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.
33 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
35 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
37 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
38 Merged patch from Sebastian Wiedenroth.
40 JH/05 Fix results-pipe from transport process. Several recipients, combined
41 with certificate use, exposed issues where response data items split
42 over buffer boundaries were not parsed properly. This eventually
43 resulted in duplicates being sent. This issue only became common enough
44 to notice due to the introduction of conection certificate information,
45 the item size being so much larger. Found and fixed by Wolfgang Breyha.
47 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
48 size buffer was used, resulting in syntax errors when an expansion
51 JH/07 Add support for directories of certificates when compiled with a GnuTLS
52 version 3.3.6 or later.
57 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
58 checkers that were complaining about end of non-void function with no
61 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
62 This was a regression intruduced in 4.83 by another bugfix.
64 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
66 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
67 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
73 TF/01 Correctly close the server side of TLS when forking for delivery.
75 When a message was received over SMTP with TLS, Exim failed to clear up
76 the incoming connection properly after forking off the child process to
77 deliver the message. In some situations the subsequent outgoing
78 delivery connection happened to have the same fd number as the incoming
79 connection previously had. Exim would try to use TLS and fail, logging
80 a "Bad file descriptor" error.
82 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
83 utilities have not been installed.
85 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
86 temporary space as the ACL may create new global variables.
88 TL/01 LDAP support uses per connection or global context settings, depending
89 upon the detected version of the libraries at build time.
91 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
92 to extract and use the src ip:port in logging and expansions as if it
93 were a direct connection from the outside internet. PPv2 support was
94 updated based on HAProxy spec change in May 2014.
96 JH/02 Add ${listextract {number}{list}{success}{fail}}.
98 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
99 Properly escape header and check for NULL return.
101 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
104 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
106 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
107 characters in header names, implemented as a verify condition.
108 Contributed by Michael Fischer v. Mollard.
110 TL/05 Rename SPF condition results err_perm and err_temp to standardized
111 results permerror and temperror. Previous values are deprecated but
112 still accepted. In a future release, err_perm and err_temp will be
113 completely removed, which will be a backward incompatibility if the
114 ACL tests for either of these two old results. Patch contributed by
115 user bes-internal on the mailing list.
117 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
119 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
120 selectors, in both main and reject logs.
122 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
125 JH/07 Add malware type "sock" for talking to simple daemon.
127 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
129 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
130 routers/transports under cutthrough routing.
132 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
133 numbers. Touch up "bool" conditional to keep the same definition.
135 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
137 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
139 JH/12 Expand items in router/transport headers_add or headers_remove lists
140 individually rather than the list as a whole. Bug 1452.
142 Required for reasonable handling of multiple headers_ options when
143 they may be empty; requires that headers_remove items with embedded
144 colons must have them doubled (or the list-separator changed).
146 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
147 view the policy declared in the DMARC record. Currently, $dmarc_status
148 is a combined value of both the record presence and the result of the
151 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
153 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
154 dnslookup router and the smtp transport (applying to the forward
157 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
158 of ldap servers used for a specific lookup. Patch provided by Heiko
161 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
162 New variable $lookup_dnssec_authenticated for observability.
164 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
165 Patch submitted by Lars Timman.
167 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
169 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
170 Requires trusted mode and valid format message id, aborts otherwise.
171 Patch contributed by Heiko Schlichting.
173 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
174 certextract with support for various fields. Bug 1358.
176 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
177 is requested by default, modifiable by smtp transport option
180 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
181 operate on certificate variables to give certificate fingerprints
182 Also new ${sha256:cert_variable}.
184 JH/23 The PRDR feature is moved from being Experimental into the mainline.
186 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
187 Christian Aistleitner.
189 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
191 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
192 file. Patch from Wolfgang Breyha.
194 JH/25 Expand the coverage of the delivery $host and $host_address to
195 client authenticators run in verify callout. Bug 1476.
197 JH/26 Port service names are now accepted for tls_on_connect_ports, to
198 align with daemon_smtp_ports. Bug 72.
200 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
201 support and error reporting did not work properly.
203 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
204 and is readable. Patch from Andrew Colin Kissa.
206 TL/14 Enhance documentation of ${run expansion and how it parses the
207 commandline after expansion, particularly in the case when an
208 unquoted variable expansion results in an empty value.
210 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
212 PP/02 Fix internal collision of T_APL on systems which support RFC3123
213 by renaming away from it. Addresses GH issue 15, reported by
216 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
218 TL/15 SECURITY: prevent double expansion in math comparison functions
219 (can expand unsanitized data). Not remotely exploitable.
226 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
228 PP/02 Make -n do something, by making it not do something.
229 When combined with -bP, the name of an option is not output.
231 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
234 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
235 $sender_host_name and config options to manage this, and basic check
238 PP/05 DSCP support for outbound connections and control modifier for inbound.
240 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
241 (Only plugin which currently uses this is kerberos4, which nobody should
242 be using, but we should make it available and other future plugins might
243 conceivably use it, even though it would break NAT; stuff *should* be
244 using channel bindings instead).
246 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
247 name; added for Sendmail compatibility; requires admin caller.
248 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
249 just ignore it); requires trusted caller.
250 Also parse but ignore: -Ac -Am -X<logfile>
253 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
255 TL/02 Add +smtp_confirmation as a default logging option.
257 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
258 Patch by Magnus Holmgren from 2007-02-20.
260 TL/04 Bugzilla 1281 - Spec typo.
261 Bugzilla 1283 - Spec typo.
262 Bugzilla 1290 - Spec grammar fixes.
264 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
266 TL/06 Add Experimental DMARC support using libopendmarc libraries.
268 TL/07 Fix an out of order global option causing a segfault. Reported to dev
269 mailing list by by Dmitry Isaikin.
271 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
273 JH/02 Support "G" suffix to numbers in ${if comparisons.
275 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
277 NM/01 Bugzilla 1197 - Spec typo
278 Bugzilla 1196 - Spec examples corrections
280 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
282 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
283 gnutls_enable_pkcs11, but renamed to more accurately indicate its
286 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
287 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
289 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
290 "acl {{name}{arg}...}", and optional args on acl condition
293 JH/05 Permit multiple router/transport headers_add/remove lines.
295 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
297 JH/07 Avoid using a waiting database for a single-message-only transport.
298 Performance patch from Paul Fisher. Bugzilla 1262.
300 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
303 JH/09 Add $headers_added variable, with content from use of ACL modifier
304 add_header (but not yet added to the message). Bugzilla 199.
306 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
307 Pulled from Bugzilla 817 by Wolfgang Breyha.
309 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
311 (nb: this is the same fix as in Exim 4.80.1)
313 JH/11 Add A= logging on delivery lines, and a client_set_id option on
316 JH/12 Add optional authenticated_sender logging to A= and a log_selector
319 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
321 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
322 advertise SMTP AUTH mechanism to us, instead of a generic
323 protocol violation error. Also, make Exim more robust to bad
324 data from the Dovecot auth socket.
326 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
328 When a queue runner is handling a message, Exim first routes the
329 recipient addresses, during which it prunes them based on the retry
330 hints database. After that it attempts to deliver the message to
331 any remaining recipients. It then updates the hints database using
334 So if a recipient address works intermittently, it can get repeatedly
335 deferred at routing time. The retry hints record remains fresh so the
336 address never reaches the final cutoff time.
338 This is a fairly common occurrence when a user is bumping up against
339 their storage quota. Exim had some logic in its local delivery code
340 to deal with this. However it did not apply to per-recipient defers
341 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
343 This change adds a proper retry rule check during routing so that the
344 final cutoff time is checked against the message's age. We only do
345 this check if there is an address retry record and there is not a
346 domain retry record; this implies that previous attempts to handle
347 the address had the retry_use_local_parts option turned on. We use
348 this as an approximation for the destination being like a local
349 delivery, as in LMTP.
351 I suspect this new check makes the old local delivery cutoff check
352 redundant, but I have not verified this so I left the code in place.
354 TF/02 Correct gecos expansion when From: is a prefix of the username.
356 Test 0254 submits a message to Exim with the header
360 When I ran the test suite under the user fanf2, Exim expanded
361 the header to contain my full name, whereas it should have added
362 a Resent-Sender: header. It erroneously treats any prefix of the
363 username as equal to the username.
365 This change corrects that bug.
367 GF/01 DCC debug and logging tidyup
368 Error conditions log to paniclog rather than rejectlog.
369 Debug lines prefixed by "DCC: " to remove any ambiguity.
371 TF/03 Avoid unnecessary rebuilds of lookup-related code.
373 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
374 Bug spotted by Jeremy Harris; was flawed since initial commit.
375 Would have resulted in OCSP responses post-SNI triggering an Exim
376 NULL dereference and crash.
378 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
380 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
381 Bug detection, analysis and fix by Samuel Thibault.
382 Bugzilla 1331, Debian bug #698092.
384 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
386 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
387 Server implementation by Todd Lyons, client by JH.
388 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
389 config variable "prdr_enable" controls whether the server
390 advertises the facility. If the client requests PRDR a new
391 acl_data_smtp_prdr ACL is called once for each recipient, after
392 the body content is received and before the acl_smtp_data ACL.
393 The client is controlled by bolth of: a hosts_try_prdr option
394 on the smtp transport, and the server advertisement.
395 Default client logging of deliveries and rejections involving
396 PRDR are flagged with the string "PRDR".
398 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
399 fclose(). Diagnosis by Todd Lyons.
401 PP/17 Update configure.default to handle IPv6 localhost better.
402 Patch by Alain Williams (plus minor tweaks).
405 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
406 This is now consistent with GnuTLS, and is now documented: the
407 previous undocumented portable approach to treating the option as
408 unset was to force an expansion failure. That still works, and
409 an empty string is now equivalent.
411 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
412 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
413 not performing validation itself.
415 PP/20 Added force_command boolean option to pipe transport.
416 Patch from Nick Koston, of cPanel Inc.
418 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
421 TF/04 Added udpsend ACL modifer and hexquote expansion operator
423 PP/21 Fix eximon continuous updating with timestamped log-files.
424 Broken in a format-string cleanup in 4.80, missed when I repaired the
425 other false fix of the same issue.
426 Report and fix from Heiko Schlichting.
429 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
430 Report from Prashanth Katuri.
432 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
433 It's SecureTransport, so affects any MacOS clients which use the
434 system-integrated TLS libraries, including email clients.
436 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
437 using a MIME ACL for non-SMTP local injection.
438 Report and assistance in diagnosis by Warren Baker.
440 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
442 JH/16 Fix comparisons for 64b. Bugzilla 1385.
444 TL/09 Add expansion variable $authenticated_fail_id to keep track of
445 last id that failed so it may be referenced in subsequent ACL's.
447 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
450 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
451 ldap library initialization, allowing self-signed CA's to be
452 used. Also properly sets require_cert option later in code by
453 using NULL (global ldap config) instead of ldap handle (per
454 session). Bug diagnosis and testing by alxgomz.
456 TL/12 Enhanced documentation in the ratelimit.pl script provided in
457 the src/util/ subdirectory.
459 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
460 renamed to Transport Post Delivery Action by Jeremy Harris, as
463 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
464 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
465 redis_servers = needs to be configured which will be used by the redis
466 lookup. Patch from Warren Baker, of The Packet Hub.
468 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
470 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
471 hostname or reverse DNS when processing a host list. Used suggestions
472 from multiple comments on this bug.
474 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
476 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
477 Missed a few lines, added it to make the runtest require no keyboard
480 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
481 contains upper case chars. Make router use caseful_local_part.
483 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
484 support when GnuTLS has been built with p11-kit.
490 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
492 This, or similar/improved, will also be change PP/11 of 4.82.
498 PP/01 Handle short writes when writing local log-files.
499 In practice, only affects FreeBSD (8 onwards).
500 Bugzilla 1053, with thanks to Dmitry Isaikin.
502 NM/01 Bugzilla 949 - Documentation tweak
504 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
507 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
509 PP/02 Implemented gsasl authenticator.
511 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
513 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
514 `pkg-config foo` for cflags/libs.
516 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
517 with rest of GSASL and with heimdal_gssapi.
519 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
520 `pkg-config foo` for cflags/libs for the TLS implementation.
522 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
523 properties get this fed in as external SSF. A number of robustness
524 and debugging improvements to the cyrus_sasl authenticator.
526 PP/08 cyrus_sasl server now expands the server_realm option.
528 PP/09 Bugzilla 1214 - Log authentication information in reject log.
529 Patch by Jeremy Harris.
531 PP/10 Added dbmjz lookup type.
533 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
535 PP/12 MAIL args handles TAB as well as SP, for better interop with
536 non-compliant senders.
537 Analysis and variant patch by Todd Lyons.
539 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
540 Bug report from Lars Müller <lars@samba.org> (via SUSE),
541 Patch from Dirk Mueller <dmueller@suse.com>
543 PP/13 tls_peerdn now print-escaped for spool files.
544 Observed some $tls_peerdn in wild which contained \n, which resulted
545 in spool file corruption.
547 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
548 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
549 or write after TLS renegotiation, which otherwise led to messages
552 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
553 as a tracking header (ie: a signed header comes before the signature).
554 Patch from Wolfgang Breyha.
556 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
557 comma-sep list; embedded commas doubled.
559 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
561 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
563 Report and patch from Dmitry Banschikov.
565 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
566 Removed SSL_clear() after SSL_new() which led to protocol negotiation
567 failures. We appear to now support TLS1.1+ with Exim.
569 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
570 lets Exim select keys and certificates based upon TLS SNI from client.
571 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
572 before an outbound SMTP session. New log_selector, +tls_sni.
574 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
575 NULL dereference. Report and patch from Alun Jones.
577 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
578 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
579 Not seeing resolver debug output on NetBSD, but suspect this is a
580 resolver implementation change.
582 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
583 Left warnings. Added "eximon gdb" invocation mode.
585 PP/21 Defaulting "accept_8bitmime" to true, not false.
587 PP/22 Added -bw for inetd wait mode support.
589 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
590 locate the relevant includes and libraries. Made this the default.
592 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
593 Bugzilla 1246, report and most of solution from Tomasz Kusy.
595 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
596 This may cause build issues on older platforms.
598 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
599 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
600 gnutls_require_mac & gnutls_require_protocols (no longer supported).
601 Added SNI support via GnuTLS too.
602 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
604 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
606 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
607 Applied second patch from Janne, implementing suggestion to default
608 multiple-strings-in-record handling to match SPF spec.
610 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
612 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
613 read-only, out of scope).
614 Patch from Wolfgang Breyha, report from Stuart Northfield.
616 PP/29 Fix three issues highlighted by clang analyser static analysis.
617 Only crash-plausible issue would require the Cambridge-specific
618 iplookup router and a misconfiguration.
619 Report from Marcin Mirosław.
621 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
623 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
624 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
625 As part of this, removing so much warning spew let me fix some minor
626 real issues in debug logging.
628 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
629 assignment on my part. Fixed.
631 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
632 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
633 Janne Snabb (who went above and beyond: thank you).
635 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
636 string otherwise requires a connection and a bunch more work and it's
637 relatively easy to get wrong. Should also expose TLS library linkage
640 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
641 64-bit ${eval} (JH/03).
643 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
644 GNU libc to support some of the 64-bit stuff, should not lead to
645 conflicts. Defined before os.h is pulled in, so if a given platform
646 needs to override this, it can.
648 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
649 protection layer was required, which is not implemented.
650 Bugzilla 1254, patch from Wolfgang Breyha.
652 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
653 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
654 tls_dhparam take prime identifiers. Also unbreak combination of
655 OpenSSL+DH_params+TLSSNI.
657 PP/39 Disable SSLv2 by default in OpenSSL support.
663 PP/01 Solaris build fix for Oracle's LDAP libraries.
664 Bugzilla 1109, patch from Stephen Usher.
666 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
668 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
671 TF/02 Fix a couple more cases where we did not log the error message
672 when unlink() failed. See also change 4.74-TF/03.
674 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
675 lock up or crash if it happened to be inside a call to libc when it
676 got a SIGUSR1 from exiwhat.
678 The SIGUSR1 handler appends the current process status to the process
679 log which is later printed by exiwhat. It used to use the general
680 purpose logging code to do this, but several functions it calls are
681 not safe for signals.
683 The new output code in the SIGUSR1 handler is specific to the process
684 log, and simple enough that it's easy to inspect for signal safety.
685 Removing some special cases also simplifies the general logging code.
686 Removing the spurious timestamps from the process log simplifies
689 TF/04 Improved ratelimit ACL condition.
691 The /noupdate option has been deprecated in favour of /readonly which
692 has clearer semantics. The /leaky, /strict, and /readonly update modes
693 are mutually exclusive. The update mode is no longer included in the
694 database key; it just determines when the database is updated. (This
695 means that when you upgrde Exim will forget old rate measurements.)
697 Exim now checks that the per_* options are used with an update mode that
698 makes sense for the current ACL. For example, when Exim is processing a
699 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
700 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
701 must specify per_mail/readonly. If you omit the update mode it defaults to
702 /leaky where that makes sense (as before) or /readonly where required.
704 The /noupdate option is now undocumented but still supported for
705 backwards compatibility. It is equivalent to /readonly except that in
706 ACLs where /readonly is required you may specify /leaky/noupdate or
707 /strict/noupdate which are treated the same as /readonly.
709 A useful new feature is the /count= option. This is a generalization
710 of the per_byte option, so that you can measure the throughput of other
711 aggregate values. For example, the per_byte option is now equivalent
712 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
714 The per_rcpt option has been generalized using the /count= mechanism
715 (though it's more complicated than the per_byte equivalence). When it is
716 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
717 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
718 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
719 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
720 Note that using per_rcpt with a non-readonly update mode in more than
721 one ACL will cause the recipients to be double-counted. (The per_mail
722 and per_byte options don't have this problem.)
724 The handling of very low rates has changed slightly. If the computed rate
725 is less than the event's count (usually one) then this event is the first
726 after a long gap. In this case the rate is set to the same as this event's
727 count, so that the first message of a spam run is counted properly.
729 The major new feature is a mechanism for counting the rate of unique
730 events. The new per_addr option counts the number of different
731 recipients that someone has sent messages to in the last time period. It
732 behaves like per_rcpt if all the recipient addresses are different, but
733 duplicate recipient addresses do not increase the measured rate. Like
734 the /count= option this is a general mechanism, so the per_addr option
735 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
736 example, measure the rate that a client uses different sender addresses
737 with the options per_mail/unique=$sender_address. There are further
738 details in the main documentation.
740 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
742 TF/06 Removed a few PCRE remnants.
744 TF/07 Automatically extract Exim's version number from tags in the git
745 repository when doing development or release builds.
747 PP/02 Raise smtp_cmd_buffer_size to 16kB.
748 Bugzilla 879. Patch from Paul Fisher.
750 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
751 Heavily based on revision 40f9a89a from Simon Arlott's tree.
754 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
756 PP/05 Variable $av_failed, true if the AV scanner deferred.
757 Bugzilla 1078. Patch from John Horne.
759 PP/06 Stop make process more reliably on build failure.
760 Bugzilla 1087. Patch from Heiko Schlittermann.
762 PP/07 Make maildir_use_size_file an _expandable_ boolean.
763 Bugzilla 1089. Patch from Heiko Schlittermann.
765 PP/08 Handle ${run} returning more data than OS pipe buffer size.
766 Bugzilla 1131. Patch from Holger Weiß.
768 PP/09 Handle IPv6 addresses with SPF.
769 Bugzilla 860. Patch from Wolfgang Breyha.
771 PP/10 GnuTLS: support TLS 1.2 & 1.1.
773 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
776 PP/11 match_* no longer expand right-hand-side by default.
777 New compile-time build option, EXPAND_LISTMATCH_RHS.
778 New expansion conditions, "inlist", "inlisti".
780 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
782 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
784 PP/14 fix log_write() format string regression from TF/03.
785 Bugzilla 1152. Patch from Dmitry Isaikin.
791 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
793 PP/02 Harmonised TLS library version reporting; only show if debugging.
794 Layout now matches that introduced for other libraries in 4.74 PP/03.
796 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
798 PP/04 New "dns_use_edns0" global option.
800 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
803 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
804 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
806 TK/01 Updated PolarSSL code to 0.14.2.
807 Bugzilla 1097. Patch from Andreas Metzler.
809 PP/07 Catch divide-by-zero in ${eval:...}.
812 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
815 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
816 format-string attack -- SECURITY: remote arbitrary code execution.
818 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
819 time unintentionally subject to list matching rules, letting the header
820 cause arbitrary Exim lookups (of items which can occur in lists, *not*
821 arbitrary string expansion). This allowed for information disclosure.
823 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
824 INT_MIN/-1 -- value coerced to INT_MAX.
830 NM/01 Workround for PCRE version dependency in version reporting
833 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
834 This fixes portability to compilers other than gcc, notably
835 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
837 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
838 makefiles for portability to HP-UX and POSIX correctness.
840 PP/01 Permit LOOKUP_foo enabling on the make command-line.
841 Also via indented variable definition in the Makefile.
842 (Debugging by Oliver Heesakkers).
844 PP/02 Restore caching of spamd results with expanded spamd_address.
845 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
847 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
848 Improves build reliability. Fix from: Frank Elsner
850 NM/02 Fix wide character breakage in the rfc2047 coding
851 Fixes bug 1064. Patch from Andrey N. Oktyabrski
853 NM/03 Allow underscore in dnslist lookups
854 Fixes bug 1026. Patch from Graeme Fowler
856 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
857 Code patches from Adam Ciarcinski of NetBSD.
859 NM/04 Fixed exiqgrep to cope with mailq missing size issue
862 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
863 is logged, to avoid truncation. Patch from John Horne.
865 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
866 Patch from Jakob Hirsch.
868 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
869 SQL string expansion failure details.
870 Patch from Andrey Oktyabrski.
872 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
873 Patch from Simon Arlott.
875 PP/09 New lookups functionality failed to compile on old gcc which rejects
876 extern declarations in function scope.
877 Patch from Oliver Fleischmann
879 PP/10 Use sig_atomic_t for flags set from signal handlers.
880 Check getgroups() return and improve debugging.
881 Fixed developed for diagnosis in bug 927 (which turned out to be
884 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
885 Patch from Mark Zealey.
887 PP/12 Bugzilla 1056: Improved spamd server selection.
888 Patch from Mark Zealey.
890 PP/13 Bugzilla 1086: Deal with maildir quota file races.
891 Based on patch from Heiko Schlittermann.
893 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
894 Patch from Uwe Doering, sign-off by Michael Haardt.
896 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
897 variable declaration deep within a block. Bug and patch from
900 PP/15 lookups-Makefile IRIX compatibilty coercion.
902 PP/16 Make DISABLE_DKIM build knob functional.
904 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
905 Patch by Simon Arlott
907 TF/03 Fix valgrind.h portability to C89 compilers that do not support
908 variable argument macros. Our copy now differs from upstream.
914 TF/01 Failure to get a lock on a hints database can have serious
915 consequences so log it to the panic log.
917 TF/02 Log LMTP confirmation messages in the same way as SMTP,
918 controlled using the smtp_confirmation log selector.
920 TF/03 Include the error message when we fail to unlink a spool file.
922 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
923 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
924 for maintaining out-of-tree patches for some time.
926 PP/01 Bugzilla 139: Documentation and portability issues.
927 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
928 Handle per-OS dynamic-module compilation flags.
930 PP/02 Let /dev/null have normal permissions.
931 The 4.73 fixes were a little too stringent and complained about the
932 permissions on /dev/null. Exempt it from some checks.
933 Reported by Andreas M. Kirchwitz.
935 PP/03 Report version information for many libraries, including
936 Exim version information for dynamically loaded libraries. Created
937 version.h, now support a version extension string for distributors
938 who patch heavily. Dynamic module ABI change.
940 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
941 privilege escalation vulnerability whereby the Exim run-time user
942 can cause root to append content of the attacker's choosing to
945 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
948 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
949 If dropping privileges for untrusted macros, we disabled normal logging
950 on the basis that it would fail; for the Exim run-time user, this is not
951 the case, and it resulted in successful deliveries going unlogged.
952 Fixed. Reported by Andreas Metzler.
958 PP/01 Date: & Message-Id: revert to normally being appended to a message,
959 only prepend for the Resent-* case. Fixes regression introduced in
960 Exim 4.70 by NM/22 for Bugzilla 607.
962 PP/02 Include check_rfc2047_length in configure.default because we're seeing
963 increasing numbers of administrators be bitten by this.
965 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
967 PP/03 Bugzilla 994: added openssl_options main configuration option.
969 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
971 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
973 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
975 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
976 without a peer certificate, leading to a segfault because of an
977 assumption that peers always have certificates. Be a little more
978 paranoid. Problem reported by Martin Tscholak.
980 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
981 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
982 NB: ClamAV planning to remove STREAM in "middle of 2010".
983 CL also introduces -bmalware, various -d+acl logging additions and
984 more caution in buffer sizes.
986 PP/09 Implemented reverse_ip expansion operator.
988 PP/10 Bugzilla 937: provide a "debug" ACL control.
990 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
992 PP/12 Bugzilla 973: Implement --version.
994 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
996 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
998 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1000 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1001 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1002 ignore trailing whitespace.
1004 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1006 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1009 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1010 Notification from Dr Andrew Aitchison.
1012 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1013 ExtendedDetectionInfo response format.
1014 Notification from John Horne.
1016 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1019 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1020 XSL and documented dependency on system catalogs, with examples of how
1023 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1026 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1027 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1028 configuration file which is writeable by the Exim user or group.
1030 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1031 of configuration files to cover files specified with the -C option if
1032 they are going to be used with root privileges, not just the default
1035 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1036 option (effectively making it always true).
1038 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1039 files to be used while preserving root privileges.
1041 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1042 that rogue child processes cannot use them.
1044 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1045 run-time user, instead of root.
1047 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1048 Exim run-time user without dropping privileges.
1050 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1051 result string, instead of calling string_vformat() twice with the same
1054 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1055 for other users. Others should always drop root privileges if they use
1056 -C on the command line, even for a whitelisted configure file.
1058 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1060 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1066 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1067 $data_path, and $header_path variables; fixed documentation bugs and
1070 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1071 exipick to access non-standard spools, including the "frozen" queue
1074 NM/01 Bugzilla 965: Support mysql stored procedures.
1075 Patch from Alain Williams
1077 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1079 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1080 Patch from Andreas Metzler
1082 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1083 Patch from Kirill Miazine
1085 NM/05 Bugzilla 671: Added umask to procmail example.
1087 JJ/03 installed exipick 20100323.0, fixing doc bug
1089 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1090 directory. Notification and patch from Dan Rosenberg.
1092 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1094 TK/02 Improve log output when DKIM signing operation fails.
1096 MH/01 Treat the transport option dkim_domain as a colon separated
1097 list, not as a single string, and sign the message with each element,
1098 omitting multiple occurences of the same signer.
1100 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1101 Bugzilla 985, 986. Patch by Simon Arlott
1103 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1104 Patch by Simon Arlott
1106 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1107 MBX locking. Notification from Dan Rosenberg.
1113 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1115 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1117 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1119 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1121 NM/04 Bugzilla 915: Flag broken perl installation during build.
1127 TK/01 Added patch by Johannes Berg that expands the main option
1128 "spamd_address" if it starts with a dollar sign.
1130 TK/02 Write list of recipients to X-Envelope-Sender header when building
1131 the mbox-format spool file for content scanning (suggested by Jakob
1134 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1135 (http://www.dcc-servers.net/) support via dccifd. Activated by
1136 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1138 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1139 by Mark Daniel Reidel <mr@df.eu>.
1141 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1142 When building exim an external PCRE library is now needed -
1143 PCRE is a system library on the majority of modern systems.
1144 See entry on PCRE_LIBS in EDITME file.
1146 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1147 conversation. Added nologin parameter to request.
1148 Patch contributed by Kirill Miazine.
1150 TF/01 Do not log submission mode rewrites if they do not change the address.
1152 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1154 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1155 log files in place. Contributed by Roberto Lima.
1157 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1159 TF/03 Bugzilla 615: When checking the local_parts router precondition
1160 after a local_part_suffix or local_part_prefix option, Exim now
1161 does not use the address's named list lookup cache, since this
1162 contains cached lookups for the whole local part.
1164 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1165 Robert Millan. Documentation is in experimental-spec.txt.
1167 TF/04 Bugzilla 668: Fix parallel build (make -j).
1169 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1171 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1172 Patch provided by Jan Srzednicki.
1174 TF/05 Leading white space used to be stripped from $spam_report which
1175 wrecked the formatting. Now it is preserved.
1177 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1178 that they are available at delivery time.
1180 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1182 TF/08 TLS error reporting now respects the incoming_interface and
1183 incoming_port log selectors.
1185 TF/09 Produce a more useful error message if an SMTP transport's hosts
1186 setting expands to an empty string.
1188 NM/06 Bugzilla 744: EXPN did not work under TLS.
1189 Patch provided by Phil Pennock.
1191 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1192 Patch provided by Richard Godbee.
1194 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1195 acl_smtp_notquit, added index entry.
1197 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1198 Patch provided by Eugene Bujak.
1200 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1201 accept(). Patch provided by Maxim Dounin.
1203 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1204 Patch provided by Phil Pennock.
1206 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1208 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1209 Patch provided by Brad "anomie" Jorsch.
1211 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1212 Patch provided by Dean Brooks.
1214 TK/05 Add native DKIM support (does not depend on external libraries).
1216 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1217 Patch provided by Graeme Fowler.
1219 NM/16 Bugzilla 851: Documentation example syntax fix.
1221 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1223 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1226 NM/19 Bugzilla 745: TLS version reporting.
1227 Patch provided by Phil Pennock.
1229 NM/20 Bugzilla 167: bool: condition support.
1230 Patch provided by Phil Pennock.
1232 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1233 clients. Patch provided by Phil Pennock.
1235 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1236 Patch provided by Brad "anomie" Jorsch.
1238 NM/23 Bugzilla 687: Fix misparses in eximstats.
1239 Patch provided by Heiko Schlittermann.
1241 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1242 Patch provided by Heiko Schlittermann.
1244 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1245 plus update to original patch.
1247 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1249 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1250 Patch provided by David Brownlee.
1252 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1254 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1256 NM/30 Bugzilla 888: TLS documentation bugfixes.
1258 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1260 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1261 Unlike the original bugzilla I have changed all shell scripts in src tree.
1263 NM/33 Bugzilla 898: Transport filter timeout fix.
1264 Patch by Todd Rinaldo.
1266 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1267 Patch by Serge Demonchaux.
1269 NM/35 Bugzilla 39: Base64 decode bug fixes.
1270 Patch by Jakob Hirsch.
1272 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1274 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1276 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1278 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1284 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1285 ALT-N's libdkim that I have put here:
1286 http://duncanthrax.net/exim-experimental/
1288 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1289 were called 'true' and it seems that C99 defines that as a reserved
1290 keyword to be used with 'bool' variable types. That means you could
1291 not include C99-style headers which use bools without triggering
1292 build errors in sieve.c.
1294 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1295 as mailq or other aliases. Changed the --help handling significantly
1296 to do whats expected. exim_usage() emits usage/help information.
1298 SC/01 Added the -bylocaldomain option to eximstats.
1300 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1302 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1304 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1310 PH/01 Another patch from the Sieve maintainer.
1312 PH/02 When an IPv6 address is converted to a string for single-key lookup
1313 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1314 dots are used instead of colons so that keys in lsearch files need not
1315 contain colons. This was done some time before quoting was made available
1316 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1317 (notated using the quote facility) so as to distinguish them from IPv4
1318 keys. This meant that lookups for IP addresses in host lists did not work
1319 for iplsearch lookups.
1321 This has been fixed by arranging for IPv6 addresses to be expressed with
1322 colons if the lookup type is iplsearch. This is not incompatible, because
1323 previously such lookups could never work.
1325 The situation is now rather anomolous, since one *can* have colons in
1326 ordinary lsearch keys. However, making the change in all cases is
1327 incompatible and would probably break a number of configurations.
1329 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1332 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1333 conversion specification without a maximum field width, thereby enabling
1334 a rogue spamd server to cause a buffer overflow. While nobody in their
1335 right mind would setup Exim to query an untrusted spamd server, an
1336 attacker that gains access to a server running spamd could potentially
1337 exploit this vulnerability to run arbitrary code as the Exim user.
1339 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1340 $primary_hostname instead of what libspf2 thinks the hosts name is.
1342 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1343 a directory entry by the name of the lookup key. Previously, if a
1344 symlink pointed to a non-existing file or a file in a directory that
1345 Exim lacked permissions to read, a lookup for a key matching that
1346 symlink would fail. Now it is enough that a matching directory entry
1347 exists, symlink or not. (Bugzilla 503.)
1349 PH/03 The body_linecount and body_zerocount variables are now exported in the
1352 PH/04 Added the $dnslist_matched variable.
1354 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1355 This means they are set thereafter only if the connection becomes
1358 PH/06 Added the client_condition to authenticators so that some can be skipped
1359 by clients under certain conditions.
1361 PH/07 The error message for a badly-placed control=no_multiline_responses left
1362 "_responses" off the end of the name.
1364 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1366 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1367 (without spaces) instead of just copying the configuration text.
1369 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1371 PH/11 Added $max_received_linelength.
1373 PH/12 Added +ignore_defer and +include_defer to host lists.
1375 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1376 way in which PCRE > 7.0 is built.
1378 PH/14 Implemented queue_only_load_latch.
1380 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1381 MAIL command. The effect was to mangle the value on 64-bit systems.
1383 PH/16 Another patch from the Sieve maintainer.
1385 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1387 PH/18 If a system quota error occurred while trying to create the file for
1388 a maildir delivery, the message "Mailbox is full" was not appended to the
1389 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1390 only to a quota excession during the actual writing of the file.
1392 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1393 characters?) which causes problems in log lines. The DN values are now
1394 passed through string_printing() before being added to log lines.
1396 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1397 and InterBase are left for another time.)
1399 PH/21 Added message_body_newlines option.
1401 PH/22 Guard against possible overflow in moan_check_errorcopy().
1403 PH/23 POSIX allows open() to be a macro; guard against that.
1405 PH/24 If the recipient of an error message contained an @ in the local part
1406 (suitably quoted, of course), incorrect values were put in $domain and
1407 $local_part during the evaluation of errors_copy.
1413 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1414 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1417 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1418 issue a MAIL command.
1420 PH/02 In an ACL statement such as
1422 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1424 if a client was not listed at all, or was listed with a value other than
1425 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1426 the condition was not true (as it should be), so access was not denied.
1427 The bug was that the ! inversion was incorrectly passed on to the second
1428 item. This has been fixed.
1430 PH/03 Added additional dnslists conditions == and =& which are different from
1431 = and & when the dns lookup returns more than one IP address.
1433 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1434 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1436 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1437 FSYNC, which compiles an option called disable_fsync that allows for
1438 bypassing fsync(). The documentation is heavily laced with warnings.
1440 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1442 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1443 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1444 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1445 including adding "make clean"; (3) Added -fPIC when compiling the test
1446 dynamically loaded module, to get rid of a warning.
1448 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1449 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1450 = 0s. The bug is otherwise harmless.
1452 PH/07 There was a bug in the dovecot authenticator such that the value of
1453 $auth1 could be overwritten, and so not correctly preserved, after a
1454 successful authentication. This usually meant that the value preserved by
1455 the server_setid option was incorrect.
1457 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1459 PH/09 Installed PCRE release 7.0.
1461 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1462 run for batched SMTP input. It is now run at the start of every message
1463 in the batch. While fixing this I discovered that the process information
1464 (output by running exiwhat) was not always getting set for -bs and -bS
1465 input. This is fixed, and it now also says "batched" for BSMTP.
1467 PH/11 Added control=no_pipelining.
1469 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1470 patch, slightly modified), and move the expansion of helo_data till after
1471 the connection is made in the smtp transport (so it can use these
1474 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1476 PH/14 Added log_selector = +pid.
1478 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1480 PH/16 Add ${if forany and ${if forall.
1482 PH/17 Added dsn_from option to vary the From: line in DSNs.
1484 PH/18 Flush SMTP output before performing a callout, unless control =
1485 no_callout_flush is set.
1487 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1488 was true (the default) a successful delivery failed to delete the retry
1489 item, thus causing premature timeout of the address. The bug is now
1492 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1494 PH/21 Long custom messages for fakedefer and fakereject are now split up
1495 into multiline reponses in the same way that messages for "deny" and
1496 other ACL rejections are.
1498 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1499 with slight modification.
1501 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1502 draft, changing the syntax and factoring some duplicate code".
1504 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1505 for deliveries of the second and subsequent messages over the same SMTP
1508 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1509 ${reduce, with only minor "tidies".
1511 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1513 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1514 expansion side effects.
1516 PH/27 When a message times out after an over-quota error from an Exim-imposed
1517 quota, the bounce message says "mailbox is full". This message was not
1518 being given when it was a system quota that was exceeded. It now should
1521 MH/03 Made $recipients available in local_scan(). local_scan() already has
1522 better access to the recipient list through recipients_list[], but
1523 $recipients can be useful in postmaster-provided expansion strings.
1525 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1526 in the case of a MAIL command with additional options following the
1527 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1528 were accidentally chopped off.
1530 PH/29 SMTP synchronization checks are implemented when a command is read -
1531 there is a check that no more input is waiting when there shouldn't be
1532 any. However, for some commands, a delay in an ACL can mean that it is
1533 some time before the response is written. In this time, more input might
1534 arrive, invalidly. So now there are extra checks after an ACL has run for
1535 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1536 pipelining has not been advertised.
1538 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1540 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1541 returned in the SMTP response when smtp_return_error_details was false.
1542 This has been fixed.
1544 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1545 instead of the C I/O that was originally supplied, because problems were
1546 reported on Solaris.
1548 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1549 Exim which did not show up earlier: it was assuming that a call to
1550 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1551 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1552 was a macro that became an assignment, so it seemed to work. This has
1553 changed to a proper function call with a void return, hence the compile
1554 error. Exim's code has been fixed.
1556 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1559 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1561 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1563 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1564 because it turns out to be too big or there is a timeout) there is no
1565 "Completed" line in the log. When some messages of this type were
1566 selected by exigrep, they were listed as "not completed". Others were
1567 picked up by some special patterns. I have improved the selection
1568 criteria to be more general.
1570 PH/38 The host_find_failed option in the manualroute router can now be set
1571 to "ignore", to completely ignore a host whose IP address cannot be
1572 found. If all hosts are ignored, the behaviour is controlled by the new
1573 host_all_ignored option.
1575 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1576 homing or because of multiple MX records with /mx) generated more than
1577 one IP address, and the following item turned out to be the local host,
1578 all the secondary addresses of the first item were incorrectly removed
1579 from the list, along with the local host and any following hosts (which
1580 is what is supposed to happen).
1582 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1583 whoever called Exim into the -H file. In the case of the daemon it was
1584 behaving confusingly. When first started, it used values for whoever
1585 started the daemon, but after a SIGHUP it used the Exim user (because it
1586 calls itself on a restart). I have changed the code so that it now always
1589 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1590 message are rejected with the same error (e.g. no authentication or bad
1591 sender address), and a DATA command is nevertheless sent (as can happen
1592 with PIPELINING or a stupid MUA), the error message that was given to the
1593 RCPT commands is included in the rejection of the DATA command. This is
1594 intended to be helpful for MUAs that show only the final error to their
1597 PH/42 Another patch from the Sieve maintainer.
1599 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1600 Eximstats - Fixed some broken HTML links and added missing column headers
1602 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1603 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1605 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1607 PH/43 Yet another patch from the Sieve maintainer.
1609 PH/44 I found a way to check for a TCP/IP connection going away before sending
1610 the response to the final '.' that terminates a message, but only in the
1611 case where the client has not sent further data following the '.'
1612 (unfortunately, this is allowed). However, in many cases there won't be
1613 any further data because there won't be any more messages to send. A call
1614 to select() can be used: if it shows that the input is "ready", there is
1615 either input waiting, or the socket has been closed. An attempt to read
1616 the next input character can distinguish the two cases. Previously, Exim
1617 would have sent an OK response which the client would never have see.
1618 This could lead to message repetition. This fix should cure that, at
1619 least in a lot of common cases.
1621 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1622 advertised in response to EHLO.
1628 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1629 fixed by 4.65/MH/01 (is this a record?) are fixed:
1631 (i) An empty string was always treated as zero by the numeric comparison
1632 operators. This behaviour has been restored.
1634 (ii) It is documented that the numeric comparison operators always treat
1635 their arguments as decimal numbers. This was broken in that numbers
1636 starting with 0 were being interpreted as octal.
1638 While fixing these problems I realized that there was another issue that
1639 hadn't been noticed. Values of message_size_limit (both the global option
1640 and the transport option) were treated as octal if they started with 0.
1641 The documentation was vague. These values are now always treated as
1642 decimal, and I will make that clear in the documentation.
1648 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1649 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1652 MH/01 Don't check that the operands of numeric comparison operators are
1653 integers when their expansion is in "skipping" mode (fixes bug
1654 introduced by 4.64-PH/07).
1656 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1657 child addresses, Exim now panics and dies. Previously, because the count
1658 is held in a short int, deliveries were likely to be lost. As such a
1659 large number of recipients for a single message is ridiculous
1660 (performance will be very, very poor), I have chosen to impose a limit
1661 rather than extend the field.
1667 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1668 leftover -K file (the existence of which was triggered by #402).
1669 While we were at it, introduced process PID as part of the -K
1670 filename. This should rule out race conditions when creating
1673 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1674 processing considerably. Previous code took too long for large mails,
1675 triggering a timeout which in turn triggers #401.
1677 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1678 in the DK code in transports.c. sendfile() is not really portable,
1679 hence the _LINUX specificness.
1681 TF/01 In the add_headers option to the mail command in an Exim filter,
1682 there was a bug that Exim would claim a syntax error in any
1683 header after the first one which had an odd number of characters
1686 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1687 callout verification, Exim cached a "reject" for the entire domain. This
1688 is correct for most verifications, but it is not correct for a recipient
1689 verification with use_sender or use_postmaster set, because in that case
1690 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1691 case of MAIL FROM:<> rejection from other early rejections (e.g.
1692 rejection of HELO). When verifying a recipient using a non-null MAIL
1693 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1694 Whatever the result of the callout, the value of the domain cache is
1695 left unchanged (for any other kind of callout, getting as far as trying
1696 RCPT means that the domain itself is ok).
1698 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1701 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1702 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1703 session, the daemon ignores ECONNECT errors and logs others; it now
1704 ignores EPIPE as well.
1706 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1707 (quoted-printable decoding).
1709 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1710 later the small subsequent patch to fix an introduced bug.
1712 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1714 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1716 PH/08 An error is now given if message_size_limit is specified negative.
1718 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1719 to be given (somewhat) arbitrary names.
1721 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1724 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1725 miscellaneous code fixes
1727 PH/10 Added the log_reject_target ACL modifier to specify where to log
1730 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1731 hostname. This is wrong, because it relates to the incoming message (and
1732 probably the interface on which it is arriving) and not to the outgoing
1733 callout (which could be using a different interface). This has been
1734 changed to use the value of the helo_data option from the smtp transport
1735 instead - this is what is used when a message is actually being sent. If
1736 there is no remote transport (possible with a router that sets up host
1737 addresses), $smtp_active_hostname is used.
1739 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1740 tweaks were necessary in order to get it to work (see also 21 below):
1741 (a) The code assumed that strncpy() returns a negative number on buffer
1742 overflow, which isn't the case. Replaced with Exim's string_format()
1744 (b) There were several signed/unsigned issues. I just did the minimum
1745 hacking in of casts. There is scope for a larger refactoring.
1746 (c) The code used strcasecmp() which is not a standard C function.
1747 Replaced with Exim's strcmpic() function.
1748 (d) The code set only $1; it now sets $auth1 as well.
1749 (e) A simple test gave the error "authentication client didn't specify
1750 service in request". It would seem that Dovecot has changed its
1751 interface. Fortunately there's a specification; I followed it and
1752 changed what the client sends and it appears to be working now.
1754 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1757 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1758 address A is aliased to B and C, where B exists and C does not. Without
1759 -v the output is "A verified" because verification stops after a
1760 successful redirection if more than one address is generated. However,
1761 with -v the child addresses are also verified. Exim was outputting "A
1762 failed to verify" and then showing the successful verification for C,
1763 with its parentage. It now outputs "B failed to verify", showing B's
1764 parentage before showing the successful verification of C.
1766 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1767 look up a TXT record in a specific list after matching in a combined
1770 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1771 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1772 they consult the DNS. I had assumed they would set it the way they
1773 wanted; and indeed my experiments on Linux seem to show that in some
1774 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1775 To be on the safe side, however, I have now made the interface to
1776 host_find_byname() similar to host_find_bydns(), with an argument
1777 containing the DNS resolver options. The host_find_byname() function now
1778 sets these options at its start, just as host_find_bydns() does. The smtp
1779 transport options dns_qualify_single and dns_search_parents are passed to
1780 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1781 of host_find_byname() use the default settings of RES_DEFNAMES
1782 (qualify_single) but not RES_DNSRCH (search_parents).
1784 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1785 spool_read_header() do less string testing, by means of a preliminary
1786 switch on the second character of optional "-foo" lines. (This is
1787 overdue, caused by the large number of possibilities that now exist.
1788 Originally there were few.) While I was there, I also converted the
1789 str(n)cmp tests so they don't re-test the leading "-" and the first
1790 character, in the hope this might squeeze out yet more improvement.
1792 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1793 flag allowing group syntax was set by the header_syntax check but not
1794 turned off, possible causing trouble later; (2) The flag was not being
1795 set at all for the header_verify test, causing "group"-style headers to
1796 be rejected. I have now set it in this case, and also caused header_
1797 verify to ignore an empty address taken from a group. While doing this, I
1798 came across some other cases where the code for allowing group syntax
1799 while scanning a header line wasn't quite right (mostly, not resetting
1800 the flag correctly in the right place). These bugs could have caused
1801 trouble for malformed header lines. I hope it is now all correct.
1803 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1804 with the "reply" argument non-NULL. The code, however (which originally
1805 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1806 but it didn't always do it. This confused somebody who was copying the
1807 code for some other use. I have removed all the tests.
1809 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1810 feature that was used to support insecure browsers during the U.S. crypto
1811 embargo. It requires special client support, and Exim is probably the
1812 only MTA that supported it -- and would never use it because real RSA is
1813 always available. This code has been removed, because it had the bad
1814 effect of slowing Exim down by computing (never used) parameters for the
1815 RSA_EXPORT functionality.
1817 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1818 authenticator to fail if there's a tab character in the incoming data
1819 (there should never be unless someone is messing about, as it's supposed
1820 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1821 if the connection is using TLS or if the remote IP is the same as the
1822 local IP, and the "valid-client-cert option" if a client certificate has
1825 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1826 authenticators. This can be used for authorization after authentication
1827 succeeds. (In the case of plaintext, it servers for both authentication
1830 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1831 if any retry times were supplied.
1833 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1834 connection, where there is no client IP address to check. In this
1835 situation, the verify now always succeeds.
1837 PH/25 Applied John Jetmore's -Mset patch.
1839 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1841 PH/27 In a string expansion for a processed (not raw) header when multiple
1842 headers of the same name were present, leading whitespace was being
1843 removed from all of them, but trailing whitespace was being removed only
1844 from the last one. Now trailing whitespace is removed from each header
1845 before concatenation. Completely empty headers in a concatenation (as
1846 before) are ignored.
1848 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1849 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1851 PH/29 [Removed. This was a change that I later backed out, and forgot to
1852 correct the ChangeLog entry (that I had efficiently created) before
1853 committing the later change.]
1855 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1856 address errors (4xx response to RCPT) over the same connection as other
1857 messages routed to the same hosts. Such deliveries are always "forced",
1858 so retry times are not inspected. This resulted in far too many retries
1859 for the affected addresses. The effect occurred only when there were more
1860 hosts than the hosts_max_try setting in the smtp transport when it had
1861 the 4xx errors. Those hosts that it had tried were not added to the list
1862 of hosts for which the message was waiting, so if all were tried, there
1863 was no problem. Two fixes have been applied:
1865 (i) If there are any address or message errors in an SMTP delivery, none
1866 of the hosts (tried or untried) are now added to the list of hosts
1867 for which the message is waiting, so the message should not be a
1868 candidate for sending over the same connection that was used for a
1869 successful delivery of some other message. This seems entirely
1870 reasonable: after all the message is NOT "waiting for some host".
1871 This is so "obvious" that I'm not sure why it wasn't done
1872 previously. Hope I haven't missed anything, but it can't do any
1873 harm, as the worst effect is to miss an optimization.
1875 (ii) If, despite (i), such a delivery is accidentally attempted, the
1876 routing retry time is respected, so at least it doesn't keep
1877 hammering the server.
1879 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1880 in ${readsocket because some servers need this prod.
1882 PH/32 Added some extra debug output when updating a wait-xxx database.
1884 PH/33 The hint "could be header name not terminated by colon", which has been
1885 given for certain expansion errors for a long time, was not being given
1886 for the ${if def:h_colon_omitted{... case.
1888 PH/34 The spec says: "With one important exception, whenever a domain list is
1889 being scanned, $domain contains the subject domain." There was at least
1890 one case where this was not true.
1892 PH/35 The error "getsockname() failed: connection reset by peer" was being
1893 written to the panic log as well as the main log, but it isn't really
1894 panic-worthy as it just means the connection died rather early on. I have
1895 removed the panic log writing for the ECONNRESET error when getsockname()
1898 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1899 runs only) independently of the message's sender address. This meant
1900 that, if the 4xx error was in fact related to the sender, a different
1901 message to the same recipient with a different sender could confuse
1902 things. In particualar, this can happen when sending to a greylisting
1903 server, but other circumstances could also provoke similar problems.
1904 I have changed the default so that the retry time for these errors is now
1905 based a combination of the sender and recipient addresses. This change
1906 can be overridden by setting address_retry_include_sender=false in the
1909 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1910 remote server are returned as part of bounce messages. This was not
1911 happening for LMTP over a pipe (the lmtp transport), but now it is the
1912 same for both kinds of LMTP.
1914 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1915 in header lines that were in fact CNAMEs. This is no longer the case.
1917 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1918 and queue runs started by the daemon processed all messages. This has
1919 been fixed so that -R and -S can now usefully be given with -q<time>.
1921 PH/40 Import PCRE release 6.7 (fixes some bugs).
1923 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1925 PH/42 Give an error if -q is specified more than once.
1927 PH/43 Renamed the variables $interface_address and $interface_port as
1928 $received_ip_address and $received_port, to make it clear that these
1929 values apply to message reception, and not to the outgoing interface when
1930 a message is delivered. (The old names remain recognized, of course.)
1932 PH/44 There was no timeout on the connect() call when using a Unix domain
1933 socket in the ${readsocket expansion. There now is.
1935 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1936 be meaningful with "accept".
1938 SC/01 Eximstats V1.43
1939 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1941 SC/02 Eximstats V1.44
1942 Use a glob alias rather than an array ref in the generated
1943 parser. This improves both readability and performance.
1945 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1946 Collect SpamAssassin and rejection statistics.
1947 Don't display local sender or destination tables unless
1948 there is data to show.
1949 Added average volumes into the top table text output.
1951 SC/04 Eximstats V1.46
1952 Collect data on the number of addresses (recipients)
1953 as well as the number of messages.
1955 SC/05 Eximstats V1.47
1956 Added 'Message too big' to the list of mail rejection
1957 reasons (thanks to Marco Gaiarin).
1959 SC/06 Eximstats V1.48
1960 Mainlog lines which have GMT offsets and are too short to
1961 have a flag are now skipped.
1963 SC/07 Eximstats V1.49 (Alain Williams)
1964 Added the -emptyok flag.
1966 SC/08 Eximstats V1.50
1967 Fixes for obtaining the IP address from reject messages.
1969 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1970 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1971 whitesspace changes from 4.64-PH/27
1973 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1976 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1977 are found, allow negative numbers in numeric criteria)
1979 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1981 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1982 to match changes made in 4.64-PH/43
1984 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1986 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1987 "domain" is actually an IP address, and force a failure. This locks out
1988 those revolvers/nameservers that support "A-for-A" lookups, in
1989 contravention of the specifications.
1991 PH/48 When a host name was looked up from an IP address, and the subsequent
1992 forward lookup of the name timed out, the host name was left in
1993 $sender_host_name, contrary to the specification.
1995 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1996 restricted to single-key lookups, Exim was not diagnosing an error if
1997 * or *@ was used with a query-style lookup.
1999 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2001 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2002 long ago, but noone interested enough thought of it. Let's just say that
2003 the "1.1" means that there are some new functions that weren't there at
2004 some point in the past.
2006 PH/51 Error processing for expansion failure of helo_data from an smtp
2007 transport during callout processing was broken.
2009 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2010 tested/used via the -bh/-bhc/-bs options.
2012 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2013 bug, fixed in subsequent PCRE releases).
2015 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2016 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2018 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2024 SC/01 Use a glob alias rather than an array ref in eximstats generated
2025 parser. This improves both readability and performance.
2027 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2028 Don't display local sender or destination tables in eximstats unless
2029 there is data to show.
2030 Added average volumes into the eximstats top table text output.
2032 SC/03 Collect data on the number of addresses (recipients) as well
2033 as the number of messages in eximstats.
2035 TF/01 Correct an error in the documentation for the redirect router. Exim
2036 does (usually) call initgroups() when daemonizing.
2038 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2039 with consistent privilege compared to when running as a daemon.
2041 TF/03 Note in the spec that $authenticated_id is not set for local
2042 submissions from trusted users.
2044 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2045 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2047 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2048 by adding some example configuration directives to the default
2049 configuration file. A little bit of work is required to uncomment the
2050 directives and define how usernames and passwords are checked, but
2051 there is now a framework to start from.
2053 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2054 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2055 without this. I don't know how relevant this is to other LDAP libraries.
2057 PH/02 Add the verb name to the "unknown ACL verb" error.
2059 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2061 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2063 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2064 directory not expanded when it should be if an expanded home directory
2065 was set for the address (which is overridden by the transport).
2067 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2070 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2071 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2072 because it is too late at that time, and has no effect.
2074 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2075 security issue with \' (bugzilla #107). I could not use the
2076 PQescapeStringConn() function, because it needs a PGconn value as one of
2079 PH/08 When testing addresses using -bt, indicate those final addresses that
2080 are duplicates that would not cause an additional delivery. At least one
2081 person was confused, thinking that -bt output corresponded to deliveries.
2082 (Suppressing duplicates isn't a good idea as you lose the information
2083 about possibly different redirections that led to the duplicates.)
2085 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2086 systems where poll() doesn't work, in particular OS X.
2088 PH/10 Added more information to debugging output for retry time not reached.
2090 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2091 operations in malware.c.
2093 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2096 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2097 syslog_duplication set false, log lines that would normally be written
2098 both the the main log and to the reject log were not written to syslog at
2101 PH/14 In the default configuration, change the use of "message" in ACL warn
2102 statements to "add_header".
2104 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2105 not followed by a command (e.g. "seen endif").
2107 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2108 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2111 PH/17 Added extra conditions to the default value of delay_warning_condition
2115 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2116 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2117 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2120 The Auto-Submitted: and various List- headers are standardised, whereas I
2121 don't think Precedence: ever was.
2123 PH/18 Refactored debugging code in route_finduser() to show more information,
2124 in particular, the error code if getpwnam() issues one.
2126 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2127 This is apparently needed in addition to the PH/07 change above to avoid
2128 any possible encoding problems.
2130 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2131 but not after initializing Perl.
2133 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2134 output them only if debugging. By default they are written stderr,
2135 apparently, which is not desirable.
2137 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2140 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2143 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2145 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2146 authenticated or an ident call has been made. Suppress the default
2147 values for $authenticated_id and $authenticated_sender (but permit -oMai
2148 and -oMas) when testing with -bh.
2150 PH/24 Re-jigged the order of the tests in the default configuration so that the
2151 tests for valid domains and recipients precede the DNS black list and CSA
2152 tests, on the grounds that those ones are more expensive.
2154 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2155 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2156 This bug exists in every version of Exim that I still have, right back to
2159 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2160 However, an attempt to turn on case-sensitivity in a regex key by
2161 including (?-i) didn't work because the subject string was already
2162 lowercased, and the effects were non-intuitive. It turns out that a
2163 one-line patch can be used to allow (?-i) to work as expected.
2169 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2170 other effects) broke the use of negated acl sub-conditions.
2172 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2175 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2176 "Deny" causes Exim to reject the incoming connection with a 554 error.
2177 Unfortunately, if there is a major crisis, such as a disk failure,
2178 tcp-wrappers gives "deny", whereas what one would like would be some
2179 kind of temporary error. A kludge has been added to help with this.
2180 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2181 554 error is used if errno is still zero or contains ENOENT (which occurs
2182 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2185 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2187 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2188 errors. Otherwise a message that provokes a temporary error (when other
2189 messages do not) can cause a whole host to time out.
2191 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2192 addresses were routed directly to files or pipes from a redirect router.
2193 File deliveries just didn't batch; pipe deliveries might have suffered
2196 PH/06 A failure to get a lock for a hints database would erroneously always say
2197 "Failed to get write lock", even when it was really a read lock.
2199 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2200 of 0600. This has been changed to use the value of the lockfile_mode
2201 option (which defaults to 0600).
2203 PH/08 Applied small patch from the Sieve maintainer.
2205 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2206 folder from quota calculations, a direct delivery into this folder messed
2207 up the contents of the maildirsize file. This was because the regex was
2208 used only to exclude .Trash (or whatever) when the size of the mailbox
2209 was calculated. There was no check that a delivery was happening into an
2210 excluded directory. This bug has been fixed by ignoring all quota
2211 processing for deliveries into excluded directories.
2213 PH/10 Added the maildirfolder_create_regex option to appendfile.
2219 PH/01 The code for finding all the local interface addresses on a FreeBSD
2220 system running IPv6 was broken. This may well have applied to all BSD
2221 systems, as well as to others that have similar system calls. The broken
2222 code found IPv4 interfaces correctly, but gave incorrect values for the
2223 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2224 that it would not match correctly against @[] and not recognize the IPv6
2227 PH/02 The ipliteral router was not recognizing addresses of the form user@
2228 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2230 PH/03 Added disable_ipv6.
2232 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2233 decoded form, because it is most often used to construct To: headers
2234 lines in autoreplies, and the decoded form may well be syntactically
2235 invalid. However, $reply_address has leading white space removed, and all
2236 newlines turned into spaces so that the autoreply transport does not
2239 PH/05 If group was specified without a user on a router, and no group or user
2240 was specified on a transport, the group from the router was ignored.
2242 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2243 for visible compile-time settings that can be used to change these
2244 numbers, for those that want even more. Backwards compatibility with old
2245 spool files has been maintained. However, going back to a previous Exim
2246 release will lost any variables that are in spool files.
2248 PH/07 Two small changes when running in the test harness: increase delay when
2249 passing a TCP/IP connection to a new process, in case the original
2250 process has to generate a bounce, and remove special handling of
2251 127.0.0.2 (sic), which is no longer necessary.
2253 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2254 be the same on different OS.
2256 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2259 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2260 whether --show-vars was specified or not
2262 JJ/02 exipick: Added support for new ACL variable spool format introduced
2265 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2266 syntactically invalid From: or Reply-to: line, and a filter used this to
2267 generate an autoreply, and therefore failed to obtain an address for the
2268 autoreply, Exim could try to deliver to a non-existent relative file
2269 name, causing unrelated and misleading errors. What now happens is that
2270 it logs this as a hard delivery error, but does not attempt to create a
2273 PH/11 The exinext utility has a -C option for testing purposes, but although
2274 the given file was scanned by exinext itself; it wasn't being passed on
2275 when Exim was called.
2277 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2278 an end-of-file indication when reading a command response.
2280 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2281 compiled. In many other places in Exim, IPv6 addresses are always
2282 recognized, so I have changed this. It also means that IPv4 domain
2283 literals of the form [IPV4:n.n.n.n] are now always recognized.
2285 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2286 used if the router is not running as root, for example, when verifying at
2287 ACL time, or when using -bh. The debugging output from this situation was
2288 non-existent - all you got was a failure to exec. I have made two
2291 (a) Failures to set uid/gid, the current directory, or a process leader
2292 in a subprocess such as that created by queryprogram now generate
2293 suitable debugging ouput when -d is set.
2295 (b) The queryprogram router detects when it is not running as root,
2296 outputs suitable debugging information if -d is set, and then runs
2297 the subprocess without attempting to change uid/gid.
2299 PH/15 Minor change to Makefile for building test_host (undocumented testing
2302 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2303 additional section of a DNS packet that returns MX or SRV records.
2304 Instead, it always explicitly searches for A/AAAA records. This avoids
2305 major problems that occur when a DNS server includes only records of one
2306 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2307 fixed another bug: if SRV records were looked up and the corresponding
2308 address records were *not* found in the additional section, the port
2309 values from the SRV records were lost.
2311 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2312 using the correct key (the original address) when searching the retry
2313 rules in order to find which one to use for generating the retry hint.
2315 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2316 adding the default one. Similarly, if it contains a Reply-To: header, the
2317 errors_reply_to option, if set, is not used.
2319 PH/19 When calculating a retry time, Exim used to measure the "time since
2320 failure" by looking at the "first failed" field in the retry record. Now
2321 it does not use this if it is later than than the arrival time of the
2322 message. Instead it uses the arrival time. This makes for better
2323 behaviour in cases where some deliveries succeed, thus re-setting the
2324 "first failed" field. An example is a quota failure for a huge message
2325 when small messages continue to be delivered. Without this change, the
2326 "time since failure" will always be short, possible causing more frequent
2327 delivery attempts for the huge message than are intended.
2328 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2330 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2331 $1, $2, $3) because the numerical variables can be reset during some
2332 expansion items (e.g. "match"), thereby losing the authentication data.
2334 PH/21 Make -bV show the size of off_t variables so that the test suite can
2335 decide whether to run tests for quotas > 2G.
2337 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2338 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2339 filecount value is greater than 2G or if a quota value is greater than 2G
2340 on a system where the size of off_t is not greater than 4, a panic error
2343 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2344 never match. The debug and -bh output now contains an explicit error
2345 message indicating a malformed IPv4 address or mask.
2347 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2348 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2349 PH/23 above applies.
2351 PH/25 Do not write to syslog when running in the test harness. The only
2352 occasion when this arises is a failure to open the main or panic logs
2353 (for which there is an explicit test).
2355 PH/26 Added the /no_tell option to "control=freeze".
2357 PH/27 If a host name lookup failed very early in a connection, for example, if
2358 the IP address matched host_lookup and the reverse lookup yielded a name
2359 that did not have a forward lookup, an error message of the form "no IP
2360 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2361 could be logged. Now it outputs the IP address instead of "NULL".
2363 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2364 allows the sender and the authenticated sender to be set when
2365 submitting a message from within Exim. Since child_open_exim() is
2366 documented for local_scan(), the new function should be too.
2368 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2369 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2370 results in an empty string is now treated as unset.
2372 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2374 PH/31 Added sender_verify_fail logging option.
2376 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2377 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2378 tidied the source and removed it altogether.
2380 PH/33 When a queue run was abandoned because the load average was too high, a
2381 log line was always written; now it is written only if the queue_run log
2382 selector is set. In addition, the log line for abandonment now contains
2383 information about the queue run such as the pid. This is always present
2384 in "start" and "stop" lines but was omitted from the "abandon" line.
2386 PH/34 Omit spaces between a header name and the colon in the error message that
2387 is given when verify = headers_syntax fails (if there are lots of them,
2388 the message gets confusing).
2390 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2391 names, as there are now some PTR records that contain slashes. This check
2392 is only to protect against broken name servers that fall over on strange
2393 characters, so the fact that it applies to all lookups doesn't matter.
2395 PH/36 Now that the new test suite is complete, we can remove some of the
2396 special code in Exim that was needed for the old test suite. For example,
2397 sorting DNS records because real resolvers return them in an arbitrary
2398 order. The new test suite's fake resolver always returns records in the
2401 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2402 bounces) except when queue_only is set, to avoid logging races between
2403 the different processes.
2405 PH/38 Panic-die if .include specifies a non-absolute path.
2407 PH/39 A tweak to the "H" retry rule from its user.
2409 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2410 a label. They prevented compilation on older perls.
2412 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2413 a warning to be raised on newish perls.
2415 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2416 on queue. Changes to match documented behaviour of showing count of
2417 messages matching specified criteria.
2419 PH/40 Changed the default ident timeout from 30s to 5s.
2421 PH/41 Added support for the use of login_cap features, on those BSD systems
2422 that have them, for controlling the resources used by pipe deliveries.
2424 PH/42 The content-scanning code uses fopen() to create files in which to put
2425 message data. Previously it was not paying any attention to the mode of
2426 the files. Exim runs with umask(0) because the rest of the code creates
2427 files with open(), and sets the required mode explicitly. Thus, these
2428 files were ending up world-writeable. This was not a big issue, because,
2429 being within the spool directory, they were not world-accessible. I have
2430 created a function called modefopen, which takes an additional mode
2431 argument. It sets umask(777), creates the file, chmods it to the required
2432 mode, then resets the umask. All the relevant calls to fopen() in the
2433 content scanning code have been changed to use this function.
2435 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2436 to 24 hours. This avoids potential overflow problems when processing G
2437 and H retry rules. I suspect nobody ever tinkers with this value.
2439 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2441 PH/45 When the plaintext authenticator is running as a client, the server's
2442 challenges are checked to ensure they are valid base64 strings. By
2443 default, the authentication attempt is cancelled if an invalid string is
2444 received. Setting client_ignore_invalid_base64 true ignores these errors.
2445 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2446 they are received. Thus, the responses can be made to depend on the
2447 challenges. If an invalid string is ignored, an empty string is placed in
2450 PH/46 Messages that are created by the autoreply transport now contains a
2451 References: header, in accordance with RFCs 2822 and 3834.
2453 PH/47 Added authenticated_sender_force to the smtp transport.
2455 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2457 PH/49 Installed latest patch from the Sieve maintainer.
2459 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2460 was also set, the appendfile transport was unnecessarily scanning a
2461 directory of message files (e.g. for maildir delivery) to find the count
2462 of files (along with the size), even though it did not need this
2463 information. It now does the scan only if it needs to find either the
2464 size of the count of files.
2466 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2468 PH/52 Two bugs concerned with error handling when the smtp transport is
2471 (i) Exim was not creating retry information for temporary errors given
2472 for individual recipients after the DATA command when the smtp transport
2473 was used in LMTP mode. This meant that they could be retried too
2474 frequently, and not timed out correctly.
2476 (ii) Exim was setting the flag that allows error details to be returned
2477 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2478 recipients that were returned after the DATA command.
2480 PH/53 This is related to PH/52, but is more general: for any failing address,
2481 when detailed error information was permitted to be returned to the
2482 sender, but the error was temporary, then after the final timeout, only
2483 "retry timeout exceeded" was returned. Now it returns the full error as
2484 well as "retry timeout exceeded".
2486 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2487 do this, and (what is worse) MTAs that accept it.
2489 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2490 will now be deprecated.
2492 PH/56 New os.c-cygwin from the Cygwin maintainer.
2494 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2495 formats (previously only available in exim formats via -bpr, -bpru,
2496 and -bpra. Now also available in native and exiqgrep formats)
2498 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2499 with very large, slow to parse queues
2501 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2503 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2505 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2506 responses to authentication challenges, though it was showing the
2507 challenges; (ii) I've removed the CR characters from the debug output for
2510 PH/58 Allow for the insertion of a newline as well as a space when a string
2511 is turned into more than one encoded-word during RFC 2047 encoding. The
2512 Sieve code now uses this.
2514 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2515 data_4xx, lost_connection, tls_required.
2517 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2518 message was being sent as an SMTP response.
2520 PH/61 Add -l and -k options to exicyclog.
2522 PH/62 When verifying, if an address was redirected to one new address, so that
2523 verification continued, and the new address failed or deferred after
2524 having set something in $address_data, the value of $address_data was not
2525 passed back to the ACL. This was different to the case when no
2526 redirection occurred. The value is now passed back in both cases.
2528 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2529 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2530 use login_cap.h, so on its own it isn't the distinguishing feature. The
2531 new name refers directly to the setclassresources() function.
2533 PH/65 Added configuration files for NetBSD3.
2535 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2537 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2538 is preferred over IPv4.
2540 PH/68 The bounce_return_message and bounce_return_body options were not being
2541 honoured for bounces generated during the reception of non-SMTP messages.
2542 In particular, this applied to messages rejected by the ACL. This bug has
2543 been fixed. However, if bounce_return_message is true and bounce_return_
2544 body is false, the headers that are returned for a non-SMTP message
2545 include only those that have been read before the error was detected.
2546 (In the case of an ACL rejection, they have all been read.)
2548 PH/69 The HTML version of the specification is now built in a directory called
2549 spec_html instead of spec.html, because the latter looks like a path with
2550 a MIME-type, and this confuses some software.
2552 PH/70 Catch two compiler warnings in sieve.c.
2554 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2555 function verify_get_ident() calls ip_connect() to connect a socket, but
2556 if the "connect()" function timed out, ip_connect() used to close the
2557 socket. However, verify_get_ident() also closes the socket later, and in
2558 between Exim writes to the log, which may get opened at this point. When
2559 the socket was closed in ip_connect(), the log could get the same file
2560 descriptor number as the socket. This naturally causes chaos. The fix is
2561 not to close the socket in ip_connect(); the socket should be closed by
2562 the function that creates it. There was only one place in the code where
2563 this was missing, in the iplookup router, which I don't think anybody now
2564 uses, but I've fixed it anyway.
2566 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2567 well as to direct DNS lookups. Otherwise the handling of names in host
2568 lists is inconsistent and therefore confusing.
2574 PH/01 Two changes to the default runtime configuration:
2576 (1) Move the checks for relay_from_hosts and authenticated clients from
2577 after to before the (commented out) DNS black list checks.
2579 (2) Add control=submission to the relay_from_hosts and authenticated
2580 clients checks, on the grounds that messages accepted by these
2581 statements are most likely to be submissions.
2583 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2585 (1) Generate an error if the third argument for the ${prvs expansion is
2588 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2591 (3) Reset the variables that are obtained from the first argument of
2592 ${prvscheck and used in the second argument before leaving the code,
2593 because their memory is reclaimed, so using them afterwards may do
2596 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2597 one (it's much easier than Tom thought :-).
2599 (5) Because of (4), we can now allow for the use of $prvscheck_result
2600 inside the third argument.
2602 PH/03 For some reason, the default setting of PATH when running a command from
2603 a pipe transport was just "/usr/bin". I have changed it to
2606 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2607 anything to be listed in the output from -bV.
2609 PH/05 When a filter generated an autoreply, the entire To: header line was
2610 quoted in the delivery log line, like this:
2612 => >A.N.Other <ano@some.domain> <original@ddress> ...
2614 This has been changed so that it extracts the operative address. There
2615 may be more than one such address. If so, they are comma-separated, like
2618 => >ano@some.domain,ona@other.domain <original@ddress> ...
2620 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2621 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2622 not being looked up in the rDNS to get a host name, Exim was showing the
2623 IP address twice in Received: lines, even though the IP addresses were
2624 identical. For example:
2626 Received: from [1.2.3.4] (helo=[1.2.3.4])
2628 However, if the real host name was known, it was omitting the HELO data
2629 if it matched the actual IP address. This has been tidied up so that it
2630 doesn't show the same IP address twice.
2632 PH/07 When both +timestamp and +memory debugging was on, the value given by
2633 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2634 called by the debug printing, thereby overwriting the timestamp buffer.
2635 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2637 PH/08 When the original message was included in an autoreply transport, it
2638 always said "this is a copy of the message, including all the headers",
2639 even if body_only or headers_only was set. It now gives an appropriate
2642 PH/09 Applied a patch from the Sieve maintainer which:
2644 o fixes some comments
2645 o adds the (disabled) notify extension core
2646 o adds some debug output for the result of if/elsif tests
2647 o points to the current vacation draft in the documentation
2648 and documents the missing references header update
2652 o fixes a bug in processing the envelope test (when testing
2653 multiple envelope elements, the last element determinted the
2656 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2657 Electronic Mail") by including:
2659 Auto-submitted: auto-generated
2661 in the messages that it generates (bounce messages and others, such as
2662 warnings). In the case of bounce messages for non-SMTP mesages, there was
2663 also a typo: it was using "Auto_submitted" (underscore instead of
2664 hyphen). Since every message generated by Exim is necessarily in response
2665 to another message, thes have all been changed to:
2667 Auto-Submitted: auto-replied
2669 in accordance with these statements in the RFC:
2671 The auto-replied keyword:
2673 - SHOULD be used on messages sent in direct response to another
2674 message by an automatic process,
2676 - MUST NOT be used on manually-generated messages,
2678 - MAY be used on Delivery Status Notifications (DSNs) and Message
2679 Disposition Notifications (MDNs),
2681 - MUST NOT be used on messages generated by automatic or periodic
2682 processes, except for messages which are automatic responses to
2685 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2686 to the default Received: header definition.
2688 PH/12 Added log selector acl_warn_skipped (default on).
2690 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2691 variables because (a) they are in the wrong storage pool and (b) even if
2692 they were copied, it wouldn't work properly because of the caching.
2694 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2695 checking when decoding. Apparently there are clients that generate
2696 overlong encoded strings. Why am I not surprised?
2698 PH/15 If the first argument of "${if match_address" was not empty, but did not
2699 contain an "@" character, Exim crashed. Now it writes a panic log message
2700 and treats the condition as false.
2702 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2704 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2705 extension "envelope-auth". The code is finished and in agreement with
2706 other implementations, but there is no documentation so far and in fact,
2707 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2708 not changing the active code.
2710 Print executed "if" and "elsif" statements when debugging is used. This
2711 helps a great deal to understand what a filter does.
2713 Document more things not specified clearly in RFC3028. I had all this
2714 sorted out, when out of a sudden new issues came to my mind. Oops."
2716 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2719 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2720 canonical form (as documented). However, after a host name lookup from
2721 the IP address, check_host() was doing a simple string comparison with
2722 addresses acquired from the DNS when checking that the found name did
2723 have the original IP as one of its addresses. Since any found IPv6
2724 addresses are likely to be in abbreviated form, the comparison could
2725 fail. Luckily, there already exists a function for doing the comparison
2726 by converting both addresses to binary, so now that is used instead of
2727 the text comparison.
2729 PH/20 There was another similar case to PH/19, when a complete host name was
2730 given in a host list; looking up its IP address could give an abbreviated
2731 form, whereas the current host's name might or might not be abbreviated.
2732 The same fix has been applied.
2738 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2739 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2742 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2743 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2745 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2747 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2748 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2749 Also, for "Auto-Submitted", treat anything other than "no" as
2750 non-personal, in accordance with RFC 3834. (Previously it treated
2751 anything starting "auto-" as non-personal.)
2753 TF/01 The control=submission/name=... option had a problem with syntax
2754 errors if the name included a slash character. The /name= option
2755 now slurps the rest of the string, so it can include any characters
2756 but it must come last in the list of options (after /sender_retain
2759 PH/05 Some modifications to the interface to the fake nameserver for the new
2767 TK/01 Added the "success_on_redirect" address verification option. See
2768 NewStuff for rationale and an example.
2770 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2772 PH/02 Patch to exigrep to allow it to work on syslog lines.
2774 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2775 fread() to skip over the body file's header line, because in Cygwin the
2776 header line is locked and is inaccessible.
2778 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2779 co-exist for some time) to make it clear that it is the Exim ID that is
2780 referenced, not the Message-ID: header line.
2782 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2783 string_format() function, because snprintf() does not exist on all
2786 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2787 ${stat: expansion item.
2789 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2790 protocol synchronization error", to keep the pedants happy.
2792 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2793 well as for IRIX systems, when gcc is being used. See the host.c source
2796 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2798 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2801 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2802 transport and to the smtp transport in LMTP mode.
2804 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2806 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2807 run a filter in a subprocess. This could lead to confusion in subsequent
2808 lookups in the parent process. There should also be a search_tidyup() at
2809 the end of the subprocess.
2811 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2812 only if the host matched helo_try_verify_hosts, which caused the
2813 verification to occur when the EHLO/HELO command was issued. The ACL just
2814 tested the remembered result. Now, if a previous verification attempt has
2815 not happened, "verify = helo" does it there and then.
2817 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2819 TK/03 Fix log output including CR from clamd.
2821 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2822 exist provoked a memory error which could cause a segfault.
2824 PH/15 Installed PCRE 6.2
2826 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2828 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2829 of the problem. Specifically, suggested +O2 rather than +O1 for the
2832 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2834 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2835 record, the port was not correctly logged when the outgoing_port log
2836 selector was set (it logged the transort's default port).
2838 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2839 fallback_hosts, and "hosts" in the smtp transport.
2841 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2842 host errors such as "Connection refused".
2844 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2845 authentication with radiusclient 0.4.9:
2847 - Error returned from rc_read_config was caught wrongly
2848 - Username/password not passed on to radius server due to wrong length.
2850 The presumption is that some radiusclient API changes for 4.51/PH/17
2851 were not taken care of correctly. The code is still untested by me (my
2852 Linux distribution still has 0.3.2 of radiusclient), but it was
2853 contributed by a Radius user.
2855 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2856 expanding the "port" option of the smtp transport.
2858 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2859 while reading a MIME header. Thanks to Tom Hughes for a patch.
2861 PH/24 Include config.h inside local_scan.h so that configuration settings are
2864 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2865 that in an ACL for RCPT (for example), you can examine exactly what was
2868 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2869 commands, but it was not correctly comparing the address with the actual
2870 client host address. Thus, it would show the EHLO address in Received:
2871 header lines when this was not necessary.
2873 PH/27 Added the % operator to ${eval:}.
2875 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2876 it should be ignoring failures (because with -C, for example, it has lost
2877 privilege). It wasn't ignoring creation failures other than "already
2880 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2883 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2884 because an input line was too long, either on its own, or by virtue of
2885 too many continuations, the temporary file was not being removed, and the
2886 return code was incorrect.
2888 PH/31 Missing "BOOL" in function definition in filtertest.c.
2890 PH/32 Applied Sieve patches from the maintainer.
2892 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2894 PH/33 Added "verify = not_blind".
2896 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2897 Local/Makefile (with some defaults set). These are used in built scripts
2898 such as exicyclog, but they have never been used in the exim_install
2899 script (though there are many overriding facilities there). I have
2900 arranged that the exim_install script now takes note of these two
2903 PH/35 Installed configuration files for Dragonfly.
2905 PH/36 When a locally submitted message by a trusted user did not contain a
2906 From: header, and the sender address was obtained from -f or from an SMTP
2907 MAIL command, and the trusted user did not use -F to supply a sender
2908 name, $originator_name was incorrectly used when constructing a From:
2909 header. Furthermore, $originator_name was used for submission mode
2910 messages from external hosts without From: headers in a similar way,
2911 which is clearly wrong.
2913 PH/37 Added control=suppress_local_fixups.
2915 PH/38 When log_selector = +received_sender was set, and the addition of the
2916 sender made the log line's construction buffer exactly full, or one byte
2917 less than full, an overflow happened when the terminating "\n" was
2920 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2921 when the result of a list match is failure because a DNS lookup failed.
2923 PH/40 RM_COMMAND is now used in the building process.
2925 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2926 the "build-* directories that it finds.
2928 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2929 address in a domain literal was a prefix of an interface address.
2931 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2932 when verifying a sender address, unless rewrite_headers is false.
2934 PH/44 Wrote a long comment about why errors_to addresses are verified as
2935 recipients, not senders.
2937 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2938 the ratelimit ACL was added.
2940 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2942 PH/46 Added extra information about PostgreSQL errors to the error string.
2944 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2945 avoiding the need to install special zones in a real server. This is
2946 backwards compatible; if it can't find the fake resolver, it drops back.
2947 Thus, both old and new test suites can be run.
2949 TF/02 Added util/ratelimit.pl
2951 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2952 clock is set back in time.
2954 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2955 Candler <B.Candler@pobox.com>.
2957 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2958 OK for addresses that are the result of redirections.
2960 PH/48 A number of further additions for the benefit of the new test suite,
2961 including a fake gethostbyname() that interfaces to the fake DNS resolver
2964 TF/06 The fix for widen_domains has also been applied to qualify_single and
2965 search_parents which are the other dnslookup options that can cause
2968 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2971 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2973 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2974 service" if it encountered an MX record with an empty target hostname.
2975 The message is now "an MX or SRV record indicated no SMTP service".
2977 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2978 if the DNS of the sending site is misconfigured. This is quite a
2979 common situation. This change restores the behaviour of treating a
2980 helo verification defer as a failure.
2982 PH/51 If self=fail was set on a router, the bounce message did not include the
2983 actual error message.
2989 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2991 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2992 command itself ended in error, the underlying message about the transport
2993 filter timeout was being overwritten with the pipe command error. Now the
2994 underlying error message should be appended to the second error message.
2996 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2998 PH/02 Reduce the amount of output that "make" produces by default. Full output
2999 can still be requested.
3001 PH/03 The warning log line about a condition test deferring for a "warn" verb
3002 was being output only once per connection, rather than after each
3003 occurrence (because it was using the same function as for successful
3004 "warn" verbs). This seems wrong, so I have changed it.
3006 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3007 it should not have, which might have caused a crash in the right
3008 circumstances, but probably never did.
3010 PH/04 Installed a modified version of Tony Finch's patch to make submission
3011 mode fix the return path as well as the Sender: header line, and to
3012 add a /name= option so that you can make the user's friendly name appear
3015 TF/03 Added the control = fakedefer ACL modifier.
3017 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3018 Mark Lowes for thorough testing.
3020 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3022 TK/03 Merged latest SRS patch from Miles Wilton.
3024 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3025 with the definition in sysexits.h (which is #included earlier).
3026 Fortunately, Exim does not actually use EX_OK. The code used to try to
3027 preserve the sysexits.h value, by assumimg that macro definitions were
3028 scanned for macro replacements. I have been disabused of this notion,
3029 so now the code just undefines EX_OK before #including unistd.h.
3031 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3032 in the smtp transport. When a block could not be written in a single
3033 write() function, the timeout was being re-applied to each part-write.
3034 This seems wrong - if the receiver was accepting one byte at a time it
3035 would take for ever. The timeout is now adjusted when this happens. It
3036 doesn't have to be particularly precise.
3038 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3039 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3041 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3042 without a domain if the check to <postmaster@domain> fails.
3044 SC/01 Eximstats: added -xls and the ability to specify output files
3045 (patch written by Frank Heydlauf).
3047 SC/02 Eximstats: use FileHandles for outputing results.
3049 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3051 SC/04 Eximstats: fixed display of large numbers with -nvr option
3053 SC/05 Eximstats: fixed merging of reports with empty tables.
3055 SC/06 Eximstats: added the -include_original_destination flag
3057 SC/07 Eximstats: removed tabs and trailing whitespace.
3059 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3061 TK/06 MBOX spool code: Add real "From " MBOX separator line
3062 so the .eml file is really in mbox format (even though
3063 most programs do not really care). Patch from Alex Miller.
3065 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3066 The latter is generated from $received_to and is only set if the
3067 message has one envelope recipient. SA can use these headers,
3068 obviously out-of-the-box. Patch from Alex Miller.
3070 PH/08 The ${def test on a variable was returning false if the variable's
3071 value was "0", contrary to what the specification has always said!
3072 The result should be true unless the variable is empty.
3074 PH/09 The syntax error of a character other than { following "${if
3075 def:variable_name" (after optional whitespace) was not being diagnosed.
3076 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3077 accidental colon was present, for example, could give incorrect results.
3079 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3080 result is used (not including appendfile, where other changes are about
3083 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3084 This involved changing a lot of size variables from int to off_t. It
3085 should work with maildirs and everything.
3087 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3088 spamd dying while we are connected to it.
3090 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3093 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3094 function for BDB 4.3.
3096 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3098 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3099 the sources for PCRE have been re-arranged and more files are now
3102 PH/15 The code I had for printing potentially long long variables in PH/11
3103 above was not the best (it lost precision). The length of off_t variables
3104 is now inspected at build time, and an appropriate printing format (%ld
3105 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3106 to be "long long int" or "long int". This is needed for the internal
3107 formatting function string_vformat().
3109 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3110 the configuration file to be ":syslog", then the script "guesses" where
3111 the logs files are, rather than using the compiled in default. In our
3112 case the guess is not the same as the compiled default, so the script
3113 suddenly stopped working when I started to use syslog. The patch checks
3114 to see if log_file_path is "". If so, it attempts to read it from exim
3115 with no configuration file to get the compiled in version, before it
3116 falls back to the previous guessing code."
3118 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3119 implementing BATV in an Exim configuration. See NewStuff for the gory
3122 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3123 Makefile that are specific to HP-UX.
3125 PH/18 If the "use_postmaster" option was set for a recipient callout together
3126 with the "random" option, the postmaster address was used as the MAIL
3127 FROM address for the random test, but not for the subsequent recipient
3128 test. It is now used for both.
3130 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3131 patch removes a few documentation additions to RFC 3028, because the
3132 latest draft now contains them. It adds the new en;ascii-case comparator
3133 and a new error check for 8bit text in MIME parts. Comparator and
3134 require names are now matched exactly. I enabled the subaddress
3135 extension, but it is not well tested yet (read: it works for me)."
3137 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3138 rework some of the code of TK/09 above to avoid the hardwired use of
3139 "%lld" and "long long". Replaced the call to snprintf() with a call to
3142 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3143 records point to non-existent hosts", "retry timeout exceeded", and
3144 "retry time not reached for any host after a long failure period".
3146 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3147 experimental DomainKeys support:
3149 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3150 (2) On an error such as an illegally used "control", the wrong name for
3151 the control was given.
3153 These problems did NOT occur unless DomainKeys support was compiled.
3155 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3157 PH/24 Added ${if match_ip condition.
3159 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3160 databases so that it will be absolutely obvious if a crash occurs in the
3161 DB library. This is a regular occurrence (often caused by mis-matched
3164 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3165 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3166 picked up on a user's system that detects such things. There doesn't seem
3167 to be a gcc warning option for this - only an attribute that has to be
3168 put on the function's prototype. It seems that in Fedora Core 4 they have
3169 set this on a number of new functions. No doubt there will be more in due
3172 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3173 specify a transport. However, if an address that was verified by such a
3174 router was the subject of a callout, Exim crashed because it tried to
3175 read the rcpt_include_affixes from the non-existent transport. Now it
3176 just assumes that the setting of that option is false. This bug was
3177 introduced by 4.51/PH/31.
3179 PH/28 Changed -d+all to exclude +memory, because that information is very
3180 rarely of interest, but it makes the output a lot bigger. People tend to
3181 do -d+all out of habit.
3183 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3184 code in os-type was giving problems when libc.so lives in lib64, like on
3187 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3188 aren't the modern standard, and it seems that some systems' include files
3189 don't always have them. Exim was already checking for some of the newer
3190 ones like T_AAAA, and defining it itself. I've added checks for all the
3191 record types that Exim uses.
3193 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3194 not automatically generating a new one, as it is supposed to. This
3195 prevented TLS from working. If the file did exist, but contained invalid
3196 data, a new version was generated, as expected. It was only the case of a
3197 non-existent file that was broken.
3199 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3200 with a change in libdomainkeys > 0.64.
3202 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3203 from DNS. If the selector record carries the flag, it now has
3204 precedence over the domain-wide flag.
3206 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3208 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3209 the use of a port name that isn't defined in /etc/services) occurred, the
3210 message was deferred as in a normal delivery, and thus remained on the
3211 spool, instead of being failed because of the mua_wrapper setting. This
3212 is now fixed, and I tidied up some of the mua_wrapper messages at the
3215 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3216 the messages in a hash of arrays rather than using individual hashes.
3217 This is a bit cleaner and results in dramatic memory savings, albeit
3218 at a slight CPU cost.
3220 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3221 as requested by Marc Sherman.
3223 SC/10 Eximstats: added histograms for user specified patterns as requested
3226 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3228 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3229 fopen() in the content-scanning modules that did not already have it.
3235 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3236 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3238 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3240 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3242 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3243 to test Sieve filters that use "vacation".
3245 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3246 that changes the way the GnuTLS parameters are stored in the cache file.
3247 The new format can be generated externally. For backward compatibility,
3248 if the data in the cache doesn't make sense, Exim assumes it has read an
3249 old-format file, and it generates new data and writes a new file. This
3250 means that you can't go back to an older release without removing the
3253 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3254 work if there are any delivery delays because "one_time" forces the
3255 parent to be marked "delivered", so its unseen clone is never tried
3256 again. For this reason, Exim now forbids the simultaneous setting of
3259 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3260 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3261 entry if you want to know the details. The fix, however, neglected to
3262 consider the case when local delivery batching is involved. The test for
3263 "previously delivered" was not happening when checking to see if an
3264 address could be batched with a previous (undelivered) one; under
3265 certain circumstances this could lead to multiple deliveries to the same
3268 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3269 in its include files, and this causes problems building Exim.
3271 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3272 header_syntax) but Exim was just ignoring anything given after a slash.
3273 In particular, this caused confusion with an attempt to use "verify =
3274 reverse_host_lookup/defer_ok". An error is now given when options are
3275 supplied for verify items that do not have them. (Maybe reverse_host_
3276 lookup should have a defer_ok option, but that's a different point.)
3278 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3279 defined by RFC 821) to 2048, because there were problems with some AUTH
3280 commands, and RFC 1869 says the size should be increased for extended
3281 SMTP commands that take arguments.
3283 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3286 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3287 "unknown" error; now it says that the functionality isn't in the binary.
3289 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3290 an address' error message when a string expansion fails (syntax or
3291 whatever). Otherwise the password may appear in the log. Following change
3292 PH/42 below, there is no longer a chance of it appearing in a bounce
3295 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3297 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3298 of its addresses was ever tried. (Bugzilla bug #2.)
3300 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3301 the result incorrectly in the debug output. (It correctly added a newline
3302 to what was transported.)
3304 TF/01 Added $received_time.
3306 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3307 commented out examples of how to interface to a virus scanner and to
3308 SpamAssassin. Also added commented examples of av_scanner and
3309 spamd_address settings.
3311 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3312 and controls are allowed in which ACLs. There were a couple of minor
3313 errors. Some of the entries in the conditions table (which is a table of
3314 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3315 negation of where the condition IS allowed.
3317 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3319 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3320 header file does not have a version number, so I've had to invent a new
3321 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3322 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3323 radiusclient), but it was contributed by a Radius user.
3325 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3326 files or directories, for OpenSSL.
3328 PH/19 When an Exim process that is running as root has to create an Exim log
3329 file, it does so in a subprocess that runs as exim:exim so as to get the
3330 ownership right at creation (otherwise, other Exim processes might see
3331 the file with the wrong ownership). There was no test for failure of this
3332 fork() call, which would lead to the process getting stuck as it waited
3333 for a non-existent subprocess. Forks do occasionally fail when resources
3334 run out. I reviewed all the other calls to fork(); they all seem to check
3337 PH/20 When checking for unexpected SMTP input at connect time (before writing
3338 the banner), Exim was not dealing correctly with a non-positive return
3339 from the read() function. If the client had disconnected by this time,
3340 the result was a log entry for a synchronization error with an empty
3341 string after "input=" when read() returned zero. If read() returned -1
3342 (an event I could not check), uninitialized data bytes were printed.
3343 There were reports of junk text (parts of files, etc) appearing after
3346 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3348 PH/22 Added support for macro redefinition, and (re)definition in between
3349 driver and ACL definitions.
3351 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3352 forgetting to use the resulting value; it was using the unexpanded value.
3354 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3355 hadn't been configured. The fix is from Juergen Kreileder, who
3356 understands it better than I do:
3358 "Here's what I see happening with three configured cyrus_sasl
3359 authenticators configured (plain, login, cram-md5):
3361 On startup auth_cyrus_sasl_init() gets called for each of these.
3362 This means three calls to sasl_listmech() without a specified mech_list.
3363 => SASL tests which mechs of all available mechs actually work
3364 => three warnings about OTP not working
3365 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3367 With the patch, sasl_listmech() also gets called three times. But now
3368 SASL's mech_list option is set to the server_mech specified in the the
3369 authenticator. Or in other words, the answer from sasl_listmech()
3370 gets limited to just the mech you're testing for (which is different
3372 => the return list contains just 'plain' or 'login', 'cram-md5' or
3373 nothing depending on the value of ob->server_mech.
3375 I've just tested the patch: Authentication still works fine,
3376 unavailable mechs specified in the exim configuration are still
3377 caught, and the auth.log warnings about OTP are gone."
3379 PH/25 When debugging is enabled, the contents of the command line are added
3380 to the debugging output, even when log_selector=+arguments is not
3383 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3384 answer is "GNU", and only if the return is "GNU/something" is the answer
3387 PH/27 $acl_verify_message is now set immediately after the failure of a
3388 verification in an ACL, and so is available in subsequent modifiers. In
3389 particular, the message can be preserved by coding like this:
3391 warn !verify = sender
3392 set acl_m0 = $acl_verify_message
3394 Previously, $acl_verify_message was set only while expanding "message"
3395 and "log_message" when a very denied access.
3397 PH/28 Modified OS/os.c-Linux with
3399 -#ifndef OS_LOAD_AVERAGE
3400 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3402 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3403 nomenclature these days.)
3405 PH/29 Installed patch from the Sieve maintainer that adds the options
3406 sieve_useraddress and sieve_subaddress to the redirect router.
3408 PH/30 In these circumstances:
3409 . Two addresses routed to the same list of hosts;
3410 . First host does not offer TLS;
3411 . First host accepts first address;
3412 . First host gives temporary error to second address;
3413 . Second host offers TLS and a TLS session is established;
3414 . Second host accepts second address.
3415 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3416 and peerdn, if requested) that were in fact used only for the second
3419 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3420 attention to any local part prefix or suffix that was matched by the
3421 router that accepted the address. It now behaves in the same way as it
3422 does for delivery: the affixes are removed from the local part unless
3423 rcpt_include_affixes is set on the transport.
3425 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3426 timeout during the DATA phase of an incoming message.
3428 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3429 applied a patch sanctioned by the Sieve maintainer.
3431 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3432 the uid or gid is negative. A case of a negative gid caused this to be
3433 noticed. The fix allows for either to be negative.
3435 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3436 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3439 PH/36 The change PH/12 above was broken. Fixed it.
3441 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3442 the grounds that routing the same address twice would always produce the
3443 same answer. This might have been true once, but it is certainly no
3444 longer true now. Routing a child address may depend on the previous
3445 routing that produced that child. Some complicated redirection strategies
3446 went wrong when messages had multiple recipients, and made Exim's
3447 behaviour dependent on the order in which the addresses were given.
3449 I have moved the duplicate checking until after the routing is complete.
3450 Exim scans the addresses that are assigned to local and remote
3451 transports, and removes any duplicates. This means that more work will be
3452 done, as duplicates will always all be routed, but duplicates are
3453 presumably rare, so I don't expect this is of any significance.
3455 For deliveries to pipes, files, and autoreplies, the duplicate checking
3456 still happens during the routing process, since they are not going to be
3459 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3460 It corrects a timeout issue with spamd. This is Ian's comment: "The
3461 background is that sometimes spamd either never reads data from a
3462 connection it has accepted, or it never writes response data. The exiscan
3463 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3464 blindly assumes that writes won't block so it may never time out."
3466 PH/39 Allow G after quota size as well as K and M.
3468 PH/40 The value set for $authenticated_id in an authenticator may not contain
3469 binary zeroes or newlines because the value is written to log lines and
3470 to spool files. There was no check on this. Now the value is run through
3471 the string_printing() function so that such characters are converted to
3472 printable escape sequences.
3474 PH/41 $message_linecount is a new variable that contains the total number of
3475 lines in the message. Compare $body_linecount, which is the count for the
3478 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3479 bounce and delay warning messages, except in certain special cases, which
3482 (a) An SMTP error message from a remote host;
3483 (b) A message specified in a :fail: redirection;
3484 (c) A message specified in a "fail" command in a system filter;
3485 (d) A message specified in a FAIL return from the queryprogram router;
3486 (e) A message specified by the cannot_route_message router option.
3488 In these cases only, Exim does include the error details in bounce and
3489 warning messages. There are also a few cases where bland messages such
3490 as "unrouteable address" or "local delivery error" are given.
3492 PH/43 $value is now also set for the "else" part of a ${run expansion.
3494 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3495 being worked on, but at least Exim now implements the latest version to
3498 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3499 process to complete was treated as a delivery failure, a timeout while
3500 writing the message to the pipe was logged, but erroneously treated as a
3501 successful delivery. Such timeouts include transport filter timeouts. For
3502 consistency with the overall process timeout, these timeouts are now
3503 treated as errors, giving rise to delivery failures by default. However,
3504 there is now a new Boolean option for the pipe transport called
3505 timeout_defer, which, if set TRUE, converts the failures into defers for
3506 both kinds of timeout. A transport filter timeout is now identified in
3509 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3510 systems where "make" and "gmake" are different, calling "gmake" at top
3511 level broke things. I've arranged for the value of $(MAKE) to be passed
3512 from the Makefile to this script so that it can call the same version of
3516 A note about Exim versions 4.44 and 4.50
3517 ----------------------------------------
3519 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3520 changes of various kinds. As a consequence, a big documentation update was
3521 needed. This delayed the release for rather longer than seemed good, especially
3522 in the light of a couple of (minor) security issues. Therefore, the changes
3523 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3524 release. So 4.44 and 4.50 are in effect two different branches that both start
3527 I have left the 4.50 change log unchanged; it contains all the changes since
3528 4.43. The change log for 4.44 is below; many of its items are identical to
3529 those for 4.50. This seems to be the most sensible way to preserve the
3530 historical information.
3536 1. Minor wording change to the doc/README.SIEVE file.
3538 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3539 computation of the current number of files was incorrect.
3541 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3542 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3545 4. Give more explanation in the error message when the command for a transport
3546 filter fails to execute.
3548 5. There are several places where Exim runs a non-Exim command in a
3549 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3550 was being done only for the command run by the queryprogram router. It is
3551 now done for all such subprocesses. The other cases are: ${run, transport
3552 filters, and the commands run by the lmtp and pipe transports.
3554 6. Added CONFIGURE_GROUP build-time option.
3556 7. Some older OS have a limit of 256 on the maximum number of file
3557 descriptors. Exim was using setrlimit() to set 1000 as a large value
3558 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3559 systems. I've change it so that if it can't get 1000, it tries for 256.
3561 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3562 was an oversight, and furthermore, ever since the addition of extra
3563 controls (e.g. 4.43/32), the checks on when to allow different forms of
3564 "control" were broken. There should now be diagnostics for all cases when a
3565 control that does not make sense is encountered.
3567 9. Added the /retain_sender option to "control=submission".
3569 10. $recipients is now available in the predata ACL (oversight).
3571 11. Tidy the search cache before the fork to do a delivery from a message
3572 received from the command line. Otherwise the child will trigger a lookup
3573 failure and thereby defer the delivery if it tries to use (for example) a
3574 cached ldap connection that the parent has called unbind on.
3576 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3577 of $address_data from the recipient verification was clobbered by the
3578 sender verification.
3580 13. The value of address_data from a sender verification is now available in
3581 $sender_address_data in subsequent conditions in the ACL statement.
3583 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3585 15. Added a new option "connect=<time>" to callout options, to set a different
3588 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3589 was its contents. (It was OK if the option was not defined at all.)
3591 17. A "Completed" log line is now written for messages that are removed from
3592 the spool by the -Mrm option.
3594 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3595 information about exactly what failed.
3597 19. Added -dd to debug only the daemon process.
3599 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3600 handles timeouts, both on the server side and network timeouts. Renamed the
3601 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3603 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3604 It is now set to "smtps".
3606 22. $host_address is now set to the target address during the checking of
3607 ignore_target_hosts.
3609 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3610 being passed; this would have caused $sender_host_name to have been used if
3611 matching the list had actually called for a host name (not very likely,
3612 since this list is usually IP addresses). A host name is now passed as
3615 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3616 code that specifies a non-restarting handler (typically sigaction() in
3617 modern systems) in an attempt to fix a rare and obscure crash bug.
3619 25. Narrowed the window for a race in the daemon that could cause it to ignore
3620 SIGCHLD signals. This is not a major problem, because they are used only to
3621 wake it up if nothing else does.
3623 26. A malformed maildirsize file could cause Exim to calculate negative values
3624 for the mailbox size or file count. Odd effects could occur as a result.
3625 The maildirsize information is now recalculated if the size or filecount
3628 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3629 support for a long time. Removed HAVE_SYS_VFS_H.
3631 28. Installed the latest version of exipick from John Jetmore.
3633 29. In an address list, if the pattern was not a regular expression, an empty
3634 subject address (from a bounce message) matched only if the pattern was an
3635 empty string. Non-empty patterns were not even tested. This was the wrong
3636 because it is perfectly reasonable to use an empty address as part of a
3637 database query. An empty address is now tested by patterns that are
3638 lookups. However, all the other forms of pattern expect the subject to
3639 contain a local part and a domain, and therefore, for them, an empty
3640 address still always fails if the pattern is not itself empty.
3642 30. Exim went into a mad DNS loop when attempting to do a callout where the
3643 host was specified on an smtp transport, and looking it up yielded more
3644 than one IP address.
3646 31. Re-factored the code for checking spool and log partition space into a
3647 function that finds that data and another that does the check. The former
3648 is then used to implement four new variables: $spool_space, $log_space,
3649 $spool_inodes, and $log_inodes.
3651 32. The RFC2047 encoding function was originally intended for short strings
3652 such as real names; it was not keeping to the 75-character limit for
3653 encoded words that the RFC imposes. It now respects the limit, and
3654 generates multiple encoded words if necessary. To be on the safe side, I
3655 have increased the buffer size for the ${rfc2047: expansion operator from
3658 33. It is now permitted to omit both strings after an "if" condition; if the
3659 condition is true, the result is "true". As before, when the second string
3660 is omitted, a false condition yields an empty string. This makes it less
3661 cumbersome to write custom ACL and router conditions.
3663 34. Failure to deliver a bounce message always caused it to be frozen, even if
3664 there was an errors_to setting on the router. The errors_to setting is now
3667 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3668 canonical form (fully expanded) before being placed in
3669 $sender_host_address.
3671 36. The table in the code that translates DNS record types into text (T_A to
3672 "A" for instance) was missing entries for NS and CNAME. It is just possible
3673 that this could have caused confusion if both these types were looked up
3674 for the same domain, because the text type is used as part of Exim's
3675 per-process caching. But the chance of anyone hitting this buglet seems
3678 37. The dnsdb lookup has been extended in a number of ways.
3680 (1) There is a new type, "zns", which walks up the domain tree until it
3681 finds some nameserver records. It should be used with care.
3683 (2) There is a new type, "mxh", which is like "mx" except that it returns
3684 just the host names, not the priorities.
3686 (3) It is now possible to give a list of domains (or IP addresses) to be
3687 looked up. The behaviour when one of the lookups defers can be
3688 controlled by a keyword.
3690 (4) It is now possible to specify the separator character for use when
3691 multiple records are returned.
3693 38. The dnslists ACL condition has been extended: it is now possible to supply
3694 a list of IP addresses and/or domains to be looked up in a particular DNS
3697 39. Added log_selector=+queue_time_overall.
3699 40. When running the queue in the test harness, wait just a tad after forking a
3700 delivery process, to get repeatability of debugging output.
3702 41. Include certificate and key file names in error message when GnuTLS fails
3703 to set them up, because the GnuTLS error message doesn't include the name
3704 of the failing file when there is a problem reading it.
3706 42. Allow both -bf and -bF in the same test run.
3708 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3710 44. The "Exiscan patch" is now merged into the mainline Exim source.
3712 45. Sometimes the final signoff response after QUIT could fail to get
3713 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3714 < 0 before doing a fflush(). This bug looks as though it goes back to the
3715 introduction of TLS in release 3.20, but "sometimes" must have been rare
3716 because the tests only now provoked it.
3718 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3719 (this can affect the format of dates).
3721 47. exim_tidydb, when checking for the continued existence of a message for
3722 which it has found a message-specific retry record, was not finding
3723 messages that were in split spool directories. Consequently, it was
3724 deleting retry records that should have stayed in existence.
3726 48. Steve fixed some bugs in eximstats.
3728 49. The SPA authentication driver was not abandoning authentication and moving
3729 on to the next authenticator when an expansion was forced to fail,
3730 contradicting the general specification for all authenticators. Instead it
3731 was generating a temporary error. It now behaves as specified.
3733 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3734 (the order specifies the preference for clients). The order is now AES256,
3735 AES128, 3DES, ARCFOUR128.
3737 51. Small patch to Sieve code - explicitly set From: when generating an
3740 52. Exim crashed if a remote delivery caused a very long error message to be
3741 recorded - for instance if somebody sent an entire SpamAssassin report back
3742 as a large number of 550 error lines. This bug was coincidentally fixed by
3743 increasing the size of one of Exim's internal buffers (big_buffer) that
3744 happened as part of the Exiscan merge. However, to be on the safe side, I
3745 have made the code more robust (and fixed the comments that describe what
3748 53. Now that there can be additional text after "Completed" in log lines (if
3749 the queue_time_overall log selector is set), a one-byte patch to exigrep
3750 was needed to allow it to recognize "Completed" as not the last thing in
3753 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3754 patch that reportedly fixes this has been added. I am not expert enough to
3755 create a test for it. This is what the patch creator wrote:
3757 "I found a little strange behaviour of ldap code when working with
3758 Windows 2003 AD Domain, where users was placed in more than one
3759 Organization Units. When I tried to give exim partial DN, the exit code
3760 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3761 But simultaneously result of request was absolutely normal ldap result,
3762 so I produce this patch..."
3764 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3765 REFERENCE, so I have modified the code to exclude the patch when that macro
3768 55. Some experimental protocols are using DNS PTR records for new purposes. The
3769 keys for these records are domain names, not reversed IP addresses. The
3770 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3771 leaves it alone. Component reversal etc. now happens only for IP addresses.
3774 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3776 57. Double the size of the debug message buffer (to 2048) so that more of very
3777 long debug lines gets shown.
3779 58. The exicyclog utility now does better if the number of log files to keep
3780 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3782 59. Two changes related to the smtp_active_hostname option:
3784 (1) $smtp_active_hostname is now available as a variable.
3785 (2) The default for smtp_banner uses $smtp_active_hostname instead
3786 of $primary_hostname.
3788 60. The host_aton() function is supposed to be passed a string that is known
3789 to be a valid IP address. However, in the case of IPv6 addresses, it was
3790 not checking this. This is a hostage to fortune. Exim now panics and dies
3791 if the condition is not met. A case was found where this could be provoked
3792 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3793 components; fortuitously, this particular loophole had already been fixed
3794 by change 4.50/55 above.
3796 If there are any other similar loopholes, the new check in host_aton()
3797 itself should stop them being exploited. The report I received stated that
3798 data on the command line could provoke the exploit when Exim was running as
3799 exim, but did not say which command line option was involved. All I could
3800 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3801 running as the user.
3804 61. There was a buffer overflow vulnerability in the SPA authentication code
3805 (which came originally from the Samba project). I have added a test to the
3806 spa_base64_to_bits() function which I hope fixes it.
3809 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3810 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3812 63. The daemon start-up calls getloadavg() while still root for those OS that
3813 need the first call to be done as root, but it missed one case: when
3814 deliver_queue_load_max is set with deliver_drop_privilege. This is
3815 necessary for the benefit of the queue runner, because there is no re-exec
3816 when deliver_drop_privilege is set.
3818 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3819 This has been fixed.
3821 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3822 was in use, was not putting the data itself into the right store pool;
3823 consequently, it could be overwritten for a subsequent message in the same
3824 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3827 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3829 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3830 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3831 the function were treating the return as a boolean value, which happened to
3832 work because 0=false and not-0=true, but is not correct code.
3834 68. The host_aton() function was not handling scoped IPv6 addresses (those
3835 with, for example, "%eth0" on the end) correctly.
3837 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3838 negated items (that is, ~something) in unsigned ints. Some compilers
3839 apparently mutter when there is no cast.
3841 70. If an address verification called from an ACL failed, and did not produce a
3842 user-specific message (i.e. there was only a "system" message), nothing was
3843 put in $acl_verify_message. In this situation, it now puts the system
3846 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3847 session; change 4.31/43 added the unwanted input to the log line - except
3848 that it did not do this in the start of session case. It now does.
3850 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3851 This is wrong and can cause the other end to generate a synchronization
3852 error if it is another Exim or anything else that does the synchronization
3853 check. A QUIT command is no longer sent after a timeout.
3855 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3856 during host lookups.
3858 74. The defer_ok option of callout verification was not working if it was used
3859 when verifying addresses in header lines, that is, for this case:
3861 verify = header_sender/callout=defer_ok
3863 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3864 those file descriptors could be used for SMTP connections. If anything
3865 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3866 could be sent to the SMTP client, causing chaos. The daemon now opens
3867 stdin, stdout, and stderr to /dev/null when it puts itself into the
3870 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3871 log by default. The user can override this with suitable Perl magic.
3873 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3874 the log message when discard triggers, was not working for the DATA ACL or
3875 for the non-SMTP ACL.
3877 78. Error message wording change in sieve.c.
3879 79. If smtp_accept_max_per_host was set, the number of connections could be
3880 restricted to fewer than expected, because the daemon was trying to set up
3881 a new connection before checking whether the processes handling previous
3882 connections had finished. The check for completed processes is now done
3883 earlier. On busy systems, this bug wouldn't be noticed because something
3884 else would have woken the daemon, and it would have reaped the completed
3887 80. If a message was submitted locally by a user whose login name contained one
3888 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3889 It caused a spool format error. I have fixed the spool reading code. A
3890 related problem was that the "from" clause in the Received: line became
3891 illegal because of the space(s). It is now covered by ${quote_local_part.
3893 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3896 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3898 83. Patch from Sieve maintainer to fix unterminated string problem in
3899 "vacation" handling.
3901 84. Some minor changes to the Linux configuration files to help with other
3902 OS variants using glibc.
3904 85. One more patch for Sieve to update vacation handling to latest spec.
3907 ----------------------------------------------------
3908 See the note above about the 4.44 and 4.50 releases.
3909 ----------------------------------------------------
3915 1. Change 4.43/35 introduced a bug that caused file counts to be
3916 incorrectly computed when quota_filecount was set in an appendfile
3919 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3920 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3923 3. Give more explanation in the error message when the command for a transport
3924 filter fails to execute.
3926 4. There are several places where Exim runs a non-Exim command in a
3927 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3928 was being done only for the command run by the queryprogram router. It is
3929 now done for all such subprocesses. The other cases are: ${run, transport
3930 filters, and the commands run by the lmtp and pipe transports.
3932 5. Some older OS have a limit of 256 on the maximum number of file
3933 descriptors. Exim was using setrlimit() to set 1000 as a large value
3934 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3935 systems. I've change it so that if it can't get 1000, it tries for 256.
3937 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3938 was an oversight, and furthermore, ever since the addition of extra
3939 controls (e.g. 4.43/32), the checks on when to allow different forms of
3940 "control" were broken. There should now be diagnostics for all cases when a
3941 control that does not make sense is encountered.
3943 7. $recipients is now available in the predata ACL (oversight).
3945 8. Tidy the search cache before the fork to do a delivery from a message
3946 received from the command line. Otherwise the child will trigger a lookup
3947 failure and thereby defer the delivery if it tries to use (for example) a
3948 cached ldap connection that the parent has called unbind on.
3950 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3951 of $address_data from the recipient verification was clobbered by the
3952 sender verification.
3954 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3955 was its contents. (It was OK if the option was not defined at all.)
3957 11. A "Completed" log line is now written for messages that are removed from
3958 the spool by the -Mrm option.
3960 12. $host_address is now set to the target address during the checking of
3961 ignore_target_hosts.
3963 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3964 being passed; this would have caused $sender_host_name to have been used if
3965 matching the list had actually called for a host name (not very likely,
3966 since this list is usually IP addresses). A host name is now passed as
3969 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3970 code that specifies a non-restarting handler (typically sigaction() in
3971 modern systems) in an attempt to fix a rare and obscure crash bug.
3973 15. Narrowed the window for a race in the daemon that could cause it to ignore
3974 SIGCHLD signals. This is not a major problem, because they are used only to
3975 wake it up if nothing else does.
3977 16. A malformed maildirsize file could cause Exim to calculate negative values
3978 for the mailbox size or file count. Odd effects could occur as a result.
3979 The maildirsize information is now recalculated if the size or filecount
3982 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3983 support for a long time. Removed HAVE_SYS_VFS_H.
3985 18. Updated exipick to current release from John Jetmore.
3987 19. Allow an empty sender to be matched against a lookup in an address list.
3988 Previously the only cases considered were a regular expression, or an
3991 20. Exim went into a mad DNS lookup loop when doing a callout where the
3992 host was specified on the transport, if the DNS lookup yielded more than
3995 21. The RFC2047 encoding function was originally intended for short strings
3996 such as real names; it was not keeping to the 75-character limit for
3997 encoded words that the RFC imposes. It now respects the limit, and
3998 generates multiple encoded words if necessary. To be on the safe side, I
3999 have increased the buffer size for the ${rfc2047: expansion operator from
4002 22. Failure to deliver a bounce message always caused it to be frozen, even if
4003 there was an errors_to setting on the router. The errors_to setting is now
4006 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4007 canonical form (fully expanded) before being placed in
4008 $sender_host_address.
4010 24. Updated eximstats to version 1.33
4012 25. Include certificate and key file names in error message when GnuTLS fails
4013 to set them up, because the GnuTLS error message doesn't include the name
4014 of the failing file when there is a problem reading it.
4016 26. Expand error message when OpenSSL has problems setting up cert/key files.
4019 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4020 (this can affect the format of dates).
4022 28. exim_tidydb, when checking for the continued existence of a message for
4023 which it has found a message-specific retry record, was not finding
4024 messages that were in split spool directories. Consequently, it was
4025 deleting retry records that should have stayed in existence.
4027 29. eximstats updated to version 1.35
4028 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4029 1.35 - bugfix such that pie charts by volume are generated correctly
4031 30. The SPA authentication driver was not abandoning authentication and moving
4032 on to the next authenticator when an expansion was forced to fail,
4033 contradicting the general specification for all authenticators. Instead it
4034 was generating a temporary error. It now behaves as specified.
4036 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4037 (the order specifies the preference for clients). The order is now AES256,
4038 AES128, 3DES, ARCFOUR128.
4040 31. Small patch to Sieve code - explicitly set From: when generating an
4043 32. Exim crashed if a remote delivery caused a very long error message to be
4044 recorded - for instance if somebody sent an entire SpamAssassin report back
4045 as a large number of 550 error lines. This bug was coincidentally fixed by
4046 increasing the size of one of Exim's internal buffers (big_buffer) that
4047 happened as part of the Exiscan merge. However, to be on the safe side, I
4048 have made the code more robust (and fixed the comments that describe what
4051 33. Some experimental protocols are using DNS PTR records for new purposes. The
4052 keys for these records are domain names, not reversed IP addresses. The
4053 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4054 leaves it alone. Component reversal etc. now happens only for IP addresses.
4057 34. The host_aton() function is supposed to be passed a string that is known
4058 to be a valid IP address. However, in the case of IPv6 addresses, it was
4059 not checking this. This is a hostage to fortune. Exim now panics and dies
4060 if the condition is not met. A case was found where this could be provoked
4061 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4062 components; fortuitously, this particular loophole had already been fixed
4063 by change 4.50/55 or 4.44/33 above.
4065 If there are any other similar loopholes, the new check in host_aton()
4066 itself should stop them being exploited. The report I received stated that
4067 data on the command line could provoke the exploit when Exim was running as
4068 exim, but did not say which command line option was involved. All I could
4069 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4070 running as the user.
4073 35. There was a buffer overflow vulnerability in the SPA authentication code
4074 (which came originally from the Samba project). I have added a test to the
4075 spa_base64_to_bits() function which I hope fixes it.
4078 36. The daemon start-up calls getloadavg() while still root for those OS that
4079 need the first call to be done as root, but it missed one case: when
4080 deliver_queue_load_max is set with deliver_drop_privilege. This is
4081 necessary for the benefit of the queue runner, because there is no re-exec
4082 when deliver_drop_privilege is set.
4084 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4085 was in use, was not putting the data itself into the right store pool;
4086 consequently, it could be overwritten for a subsequent message in the same
4087 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4090 38. Sometimes the final signoff response after QUIT could fail to get
4091 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4092 < 0 before doing a fflush(). This bug looks as though it goes back to the
4093 introduction of TLS in release 3.20, but "sometimes" must have been rare
4094 because the tests only now provoked it.
4100 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4101 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4102 Now it returns an integer. A number of calls were still expecting a T/F
4103 return. Fortuitously, in all cases, the tests worked in OK situations,
4104 which is the norm. However, things would have gone wrong on any write
4105 failures on the smtp file descriptor. This function is used when sending
4106 messages over SMTP and also when doing verify callouts.
4108 2. When Exim is called to do synchronous delivery of a locally submitted
4109 message (the -odf or -odi options), it no longer closes stderr before doing
4112 3. Implemented the mua_wrapper option.
4114 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4116 5. Implemented the functions header_remove(), header_testname(),
4117 header_add_at_position(), and receive_remove_recipient(), and exported them
4120 6. If an ACL "warn" statement specified the addition of headers, Exim already
4121 inserted X-ACL-Warn: at the start if there was no header name. However, it
4122 was not making this test for the second and subsequent header lines if
4123 there were newlines in the string. This meant that an invalid header could
4124 be inserted if Exim was badly configured.
4126 7. Allow an ACL "warn" statement to add header lines at the start or after all
4127 the Received: headers, as well as at the end.
4129 8. Added the rcpt_4xx retry error code.
4131 9. Added postmaster_mailfrom=xxx to callout verification option.
4133 10. Added mailfrom=xxxx to the callout verification option, for verify=
4136 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4137 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4139 12. Inserted some casts to stop certain compilers warning when using pointer
4140 differences as field lengths or precisions in printf-type calls (mostly
4141 affecting debugging statements).
4143 13. Added optional readline() support for -be (dynamically loaded).
4145 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4146 same clock tick as a message's arrival, so that its received time was the
4147 same as the "first fail" time on the retry record, and that message
4148 remained on the queue past the ultimate address timeout, every queue runner
4149 would try a delivery (because it was past the ultimate address timeout) but
4150 after another failure, the ultimate address timeout, which should have then
4151 bounced the address, did not kick in. This was a "< instead of <=" error;
4152 in most cases the first failure would have been in the next clock tick
4153 after the received time, and all would be well.
4155 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4156 being recognized when the domain list was tested by the match_domain
4157 condition in an expansion string.
4159 16. Added the ${str2b64: operator.
4161 17. Exim was always calling setrlimit() to set a large limit for the number of
4162 processes, without checking whether the existing limit was already
4163 adequate. (It did check for the limit on file descriptors.) Furthermore,
4164 errors from getrlimit() and setrlimit() were being ignored. Now they are
4165 logged to the main and panic logs, but Exim does carry on, to try to do its
4166 job under whatever limits there are.
4168 18. Imported PCRE 5.0.
4170 19. Trivial typo in log message " temporarily refused connection" (the leading
4173 20. If the log selector return_path_on_delivery was set and an address was
4174 redirected to /dev/null, the delivery process crashed because it assumed
4175 that a return path would always be set for a "successful" delivery. In this
4176 case, the whole delivery is bypassed as an optimization, and therefore no
4179 21. Internal re-arrangement: the function for sending a challenge and reading
4180 a response while authentication was assuming a zero-terminated challenge
4181 string. It's now changed to take a pointer and a length, to allow for
4182 binary data in such strings.
4184 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4186 23. Exim was not respecting finduser_retries when seeking the login of the
4187 uid under which it was called; it was always trying 10 times. (The default
4188 setting of finduser_retries is zero.) Also, it was sleeping after the final
4189 failure, which is pointless.
4191 24. Implemented tls_on_connect_ports.
4193 25. Implemented acl_smtp_predata.
4195 26. If the domain in control=submission is set empty, Exim assumes that the
4196 authenticated id is a complete email address when it generates From: or
4197 Sender: header lines.
4199 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4200 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4201 chown and chgrp in /bin and hostname in /usr/bin.
4203 28. Exim was keeping the "process log" file open after each use, just as it
4204 does for the main log. This opens the possibility of it remaining open for
4205 long periods when the USR1 signal hits a daemon. Occasional processlog
4206 errors were reported, that could have been caused by this. Anyway, it seems
4207 much more sensible not to leave this file open at all, so that is what now
4210 29. The long-running daemon process does not normally write to the log once it
4211 has entered its main loop, and it closes the log before doing so. This is
4212 so that log files can straightforwardly be renamed and moved. However,
4213 there are a couple of unusual error situations where the daemon does write
4214 log entries, and I had neglected to close the log afterwards.
4216 30. The text of an SMTP error response that was received during a remote
4217 delivery was being truncated at 512 bytes. This is too short for some of
4218 the long messages that one sometimes sees. I've increased the limit to
4221 31. It is now possible to make retry rules that apply only when a message has a
4222 specific sender, in particular, an empty sender.
4224 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4225 it possible to be selective about when SMTP synchronization is enforced.
4227 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4229 32. Implemented hosts_connection_nolog.
4231 33. Added an ACL for QUIT.
4233 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4236 35. Added mailbox_size and mailbox_filecount to appendfile.
4238 36. Added control = no_multiline_responses to ACLs.
4240 37. There was a bug in the logic of the code that waits for the clock to tick
4241 in the case where the clock went backwards by a substantial amount such
4242 that the microsecond fraction of "now" was more than the microsecond
4243 fraction of "then" (but the whole seconds number was less).
4245 38. Added support for the libradius Radius client library this is found on
4246 FreeBSD (previously only the radiusclient library was supported).
4252 1. When certain lookups returned multiple values in the form name=value, the
4253 quoting of the values was not always being done properly. Specifically:
4254 (a) If the value started with a double quote, but contained no whitespace,
4256 (b) If the value contained whitespace other than a space character (i.e.
4257 tabs or newlines or carriage returns) it was not quoted.
4258 This fix has been applied to the mysql and pgsql lookups by writing a
4259 separate quoting function and calling it from the lookup code. The fix
4260 should probably also be applied to nisplus, ibase and oracle lookups, but
4261 since I cannot test any of those, I have not disturbed their existing code.
4263 2. A hit in the callout cache for a specific address caused a log line with no
4264 reason for rejecting RCPT. Now it says "Previous (cached) callout
4265 verification failure".
4267 3. There was an off-by-one bug in the queryprogram router. An over-long
4268 return line was truncated at 256 instead of 255 characters, thereby
4269 overflowing its buffer with the terminating zero. As well as fixing this, I
4270 have increased the buffer size to 1024 (and made a note to document this).
4272 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4273 when Exim is waiting for an SMTP response from a remote server, Exim
4274 restarts its select() call on the socket, thereby resetting its timeout.
4275 This is not a problem when such interrupts are rare. Somebody set up a cron
4276 job to run exiwhat every 2 minutes, which is less than the normal select()
4277 timeout (5 or 10 minutes). This meant that the select() timeout never
4278 kicked in because it was always reset. I have fixed this by comparing the
4279 time when an interrupt arrives with the time at the start of the first call
4280 to select(). If more time than the timeout has elapsed, the interrupt is
4281 treated as a timeout.
4283 5. Some internal re-factoring in preparation for the addition of Sieve
4284 extensions (by MH). In particular, the "personal" test is moved to a
4285 separate function, and given an option for scanning Cc: and Bcc: (which is
4286 not set for Exim filters).
4288 6. When Exim created an email address using the login of the caller as the
4289 local part (e.g. when creating a From: or Sender: header line), it was not
4290 quoting the local part when it contained special characters such as @.
4292 7. Installed new OpenBSD configuration files.
4294 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4295 try to make them clearer.
4297 9. Callout options, other than the timeout value, were being ignored when
4298 verifying sender addresses in header lines. For example, when using
4300 verify = header_sender/callout=no_cache
4302 the cache was (incorrectly) being used.
4304 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4305 only the Cygwin environment.
4307 11. When return_path_on_delivery was set as a log selector, if different remote
4308 addresses in the same message used different return paths and parallel
4309 remote delivery occurred, the wrong values would sometimes be logged.
4310 (Whenever a remote delivery process finished, the return path value from
4311 the most recently started remote delivery process was logged.)
4313 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4314 lines when AUTH and/or TLS are in use. This is the "received protocol"
4315 field. Exim used to use "asmtp" for authenticated SMTP, without any
4316 indication (in the protocol name) for TLS use. Now it follows the RFC and
4317 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4318 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4319 names appear in log lines as well as in Received: header lines.
4321 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4322 extensions, and comparison tests, and to fix some bugs.
4324 14. Changes to the "personal" filter test:
4326 (1) The test was buggy in that it was just doing the equivalent of
4327 "contains" tests on header lines. For example, if a user's address was
4328 anne@some.where, the "personal" test would incorrectly be true for
4330 To: susanne@some.where
4332 This test is now done by extracting each address from the header in turn,
4333 and checking the entire address. Other tests that are part of "personal"
4334 are now done using regular expressions (for example, to check local parts
4335 of addresses in From: header lines).
4337 (2) The list of non-personal local parts in From: addresses has been
4338 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4339 taken from the Sieve specification recommendations.
4341 (3) If the message contains any header line starting with "List-" it is
4342 treated as non-personal.
4344 (4) The test for "circular" in the Subject: header line has been removed
4345 because it now seems ill-conceived.
4347 15. Minor typos in src/EDITME comments corrected.
4349 16. Installed latest exipick from John Jetmore.
4351 17. If headers_add on a router specified a text string that was too long for
4352 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4353 of string_sprintf() is now avoided.
4355 18. $message_body_size was not set (it was always zero) when running the DATA
4356 ACL and the local_scan() function.
4358 19. For the "mail" command in an Exim filter, no default was being set for
4359 the once_repeat time, causing a random time value to be used if "once" was
4360 specified. (If the value happened to be <= 0, no repeat happened.) The
4361 default is now 0s, meaning "never repeat". The "vacation" command was OK
4362 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4363 (I found it when inspecting the code).
4365 20. There is now an overall timeout for performing a callout verification. It
4366 defaults to 4 times the callout timeout, which applies to individual SMTP
4367 commands during the callout. The overall timeout applies when there is more
4368 than one host that can be tried. The timeout is checked before trying the
4369 next host. This prevents very long delays if there are a large number of
4370 hosts and all are timing out (e.g. when the network connections are timing
4371 out). The value of the overall timeout can be changed by specifying an
4372 additional sub-option for "callout", called "maxwait". For example:
4374 verify = sender/callout=5s,maxwait=20s
4376 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4377 to the end before writing, but this should make it even safer).
4379 22. Exim was forgetting that it had advertised PIPELINING for the second and
4380 subsequent messages on an SMTP connection. It was also not resetting its
4381 memory on STARTTLS and an internal HELO.
4383 23. When Exim logs an SMTP synchronization error within a session, it now
4384 records whether PIPELINING has been advertised or not.
4386 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4387 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4388 rather than long int.
4390 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4392 26. Added the never_mail option to autoreply.
4398 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4399 crash if the getsockname() call failed; this can happen if a connection is
4400 closed very soon after it is established. The problem was simply in the
4401 order in which certain operations were done, causing Exim to try to write
4402 to the SMTP stream before it had set up the file descriptor. The bug has
4403 been fixed by making things happen in the correct order.
4409 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4410 before the connection was closed, thus losing the rejection response.
4412 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4413 some early Solaris releases, but causes trouble in current releases where
4414 socklen_t is defined.
4416 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4419 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4422 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4423 that releasing the top of it at the end releases what was used for sub-
4424 expansions (unless the block got too big). However, discard this block if
4425 the first thing is a variable or header, so that we can use its block when
4426 it is dynamic (useful for very large $message_headers, for example).
4428 6. Lookups now cache *every* query, not just the most recent. A new, separate
4429 store pool is used for this. It can be recovered when all lookup caches are
4430 flushed. Lookups now release memory at the end of their result strings.
4431 This has involved some general refactoring of the lookup sources.
4433 7. Some code has been added to the store_xxx() functions to reduce the amount
4434 of flapping under certain conditions.
4436 8. log_incoming_interface used to affect only the <= reception log lines. Now
4437 it causes the local interface and port to be added to several more SMTP log
4438 lines, for example "SMTP connection from", and rejection lines.
4440 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4442 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4444 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4445 could be overwritten at the end of the current message (or the start of a
4446 new message if it was set in a HELO ACL). The value is now preserved for
4447 the duration of the SMTP connection.
4449 12. If a transport had a headers_rewrite setting, and a matching header line
4450 contained an unqualified address, that address was qualified, even if it
4451 did not match any rewriting rules. The underlying bug was that the values
4452 of the flags that permit the existence of unqualified sender and recipient
4453 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4454 non-local messages, and by -bnq for local messages) were not being
4455 preserved with the message after it was received.
4457 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4458 "next input=" as part of the text comprising the host identity instead of
4459 the correct text. The code was using the same buffer for two different
4460 strings. However, depending on which order the printing function evaluated
4461 its arguments, the bug did not always show up. Under Linux, for example, my
4462 test suite worked just fine.
4464 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4465 doesn't work with some older versions of Perl. It has been changed to "my",
4466 which in any case is probably the better facility to use.
4468 15. A really picky compiler found some instances of statements for creating
4469 error messages that either had too many or two few arguments for the format
4472 16. The size of the buffer for calls to the DNS resolver has been increased
4473 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4474 for addresses that have a lot of PTR records. This alleviates a problem; it
4475 does not fully solve it.
4477 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4478 buffer now truncates the list and logs the incident, which is the same
4479 action as happens when Exim is looking up a host name and its aliases.
4480 Previously in this situation something unpredictable would happen;
4481 sometimes it was "internal error: store_reset failed".
4483 18. If a server dropped the connection unexpectedly when an Exim client was
4484 using GnuTLS and trying to read a response, the client delivery process
4485 crashed while trying to generate an error log message.
4487 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4488 string, for example:
4490 warn message = H1: something\nH2: something
4492 the text was added as a single header line from Exim's point of view
4493 though it ended up OK in the delivered message. However, searching for the
4494 second and subsequent header lines using $h_h2: did not work. This has been
4495 fixed. Similarly, if a system filter added multiple headers in this way,
4496 the routers could not see them.
4498 20. Expanded the error message when iplsearch is called with an invalid key to
4499 suggest using net-iplsearch in a host list.
4501 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4502 ACLs are no longer actually imposed (and a message to that effect is
4505 22. If a "gecos" field in a passwd entry contained escaped characters, in
4506 particular, if it contained a \" sequence, Exim got it wrong when building
4507 a From: or a Sender: header from that name. A second bug also caused
4508 incorrect handling when an unquoted " was present following a character
4509 that needed quoting.
4511 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4512 was not being matched caselessly.
4514 24. Arranged for all hyphens in the exim.8 source to be escaped with
4517 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4518 properly. Recipient callout cache records were still being keyed to include
4519 the sender, even when use_sender was set false. This led to far more
4520 callouts that were necessary. The sender is no longer included in the key
4521 when use_sender is false.
4523 26. Added "control = submission" modifier to ACLs.
4525 27. Added the ${base62d: operator to decode base 62 numbers.
4527 28. dnsdb lookups can now access SRV records.
4529 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4530 the configuration file.
4532 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4533 (-v) mode. This makes the output for a verbose queue run more intelligible.
4535 31. Added a use_postmaster feature to recipient callouts.
4537 32. Added the $body_zerocount variable, containing the number of binary zero
4538 bytes in the message body.
4540 33. The time of last modification of the "new" subdirectory is now used as the
4541 "mailbox time last read" when there is a quota error for a maildir
4544 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4546 35. Added +ignore_unknown as a special item in host lists.
4548 36. Code for decoding IPv6 addresses in host lists is now included, even if
4549 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4550 address was recognized as an IP address, but was then not correctly decoded
4551 into binary, causing unexpected and incorrect effects when compared with
4558 1. Very minor rewording of debugging text in manualroute to say "list of
4559 hosts" instead of "hostlist".
4561 2. If verify=header_syntax was set, and a header line with an unqualified
4562 address (no domain) and a large number of spaces between the end of the
4563 name and the colon was received, the reception process suffered a buffer
4564 overflow, and (when I tested it) crashed. This was caused by some obsolete
4565 code that should have been removed. The fix is to remove it!
4567 3. When running in the test harness, delay a bit after writing a bounce
4568 message to get a bit more predictability in the log output.
4570 4. Added a call to search_tidyup() just before forking a reception process. In
4571 theory, someone could use a lookup in the expansion of smtp_accept_max_
4572 per_host which, without the tidyup, could leave open a database connection.
4574 5. Added the variables $recipient_data and $sender_data which get set from a
4575 lookup success in an ACL "recipients" or "senders" condition, or a router
4576 "senders" option, similar to $domain_data and $local_part_data.
4578 6. Moved the writing of debug_print from before to after the "senders" test
4581 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4582 problems for message scanning, either using a data ACL, or using
4583 local_scan() because the Received: header was not generated till after they
4584 were called (in order to set the time as the time of reception completion).
4585 I have revised the way this works. The header is now generated after the
4586 body is received, but before the ACL or local_scan() are called. After they
4587 are run, the timestamp in the header is updated.
4593 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4594 before starting a queue runner without re-exec. This happened only when
4595 deliver_drop_privilege was set or when the Exim user was set to root. The
4596 effect of the bug was that timeouts during subsequent deliveries caused
4597 crashes instead of being properly handled. The handler is now left at its
4598 default (and expected) setting.
4600 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4601 message, again when deliver_drop_privilege is set or Exim is run as root.
4602 The bug described in (1) was not present in this case, but the tidying up
4603 of the other signals was missing. I have made the two cases consistent.
4605 3. The ignore_target_hosts setting on a manualroute router was being ignored
4606 for hosts that were looked up using the /MX notation.
4608 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4611 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4612 operated on the sender address. After changing the $sender_address to <>
4613 for the sender address verify, Exim was re-instated it as the original
4614 (before rewriting) address, but remembering that it had rewritten it, so it
4615 wasn't rewriting it again. This bug also had the effect of breaking the
4616 sender address verification caching when the sender address was rewritten.
4618 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4619 This has been changed so that if the ip literal address matches
4620 ignore_target_hosts, the router declines.
4622 7. Added expansion conditions match_domain, match_address, and match_local_
4623 part (NOT match_host).
4625 8. The placeholder for the Received: header didn't have a length field set.
4627 9. Added code to Exim itself and to exim_lock to test for a specific race
4628 condition that could lead to file corruption when using MBX delivery. The
4629 issue is with the lockfile that is created in /tmp. If this file is removed
4630 after a process has opened it but before that process has acquired a lock,
4631 there is the potential for a second process to recreate the file and also
4632 acquire a lock. This could lead to two Exim processes writing to the file
4633 at the same time. The added code performs the same test as UW imapd; it
4634 checks after acquiring the lock that its file descriptor still refers to
4635 the same named file.
4637 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4638 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4641 11. Added the smtp_active_hostname option. If used, this will typically be made
4642 to depend on the incoming interface address. Because $interface_address is
4643 not set up until the daemon has forked a reception process, error responses
4644 that can happen earlier (such as "too many connections") no longer contain
4647 12. If an expansion in a condition on a "warn" statement fails because a lookup
4648 defers, the "warn" statement is abandoned, and the next ACL statement is
4649 processed. Previously this caused the whole ACL to be aborted.
4651 13. Added the iplsearch lookup type.
4653 14. Added ident_timeout as a log selector.
4655 15. Added tls_certificate_verified as a log selector.
4657 16. Added a global option tls_require_ciphers (compare the smtp transport
4658 option of the same name). This controls incoming TLS connections.
4660 17. I finally figured out how to make tls_require_ciphers do a similar thing
4661 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4662 before starting the TLS session.
4664 18. Tabs are now shown as \t in -bP output.
4666 19. If the log selector return_path_on_delivery was set, Exim crashed when
4667 bouncing a message because it had too many Received: header lines.
4669 20. If two routers both had headers_remove settings, and the first one included
4670 a superfluous trailing colon, the final name in the first list and the
4671 first name in the second list were incorrectly joined into one item (with a
4672 colon in the middle).
4678 1. Added -C and -D options to the exinext utility, mainly to make it easier
4679 to include in the automated testing, but these could be helpful when
4680 multiple configurations are in use.
4682 2. The exinext utility was not formatting the output nicely when there was
4683 an alternate port involved in the retry record key, nor when there was a
4684 message id as well (for retries that were specific to a specific message
4685 and a specific host). It was also confused by IPv6 addresses, because of
4686 the additional colons they contain. I have fixed the IPv4 problem, and
4687 patched it up to do a reasonable job for IPv6.
4689 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4690 delivery, the log line now contains "pipelined" if PIPELINING was used.
4692 4. An SMTP transport process used to panic and die if the bind() call to set
4693 an explicit outgoing interface failed. This has been changed; it is now
4694 treated in the same way as a connect() failure.
4696 5. A reference to $sender_host_name in the part of a conditional expansion
4697 that was being skipped was still causing a DNS lookup. This no longer
4700 6. The def: expansion condition was not recognizing references to header lines
4701 that used bh_ and bheader_.
4703 7. Added the _cache feature to named lists.
4705 8. The code for checking quota_filecount in the appendfile transport was
4706 allowing one more file than it should have been.
4708 9. For compatibility with Sendmail, the command line option
4716 and sets the incoming protocol and host name (for trusted callers). The
4717 host name and its colon can be omitted when only the protocol is to be set.
4718 Note the Exim already has two private options, -pd and -ps, that refer to
4719 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4720 "s", but I don't think that's a major issue.
4722 10. A number of refactoring changes to the code, none of which should affect
4725 (a) The number of logging options was getting close to filling up the
4726 32-bit word that was used as a bit map. I have split them into two classes:
4727 those that are passed in the argument to log_write(), and those that are
4728 only ever tested independently outside of that function. These are now in
4729 separate 32-bit words, so there is plenty of room for expansion again.
4730 There is no change in the user interface or the logging behaviour.
4732 (b) When building, for example, log lines, the code previously used a
4733 macro that called string_cat() twice, in order to add two strings. This is
4734 not really sufficiently general. Furthermore, there was one instance where
4735 it was actually wrong because one of the argument was used twice, and in
4736 one call a function was used. (As it happened, calling the function twice
4737 did not affect the overall behaviour.) The macro has been replaced by a
4738 function that can join an arbitrary number of extra strings onto a growing
4741 (c) The code for expansion conditions now uses a table and a binary chop
4742 instead of a serial search (which was left over from when there were very
4743 few conditions). Also, it now recognizes conditions like "pam" even when
4744 the relevant support is not compiled in: a suitably worded error message is
4745 given if an attempt is made to use such a condition.
4747 11. Added ${time_interval:xxxxx}.
4749 12. A bug was causing one of the ddress fields not to be passed back correctly
4750 from remote delivery subprocesses. The field in question was not being
4751 subsequently used, so this caused to problems in practice.
4753 13. Added new log selectors queue_time and deliver_time.
4755 14. Might have fixed a bug in maildirsizefile handling that threw up
4756 "unexpected character" debug warnings, and recalculated the data
4757 unnecessarily. In any case, I expanded the warning message to give more
4760 15. Added the message "Restricted characters in address" to the statements in
4761 the default ACL that block characters like @ and % in local parts.
4763 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4764 Three changes have been made:
4766 (a) There was a serious bug; a negative response to MAIL caused the whole
4767 recipient domain to be cached as invalid, thereby blocking all messages
4768 to all local parts at the same domain, from all senders. This bug has
4769 been fixed. The domain is no longer cached after a negative response to
4770 MAIL if the sender used is not empty.
4772 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4775 (c) A new callout option, "use_sender" has been added for people who want
4776 the modified behaviour.
4782 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4785 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4786 indeed breaks things for older releases.
4788 3. Added additional logging to the case where there is a problem reading data
4789 from a filter that is running in a subprocess using a pipe, in order to
4790 try to track down a specific problem.
4792 4. Testing facility fudge: when running in the test harness and attempting
4793 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4794 getting "No route to host". Convert this to a timeout.
4796 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4799 6. Some OS don't have socklen_t but use size_t instead. This affects the
4800 fifth argument of getsockopt() amongst other things. This is now
4801 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4802 can be set for individual OS. I have set it for SunOS5, OSF1, and
4803 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4804 some earlier ones do not.
4806 7. Change 4.30/15 was not doing the test caselessly.
4808 8. The standard form for an IPv6 address literal was being rejected by address
4809 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4810 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4811 this, as well as the form without the "IPv6" on the front (but only when
4812 address literals are enabled, of course).
4814 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4816 10. Exim crashed if a message with an empty sender address specified by -f
4817 encountered a router with an errors_to setting. This could be provoked only
4818 by a command such as
4822 where an empty string was supplied; "<>" did not hit this bug.
4824 11. Installed PCRE release 4.5.
4826 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4827 remained set. It is now erased.
4829 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4830 times from message ids (which are base 36 rather than the normal 62).
4832 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4833 were being counted as actual protocol errors, and logged if the log
4834 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4835 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4836 and DATA following a set of rejected RCPTs do not count as protocol errors.
4837 In other words, Exim assumes they were pipelined, though this may not
4838 actually be the case. Of course, in all cases the client gets an
4839 appropriate error code.
4841 15. If a lookup fails in an ACL condition, a message about the failure may
4842 be available; it is used if testing the ACL cannot continue, because most
4843 such messages specify what the cause of the deferral is. However, some
4844 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4845 that caused an old message to be retained and used if a later statement
4846 caused a defer, replacing the real cause of the deferral.
4848 16. If an IP address had so many PTR records that the DNS lookup buffer
4849 was not large enough to hold them, Exim could crash while trying to process
4850 the truncated data. It now detects and logs this case.
4852 17. Further to 4.21/58, another change has been made: if (and only if) the
4853 first line of a message (the first header line) ends with CRLF, a bare LF
4854 in a subsequent header line has a space inserted after it, so as not to
4855 terminate the header.
4857 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4858 unnecessarily, used atoi() instead of strtol(), and didn't check the
4859 termination when getting file sizes from file names by regex.
4861 19. Completely re-implemented the support for maildirsize files, in the light
4862 of a number of problems with the previous contributed implementation
4863 (4.30/29). In particular:
4865 . If the quota is zero, the maildirsize file is maintained, but no quota is
4868 . If the maildir directory does not exist, it is created before any attempt
4869 to write a maildirsize file.
4871 . The quota value in the file is just a cache; if the quota is changed in
4872 the transport, the new value overrides.
4874 . A regular expression is available for excluding directories from the
4877 20. The autoreply transport checks the characters in options that define the
4878 message's headers; it allows continued headers, but it was checking with
4879 isspace() after an embedded newline instead of explicitly looking for a
4882 21. If all the "regular" hosts to which an address was routed had passed their
4883 expiry times, and had not reached their retry times, the address was
4884 bounced, even if fallback hosts were defined. Now Exim should go on to try
4887 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4888 equivalent code in the SMTP transport. Some hosts send humungous responses
4889 to HELO/EHLO, more than 1024 it seems.
4891 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4892 gives compiler warnings in some environments. I've now done it "properly",
4895 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4896 (because of problems with the built-in one) was declared to return uschar *
4897 instead of char *, causing compiler failure.
4899 25. Fixed a file descriptor leak when processing alias/forward files.
4901 26. Fixed a minor format string issue in dbfn.c.
4903 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4905 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4906 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4909 29. When a sender address is verified, it is cached, to save repeating the test
4910 when there is more than one recipient in a message. However, when the
4911 verification involves a callout, it is possible for different callout
4912 options to be set for different recipients. It is too complicated to keep
4913 track of this in the cache, so now Exim always runs a verification when a
4914 callout is required, relying on the callout cache for the optimization.
4915 The overhead is duplication of the address routing, but this should not be
4918 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4919 to be verified with callout=postmaster, and the main callout worked but the
4920 postmaster check failed, the verification correctly failed. However, if a
4921 subsequent RCPT command asked for sender verification *without* the
4922 postmaster check, incorrect caching caused this verification also to fail,
4925 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4926 it was not caching the DNS options (qualify_single, search_parents) that
4927 were used when the lookup failed. A subsequent lookup with different
4928 options therefore always gave the same answer, though there were cases
4929 where it should not have. (Example: a "domains = !$mx_any" option on a
4930 dnslookup router: the "domains" option is always processed without any
4931 widening, but the router might have qualify_single set.) Now Exim uses the
4932 cached value only when the same options are set.
4934 32. Added John Jetmore's "exipick" utility to the distribution.
4936 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4937 than a timeout (e.g. a certificate is required, and is not provided), an
4938 Exim server now closes the connection immediately. Previously it waited for
4939 the client to close - but if the client is SSL, it seems that they each
4940 wait for each other, leading to a delay before one of them times out.
4942 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4943 maintained 0.8.x compatibility because I don't think many are using it, and
4944 it is clearly obsolete.
4946 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4949 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4950 client certificate was expired. A simple patch fixes this, though I don't
4951 understand the full logic of why the verify callback is called multiple
4954 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4955 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4956 which causes problems with some clients (such as the Certicom SSL Plus
4957 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4958 disables the coutermeasure allowing Eudora to connect."
4960 38. Exim was not checking that a write() to a log file succeeded. This could
4961 lead to Bad Things if a log got too big, in particular if it hit a file
4962 size limit. Exim now panics and dies if it cannot write to a log file, just
4963 as it does if it cannot open a log file.
4965 39. Modified OS/Makefile-Linux so that it now contains
4967 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4969 The two -D definitions ensure that Exim is compiled with large file
4970 support, which makes it possible to handle log files that are bigger than
4973 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4974 instance) a domain was checked against a named list that involved a lookup,
4975 causing $domain_data to be set, then another domain was checked against the
4976 same list, then the first domain was re-checked, the value of $domain_data
4977 after the final check could be wrong. In particular, if the second check
4978 failed, it could be set empty. This bug probably also applied to
4981 41. The strip_trailing_dot option was not being applied to the address given
4982 with the -f command-line option.
4984 42. The code for reading a message's header from the spool was incrementing
4985 $received_count, but never initializing it. This meant that the value was
4986 incorrect (doubled) while delivering a message in the same process in which
4987 it was received. In the most common configuration of Exim, this never
4988 happens - a fresh exec is done - but it can happen when
4989 deliver_drop_privilege is set.
4991 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4992 it now includes up to 150 characters of the unexpected data in the log
4995 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4996 and building data strings. The size of both of these buffers was 10 000
4997 bytes - far larger than anybody would *ever* want, thought I. Needless to
4998 say, somebody hit the limit. I have increased the maximum line length to
4999 20 000 and the maximum data length of concatenated lines to 100 000. I have
5000 also fixed two bugs, because there was no checking on these buffers. Tsk,
5001 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5002 buffer is too small.
5004 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5005 lsearch lookups. Now it does.
5007 46. When parsing a route_list item in a manualroute router, a fixed-length
5008 buffer was used for the list of hosts. I made this 1024 bytes long,
5009 thinking that nobody would ever have a list of hosts that long. Wrong.
5010 Somebody had a whole pile of complicated expansion conditions, and the
5011 string was silently truncated, leading to an expansion error. It turns out
5012 that it is easier to change to an unlimited length (owing to other changes
5013 that have happened since this code was originally written) than to build
5014 structure for giving a limitation error. The length of the item that
5015 expands into the list of hosts is now unlimited.
5017 47. The lsearch lookup could not handle data where the length of text line was
5018 more than 4095 characters. Such lines were truncated, leading to shortened
5019 data being returned. It should now handle lines of any length.
5021 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5022 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5025 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5026 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5027 no longer happens in comment lines. A list of replacements is now placed
5028 at the head of all of the source files, except those whose only change is
5029 to replace PERL_COMMAND in the very first #! line.
5031 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5032 messages on the queue, with a bottom-up merge sort, using code contributed
5033 by Michael Haardt. This should make operations like -bp somewhat faster on
5034 large queues. It won't affect queue runners, except when queue_run_in_order
5037 51. Installed eximstats 1.31 in the distribution.
5039 52. Added support for SRV lookups to the dnslookup router.
5041 53. If an ACL referred to $message_body or $message_body_end, the value was not
5042 reset for any messages that followed in the same SMTP session.
5044 54. The store-handling optimization for building very long strings was not
5045 differentiating between the different store pools. I don't think this
5046 actually made any difference in practice, but I've tidied it.
5048 55. While running the routers to verify a sender address, $sender_address
5049 was still set to the sender address. This is wrong, because when routing to
5050 send a bounce to the sender, it would be empty. Therefore, I have changed
5051 it so that, while verifying a sender address, $sender_address is set to <>.
5052 (There is no change to what happens when verifying a recipient address.)
5054 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5055 target A or AAAA records (if not already returned) without resetting the
5056 qualify_single or search_parents options of the DNS resolver. These are
5057 inappropriate in this case because the targets of MX and SRV records must
5058 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5059 target that, when qualified, matched something in the local domain. These
5060 two options are now turned off when doing these lookups.
5062 57. It seems that at least some releases of Reiserfs (which does not have the
5063 concept of a fixed number of inodes) returns zero and not -1 for the
5064 number of available inodes. This interacted badly with check_spool_inodes,
5065 which assumed that -1 was the "no such thing" setting. What I have done is
5066 to check that the total number of inodes is greater than zero before doing
5067 the test of how many are available.
5069 58. When a "warn" ACL statement has a log_message modifier, the message is
5070 remembered, and not repeated. This is to avoid a lot of repetition when a
5071 message has many recipients that cause the same warning to be written.
5072 Howewer, Exim was preserving the list of already written lines for an
5073 entire SMTP session, which doesn't seem right. The memory is now reset if a
5074 new message is started.
5076 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5077 debugging output unless log_rewrite was also set.
5079 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5080 of (int)(handle) when we know that handle contains (void *)(-1).
5082 61. The Exim daemon panic-logs an error return when it closes the incoming
5083 connection. However "connection reset by peer" seems to be common, and
5084 isn't really an error worthy of noting specially, so that particular error
5087 62. When Exim is trying to find all the local interfaces, it used to panic and
5088 die if the ioctl to get the interface flags failed. However, it seems that
5089 on at least one OS (Solaris 9) it is possible to have an interface that is
5090 included in the list of interfaces, but for which you get a failure error
5091 for this call. This happens when the interface is not "plumbed" into a
5092 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5093 failure of the "get flags" call assumes that the interface is down.
5095 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5096 makes life easier for people who are doing arithmetic on fields extracted
5097 from dates, where you often get leading zeros that should not be
5098 interpreted as octal.
5100 64. Added qualify_domain to the redirect router, to override the global
5103 65. If a pathologically long header line contained very many addresses (the
5104 report of this problem mentioned 10 000) and each of them was rewritten,
5105 Exim could use up a very large amount of memory. (It kept on making new
5106 copies of the header line as it rewrote, and never released the old ones.)
5107 At the expense of a bit more processing, the header rewriting function has
5108 been changed so that it no longer eats memory in this way.
5110 66. The generation of the Received: header has been moved from the time that a
5111 message starts to be received, to the time that it finishes. The timestamp
5112 in the Received: header should now be very close to that of the <= log
5113 line. There are two side-effects of this change:
5115 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5116 logged header lines no longer include the local Received: line, because
5117 it has not yet been created. The same applies to a copy of the message
5118 that is returned to a non-SMTP sender when a message is rejected.
5120 (b) When a filter file is tested using -bf, no additional Received: header
5121 is added to the test message. After some thought, I decided that this
5124 This change does not affect the value of $received_for. It is still set
5125 after address rewriting, but before local_scan() is called.
5127 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5129 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5130 gave an unhelpful panic error message, and a defer error. I have managed to
5131 change this behaviour so that it now rejects any supplied certificate,
5132 which seems right, as the list of acceptable certificates is empty.
5134 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5135 gave an unhelpful defer error. I have not managed to make this reject any
5136 supplied certificates, but the error message it gives is "no certificate
5137 supplied", which is not helpful.
5139 70. exigrep's output now also includes lines that are not associated with any
5140 message, but which match the given pattern. Implemented by a patch from
5141 Martin Sluka, which also tidied up the Perl a bit.
5143 71. Recipient callout verification, like sender verification, was using <> in
5144 the MAIL FROM command. This isn't really the right thing, since the actual
5145 sender may affect whether the remote host accepts the recipient or not. I
5146 have changed it to use the actual sender in the callout; this means that
5147 the cache record is now keyed on a recipient/sender pair, not just the
5148 recipient address. There doesn't seem to be a real danger of callout loops,
5149 since a callout by the remote host to check the sender would use <>.
5150 [SEE ABOVE: changed after hitting problems.]
5152 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5153 temporary errors. However, in the case of such a code being given after
5154 the end of a data transmission (i.e. after ".") Exim was failing to write
5155 a retry record for the message. (Yes, there was some broken host that was
5156 actually sending 8xx at this point.)
5158 73. An unknown lookup type in a host list could cause Exim to panic-die when
5159 the list was checked. (An example that provoked this was putting <; in the
5160 middle of a list instead of at the start.) If this happened during a DATA
5161 ACL check, a -D file could be left lying around. This kind of configuration
5162 error no longer causes Exim to die; instead it causes a defer errror. The
5163 incident is still logged to the main and panic logs.
5165 74. Buglet left over from Exim 3 conversion. The message "too many messages
5166 in one connection" was written to the rejectlog but not the mainlog, except
5167 when address rewriting (yes!) was being logged.
5169 75. Added write_rejectlog option.
5171 76. When a system filter was run not as root (that is, when system_filter_user
5172 was set), the values of the $n variables were not being returned to the
5173 main process; thus, they were not subsequently available in the $sn
5176 77. Added +return_path_on_delivery log selector.
5178 78. A connection timeout was being treated differently from recipients deferred
5179 when testing hosts_max_try with a message that was older than the host's
5180 retry timeout. (The host should not be counted, thus allowing all hosts to
5181 be tried at least once before bouncing.) This may have been the cause of an
5182 occasionally reported bug whereby a message would remain on the queue
5183 longer than the retry timeout, but would be bounced if a delivery was
5184 forced. I say "may" because I never totally pinned down the problem;
5185 setting up timeout/retry tests is difficult. See also the next item.
5187 79. The ultimate address timeout was not being applied to errors that involved
5188 a combination of host plus message (for example, a timeout on a MAIL
5189 command). When an address resolved to a number of possible hosts, and they
5190 were not all tried for each delivery (e.g. because of hosts_max_try), a
5191 message could remain on the queue longer than the retry timeout.
5193 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5196 81. Fixed an obscure SMTP outgoing bug which required at least the following
5197 conditions: (a) there was another message waiting for the same server;
5198 (b) the server returned 5xx to all RCPT commands in the first message so
5199 that the message was not completed; (c) the server dropped the connection
5200 or gave a negative response to the RSET that Exim sends to abort the
5201 transaction. The observed case was a dropped connection after DATA that had
5202 been sent in pipelining mode. That is, the server had advertised PIPELINING
5203 but was not implementing it correctly. The effect of the bug was incorrect
5204 behaviour, such as trying another host, and this could lead to a crash.
5210 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5211 and daemon.c were passed as pointers to ints; they should have been
5212 pointers to socklen_t variables (which are typically unsigned ints).
5214 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5217 3. Fixed a really odd bug that affected only the testing scheme; patching a
5218 certain fixed string in the binary changed the value of another string that
5219 happened to be identical to the end of the original first string.
5221 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5222 name", it returns that address as the IP address. On some operating
5223 systems (e.g. Solaris), it also passes back the IP address string as the
5224 "host name". However, on others (e.g. Linux), it passes back an empty
5225 string. Exim wasn't checking for this, and was changing the host name to an
5226 empty string, assuming it had been canonicized.
5228 5. Although rare, it is permitted to have more than one PTR record for a given
5229 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5230 all the names associated with an address, because they do in Solaris.
5231 However, it seems that they do not in Linux for data that comes from the
5232 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5233 I found this out when I moved to a new Linux workstation and tried to run
5234 the Exim test suite.
5236 To get round this problem I have changed the code so that it now does its
5237 own call to the DNS to look up PTR records when searching for a host name.
5238 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5239 addresses that are only in /etc/hosts are still found.
5241 This behaviour is, however, controlled by an option called host_lookup_
5242 order, which defaults to "bydns:byaddr". If people want to use the other
5243 order, or indeed, just use one or the other means of lookup, they can
5244 specify it in this variable.
5246 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5247 some operating systems, this comes back from gethostbyaddr() as an empty
5248 string, and this is what Exim used to test for. However, it seems that in
5249 other systems, "." is yielded. Exim now tests for this case too.
5251 7. The values of check_spool_space and check_log_space are now held internally
5252 as a number of kilobytes instead of an absolute number of bytes. If a
5253 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5254 kilobyte. This means that much larger values can be stored.
5256 8. Exim monitor: an attempt to get the action menu when not actually pointing
5257 at a message produces an empty menu entitled "No message selected". This
5258 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5259 no entries in it ("Shell widget menu has zero width and/or height"). So I
5260 have added a single, blank menu entry in this case.
5262 9. Added ${quote_local_part.
5264 10. MIME decoding is now applied to the contents of Subject: header lines when
5267 11. Now that a reference to $sender_host_address automatically causes a reverse
5268 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5269 host lookup before query-style lookups in lists that might use this
5270 variable. This has therefore been abolished, and the "net-" prefix is no
5271 longer necessary for query-style lookups.
5273 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5274 have been a typo for LFLAGS, so it has been changed.
5276 13. The install script calls Exim with "-C /dev/null" in order to find the
5277 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5278 to be output. Howeve, since Exim outputs its version number before the
5279 error, it didn't break the script. It just looked ugly. I fixed this by
5280 always allowing "-C /dev/null" if the caller is root.
5282 14. Ignore overlarge ACL variable number when reading spool file - insurance
5283 against a later release with more variables having written the file.
5285 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5286 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5287 this, as well as the form without the "IPv6" on the front.
5289 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5290 OS/Makefile-Darwin file.
5292 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5293 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5295 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5296 variable. However, some OS use ints for inodes. I've added cast to long int
5297 to get rid of the compiler warning.
5299 19. I had forgotten to lock out "/../" in configuration file names when
5300 ALT_CONFIG_PREFIX was set.
5302 20. Routers used for verification do not need to specify transports. However,
5303 if such a router generated a host list, and callout was configured, Exim
5304 crashed, because it could not find a port number from the (non-existent)
5305 transport. It now assumes port 25 in this circumstance.
5307 21. Added the -t option to exigrep.
5309 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5310 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5311 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5312 them is set, LOOKUP_LSEARCH is forced.
5314 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5316 24. Added sender and host information to the "rejected by local_scan()" log
5317 line; previously there was no indication of these.
5319 25. Added .include_if_exists.
5321 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5322 Linux. It turns out that not all file systems support this. Apparently some
5323 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5324 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5325 not supported on the file descriptor, is now ignored when Exim is trying to
5326 sync a directory. This applies only to Linux.
5328 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5330 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5331 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5334 29. Added support for maildirsize files from supplied patch (modified a bit).
5336 30. The use of :fail: followed by an empty string could lead Exim to respond to
5337 sender verification failures with (e.g.):
5339 550 Verification failed for <xxx>
5340 550 Sender verify failed
5342 where the first response line was missing the '-' that indicates it is not
5343 the final line of the response.
5345 31. The loop for finding the name of the user that called Exim had a hardwired
5346 limit of 10; it now uses the value of finduser_retries, which is used for
5347 all other user lookups.
5349 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5352 33. Exim was neglecting to zero errno before one call of strtol() when
5353 expanding a string and expecting an integer value. On some systems this
5354 resulted in spurious "integer overflow" errors. Also, it was casting the
5355 result into an int without checking.
5357 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5358 did not work. The code looks as if it has *never* worked, though it appears
5359 to have been documented since at least releast 1.62. I have made it work.
5361 35. The "timeout_DNS" error in retry rules, also documented since at least
5362 1.62, also never worked. As it isn't clear exactly what this means, and
5363 clearly it isn't a major issue, I have abolished the feature by treating it
5364 as "timeout", and writing a warning to the main and panic logs.
5366 36. The display of retry rules for -brt wasn't always showing the error code
5369 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5370 timeout_connect_A, timeout_connect_MX.
5372 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5373 to the empty sender.
5375 39. The daemon was not analysing the content of -oX till after it had closed
5376 stderr and disconnected from the controlling terminal. This meant that any
5377 syntax errors were only noted on the panic log, and the return code from
5378 the command was 0. By re-arranging the code a little, I've made the
5379 decoding happen first, so such errors now appear on stderr, and the return
5380 code is 1. However, the actual setting up of the sockets still happens in
5381 the disconnected process, so errors there are still only recorded on the
5384 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5385 connections (as happens on some IP stacks) was logged at start up time as
5386 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5387 it from "IPv6 and IPv4", which means that two separate sockets are being
5390 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5391 says whether AF_INET or AF_INET6 was passed as an argument.
5393 42. Exiwhat output was messed up when time zones were included in log
5396 43. Exiwhat now gives more information about the daemon's listening ports,
5397 and whether -tls-on-connect was used.
5399 44. The "port" option of the smtp transport is now expanded.
5401 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5402 silently ignored. Now an error message is written to the main and panic
5405 46. There's a new ACL modifier called "logwrite" which writes to a log file
5406 as soon as it is encountered.
5408 47. Added $local_user_uid and $local_user_gid at routing time.
5410 48. Exim crashed when trying to verify a sender address that was being
5413 49. Exim was recognizing only a space character after ".include". It now also
5414 recognizes a tab character.
5416 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5417 extracting the relevant information from the specification. The man page no
5418 longer contains scrambled data for the -d option, and I've added a section
5419 at the front about calling Exim under different names.
5421 51. Added "extra_headers" argument to the "mail" command in filter files.
5423 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5426 53. Installed eximstats 1.29.
5428 54. Added transport_filter_timeout as a generic transport option.
5430 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5431 Cc: header lines. This was required by RFC 822, but it not required by RFC
5434 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5435 incoming messages that did not have them. Now it does so only if the
5436 message originates locally, that is, if there is no associated remote host
5437 address. When Resent- header lines are present, this applies to the Resent-
5438 lines rather than the non-Resent- lines.
5440 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5441 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5443 58. Messages for configuration errors now include the name of the main
5444 configuration file - useful now that there may be more than one file in a
5445 list (.included file names were always shown).
5447 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5448 for those rare installations that do not start the daemon as root or run it
5449 setuid root. I've cut out the call to initgroups() if the daemon is not
5452 60. The Exim user and group can now be bound into the binary as text strings
5453 that are looked up at the start of Exim's processing.
5455 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5457 62. Added $mailstore_basename variable.
5459 63. Installed patch to sieve.c from Michael Haardt.
5461 64. When Exim failed to open the panic log after failing to open the main log,
5462 the original message it was trying to log was written to stderr and debug
5463 output, but if they were not available (the usual case in production), it
5464 was lost. Now it is written to syslog before the two lines that record the
5465 failures to open the logs.
5467 65. Users' Exim filters run in subprocesses under the user's uid. It is
5468 possible for a "deliver" command or an alias in a "personal" command to
5469 provoke an address rewrite. If logging of address rewriting is configured,
5470 this fails because the process is not running as root or exim. There may be
5471 a better way of dealing with this, but for the moment (because 4.30 needs
5472 to be released), I have disabled address rewrite logging when running a
5473 filter in a non-root, non-exim process.
5479 1. The buildconfig auxiliary program wasn't quoting the value set for
5480 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5481 not defined. This bug was masked in 4.22 by the effect that was fixed in
5484 2. Some messages that were rejected after a message id was allocated were
5485 shown as "incomplete" by exigrep. It no longer does this for messages that
5486 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5488 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5489 have allow_domain_literals set, the ID did not get logged in the <= line.
5490 Domain literals are now always recognized in Message-ID: header lines.
5492 4. The first argument for a ${extract expansion item is the key name or field
5493 number. Leading and trailing spaces in this item were not being ignored,
5494 causing some misleading effects.
5496 5. When deliver_drop_privilege was set, single queue runner processes started
5497 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5498 same command in the process it spins off) were not dropping privilege.
5500 6. When the daemon running as "exim" started a queue runner, it always
5501 re-executed Exim in the spun-off process. This is a waste of effort when
5502 deliver_drop_privilege is set. The new process now just calls the
5503 queue-runner function directly.
5509 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5512 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5513 ignored. Though the use of -oP was forcing the writing of a pid file, it
5514 was always written to the default place.
5516 3. If the message "no IP address found for host xxxx" is generated during
5517 incoming verification, it is now followed by identification of the incoming
5518 connection (so you can more easily find what provoked it).
5520 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5522 5. Added some features to "harden" Exim a bit more against certain attacks:
5524 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5525 be put in Local/Makefile. This is like the never_users runtime option,
5526 but it cannot be overridden. The default setting is "root".
5528 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5529 prefix string with which any file named in a -C command line option
5532 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5533 is retained for -C and -D only if the caller of Exim is root. Without
5534 it, the exim user may also use -C and -D and retain privilege.
5536 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5537 command line option is disabled.
5539 6. Macro names set by the -D option must start with an upper case letter, just
5540 like macro names defined in the configuration file.
5542 7. Added "dereference=" facility to LDAP.
5544 8. Two instances of the typo "uknown" in the source files are fixed.
5546 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5547 the Configure-Makefile script screwed up while processing it.
5549 10. Incorporated PCRE 4.4.
5551 11. The SMTP synchronization check was not operating right at the start of an
5552 SMTP session. For example, it could not catch a HELO sent before the client
5553 waited for the greeting. There is now a check for outstanding input at the
5554 point when the greeting is written. Because of the duplex, asynchronous
5555 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5556 way, but not yet received, when the check is performed.
5558 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5559 on TCP/IP sockets, because this apparently causes some broken clients to
5562 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5563 unchanged) from the Cygwin maintainer.
5565 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5566 was supported instead of testing for mbx. Effectively a typo.
5568 15. The spa authenticator server code was not checking that the input it
5569 received was valid base64.
5571 16. The debug output line for the "set" modifier in ACLs was not showing the
5572 name of the variable that was being set.
5574 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5576 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5577 lookup on its own. Something else was needed to trigger the lookup. For
5578 example, a match in host_lookup or the need for a host name in a host list.
5579 Now, if $sender_host_name is referenced and the host name has not yet been
5580 looked up, a lookup is performed. If the lookup fails, the variable remains
5581 empty, and $host_lookup_failed is set to "1".
5583 19. Added "eqi" as a case-independent comparison operator.
5585 20. The saslauthd authentication condition could segfault if neither service
5586 nor realm was specified.
5588 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5589 error message that was logged was misleading, and incoming SMTP
5590 connections were dropped. The message is now more accurate, and temporary
5591 errors are given to SMTP connections.
5593 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5594 (except RSET) with a 421 error, until QUIT is received. However, it was
5595 failing to send a response to QUIT.
5597 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5598 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5599 ACL is now run after the helo_try_verify_hosts code.
5601 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5602 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5603 case-independent, so other case variants are also recognized). Apparently
5604 some systems use these upper case variants.
5606 25. If more than two messages were waiting for the same host, and a transport
5607 filter was specified for the transport, Exim sent two messages over the
5608 same TCP/IP connection, and then failed with "socket operation on non-
5609 socket" when it tried to send the third.
5611 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5613 27. The extern definition of crypt16() in expand.c was not being excluded when
5614 the OS had its own crypt16() function.
5616 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5617 as a preferred synonym for return_size_limit, both as an option and as an
5620 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5622 30. Changed the default configuration ACL to relax the local part checking rule
5623 for addresses that are not in any local domains. For these addresses,
5624 slashes and pipe symbols are allowed within local parts, but the sequence
5625 /../ is explicitly forbidden.
5627 31. SPA server authentication was not clearing the challenge buffer before
5630 32. log_message in a "warn" ACL statement was writing to the reject log as
5631 well as to the main log, which contradicts the documentation and doesn't
5632 seem right (because no rejection is happening). So I have stopped it.
5634 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5635 However, I am unable to do any testing of this.
5637 34. Fixed an infelicity in the appendfile transport. When checking directories
5638 for a mailbox, to see if any needed to be created, it was accidentally
5639 using path names with one or more superfluous leading slashes; tracing
5640 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5642 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5643 contents are added to the log line that is written for every discarded
5644 recipient. (Previously a log_message setting was ignored.)
5646 36. The ${quote: operator now quotes the string if it is empty.
5648 37. The install script runs exim in order to find its version number. If for
5649 some reason other than non-existence or emptiness, which it checks, it
5650 could not run './exim', it was installing it with an empty version number,
5651 i.e. as "exim-". This error state is now caught, and the installation is
5654 38. An argument was missing from the function that creates an error message
5655 when Exim fails to connect to the socket for saslauthd authentication.
5656 This could cause Exim to crash, or give a corrupted message.
5658 39. Added isip, isip4, and isip6 to ${if conditions.
5660 40. The ACL variables $acl_xx are now saved with the message, and can be
5661 accessed later in routers, transports, and filters.
5663 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5664 strings in the file are not string-expanded.
5666 42. If a MAIL command specified a SIZE value that was too large to fit into an
5667 int variable, the check against message_size_limit failed. Such values are
5668 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5669 day this will have to be increased, but I don't think I want to be around
5670 when emails are that large.
5677 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5678 iconv() is not standard in FreeBSD.
5680 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5681 IPv6 enabled. The observed symptom was a segmentation fault on return from
5682 the function os_common_find_running_interfaces() in src/os.c.
5684 3. In the check_special_case() function in daemon.c I had used "errno" as an
5685 argument name, which causes warnings on some systems. This was basically a
5686 typo, since it was named "eno" in the comments!
5688 4. The code that waits for the clock to tick (at a resolution of some fraction
5689 of a second) so as to ensure message-id uniqueness was always waiting for
5690 at least one whole tick, when it could have waited for less. [This is
5691 almost certainly not relevant at current processor speeds, where it is
5692 unlikely to ever wait at all. But we try to future-proof.]
5694 5. The function that sleeps for a time interval that includes fractions of a
5695 second contained a race. It did not block SIGALRM between setting the
5696 timer, and suspending (a couple of lines later). If the interval was short
5697 and the sigsuspend() was delayed until after it had expired, the suspension
5698 never ended. On busy systems this could lead to processes getting stuck for
5701 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5702 process, before it forks any delivery processes. The open lookup caching
5703 mechanism meant that the open file or database connection was passed into
5704 the delivery process. The problem was that delivery processes always tidy
5705 up cached lookup data. This could cause a problem for the next delivery
5706 process started by the queue runner, because the external queue runner
5707 process does not know about the closure. So the next delivery process
5708 still has data in the lookup cache. In the case of a file lookup, there was
5709 no problem because closing a file descriptor in a subprocess doesn't affect
5710 the parent. However, if the lookup was caching a connection to a database,
5711 the connection was closed, and the second delivery process was likely to
5712 see errors such as "PGSQL: query failed: server closed the connection
5713 unexpectedly". The problem has been fixed by closing all cached lookups
5714 in a queue runner before running a delivery process.
5716 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5717 seem to have the "const" qualifier which it has on other OS. I've
5720 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5721 *of the same type* (client or server) with the same public name that an
5722 error should be diagnosed.
5724 9. When Exim looked up a host name for an IP address, but failed to find the
5725 original IP address when looking up the host name (a safety check), it
5726 output the message "<ip address> does not match any IP for NULL", which was
5727 confusing, to say the least. The bug was that the host name should have
5728 appeared instead of "NULL".
5730 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5731 user that is built into the binary, and the -C or -D options is used, root
5732 privilege is dropped before the configuration file is read. In addition,
5733 logging is switched to stderr instead of the normal log files. If the
5734 configuration then re-defines the Exim user, the unprivileged environment
5735 is probably not what is expected, so Exim logs a panic warning message (but
5738 However, if deliver_drop_privilege is set, the unprivileged state may well
5739 be exactly what is intended, so the warning has been cut out in that case,
5740 and Exim is allowed to try to write to its normal log files.
5746 1. smtp_return_error_details was not giving details for temporary sender
5747 or receiver verification errors.
5749 2. Diagnose a configuration error if two authenticators have the same public
5752 3. Exim used not to create the message log file for a message until the first
5753 delivery attempt. This could be confusing when incoming messages were held
5754 for policy or load reasons. The message log file is now created at the time
5755 the message is received, and an initial "Received" line is written to it.
5757 4. The automatically generated man page for command line options had a minor
5758 bug that caused no ill effects; however, a more serious problem was that
5759 the procedure for building the man page automatically didn't always
5760 operate. Consequently, release 4.20 contains an out-of-date version. This
5761 shouldn't happen again.
5763 5. When building Exim with embedded Perl support, the script that builds the
5764 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5765 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5767 6. The freeze_tell option was not being used for messages that were frozen on
5768 arrival, either by an ACL or by local_scan().
5770 7. Added the smtp_incomplete_transaction log selector.
5772 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5773 was accepting AUTH without a new EHLO.
5775 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5776 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5779 10. Logging of TCP/IP connections (when configured) now happens in the main
5780 daemon process instead of the child process, so that the TCP/IP connection
5781 count is more accurate (but it can never be perfect).
5783 11. The use of "drop" in a nested ACL was not being handled correctly in the
5784 outer ACL. Now, if condition failure induced by the nested "drop" causes
5785 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5786 or "require"), the connection is dropped.
5788 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5789 that yield "discard" can now be used with an "accept" or a "discard" verb,
5790 but an error is generated for any others (because I can't see a useful way
5791 to define what should happen).
5793 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5794 are now processed in the same way as lines in the Exim configuration file.
5795 In particular, continuation lines are supported.
5797 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5799 15. Added -ti meaning -t -i.
5801 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5802 domains, and warn by logging if others are found.
5804 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5805 returned from GIFCONF when Exim is trying to find the list of interfaces on
5806 a host. The code in os.c has been modified to copy each ifreq to an aligned
5807 structure in all cases.
5809 Also, in some cases, the returned ifreq's were being copied to a 'struct
5810 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5811 means the last couple of bytes of an IPv6 address could be chopped if the
5812 ifreq contained only a normal sockaddr (14 bytes storage).
5814 18. Named domain lists were not supported in the hosts_treat_as_local option.
5815 An entry such as +xxxx was not recognized, and was treated as a literal
5818 19. Ensure that header lines added by a DATA ACL are included in the reject log
5819 if the ACL subsequently rejects the message.
5821 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5822 MD5 (which is deprecated).
5824 21. When testing a filter file using -bf, Exim was writing a message when it
5825 took the sender from a "From " line in the message, but it was not doing so
5826 when it took $return_path from a Return-Path: header line. It now does.
5828 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5829 with a valid header line field name (a series of printing characters
5830 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5832 23. Changed "disc" in the source to "disk" to conform to the documentation and
5833 the book and for uniformity.
5835 24. Ignore Sendmail's -Ooption=value command line item.
5837 25. When execve() failed while trying to run a command in a pipe transport,
5838 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5839 could be confused with a return value of 69 from the command itself. This
5840 has been changed to 127, the value the shell returns if it is asked to run
5841 a non-existent command. The wording for the related log line suggests a
5842 non-existent command as the problem.
5844 26. If received_header_text expands to an empty string, do not add a Received:
5845 header line to the message. (Well, it adds a token one on the spool, but
5846 marks it "old" so that it doesn't get used or transmitted.)
5848 27. Installed eximstats 1.28 (addition of -nt option).
5850 28. There was no check for failure on the call to getsockname() in the daemon
5851 code. This can fail if there is a shortage of resources on the system, with
5852 ENOMEM, for example. A temporary error is now given on failure.
5854 29. Contrary to the C standard, it seems that in some environments, the
5855 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5856 program. Exim now does this explicitly; it affects the formatting of
5857 timestamps using strftime().
5859 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5860 complaints. I've now initialized all the variables, to avoid this.
5862 32. Header lines added by a system filter were not being "seen" during
5863 transport-time rewrites.
5865 33. The info_callback() function passed to OpenSSL is set up with type void
5866 (*)(SSL *, int, int), as described somewhere. However, when calling the
5867 function (actually a macro) that sets it up, the type void(*)() is
5868 expected. I've put in a cast to prevent warnings from picky compilers.
5870 34. If a DNS black list lookup found a CNAME record, but there were no A
5871 records associated with the domain it pointed at, Exim crashed.
5873 35. If a DNS black list lookup returned more than one A record, Exim ignored
5874 all but the first. It now scans all returned addresses if a particular IP
5875 value is being sought. In this situation, the contents of the
5876 $dnslist_value variable are a list of all the addresses, separated by a
5879 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5880 to accept a host name and all its aliases if the forward lookup for any of
5881 them yielded the IP address of the incoming connection. Now it accepts only
5882 those names whose forward lookup yields the correct IP address. Any other
5883 names are discarded. This closes a loophole whereby a rogue DNS
5884 administrator could create reverse DNS records to break through a
5885 wildcarded host restriction in an ACL.
5887 37. If a user filter or a system filter that ran in a subprocess used any of
5888 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5889 the wrong values were passed to the pipe command ($thisaddress had the
5890 value of $0, $0 had the value of $1, etc). This bug was introduced by
5891 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5893 38. Improved the line breaking for long SMTP error messages from ACLs.
5894 Previously, if there was no break point between 40 and 75 characters, Exim
5895 left the rest of the message alone. Two changes have been made: (a) I've
5896 reduced the minimum length to 35 characters; (b) if it can't find a break
5897 point between 35 and 75 characters, it looks ahead and uses the first one
5898 that it finds. This may give the occasional overlong line, but at least the
5899 remaining text gets split now.
5901 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5902 file descriptors might be low, and that setting 1000 would always raise it.
5903 It turns out that in some environments, the limit is already over 1000 and
5904 that lowering it causes trouble. So now Exim takes care not to decrease it.
5906 40. When delivering a message, the value of $return_path is set to $sender_
5907 address at the start of routing (routers may change the value). By an
5908 oversight, this default was not being set up when an address was tested by
5909 -bt or -bv, which affected the outcome if any router or filter referred to
5912 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5913 log only when "message" or "log_message" are set. However, if one of the
5914 conditions was an address verification, or a call to a nested ACL, the
5915 messages generated by the underlying test were being passed through. This
5916 no longer happens. The underlying message is available in $acl_verify_
5917 message for both "message" and "log_message" expansions, so it can be
5918 passed through if needed.
5920 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5921 new expansion $bh_ to give the encoded byte string without charset
5922 translation. Translation happens only if iconv() is available; HAVE_ICONV
5923 indicates this at build time. HEADERS_CHARSET gives the charset to
5924 translate to; headers_charset can change it in the configuration, and
5925 "headers charset" can change it in an individual filter file.
5927 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5928 that creates RFC 2047 encoded "words" labels them as that charset instead
5929 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5930 expansion operator; (ii) when Exim creates a From: line for a local
5931 message; (iii) when a header line is rewritten to include a "phrase" part.
5933 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5934 buggy, causing it to skip the first lines of messages whose message ID
5935 ended in 'D'. This would not have bitten before Exim release 4.14, because
5936 message IDs were unlikely to end in 'D' before then. The effect was to have
5937 incorrect size information for certain domains.
5939 45. #include "config.h" was missing at the start of the crypt16.c module. This
5940 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5943 46. If there was a timeout during a "random" callout check, Exim treated it as
5944 a failure of the random address, and carried on sending RSET and the real
5945 address. If the delay was just some slowness somewhere, the response to the
5946 original RCPT would be taken as a response to RSET and so on, causing
5947 mayhem of various kinds.
5949 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5950 when I implemented it. It didn't allow for the fact that some option values
5951 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5952 the right test for positive values.
5954 48. Domain names in DNS records are case-independent. Exim always looks them up
5955 in lower case. Some resolvers return domain names in exactly the case they
5956 appear in the zone file, that is, they may contain uppercase letters. Not
5957 all resolvers do this - some return always lower case. Exim was treating a
5958 change of case by a resolver as a change of domain, similar to a widening
5959 of a domain abbreviation. This triggered its re-routing code and so it was
5960 trying to route what was effectively the same domain again. This normally
5961 caused routing to fail (because the router wouldn't handle the domain
5962 twice). Now Exim checks for this case specially, and just changes the
5963 casing of the domain that it ultimately uses when it transmits the message
5966 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5969 50. If a filter generated a file delivery with a non-absolute name (possible if
5970 no home directory exists for the router), the forbid_file option was not
5973 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5974 the existing equality matching.
5976 52. Exim was using ints instead of ino_t variables in some places where it was
5977 dealing with inode numbers.
5979 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5980 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5981 called TMPDIR, and if it finds it is different, it changes its value.
5983 54. The smtp_printf() function is now made available to local_scan() so
5984 additional output lines can be written before returning. There is also an
5985 smtp_fflush() function to enable the detection of a dropped connection.
5986 The variables smtp_input and smtp_batched_input are exported to
5989 55. Changed the default runtime configuration: the message "Unknown user"
5990 has been removed from the ACL, and instead placed on the localuser router,
5991 using the cannot_route_message feature. This means that any verification
5992 failures that generate their own messages won't get overridden. Similarly,
5993 the "Unrouteable address" message that was in the ACL for unverifiable
5994 relay addresses has also been removed.
5996 56. Added hosts_avoid_esmtp to the smtp transport.
5998 57. The exicyclog script was not checking for the esoteric option
5999 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6000 will work only if exicyclog is run under the appropriate euid.
6002 58. Following a discussion on the list, the rules by which Exim recognises line
6003 endings on incoming messages have been changed. The -dropcr and drop_cr
6004 options are now no-ops, retained only for backwards compatibility. The
6005 following line terminators are recognized: LF CRLF CR. However, special
6006 processing applies to CR:
6008 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6009 nor a local message in the state where . is a terminator.
6011 (ii) If a bare CR is encountered in a header line, an extra space is added
6012 after the line terminator so as not to end the header. The reasoning
6013 behind this is that bare CRs in header lines are most likely either
6014 to be mistakes, or people trying to play silly games.
6016 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6017 was being incorrectly given as 18 bytes larger than it should have been.
6018 This is a VOB (very old bug).
6020 60. This may never have affected anything current, but just in case it has:
6021 When the local host is found other than at the start of a list of hosts,
6022 the local host, those with the same MX, and any that follow, are discarded.
6023 When the list in question was part of a longer list of hosts, the following
6024 hosts (not currently being processed) were also being discarded. This no
6025 longer happens. I'm not sure if this situation could ever has previously
6028 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6031 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6032 header. This is something that is recommended in a new Internet Draft, and
6033 is something that is documented as being done by Sendmail. There are two
6034 possible values. For messages generated by the autoreply transport, Exim
6037 Auto-Submitted: auto-replied
6039 whereas for all other generated messages (e.g. bounces) it adds
6041 Auto-Submitted: auto-generated
6043 63. The "personal" condition in filters now includes a test for the
6044 Auto-Submitted: header. If it contains the string "auto-" the message it
6045 not considered personal.
6047 64. Added rcpt_include_affixes as a generic transport option.
6049 65. Added queue_only_override (default true).
6051 66. Added the syslog_duplication option.
6053 67. If what should have been the first header line of a message consisted of
6054 a space followed by a colon, Exim was mis-interpreting it as a header line.
6055 It isn't of course - it is syntactically invalid and should therefore be
6056 treated as the start of the message body. The misbehaviour could have
6057 caused a number of strange effects, including loss of data in subsequent
6058 header lines, and spool format errors.
6060 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6061 client host had authenticated. This control can now be exercised by an ACL
6062 for more flexibility.
6064 69. By default, callouts do not happen when testing with -bh. There is now a
6065 variant, -bhc, which does actually run the callout code, including
6066 consulting and updating the callout cache.
6068 70. Added support for saslauthd authentication, courtesy of Alexander
6071 71. If statvfs() failed on the spool or log directories while checking their
6072 size for availability, Exim confusingly gave the error "space shortage".
6073 Furthermore, in debugging mode it crashed with a floating point exception.
6074 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6075 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6076 serious problem, Exim now writes to the main and panic logs when this
6077 happens, with details of the failure. It then refuses to accept the
6078 incoming message, giving the message "spool directory problem" or "log
6079 directory problem" with a 421 code for SMTP messages.
6081 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6082 0, 1, and 2 exist, because some OS complain for execs without them (see
6083 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6084 descriptors. However, the code omitted to check that the open succeeded,
6085 causing mysterious errors if for some reason the permissions on /dev/null
6086 got screwed. Now Exim writes a message to the main and panic logs, and
6087 bombs out if it can't open /dev/null.
6089 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6090 interact so that it is all more flexible. It is supposed to remain
6091 backwards compatible. Also added extra_local_interfaces.
6093 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6094 to bomb out with an assertion failure - to the client this appears as a
6095 connection drop. This problem occurs in the part of the code that was taken
6096 from the Samba project. Fortunately, the assertion is in a very simple
6097 function, so I have fixed this by reproducing the function inline in the
6098 one place where it is called, and arranging for authentication to fail
6099 instead of killing the process with assert().
6101 75. The SPA client code was not working when the server requested OEM rather
6102 than Unicode encoding.
6104 76. Added code to make require_files with a specific uid setting more usable in
6105 the case where statting the file as root fails - usually a non-root-mounted
6106 NFS file system. When this happens and the failure is EACCES, Exim now
6107 forks a subprocess and does the per-uid checking as the relevant uid.
6109 77. Added process_log_path.
6111 78. If log_file_path was not explicitly set, a setting of check_log_space or
6112 check_log_inodes was ignored.
6114 79. If a space check for the spool or log partitions fails, the incident is now
6115 logged. Of course, in the latter case the data may get lost...
6117 80. Added the %p formatting code to string_format() so that it can be used to
6118 print addresses in debug_print(). Adjusted all the address printing in the
6119 debugging in store.c to use %p rather than %d.
6121 81. There was a concern that a line of code in smtp_in.c could overflow a
6122 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6123 initially expressed, the concern was not well-founded, because trailing
6124 spaces are removed early. However, if the trailing spaces were followed by
6125 a NULL, they did not get removed, so the overflow was possible. Two fixes
6128 (a) I re-wrote the offending code in a cleaner fashion.
6129 (b) If an incoming SMTP command contains a NULL character, it is rejected
6132 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6133 runs initgroups(), so that if the Exim user is in any additional groups,
6134 they will be used during message reception.
6140 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob