 options therefore always gave the same answer, though there were cases
4908 where it should not have. (Example: a "domains = !$mx_any" option on a
4909 dnslookup router: the "domains" option is always processed without any
4910 widening, but the router might have qualify_single set.) Now Exim uses the
4911 cached value only when the same options are set.
4913 32. Added John Jetmore's "exipick" utility to the distribution.
4915 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4916 than a timeout (e.g. a certificate is required, and is not provided), an
4917 Exim server now closes the connection immediately. Previously it waited for
4918 the client to close - but if the client is SSL, it seems that they each
4919 wait for each other, leading to a delay before one of them times out.
4921 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4922 maintained 0.8.x compatibility because I don't think many are using it, and
4923 it is clearly obsolete.
4925 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4928 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4929 client certificate was expired. A simple patch fixes this, though I don't
4930 understand the full logic of why the verify callback is called multiple
4933 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4934 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4935 which causes problems with some clients (such as the Certicom SSL Plus
4936 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4937 disables the coutermeasure allowing Eudora to connect."
4939 38. Exim was not checking that a write() to a log file succeeded. This could
4940 lead to Bad Things if a log got too big, in particular if it hit a file
4941 size limit. Exim now panics and dies if it cannot write to a log file, just
4942 as it does if it cannot open a log file.
4944 39. Modified OS/Makefile-Linux so that it now contains
4946 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4948 The two -D definitions ensure that Exim is compiled with large file
4949 support, which makes it possible to handle log files that are bigger than
4952 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4953 instance) a domain was checked against a named list that involved a lookup,
4954 causing $domain_data to be set, then another domain was checked against the
4955 same list, then the first domain was re-checked, the value of $domain_data
4956 after the final check could be wrong. In particular, if the second check
4957 failed, it could be set empty. This bug probably also applied to
4960 41. The strip_trailing_dot option was not being applied to the address given
4961 with the -f command-line option.
4963 42. The code for reading a message's header from the spool was incrementing
4964 $received_count, but never initializing it. This meant that the value was
4965 incorrect (doubled) while delivering a message in the same process in which
4966 it was received. In the most common configuration of Exim, this never
4967 happens - a fresh exec is done - but it can happen when
4968 deliver_drop_privilege is set.
4970 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4971 it now includes up to 150 characters of the unexpected data in the log
4974 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4975 and building data strings. The size of both of these buffers was 10 000
4976 bytes - far larger than anybody would *ever* want, thought I. Needless to
4977 say, somebody hit the limit. I have increased the maximum line length to
4978 20 000 and the maximum data length of concatenated lines to 100 000. I have
4979 also fixed two bugs, because there was no checking on these buffers. Tsk,
4980 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4981 buffer is too small.
4983 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4984 lsearch lookups. Now it does.
4986 46. When parsing a route_list item in a manualroute router, a fixed-length
4987 buffer was used for the list of hosts. I made this 1024 bytes long,
4988 thinking that nobody would ever have a list of hosts that long. Wrong.
4989 Somebody had a whole pile of complicated expansion conditions, and the
4990 string was silently truncated, leading to an expansion error. It turns out
4991 that it is easier to change to an unlimited length (owing to other changes
4992 that have happened since this code was originally written) than to build
4993 structure for giving a limitation error. The length of the item that
4994 expands into the list of hosts is now unlimited.
4996 47. The lsearch lookup could not handle data where the length of text line was
4997 more than 4095 characters. Such lines were truncated, leading to shortened
4998 data being returned. It should now handle lines of any length.
5000 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5001 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5004 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5005 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5006 no longer happens in comment lines. A list of replacements is now placed
5007 at the head of all of the source files, except those whose only change is
5008 to replace PERL_COMMAND in the very first #! line.
5010 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5011 messages on the queue, with a bottom-up merge sort, using code contributed
5012 by Michael Haardt. This should make operations like -bp somewhat faster on
5013 large queues. It won't affect queue runners, except when queue_run_in_order
5016 51. Installed eximstats 1.31 in the distribution.
5018 52. Added support for SRV lookups to the dnslookup router.
5020 53. If an ACL referred to $message_body or $message_body_end, the value was not
5021 reset for any messages that followed in the same SMTP session.
5023 54. The store-handling optimization for building very long strings was not
5024 differentiating between the different store pools. I don't think this
5025 actually made any difference in practice, but I've tidied it.
5027 55. While running the routers to verify a sender address, $sender_address
5028 was still set to the sender address. This is wrong, because when routing to
5029 send a bounce to the sender, it would be empty. Therefore, I have changed
5030 it so that, while verifying a sender address, $sender_address is set to <>.
5031 (There is no change to what happens when verifying a recipient address.)
5033 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5034 target A or AAAA records (if not already returned) without resetting the
5035 qualify_single or search_parents options of the DNS resolver. These are
5036 inappropriate in this case because the targets of MX and SRV records must
5037 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5038 target that, when qualified, matched something in the local domain. These
5039 two options are now turned off when doing these lookups.
5041 57. It seems that at least some releases of Reiserfs (which does not have the
5042 concept of a fixed number of inodes) returns zero and not -1 for the
5043 number of available inodes. This interacted badly with check_spool_inodes,
5044 which assumed that -1 was the "no such thing" setting. What I have done is
5045 to check that the total number of inodes is greater than zero before doing
5046 the test of how many are available.
5048 58. When a "warn" ACL statement has a log_message modifier, the message is
5049 remembered, and not repeated. This is to avoid a lot of repetition when a
5050 message has many recipients that cause the same warning to be written.
5051 Howewer, Exim was preserving the list of already written lines for an
5052 entire SMTP session, which doesn't seem right. The memory is now reset if a
5053 new message is started.
5055 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5056 debugging output unless log_rewrite was also set.
5058 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5059 of (int)(handle) when we know that handle contains (void *)(-1).
5061 61. The Exim daemon panic-logs an error return when it closes the incoming
5062 connection. However "connection reset by peer" seems to be common, and
5063 isn't really an error worthy of noting specially, so that particular error
5066 62. When Exim is trying to find all the local interfaces, it used to panic and
5067 die if the ioctl to get the interface flags failed. However, it seems that
5068 on at least one OS (Solaris 9) it is possible to have an interface that is
5069 included in the list of interfaces, but for which you get a failure error
5070 for this call. This happens when the interface is not "plumbed" into a
5071 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5072 failure of the "get flags" call assumes that the interface is down.
5074 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5075 makes life easier for people who are doing arithmetic on fields extracted
5076 from dates, where you often get leading zeros that should not be
5077 interpreted as octal.
5079 64. Added qualify_domain to the redirect router, to override the global
5082 65. If a pathologically long header line contained very many addresses (the
5083 report of this problem mentioned 10 000) and each of them was rewritten,
5084 Exim could use up a very large amount of memory. (It kept on making new
5085 copies of the header line as it rewrote, and never released the old ones.)
5086 At the expense of a bit more processing, the header rewriting function has
5087 been changed so that it no longer eats memory in this way.
5089 66. The generation of the Received: header has been moved from the time that a
5090 message starts to be received, to the time that it finishes. The timestamp
5091 in the Received: header should now be very close to that of the <= log
5092 line. There are two side-effects of this change:
5094 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5095 logged header lines no longer include the local Received: line, because
5096 it has not yet been created. The same applies to a copy of the message
5097 that is returned to a non-SMTP sender when a message is rejected.
5099 (b) When a filter file is tested using -bf, no additional Received: header
5100 is added to the test message. After some thought, I decided that this
5103 This change does not affect the value of $received_for. It is still set
5104 after address rewriting, but before local_scan() is called.
5106 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5108 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5109 gave an unhelpful panic error message, and a defer error. I have managed to
5110 change this behaviour so that it now rejects any supplied certificate,
5111 which seems right, as the list of acceptable certificates is empty.
5113 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5114 gave an unhelpful defer error. I have not managed to make this reject any
5115 supplied certificates, but the error message it gives is "no certificate
5116 supplied", which is not helpful.
5118 70. exigrep's output now also includes lines that are not associated with any
5119 message, but which match the given pattern. Implemented by a patch from
5120 Martin Sluka, which also tidied up the Perl a bit.
5122 71. Recipient callout verification, like sender verification, was using <> in
5123 the MAIL FROM command. This isn't really the right thing, since the actual
5124 sender may affect whether the remote host accepts the recipient or not. I
5125 have changed it to use the actual sender in the callout; this means that
5126 the cache record is now keyed on a recipient/sender pair, not just the
5127 recipient address. There doesn't seem to be a real danger of callout loops,
5128 since a callout by the remote host to check the sender would use <>.
5129 [SEE ABOVE: changed after hitting problems.]
5131 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5132 temporary errors. However, in the case of such a code being given after
5133 the end of a data transmission (i.e. after ".") Exim was failing to write
5134 a retry record for the message. (Yes, there was some broken host that was
5135 actually sending 8xx at this point.)
5137 73. An unknown lookup type in a host list could cause Exim to panic-die when
5138 the list was checked. (An example that provoked this was putting <; in the
5139 middle of a list instead of at the start.) If this happened during a DATA
5140 ACL check, a -D file could be left lying around. This kind of configuration
5141 error no longer causes Exim to die; instead it causes a defer errror. The
5142 incident is still logged to the main and panic logs.
5144 74. Buglet left over from Exim 3 conversion. The message "too many messages
5145 in one connection" was written to the rejectlog but not the mainlog, except
5146 when address rewriting (yes!) was being logged.
5148 75. Added write_rejectlog option.
5150 76. When a system filter was run not as root (that is, when system_filter_user
5151 was set), the values of the $n variables were not being returned to the
5152 main process; thus, they were not subsequently available in the $sn
5155 77. Added +return_path_on_delivery log selector.
5157 78. A connection timeout was being treated differently from recipients deferred
5158 when testing hosts_max_try with a message that was older than the host's
5159 retry timeout. (The host should not be counted, thus allowing all hosts to
5160 be tried at least once before bouncing.) This may have been the cause of an
5161 occasionally reported bug whereby a message would remain on the queue
5162 longer than the retry timeout, but would be bounced if a delivery was
5163 forced. I say "may" because I never totally pinned down the problem;
5164 setting up timeout/retry tests is difficult. See also the next item.
5166 79. The ultimate address timeout was not being applied to errors that involved
5167 a combination of host plus message (for example, a timeout on a MAIL
5168 command). When an address resolved to a number of possible hosts, and they
5169 were not all tried for each delivery (e.g. because of hosts_max_try), a
5170 message could remain on the queue longer than the retry timeout.
5172 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5175 81. Fixed an obscure SMTP outgoing bug which required at least the following
5176 conditions: (a) there was another message waiting for the same server;
5177 (b) the server returned 5xx to all RCPT commands in the first message so
5178 that the message was not completed; (c) the server dropped the connection
5179 or gave a negative response to the RSET that Exim sends to abort the
5180 transaction. The observed case was a dropped connection after DATA that had
5181 been sent in pipelining mode. That is, the server had advertised PIPELINING
5182 but was not implementing it correctly. The effect of the bug was incorrect
5183 behaviour, such as trying another host, and this could lead to a crash.
5189 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5190 and daemon.c were passed as pointers to ints; they should have been
5191 pointers to socklen_t variables (which are typically unsigned ints).
5193 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5196 3. Fixed a really odd bug that affected only the testing scheme; patching a
5197 certain fixed string in the binary changed the value of another string that
5198 happened to be identical to the end of the original first string.
5200 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5201 name", it returns that address as the IP address. On some operating
5202 systems (e.g. Solaris), it also passes back the IP address string as the
5203 "host name". However, on others (e.g. Linux), it passes back an empty
5204 string. Exim wasn't checking for this, and was changing the host name to an
5205 empty string, assuming it had been canonicized.
5207 5. Although rare, it is permitted to have more than one PTR record for a given
5208 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5209 all the names associated with an address, because they do in Solaris.
5210 However, it seems that they do not in Linux for data that comes from the
5211 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5212 I found this out when I moved to a new Linux workstation and tried to run
5213 the Exim test suite.
5215 To get round this problem I have changed the code so that it now does its
5216 own call to the DNS to look up PTR records when searching for a host name.
5217 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5218 addresses that are only in /etc/hosts are still found.
5220 This behaviour is, however, controlled by an option called host_lookup_
5221 order, which defaults to "bydns:byaddr". If people want to use the other
5222 order, or indeed, just use one or the other means of lookup, they can
5223 specify it in this variable.
5225 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5226 some operating systems, this comes back from gethostbyaddr() as an empty
5227 string, and this is what Exim used to test for. However, it seems that in
5228 other systems, "." is yielded. Exim now tests for this case too.
5230 7. The values of check_spool_space and check_log_space are now held internally
5231 as a number of kilobytes instead of an absolute number of bytes. If a
5232 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5233 kilobyte. This means that much larger values can be stored.
5235 8. Exim monitor: an attempt to get the action menu when not actually pointing
5236 at a message produces an empty menu entitled "No message selected". This
5237 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5238 no entries in it ("Shell widget menu has zero width and/or height"). So I
5239 have added a single, blank menu entry in this case.
5241 9. Added ${quote_local_part.
5243 10. MIME decoding is now applied to the contents of Subject: header lines when
5246 11. Now that a reference to $sender_host_address automatically causes a reverse
5247 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5248 host lookup before query-style lookups in lists that might use this
5249 variable. This has therefore been abolished, and the "net-" prefix is no
5250 longer necessary for query-style lookups.
5252 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5253 have been a typo for LFLAGS, so it has been changed.
5255 13. The install script calls Exim with "-C /dev/null" in order to find the
5256 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5257 to be output. Howeve, since Exim outputs its version number before the
5258 error, it didn't break the script. It just looked ugly. I fixed this by
5259 always allowing "-C /dev/null" if the caller is root.
5261 14. Ignore overlarge ACL variable number when reading spool file - insurance
5262 against a later release with more variables having written the file.
5264 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5265 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5266 this, as well as the form without the "IPv6" on the front.
5268 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5269 OS/Makefile-Darwin file.
5271 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5272 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5274 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5275 variable. However, some OS use ints for inodes. I've added cast to long int
5276 to get rid of the compiler warning.
5278 19. I had forgotten to lock out "/../" in configuration file names when
5279 ALT_CONFIG_PREFIX was set.
5281 20. Routers used for verification do not need to specify transports. However,
5282 if such a router generated a host list, and callout was configured, Exim
5283 crashed, because it could not find a port number from the (non-existent)
5284 transport. It now assumes port 25 in this circumstance.
5286 21. Added the -t option to exigrep.
5288 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5289 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5290 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5291 them is set, LOOKUP_LSEARCH is forced.
5293 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5295 24. Added sender and host information to the "rejected by local_scan()" log
5296 line; previously there was no indication of these.
5298 25. Added .include_if_exists.
5300 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5301 Linux. It turns out that not all file systems support this. Apparently some
5302 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5303 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5304 not supported on the file descriptor, is now ignored when Exim is trying to
5305 sync a directory. This applies only to Linux.
5307 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5309 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5310 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5313 29. Added support for maildirsize files from supplied patch (modified a bit).
5315 30. The use of :fail: followed by an empty string could lead Exim to respond to
5316 sender verification failures with (e.g.):
5318 550 Verification failed for <xxx>
5319 550 Sender verify failed
5321 where the first response line was missing the '-' that indicates it is not
5322 the final line of the response.
5324 31. The loop for finding the name of the user that called Exim had a hardwired
5325 limit of 10; it now uses the value of finduser_retries, which is used for
5326 all other user lookups.
5328 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5331 33. Exim was neglecting to zero errno before one call of strtol() when
5332 expanding a string and expecting an integer value. On some systems this
5333 resulted in spurious "integer overflow" errors. Also, it was casting the
5334 result into an int without checking.
5336 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5337 did not work. The code looks as if it has *never* worked, though it appears
5338 to have been documented since at least releast 1.62. I have made it work.
5340 35. The "timeout_DNS" error in retry rules, also documented since at least
5341 1.62, also never worked. As it isn't clear exactly what this means, and
5342 clearly it isn't a major issue, I have abolished the feature by treating it
5343 as "timeout", and writing a warning to the main and panic logs.
5345 36. The display of retry rules for -brt wasn't always showing the error code
5348 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5349 timeout_connect_A, timeout_connect_MX.
5351 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5352 to the empty sender.
5354 39. The daemon was not analysing the content of -oX till after it had closed
5355 stderr and disconnected from the controlling terminal. This meant that any
5356 syntax errors were only noted on the panic log, and the return code from
5357 the command was 0. By re-arranging the code a little, I've made the
5358 decoding happen first, so such errors now appear on stderr, and the return
5359 code is 1. However, the actual setting up of the sockets still happens in
5360 the disconnected process, so errors there are still only recorded on the
5363 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5364 connections (as happens on some IP stacks) was logged at start up time as
5365 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5366 it from "IPv6 and IPv4", which means that two separate sockets are being
5369 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5370 says whether AF_INET or AF_INET6 was passed as an argument.
5372 42. Exiwhat output was messed up when time zones were included in log
5375 43. Exiwhat now gives more information about the daemon's listening ports,
5376 and whether -tls-on-connect was used.
5378 44. The "port" option of the smtp transport is now expanded.
5380 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5381 silently ignored. Now an error message is written to the main and panic
5384 46. There's a new ACL modifier called "logwrite" which writes to a log file
5385 as soon as it is encountered.
5387 47. Added $local_user_uid and $local_user_gid at routing time.
5389 48. Exim crashed when trying to verify a sender address that was being
5392 49. Exim was recognizing only a space character after ".include". It now also
5393 recognizes a tab character.
5395 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5396 extracting the relevant information from the specification. The man page no
5397 longer contains scrambled data for the -d option, and I've added a section
5398 at the front about calling Exim under different names.
5400 51. Added "extra_headers" argument to the "mail" command in filter files.
5402 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5405 53. Installed eximstats 1.29.
5407 54. Added transport_filter_timeout as a generic transport option.
5409 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5410 Cc: header lines. This was required by RFC 822, but it not required by RFC
5413 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5414 incoming messages that did not have them. Now it does so only if the
5415 message originates locally, that is, if there is no associated remote host
5416 address. When Resent- header lines are present, this applies to the Resent-
5417 lines rather than the non-Resent- lines.
5419 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5420 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5422 58. Messages for configuration errors now include the name of the main
5423 configuration file - useful now that there may be more than one file in a
5424 list (.included file names were always shown).
5426 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5427 for those rare installations that do not start the daemon as root or run it
5428 setuid root. I've cut out the call to initgroups() if the daemon is not
5431 60. The Exim user and group can now be bound into the binary as text strings
5432 that are looked up at the start of Exim's processing.
5434 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5436 62. Added $mailstore_basename variable.
5438 63. Installed patch to sieve.c from Michael Haardt.
5440 64. When Exim failed to open the panic log after failing to open the main log,
5441 the original message it was trying to log was written to stderr and debug
5442 output, but if they were not available (the usual case in production), it
5443 was lost. Now it is written to syslog before the two lines that record the
5444 failures to open the logs.
5446 65. Users' Exim filters run in subprocesses under the user's uid. It is
5447 possible for a "deliver" command or an alias in a "personal" command to
5448 provoke an address rewrite. If logging of address rewriting is configured,
5449 this fails because the process is not running as root or exim. There may be
5450 a better way of dealing with this, but for the moment (because 4.30 needs
5451 to be released), I have disabled address rewrite logging when running a
5452 filter in a non-root, non-exim process.
5458 1. The buildconfig auxiliary program wasn't quoting the value set for
5459 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5460 not defined. This bug was masked in 4.22 by the effect that was fixed in
5463 2. Some messages that were rejected after a message id was allocated were
5464 shown as "incomplete" by exigrep. It no longer does this for messages that
5465 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5467 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5468 have allow_domain_literals set, the ID did not get logged in the <= line.
5469 Domain literals are now always recognized in Message-ID: header lines.
5471 4. The first argument for a ${extract expansion item is the key name or field
5472 number. Leading and trailing spaces in this item were not being ignored,
5473 causing some misleading effects.
5475 5. When deliver_drop_privilege was set, single queue runner processes started
5476 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5477 same command in the process it spins off) were not dropping privilege.
5479 6. When the daemon running as "exim" started a queue runner, it always
5480 re-executed Exim in the spun-off process. This is a waste of effort when
5481 deliver_drop_privilege is set. The new process now just calls the
5482 queue-runner function directly.
5488 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5491 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5492 ignored. Though the use of -oP was forcing the writing of a pid file, it
5493 was always written to the default place.
5495 3. If the message "no IP address found for host xxxx" is generated during
5496 incoming verification, it is now followed by identification of the incoming
5497 connection (so you can more easily find what provoked it).
5499 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5501 5. Added some features to "harden" Exim a bit more against certain attacks:
5503 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5504 be put in Local/Makefile. This is like the never_users runtime option,
5505 but it cannot be overridden. The default setting is "root".
5507 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5508 prefix string with which any file named in a -C command line option
5511 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5512 is retained for -C and -D only if the caller of Exim is root. Without
5513 it, the exim user may also use -C and -D and retain privilege.
5515 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5516 command line option is disabled.
5518 6. Macro names set by the -D option must start with an upper case letter, just
5519 like macro names defined in the configuration file.
5521 7. Added "dereference=" facility to LDAP.
5523 8. Two instances of the typo "uknown" in the source files are fixed.
5525 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5526 the Configure-Makefile script screwed up while processing it.
5528 10. Incorporated PCRE 4.4.
5530 11. The SMTP synchronization check was not operating right at the start of an
5531 SMTP session. For example, it could not catch a HELO sent before the client
5532 waited for the greeting. There is now a check for outstanding input at the
5533 point when the greeting is written. Because of the duplex, asynchronous
5534 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5535 way, but not yet received, when the check is performed.
5537 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5538 on TCP/IP sockets, because this apparently causes some broken clients to
5541 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5542 unchanged) from the Cygwin maintainer.
5544 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5545 was supported instead of testing for mbx. Effectively a typo.
5547 15. The spa authenticator server code was not checking that the input it
5548 received was valid base64.
5550 16. The debug output line for the "set" modifier in ACLs was not showing the
5551 name of the variable that was being set.
5553 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5555 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5556 lookup on its own. Something else was needed to trigger the lookup. For
5557 example, a match in host_lookup or the need for a host name in a host list.
5558 Now, if $sender_host_name is referenced and the host name has not yet been
5559 looked up, a lookup is performed. If the lookup fails, the variable remains
5560 empty, and $host_lookup_failed is set to "1".
5562 19. Added "eqi" as a case-independent comparison operator.
5564 20. The saslauthd authentication condition could segfault if neither service
5565 nor realm was specified.
5567 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5568 error message that was logged was misleading, and incoming SMTP
5569 connections were dropped. The message is now more accurate, and temporary
5570 errors are given to SMTP connections.
5572 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5573 (except RSET) with a 421 error, until QUIT is received. However, it was
5574 failing to send a response to QUIT.
5576 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5577 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5578 ACL is now run after the helo_try_verify_hosts code.
5580 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5581 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5582 case-independent, so other case variants are also recognized). Apparently
5583 some systems use these upper case variants.
5585 25. If more than two messages were waiting for the same host, and a transport
5586 filter was specified for the transport, Exim sent two messages over the
5587 same TCP/IP connection, and then failed with "socket operation on non-
5588 socket" when it tried to send the third.
5590 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5592 27. The extern definition of crypt16() in expand.c was not being excluded when
5593 the OS had its own crypt16() function.
5595 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5596 as a preferred synonym for return_size_limit, both as an option and as an
5599 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5601 30. Changed the default configuration ACL to relax the local part checking rule
5602 for addresses that are not in any local domains. For these addresses,
5603 slashes and pipe symbols are allowed within local parts, but the sequence
5604 /../ is explicitly forbidden.
5606 31. SPA server authentication was not clearing the challenge buffer before
5609 32. log_message in a "warn" ACL statement was writing to the reject log as
5610 well as to the main log, which contradicts the documentation and doesn't
5611 seem right (because no rejection is happening). So I have stopped it.
5613 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5614 However, I am unable to do any testing of this.
5616 34. Fixed an infelicity in the appendfile transport. When checking directories
5617 for a mailbox, to see if any needed to be created, it was accidentally
5618 using path names with one or more superfluous leading slashes; tracing
5619 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5621 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5622 contents are added to the log line that is written for every discarded
5623 recipient. (Previously a log_message setting was ignored.)
5625 36. The ${quote: operator now quotes the string if it is empty.
5627 37. The install script runs exim in order to find its version number. If for
5628 some reason other than non-existence or emptiness, which it checks, it
5629 could not run './exim', it was installing it with an empty version number,
5630 i.e. as "exim-". This error state is now caught, and the installation is
5633 38. An argument was missing from the function that creates an error message
5634 when Exim fails to connect to the socket for saslauthd authentication.
5635 This could cause Exim to crash, or give a corrupted message.
5637 39. Added isip, isip4, and isip6 to ${if conditions.
5639 40. The ACL variables $acl_xx are now saved with the message, and can be
5640 accessed later in routers, transports, and filters.
5642 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5643 strings in the file are not string-expanded.
5645 42. If a MAIL command specified a SIZE value that was too large to fit into an
5646 int variable, the check against message_size_limit failed. Such values are
5647 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5648 day this will have to be increased, but I don't think I want to be around
5649 when emails are that large.
5656 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5657 iconv() is not standard in FreeBSD.
5659 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5660 IPv6 enabled. The observed symptom was a segmentation fault on return from
5661 the function os_common_find_running_interfaces() in src/os.c.
5663 3. In the check_special_case() function in daemon.c I had used "errno" as an
5664 argument name, which causes warnings on some systems. This was basically a
5665 typo, since it was named "eno" in the comments!
5667 4. The code that waits for the clock to tick (at a resolution of some fraction
5668 of a second) so as to ensure message-id uniqueness was always waiting for
5669 at least one whole tick, when it could have waited for less. [This is
5670 almost certainly not relevant at current processor speeds, where it is
5671 unlikely to ever wait at all. But we try to future-proof.]
5673 5. The function that sleeps for a time interval that includes fractions of a
5674 second contained a race. It did not block SIGALRM between setting the
5675 timer, and suspending (a couple of lines later). If the interval was short
5676 and the sigsuspend() was delayed until after it had expired, the suspension
5677 never ended. On busy systems this could lead to processes getting stuck for
5680 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5681 process, before it forks any delivery processes. The open lookup caching
5682 mechanism meant that the open file or database connection was passed into
5683 the delivery process. The problem was that delivery processes always tidy
5684 up cached lookup data. This could cause a problem for the next delivery
5685 process started by the queue runner, because the external queue runner
5686 process does not know about the closure. So the next delivery process
5687 still has data in the lookup cache. In the case of a file lookup, there was
5688 no problem because closing a file descriptor in a subprocess doesn't affect
5689 the parent. However, if the lookup was caching a connection to a database,
5690 the connection was closed, and the second delivery process was likely to
5691 see errors such as "PGSQL: query failed: server closed the connection
5692 unexpectedly". The problem has been fixed by closing all cached lookups
5693 in a queue runner before running a delivery process.
5695 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5696 seem to have the "const" qualifier which it has on other OS. I've
5699 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5700 *of the same type* (client or server) with the same public name that an
5701 error should be diagnosed.
5703 9. When Exim looked up a host name for an IP address, but failed to find the
5704 original IP address when looking up the host name (a safety check), it
5705 output the message "<ip address> does not match any IP for NULL", which was
5706 confusing, to say the least. The bug was that the host name should have
5707 appeared instead of "NULL".
5709 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5710 user that is built into the binary, and the -C or -D options is used, root
5711 privilege is dropped before the configuration file is read. In addition,
5712 logging is switched to stderr instead of the normal log files. If the
5713 configuration then re-defines the Exim user, the unprivileged environment
5714 is probably not what is expected, so Exim logs a panic warning message (but
5717 However, if deliver_drop_privilege is set, the unprivileged state may well
5718 be exactly what is intended, so the warning has been cut out in that case,
5719 and Exim is allowed to try to write to its normal log files.
5725 1. smtp_return_error_details was not giving details for temporary sender
5726 or receiver verification errors.
5728 2. Diagnose a configuration error if two authenticators have the same public
5731 3. Exim used not to create the message log file for a message until the first
5732 delivery attempt. This could be confusing when incoming messages were held
5733 for policy or load reasons. The message log file is now created at the time
5734 the message is received, and an initial "Received" line is written to it.
5736 4. The automatically generated man page for command line options had a minor
5737 bug that caused no ill effects; however, a more serious problem was that
5738 the procedure for building the man page automatically didn't always
5739 operate. Consequently, release 4.20 contains an out-of-date version. This
5740 shouldn't happen again.
5742 5. When building Exim with embedded Perl support, the script that builds the
5743 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5744 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5746 6. The freeze_tell option was not being used for messages that were frozen on
5747 arrival, either by an ACL or by local_scan().
5749 7. Added the smtp_incomplete_transaction log selector.
5751 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5752 was accepting AUTH without a new EHLO.
5754 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5755 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5758 10. Logging of TCP/IP connections (when configured) now happens in the main
5759 daemon process instead of the child process, so that the TCP/IP connection
5760 count is more accurate (but it can never be perfect).
5762 11. The use of "drop" in a nested ACL was not being handled correctly in the
5763 outer ACL. Now, if condition failure induced by the nested "drop" causes
5764 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5765 or "require"), the connection is dropped.
5767 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5768 that yield "discard" can now be used with an "accept" or a "discard" verb,
5769 but an error is generated for any others (because I can't see a useful way
5770 to define what should happen).
5772 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5773 are now processed in the same way as lines in the Exim configuration file.
5774 In particular, continuation lines are supported.
5776 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5778 15. Added -ti meaning -t -i.
5780 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5781 domains, and warn by logging if others are found.
5783 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5784 returned from GIFCONF when Exim is trying to find the list of interfaces on
5785 a host. The code in os.c has been modified to copy each ifreq to an aligned
5786 structure in all cases.
5788 Also, in some cases, the returned ifreq's were being copied to a 'struct
5789 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5790 means the last couple of bytes of an IPv6 address could be chopped if the
5791 ifreq contained only a normal sockaddr (14 bytes storage).
5793 18. Named domain lists were not supported in the hosts_treat_as_local option.
5794 An entry such as +xxxx was not recognized, and was treated as a literal
5797 19. Ensure that header lines added by a DATA ACL are included in the reject log
5798 if the ACL subsequently rejects the message.
5800 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5801 MD5 (which is deprecated).
5803 21. When testing a filter file using -bf, Exim was writing a message when it
5804 took the sender from a "From " line in the message, but it was not doing so
5805 when it took $return_path from a Return-Path: header line. It now does.
5807 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5808 with a valid header line field name (a series of printing characters
5809 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5811 23. Changed "disc" in the source to "disk" to conform to the documentation and
5812 the book and for uniformity.
5814 24. Ignore Sendmail's -Ooption=value command line item.
5816 25. When execve() failed while trying to run a command in a pipe transport,
5817 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5818 could be confused with a return value of 69 from the command itself. This
5819 has been changed to 127, the value the shell returns if it is asked to run
5820 a non-existent command. The wording for the related log line suggests a
5821 non-existent command as the problem.
5823 26. If received_header_text expands to an empty string, do not add a Received:
5824 header line to the message. (Well, it adds a token one on the spool, but
5825 marks it "old" so that it doesn't get used or transmitted.)
5827 27. Installed eximstats 1.28 (addition of -nt option).
5829 28. There was no check for failure on the call to getsockname() in the daemon
5830 code. This can fail if there is a shortage of resources on the system, with
5831 ENOMEM, for example. A temporary error is now given on failure.
5833 29. Contrary to the C standard, it seems that in some environments, the
5834 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5835 program. Exim now does this explicitly; it affects the formatting of
5836 timestamps using strftime().
5838 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5839 complaints. I've now initialized all the variables, to avoid this.
5841 32. Header lines added by a system filter were not being "seen" during
5842 transport-time rewrites.
5844 33. The info_callback() function passed to OpenSSL is set up with type void
5845 (*)(SSL *, int, int), as described somewhere. However, when calling the
5846 function (actually a macro) that sets it up, the type void(*)() is
5847 expected. I've put in a cast to prevent warnings from picky compilers.
5849 34. If a DNS black list lookup found a CNAME record, but there were no A
5850 records associated with the domain it pointed at, Exim crashed.
5852 35. If a DNS black list lookup returned more than one A record, Exim ignored
5853 all but the first. It now scans all returned addresses if a particular IP
5854 value is being sought. In this situation, the contents of the
5855 $dnslist_value variable are a list of all the addresses, separated by a
5858 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5859 to accept a host name and all its aliases if the forward lookup for any of
5860 them yielded the IP address of the incoming connection. Now it accepts only
5861 those names whose forward lookup yields the correct IP address. Any other
5862 names are discarded. This closes a loophole whereby a rogue DNS
5863 administrator could create reverse DNS records to break through a
5864 wildcarded host restriction in an ACL.
5866 37. If a user filter or a system filter that ran in a subprocess used any of
5867 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5868 the wrong values were passed to the pipe command ($thisaddress had the
5869 value of $0, $0 had the value of $1, etc). This bug was introduced by
5870 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5872 38. Improved the line breaking for long SMTP error messages from ACLs.
5873 Previously, if there was no break point between 40 and 75 characters, Exim
5874 left the rest of the message alone. Two changes have been made: (a) I've
5875 reduced the minimum length to 35 characters; (b) if it can't find a break
5876 point between 35 and 75 characters, it looks ahead and uses the first one
5877 that it finds. This may give the occasional overlong line, but at least the
5878 remaining text gets split now.
5880 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5881 file descriptors might be low, and that setting 1000 would always raise it.
5882 It turns out that in some environments, the limit is already over 1000 and
5883 that lowering it causes trouble. So now Exim takes care not to decrease it.
5885 40. When delivering a message, the value of $return_path is set to $sender_
5886 address at the start of routing (routers may change the value). By an
5887 oversight, this default was not being set up when an address was tested by
5888 -bt or -bv, which affected the outcome if any router or filter referred to
5891 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5892 log only when "message" or "log_message" are set. However, if one of the
5893 conditions was an address verification, or a call to a nested ACL, the
5894 messages generated by the underlying test were being passed through. This
5895 no longer happens. The underlying message is available in $acl_verify_
5896 message for both "message" and "log_message" expansions, so it can be
5897 passed through if needed.
5899 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5900 new expansion $bh_ to give the encoded byte string without charset
5901 translation. Translation happens only if iconv() is available; HAVE_ICONV
5902 indicates this at build time. HEADERS_CHARSET gives the charset to
5903 translate to; headers_charset can change it in the configuration, and
5904 "headers charset" can change it in an individual filter file.
5906 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5907 that creates RFC 2047 encoded "words" labels them as that charset instead
5908 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5909 expansion operator; (ii) when Exim creates a From: line for a local
5910 message; (iii) when a header line is rewritten to include a "phrase" part.
5912 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5913 buggy, causing it to skip the first lines of messages whose message ID
5914 ended in 'D'. This would not have bitten before Exim release 4.14, because
5915 message IDs were unlikely to end in 'D' before then. The effect was to have
5916 incorrect size information for certain domains.
5918 45. #include "config.h" was missing at the start of the crypt16.c module. This
5919 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5922 46. If there was a timeout during a "random" callout check, Exim treated it as
5923 a failure of the random address, and carried on sending RSET and the real
5924 address. If the delay was just some slowness somewhere, the response to the
5925 original RCPT would be taken as a response to RSET and so on, causing
5926 mayhem of various kinds.
5928 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5929 when I implemented it. It didn't allow for the fact that some option values
5930 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5931 the right test for positive values.
5933 48. Domain names in DNS records are case-independent. Exim always looks them up
5934 in lower case. Some resolvers return domain names in exactly the case they
5935 appear in the zone file, that is, they may contain uppercase letters. Not
5936 all resolvers do this - some return always lower case. Exim was treating a
5937 change of case by a resolver as a change of domain, similar to a widening
5938 of a domain abbreviation. This triggered its re-routing code and so it was
5939 trying to route what was effectively the same domain again. This normally
5940 caused routing to fail (because the router wouldn't handle the domain
5941 twice). Now Exim checks for this case specially, and just changes the
5942 casing of the domain that it ultimately uses when it transmits the message
5945 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5948 50. If a filter generated a file delivery with a non-absolute name (possible if
5949 no home directory exists for the router), the forbid_file option was not
5952 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5953 the existing equality matching.
5955 52. Exim was using ints instead of ino_t variables in some places where it was
5956 dealing with inode numbers.
5958 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5959 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5960 called TMPDIR, and if it finds it is different, it changes its value.
5962 54. The smtp_printf() function is now made available to local_scan() so
5963 additional output lines can be written before returning. There is also an
5964 smtp_fflush() function to enable the detection of a dropped connection.
5965 The variables smtp_input and smtp_batched_input are exported to
5968 55. Changed the default runtime configuration: the message "Unknown user"
5969 has been removed from the ACL, and instead placed on the localuser router,
5970 using the cannot_route_message feature. This means that any verification
5971 failures that generate their own messages won't get overridden. Similarly,
5972 the "Unrouteable address" message that was in the ACL for unverifiable
5973 relay addresses has also been removed.
5975 56. Added hosts_avoid_esmtp to the smtp transport.
5977 57. The exicyclog script was not checking for the esoteric option
5978 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5979 will work only if exicyclog is run under the appropriate euid.
5981 58. Following a discussion on the list, the rules by which Exim recognises line
5982 endings on incoming messages have been changed. The -dropcr and drop_cr
5983 options are now no-ops, retained only for backwards compatibility. The
5984 following line terminators are recognized: LF CRLF CR. However, special
5985 processing applies to CR:
5987 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5988 nor a local message in the state where . is a terminator.
5990 (ii) If a bare CR is encountered in a header line, an extra space is added
5991 after the line terminator so as not to end the header. The reasoning
5992 behind this is that bare CRs in header lines are most likely either
5993 to be mistakes, or people trying to play silly games.
5995 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5996 was being incorrectly given as 18 bytes larger than it should have been.
5997 This is a VOB (very old bug).
5999 60. This may never have affected anything current, but just in case it has:
6000 When the local host is found other than at the start of a list of hosts,
6001 the local host, those with the same MX, and any that follow, are discarded.
6002 When the list in question was part of a longer list of hosts, the following
6003 hosts (not currently being processed) were also being discarded. This no
6004 longer happens. I'm not sure if this situation could ever has previously
6007 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6010 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6011 header. This is something that is recommended in a new Internet Draft, and
6012 is something that is documented as being done by Sendmail. There are two
6013 possible values. For messages generated by the autoreply transport, Exim
6016 Auto-Submitted: auto-replied
6018 whereas for all other generated messages (e.g. bounces) it adds
6020 Auto-Submitted: auto-generated
6022 63. The "personal" condition in filters now includes a test for the
6023 Auto-Submitted: header. If it contains the string "auto-" the message it
6024 not considered personal.
6026 64. Added rcpt_include_affixes as a generic transport option.
6028 65. Added queue_only_override (default true).
6030 66. Added the syslog_duplication option.
6032 67. If what should have been the first header line of a message consisted of
6033 a space followed by a colon, Exim was mis-interpreting it as a header line.
6034 It isn't of course - it is syntactically invalid and should therefore be
6035 treated as the start of the message body. The misbehaviour could have
6036 caused a number of strange effects, including loss of data in subsequent
6037 header lines, and spool format errors.
6039 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6040 client host had authenticated. This control can now be exercised by an ACL
6041 for more flexibility.
6043 69. By default, callouts do not happen when testing with -bh. There is now a
6044 variant, -bhc, which does actually run the callout code, including
6045 consulting and updating the callout cache.
6047 70. Added support for saslauthd authentication, courtesy of Alexander
6050 71. If statvfs() failed on the spool or log directories while checking their
6051 size for availability, Exim confusingly gave the error "space shortage".
6052 Furthermore, in debugging mode it crashed with a floating point exception.
6053 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6054 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6055 serious problem, Exim now writes to the main and panic logs when this
6056 happens, with details of the failure. It then refuses to accept the
6057 incoming message, giving the message "spool directory problem" or "log
6058 directory problem" with a 421 code for SMTP messages.
6060 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6061 0, 1, and 2 exist, because some OS complain for execs without them (see
6062 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6063 descriptors. However, the code omitted to check that the open succeeded,
6064 causing mysterious errors if for some reason the permissions on /dev/null
6065 got screwed. Now Exim writes a message to the main and panic logs, and
6066 bombs out if it can't open /dev/null.
6068 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6069 interact so that it is all more flexible. It is supposed to remain
6070 backwards compatible. Also added extra_local_interfaces.
6072 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6073 to bomb out with an assertion failure - to the client this appears as a
6074 connection drop. This problem occurs in the part of the code that was taken
6075 from the Samba project. Fortunately, the assertion is in a very simple
6076 function, so I have fixed this by reproducing the function inline in the
6077 one place where it is called, and arranging for authentication to fail
6078 instead of killing the process with assert().
6080 75. The SPA client code was not working when the server requested OEM rather
6081 than Unicode encoding.
6083 76. Added code to make require_files with a specific uid setting more usable in
6084 the case where statting the file as root fails - usually a non-root-mounted
6085 NFS file system. When this happens and the failure is EACCES, Exim now
6086 forks a subprocess and does the per-uid checking as the relevant uid.
6088 77. Added process_log_path.
6090 78. If log_file_path was not explicitly set, a setting of check_log_space or
6091 check_log_inodes was ignored.
6093 79. If a space check for the spool or log partitions fails, the incident is now
6094 logged. Of course, in the latter case the data may get lost...
6096 80. Added the %p formatting code to string_format() so that it can be used to
6097 print addresses in debug_print(). Adjusted all the address printing in the
6098 debugging in store.c to use %p rather than %d.
6100 81. There was a concern that a line of code in smtp_in.c could overflow a
6101 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6102 initially expressed, the concern was not well-founded, because trailing
6103 spaces are removed early. However, if the trailing spaces were followed by
6104 a NULL, they did not get removed, so the overflow was possible. Two fixes
6107 (a) I re-wrote the offending code in a cleaner fashion.
6108 (b) If an incoming SMTP command contains a NULL character, it is rejected
6111 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6112 runs initgroups(), so that if the Exim user is in any additional groups,
6113 they will be used during message reception.
6119 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob