1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
12 than -2003 did; needs libidn2 in addition to libidn.
14 JH/02 The path option on a pipe transport is now expanded before use.
16 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
17 Patch provided by "Björn", documentation fix added too.
19 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
20 missing a wire-to-host endian conversion.
22 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
23 close after a BDAT command line could be taken as a following command,
24 giving a synch failure. Fix by only checking for synch immediately
25 before acknowledging the chunk.
27 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
28 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
29 macro. Patches provided by Josh Soref.
31 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
32 Previously we did not; the RFC seems ambiguous and VRFY is not listed
33 by IANA as a service extension. However, John Klensin suggests that we
36 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
37 the dkim code may be unix-mode line endings rather than smtp wire-format
38 CRLF, so prepend a CR to any bare LF.
40 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
41 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
43 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
44 the first were themselves being wrongly included in the feed into dkim
45 processing; with most chunk sizes in use this resulted in an incorrect
46 body hash calculated value.
48 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
49 DKIM signature block, for verification. Although advised against by
50 standards it is specifically not ruled illegal.
52 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
54 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
55 missing a body hash (the bh= tag).
57 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
58 It seems that HAProxy sends the Proxy Protocol information in clear and
59 only then does a TLS startup, so do the same.
61 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
62 TCP connections (such as for Spamd) unless the daemon successfully set
63 Fast Open mode on its listening sockets. This fixes breakage seen on
64 too-old kernels or those not configured for Fast Open, at the cost of
65 requiring both directions being enabled for TFO, and TFO never being used
66 by non-daemon-related Exim processes.
68 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
69 endings, at least on the first header line. Try to canonify any that get
70 past that check, despite the cost.
72 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
73 now limited to an arbitrary five deep, while parsing addresses with the
74 strip_excess_angle_brackets option enabled.
80 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
81 supports it and a size is available (ie. the sending peer gave us one).
83 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
84 years of being deprecated). The replacements are the ACLs
85 acl_smtp_mime and acl_not_smtp_mime.
87 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
88 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
89 or even an in-clear connection were permitted. Now, if the host lookup
90 was dnssec and dane was requested then the host is only used if the
91 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
92 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
93 if one fails this test.
94 This means that a poorly-configured remote DNS will make it incommunicado;
95 but it protects against a DNS-interception attack on it.
97 JH/04 Bug 1810: make continued-use of an open smtp transport connection
98 non-noisy when a race steals the message being considered.
100 JH/05 If main configuration option tls_certificate is unset, generate a
101 self-signed certificate for inbound TLS connections.
103 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
104 in rewrites and routers.
106 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
107 and logged a warning sing 4.83; now they are a configuration file error.
109 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
110 (lacking @domain). Apply the same qualification processing as RCPT.
112 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
114 JH/10 Support ${sha256:} applied to a string (as well as the previous
117 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
118 a cutthrough deliver is pending, as we always want to make a connection.
119 This also avoids re-routing the message when later placing the cutthrough
120 connection after a verify cache hit.
121 Do not update it with the verify result either.
123 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
124 when routing results in more than one destination address.
126 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
127 signing (which inhibits the cutthrough capability). Previously only
128 the presence of an option was tested; now an expansion evaluating as
129 empty is permissible (obviously it should depend only on data available
130 when the cutthrough connection is made).
132 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
133 the relevant preceding SMTP command did not note the pipelining mode.
135 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
136 Previously they were not counted.
138 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
139 as one having no matching records. Previously we deferred the message
140 that needed the lookup.
142 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
143 distinguished as "(=".
145 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
146 for missing MX records. Previously it only worked for missing A records.
148 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
150 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
151 after the data-go-ahead and data-ack. Patch from Jason Betts.
153 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
154 even for a "none" policy. Patch from Tony Meyer.
156 JH/22 Fix continued use of a connection for further deliveries. If a port was
157 specified by a router, it must also match for the delivery to be
160 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
161 When one of the recipients of a message was unsuitable for the connection
162 (has no matching addresses), we lost track of needing to mark it
163 deferred. As a result mail would be lost.
165 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
167 JH/25 Decoding ACL controls is now done using a binary search; the source code
168 takes up less space and should be simpler to maintain. Merge the ACL
169 condition decode tables also, with similar effect.
171 JH/26 Fix problem with one_time used on a redirect router which returned the
172 parent address unchanged. A retry would see the parent address marked as
173 delivered, so not attempt the (identical) child. As a result mail would
176 JH/27 Fix a possible security hole, wherein a process operating with the Exim
177 UID can gain a root shell. Credit to http://www.halfdog.net/ for
178 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
181 JH/28 Enable {spool,log} filesystem space and inode checks as default.
182 Main config options check_{log,spool}_{inodes,space} are now
183 100 inodes, 10MB unless set otherwise in the configuration.
185 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
186 Previously it only applied to the main-section connection policy
189 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
191 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
192 by me. Added RFC7919 DH primes as an alternative.
194 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
195 are not in the system include path.
197 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
198 GnuTLS, when a session startup failed (eg because the client disconnected)
199 Exim did stdio operations after fclose. This was exposed by a recent
200 change which nulled out the file handle after the fclose.
202 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
203 signed directly by the cert-signing cert, rather than an intermediate
204 OCSP-signing cert. This is the model used by LetsEncrypt.
206 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
208 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
209 an incoming connection.
211 HS/02 Bug 1802: Do not half-close the connection after sending a request
214 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
215 fallback to "prime256v1".
217 JH/34 SECURITY: Use proper copy of DATA command in error message.
218 Could leak key material. Remotely exploitable. CVE-2016-9963.
224 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
225 and 3.4.4 - once the server is enabled to respond to an OCSP request
226 it does even when not requested, resulting in a stapling non-aware
227 client dropping the TLS connection.
229 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
230 support variable-length bit vectors. No functional change.
232 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
233 The I= interface field on outgoing lines is now after the H= remote
234 host field, same as incoming lines. There is a separate
235 outgoing_interface log selector which allows you to disable the
238 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
239 If not running log_selector +smtp_connection the mainlog would be held
240 open indefinitely after a "too many connections" event, including to a
241 deleted file after a log rotate. Leave the per net connection logging
242 leaving it open for efficiency as that will be quickly detected by the
243 check on the next write.
245 HS/01 Bug 1671: Fix post transport crash.
246 Processing the wait-<transport> messages could crash the delivery
247 process if the message IDs didn't exist for some reason. When
248 using 'split_spool_directory=yes' the construction of the spool
249 file name failed already, exposing the same netto behaviour.
251 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
252 mime_regex ACL conditions.
254 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
255 to DSN fail messages (bounces): remote IP, remote greeting, remote response
256 to HELO, local diagnostic string.
258 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
259 log line to debug. Even when configured with a tls authenticator many
260 client connections are expected to not authenticate in this way, so
261 an authenticate fail is not an error.
263 HS/02 Add the Exim version string to the process info. This way exiwhat
264 gives some more detail about the running daemon.
266 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
267 matter for fast-change records such as DNSBLs.
269 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
270 retry record, even if constant. There may be multiple transports with
271 different interface settings and the retry behaviour needs to be kept
274 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
276 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
278 JH/10 Bug 840: fix log_defer_output option of pipe transport
280 JH/11 Bug 830: use same host for all RCPTS of a message, even under
281 hosts_randomize. This matters a lot when combined with mua_wrapper.
283 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
284 ${quote_pgsql:<string>} operator.
286 JH/13 Bug 1708: avoid misaligned access in cached lookup.
288 JH/14 Change header file name for freeradius-client. Relevant if compiling
289 with Radius support; from the Gentoo tree and checked under Fedora.
291 JH/15 Bug 1712: Introduce $prdr_requested flag variable
293 JH/16 Bug 1714: Permit an empty string as expansion result for transport
294 option transport_filter, meaning no filtering.
296 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
298 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
299 defaults to "*" (all hosts). The variable is now available when not built
300 with TLS, default unset, mainly to enable keeping the testsuite sane.
301 If a server certificate is not supplied (via tls_certificate) an error is
302 logged, and clients will find TLS connections fail on startup. Presumably
303 they will retry in-clear.
304 Packagers of Exim are strongly encouraged to create a server certificate
305 at installation time.
307 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
308 with the $config_file variable.
310 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
311 in transport context, after the attempt, and per-recipient. The latter type
312 is per host attempted. The event data is the error message, and the errno
313 information encodes the lookup type (A vs. MX) used for the (first) host,
314 and the trailing two digits of the smtp 4xx response.
316 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
317 to write to mainlog (or rejectlog, paniclog) in the window between file
318 creation and permissions/ownership being changed. Particularly affects
319 installations where exicyclog is run as root, rather than exim user;
320 result is that the running daemon panics and dies.
322 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
324 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
325 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
326 "pri" and "weight". Note that the previous implicit priority given by the
327 list order is no longer honoured.
329 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
332 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
333 by defining SUPPORT_SOCKS.
335 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
336 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
337 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
338 variables are renamed to proxy_{local,external}_{address,port}.
340 JH/27 Move Internationalisation support from Experimental to mainline, enabled
341 for a build by defining SUPPORT_I18N
343 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
344 of the query string, and make ${quote_redis:} do that quoting.
346 JH/29 Move Events support from Experimental to mainline, enabled by default
347 and removable for a build by defining DISABLE_EVENT.
349 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
351 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
352 cached by the daemon.
354 JH/32 Move Redis support from Experimental to mainline, enabled for a build
355 by defining LOOKUP_REDIS. The libhiredis library is required.
357 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
358 keys are given for lookup.
360 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
361 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
362 only supported when built with TLS support. The PolarSSL SHA routines
363 are still used when the TLS library is too old for convenient support.
365 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
366 openssl_options), for security. OpenSSL forces this from version 1.1.0
367 server-side so match that on older versions.
369 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
370 allocation for $value could be released as the expansion processing
371 concluded, but leaving the global pointer active for it.
373 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
374 and to use the domains and local_parts ACL conditions.
376 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
377 incorrectly not doubled on cutthrough transmission, hence seen as a
378 body-termination at the receiving system - resulting in truncated mails.
379 Commonly the sender saw a TCP-level error, and retransmitted the message
380 via the normal store-and-forward channel. This could result in duplicates
381 received - but deduplicating mailstores were liable to retain only the
382 initial truncated version.
384 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
386 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
388 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
389 we're in there, support oversigning also; bug 1309.
391 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
393 HS/04 Add support for keep_environment and add_environment options.
395 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
396 either intentional arithmetic overflow during PRNG, or testing config-
399 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
400 delivery resulted in actual delivery. Cancel cutthrough before DATA
403 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
404 rejects a recipient: pass the reject to the originator.
406 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
407 Many were false-positives and ignorable, but it's worth fixing the
410 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
411 for the new environment-manipulation done at startup. Move the routines
412 from being local to tls.c to being global via the os.c file.
414 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
415 an extract embedded as result-arg for a map, the first arg for extract
416 is unavailable so we cannot tell if this is a numbered or keyed
417 extraction. Accept either.
423 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
426 JH/02 The smtp transport option "multi_domain" is now expanded.
428 JH/03 The smtp transport now requests PRDR by default, if the server offers
431 JH/04 Certificate name checking on server certificates, when exim is a client,
432 is now done by default. The transport option tls_verify_cert_hostnames
433 can be used to disable this per-host. The build option
434 EXPERIMENTAL_CERTNAMES is withdrawn.
436 JH/05 The value of the tls_verify_certificates smtp transport and main options
437 default to the word "system" to access the system default CA bundle.
438 For GnuTLS, only version 3.0.20 or later.
440 JH/06 Verification of the server certificate for a TLS connection is now tried
441 (but not required) by default. The verification status is now logged by
442 default, for both outbound TLS and client-certificate supplying inbound
445 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
448 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
449 Status Notification (bounce) messages are now MIME format per RFC 3464.
450 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
451 under the control of the dsn_advertise_hosts option, and routers may
452 have a dsn_lasthop option.
454 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
455 default, modifiable by a malware= option. The list separator for
456 the options can now be changed in the usual way. Bug 68.
458 JH/10 The smtp_receive_timeout main option is now expanded before use.
460 JH/11 The incoming_interface log option now also enables logging of the
461 local interface on delivery outgoing connections.
463 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
464 if the interface and destination host and port all match.
466 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
469 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
470 Patch from Andrew Lewis.
472 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
473 now supports optional time-restrictions, weighting, and priority
474 modifiers per server. Patch originally by <rommer@active.by>.
476 JH/16 The spamd_address main option now supports a mixed list of local
477 and remote servers. Remote servers can be IPv6 addresses, and
478 specify a port-range.
480 JH/17 Bug 68: The spamd_address main option now supports an optional
481 timeout value per server.
483 JH/18 Bug 1581: Router and transport options headers_add/remove can
484 now have the list separator specified.
486 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
489 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
492 JH/21 Support for the A6 type of dns record is withdrawn.
494 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
495 rather than the verbs used.
497 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
498 from 255 to 1024 chars.
500 JH/24 Verification callouts now attempt to use TLS by default.
502 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
503 are generic router options now. The defaults didn't change.
505 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
506 Original patch from Alexander Shikoff, worked over by JH.
508 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
509 files if ZCAT_COMMAND is not executable.
511 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
513 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
515 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
516 Normally benign, it bites when the pair was led to by a CNAME;
517 modern usage is to not canonicalize the domain to a CNAME target
518 (and we were inconsistent anyway for A-only vs AAAA+A).
520 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
522 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
523 when evaluating $sender_host_dnssec.
525 JH/31 Check the HELO verification lookup for DNSSEC, adding new
526 $sender_helo_dnssec variable.
528 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
530 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
532 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
534 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
535 documented as working, but never had. Support all but $spam_report.
537 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
538 added for tls authenticator.
540 HS/03 Add perl_taintmode main config option
546 TL/01 When running the test suite, the README says that variables such as
547 no_msglog_check are global and can be placed anywhere in a specific
548 test's script, however it was observed that placement needed to be near
549 the beginning for it to behave that way. Changed the runtest perl
550 script to read through the entire script once to detect and set these
551 variables, reset to the beginning of the script, and then run through
552 the script parsing/test process like normal.
554 TL/02 The BSD's have an arc4random API. One of the functions to induce
555 adding randomness was arc4random_stir(), but it has been removed in
556 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
557 function when detected.
559 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
560 cause callback expansion.
562 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
563 syntax errors in an expansion can be treated as a string instead of
564 logging or causing an error, due to the internal use of bool_lax
565 instead of bool when processing it.
567 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
568 server certificates when making smtp deliveries.
570 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
572 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
574 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
576 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
577 Merged patch from Sebastian Wiedenroth.
579 JH/05 Fix results-pipe from transport process. Several recipients, combined
580 with certificate use, exposed issues where response data items split
581 over buffer boundaries were not parsed properly. This eventually
582 resulted in duplicates being sent. This issue only became common enough
583 to notice due to the introduction of connection certificate information,
584 the item size being so much larger. Found and fixed by Wolfgang Breyha.
586 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
587 size buffer was used, resulting in syntax errors when an expansion
590 JH/07 Add support for directories of certificates when compiled with a GnuTLS
591 version 3.3.6 or later.
593 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
594 is EXPERIMENTAL_EVENT, the main-configuration and transport options
595 both become "event_action", the variables become $event_name, $event_data
596 and $event_defer_errno. There is a new variable $verify_mode, usable in
597 routers, transports and related events. The tls:cert event is now also
598 raised for inbound connections, if the main configuration event_action
601 TL/06 In test suite, disable OCSP for old versions of openssl which contained
602 early OCSP support, but no stapling (appears to be less than 1.0.0).
604 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
605 server certificate names available under the smtp transport option
606 "tls_verify_cert_hostname" now do not permit multi-component wildcard
609 JH/10 Time-related extraction expansions from certificates now use the main
610 option "timezone" setting for output formatting, and are consistent
611 between OpenSSL and GnuTLS compilations. Bug 1541.
613 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
614 encoded parameter in the incoming message. Bug 1558.
616 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
617 include certificate info, eximon was claiming there were spoolfile
620 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
622 JH/14 Log delivery-related information more consistently, using the sequence
623 "H=<name> [<ip>]" wherever possible.
625 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
626 are problematic for Debian distribution, omit them from the release
629 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
631 JH/16 Fix string representation of time values on 64bit time_t architectures.
634 JH/17 Fix a null-indirection in certextract expansions when a nondefault
635 output list separator was used.
640 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
641 checkers that were complaining about end of non-void function with no
644 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
645 This was a regression introduced in 4.83 by another bugfix.
647 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
649 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
650 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
656 TF/01 Correctly close the server side of TLS when forking for delivery.
658 When a message was received over SMTP with TLS, Exim failed to clear up
659 the incoming connection properly after forking off the child process to
660 deliver the message. In some situations the subsequent outgoing
661 delivery connection happened to have the same fd number as the incoming
662 connection previously had. Exim would try to use TLS and fail, logging
663 a "Bad file descriptor" error.
665 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
666 utilities have not been installed.
668 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
669 temporary space as the ACL may create new global variables.
671 TL/01 LDAP support uses per connection or global context settings, depending
672 upon the detected version of the libraries at build time.
674 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
675 to extract and use the src ip:port in logging and expansions as if it
676 were a direct connection from the outside internet. PPv2 support was
677 updated based on HAProxy spec change in May 2014.
679 JH/02 Add ${listextract {number}{list}{success}{fail}}.
681 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
682 Properly escape header and check for NULL return.
684 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
687 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
689 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
690 characters in header names, implemented as a verify condition.
691 Contributed by Michael Fischer v. Mollard.
693 TL/05 Rename SPF condition results err_perm and err_temp to standardized
694 results permerror and temperror. Previous values are deprecated but
695 still accepted. In a future release, err_perm and err_temp will be
696 completely removed, which will be a backward incompatibility if the
697 ACL tests for either of these two old results. Patch contributed by
698 user bes-internal on the mailing list.
700 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
702 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
703 selectors, in both main and reject logs.
705 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
708 JH/07 Add malware type "sock" for talking to simple daemon.
710 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
712 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
713 routers/transports under cutthrough routing.
715 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
716 numbers. Touch up "bool" conditional to keep the same definition.
718 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
720 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
722 JH/12 Expand items in router/transport headers_add or headers_remove lists
723 individually rather than the list as a whole. Bug 1452.
725 Required for reasonable handling of multiple headers_ options when
726 they may be empty; requires that headers_remove items with embedded
727 colons must have them doubled (or the list-separator changed).
729 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
730 view the policy declared in the DMARC record. Currently, $dmarc_status
731 is a combined value of both the record presence and the result of the
734 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
736 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
737 dnslookup router and the smtp transport (applying to the forward
740 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
741 of ldap servers used for a specific lookup. Patch provided by Heiko
744 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
745 New variable $lookup_dnssec_authenticated for observability.
747 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
748 Patch submitted by Lars Timman.
750 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
752 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
753 Requires trusted mode and valid format message id, aborts otherwise.
754 Patch contributed by Heiko Schlichting.
756 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
757 certextract with support for various fields. Bug 1358.
759 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
760 is requested by default, modifiable by smtp transport option
763 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
764 operate on certificate variables to give certificate fingerprints
765 Also new ${sha256:cert_variable}.
767 JH/23 The PRDR feature is moved from being Experimental into the mainline.
769 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
770 Christian Aistleitner.
772 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
774 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
775 file. Patch from Wolfgang Breyha.
777 JH/25 Expand the coverage of the delivery $host and $host_address to
778 client authenticators run in verify callout. Bug 1476.
780 JH/26 Port service names are now accepted for tls_on_connect_ports, to
781 align with daemon_smtp_ports. Bug 72.
783 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
784 support and error reporting did not work properly.
786 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
787 and is readable. Patch from Andrew Colin Kissa.
789 TL/14 Enhance documentation of ${run expansion and how it parses the
790 commandline after expansion, particularly in the case when an
791 unquoted variable expansion results in an empty value.
793 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
795 PP/02 Fix internal collision of T_APL on systems which support RFC3123
796 by renaming away from it. Addresses GH issue 15, reported by
799 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
801 TL/15 SECURITY: prevent double expansion in math comparison functions
802 (can expand unsanitized data). Not remotely exploitable.
809 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
811 PP/02 Make -n do something, by making it not do something.
812 When combined with -bP, the name of an option is not output.
814 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
817 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
818 $sender_host_name and config options to manage this, and basic check
821 PP/05 DSCP support for outbound connections and control modifier for inbound.
823 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
824 (Only plugin which currently uses this is kerberos4, which nobody should
825 be using, but we should make it available and other future plugins might
826 conceivably use it, even though it would break NAT; stuff *should* be
827 using channel bindings instead).
829 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
830 name; added for Sendmail compatibility; requires admin caller.
831 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
832 just ignore it); requires trusted caller.
833 Also parse but ignore: -Ac -Am -X<logfile>
836 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
838 TL/02 Add +smtp_confirmation as a default logging option.
840 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
841 Patch by Magnus Holmgren from 2007-02-20.
843 TL/04 Bugzilla 1281 - Spec typo.
844 Bugzilla 1283 - Spec typo.
845 Bugzilla 1290 - Spec grammar fixes.
847 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
849 TL/06 Add Experimental DMARC support using libopendmarc libraries.
851 TL/07 Fix an out of order global option causing a segfault. Reported to dev
852 mailing list by by Dmitry Isaikin.
854 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
856 JH/02 Support "G" suffix to numbers in ${if comparisons.
858 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
860 NM/01 Bugzilla 1197 - Spec typo
861 Bugzilla 1196 - Spec examples corrections
863 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
865 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
866 gnutls_enable_pkcs11, but renamed to more accurately indicate its
869 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
870 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
872 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
873 "acl {{name}{arg}...}", and optional args on acl condition
876 JH/05 Permit multiple router/transport headers_add/remove lines.
878 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
880 JH/07 Avoid using a waiting database for a single-message-only transport.
881 Performance patch from Paul Fisher. Bugzilla 1262.
883 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
886 JH/09 Add $headers_added variable, with content from use of ACL modifier
887 add_header (but not yet added to the message). Bugzilla 199.
889 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
890 Pulled from Bugzilla 817 by Wolfgang Breyha.
892 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
894 (nb: this is the same fix as in Exim 4.80.1)
896 JH/11 Add A= logging on delivery lines, and a client_set_id option on
899 JH/12 Add optional authenticated_sender logging to A= and a log_selector
902 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
904 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
905 advertise SMTP AUTH mechanism to us, instead of a generic
906 protocol violation error. Also, make Exim more robust to bad
907 data from the Dovecot auth socket.
909 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
911 When a queue runner is handling a message, Exim first routes the
912 recipient addresses, during which it prunes them based on the retry
913 hints database. After that it attempts to deliver the message to
914 any remaining recipients. It then updates the hints database using
917 So if a recipient address works intermittently, it can get repeatedly
918 deferred at routing time. The retry hints record remains fresh so the
919 address never reaches the final cutoff time.
921 This is a fairly common occurrence when a user is bumping up against
922 their storage quota. Exim had some logic in its local delivery code
923 to deal with this. However it did not apply to per-recipient defers
924 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
926 This change adds a proper retry rule check during routing so that the
927 final cutoff time is checked against the message's age. We only do
928 this check if there is an address retry record and there is not a
929 domain retry record; this implies that previous attempts to handle
930 the address had the retry_use_local_parts option turned on. We use
931 this as an approximation for the destination being like a local
932 delivery, as in LMTP.
934 I suspect this new check makes the old local delivery cutoff check
935 redundant, but I have not verified this so I left the code in place.
937 TF/02 Correct gecos expansion when From: is a prefix of the username.
939 Test 0254 submits a message to Exim with the header
943 When I ran the test suite under the user fanf2, Exim expanded
944 the header to contain my full name, whereas it should have added
945 a Resent-Sender: header. It erroneously treats any prefix of the
946 username as equal to the username.
948 This change corrects that bug.
950 GF/01 DCC debug and logging tidyup
951 Error conditions log to paniclog rather than rejectlog.
952 Debug lines prefixed by "DCC: " to remove any ambiguity.
954 TF/03 Avoid unnecessary rebuilds of lookup-related code.
956 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
957 Bug spotted by Jeremy Harris; was flawed since initial commit.
958 Would have resulted in OCSP responses post-SNI triggering an Exim
959 NULL dereference and crash.
961 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
963 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
964 Bug detection, analysis and fix by Samuel Thibault.
965 Bugzilla 1331, Debian bug #698092.
967 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
969 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
970 Server implementation by Todd Lyons, client by JH.
971 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
972 config variable "prdr_enable" controls whether the server
973 advertises the facility. If the client requests PRDR a new
974 acl_data_smtp_prdr ACL is called once for each recipient, after
975 the body content is received and before the acl_smtp_data ACL.
976 The client is controlled by both of: a hosts_try_prdr option
977 on the smtp transport, and the server advertisement.
978 Default client logging of deliveries and rejections involving
979 PRDR are flagged with the string "PRDR".
981 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
982 fclose(). Diagnosis by Todd Lyons.
984 PP/17 Update configure.default to handle IPv6 localhost better.
985 Patch by Alain Williams (plus minor tweaks).
988 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
989 This is now consistent with GnuTLS, and is now documented: the
990 previous undocumented portable approach to treating the option as
991 unset was to force an expansion failure. That still works, and
992 an empty string is now equivalent.
994 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
995 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
996 not performing validation itself.
998 PP/20 Added force_command boolean option to pipe transport.
999 Patch from Nick Koston, of cPanel Inc.
1001 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1004 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1006 PP/21 Fix eximon continuous updating with timestamped log-files.
1007 Broken in a format-string cleanup in 4.80, missed when I repaired the
1008 other false fix of the same issue.
1009 Report and fix from Heiko Schlichting.
1012 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1013 Report from Prashanth Katuri.
1015 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1016 It's SecureTransport, so affects any MacOS clients which use the
1017 system-integrated TLS libraries, including email clients.
1019 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1020 using a MIME ACL for non-SMTP local injection.
1021 Report and assistance in diagnosis by Warren Baker.
1023 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1025 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1027 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1028 last id that failed so it may be referenced in subsequent ACL's.
1030 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1033 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1034 ldap library initialization, allowing self-signed CA's to be
1035 used. Also properly sets require_cert option later in code by
1036 using NULL (global ldap config) instead of ldap handle (per
1037 session). Bug diagnosis and testing by alxgomz.
1039 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1040 the src/util/ subdirectory.
1042 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1043 renamed to Transport Post Delivery Action by Jeremy Harris, as
1046 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1047 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1048 redis_servers = needs to be configured which will be used by the redis
1049 lookup. Patch from Warren Baker, of The Packet Hub.
1051 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1053 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1054 hostname or reverse DNS when processing a host list. Used suggestions
1055 from multiple comments on this bug.
1057 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1059 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1060 Missed a few lines, added it to make the runtest require no keyboard
1063 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1064 contains upper case chars. Make router use caseful_local_part.
1066 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1067 support when GnuTLS has been built with p11-kit.
1073 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1075 This, or similar/improved, will also be change PP/11 of 4.82.
1081 PP/01 Handle short writes when writing local log-files.
1082 In practice, only affects FreeBSD (8 onwards).
1083 Bugzilla 1053, with thanks to Dmitry Isaikin.
1085 NM/01 Bugzilla 949 - Documentation tweak
1087 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1090 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1092 PP/02 Implemented gsasl authenticator.
1094 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1096 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1097 `pkg-config foo` for cflags/libs.
1099 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1100 with rest of GSASL and with heimdal_gssapi.
1102 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1103 `pkg-config foo` for cflags/libs for the TLS implementation.
1105 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1106 properties get this fed in as external SSF. A number of robustness
1107 and debugging improvements to the cyrus_sasl authenticator.
1109 PP/08 cyrus_sasl server now expands the server_realm option.
1111 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1112 Patch by Jeremy Harris.
1114 PP/10 Added dbmjz lookup type.
1116 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1118 PP/12 MAIL args handles TAB as well as SP, for better interop with
1119 non-compliant senders.
1120 Analysis and variant patch by Todd Lyons.
1122 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1123 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1124 Patch from Dirk Mueller <dmueller@suse.com>
1126 PP/13 tls_peerdn now print-escaped for spool files.
1127 Observed some $tls_peerdn in wild which contained \n, which resulted
1128 in spool file corruption.
1130 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1131 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1132 or write after TLS renegotiation, which otherwise led to messages
1135 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1136 as a tracking header (ie: a signed header comes before the signature).
1137 Patch from Wolfgang Breyha.
1139 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1140 comma-sep list; embedded commas doubled.
1142 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1144 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1146 Report and patch from Dmitry Banschikov.
1148 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1149 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1150 failures. We appear to now support TLS1.1+ with Exim.
1152 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1153 lets Exim select keys and certificates based upon TLS SNI from client.
1154 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1155 before an outbound SMTP session. New log_selector, +tls_sni.
1157 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1158 NULL dereference. Report and patch from Alun Jones.
1160 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1161 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1162 Not seeing resolver debug output on NetBSD, but suspect this is a
1163 resolver implementation change.
1165 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1166 Left warnings. Added "eximon gdb" invocation mode.
1168 PP/21 Defaulting "accept_8bitmime" to true, not false.
1170 PP/22 Added -bw for inetd wait mode support.
1172 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1173 locate the relevant includes and libraries. Made this the default.
1175 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1176 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1178 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1179 This may cause build issues on older platforms.
1181 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1182 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1183 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1184 Added SNI support via GnuTLS too.
1185 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1187 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1189 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1190 Applied second patch from Janne, implementing suggestion to default
1191 multiple-strings-in-record handling to match SPF spec.
1193 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1195 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1196 read-only, out of scope).
1197 Patch from Wolfgang Breyha, report from Stuart Northfield.
1199 PP/29 Fix three issues highlighted by clang analyser static analysis.
1200 Only crash-plausible issue would require the Cambridge-specific
1201 iplookup router and a misconfiguration.
1202 Report from Marcin Mirosław.
1204 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1206 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1207 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1208 As part of this, removing so much warning spew let me fix some minor
1209 real issues in debug logging.
1211 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1212 assignment on my part. Fixed.
1214 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1215 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1216 Janne Snabb (who went above and beyond: thank you).
1218 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1219 string otherwise requires a connection and a bunch more work and it's
1220 relatively easy to get wrong. Should also expose TLS library linkage
1223 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1224 64-bit ${eval} (JH/03).
1226 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1227 GNU libc to support some of the 64-bit stuff, should not lead to
1228 conflicts. Defined before os.h is pulled in, so if a given platform
1229 needs to override this, it can.
1231 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1232 protection layer was required, which is not implemented.
1233 Bugzilla 1254, patch from Wolfgang Breyha.
1235 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1236 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1237 tls_dhparam take prime identifiers. Also unbreak combination of
1238 OpenSSL+DH_params+TLSSNI.
1240 PP/39 Disable SSLv2 by default in OpenSSL support.
1246 PP/01 Solaris build fix for Oracle's LDAP libraries.
1247 Bugzilla 1109, patch from Stephen Usher.
1249 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1251 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1254 TF/02 Fix a couple more cases where we did not log the error message
1255 when unlink() failed. See also change 4.74-TF/03.
1257 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1258 lock up or crash if it happened to be inside a call to libc when it
1259 got a SIGUSR1 from exiwhat.
1261 The SIGUSR1 handler appends the current process status to the process
1262 log which is later printed by exiwhat. It used to use the general
1263 purpose logging code to do this, but several functions it calls are
1264 not safe for signals.
1266 The new output code in the SIGUSR1 handler is specific to the process
1267 log, and simple enough that it's easy to inspect for signal safety.
1268 Removing some special cases also simplifies the general logging code.
1269 Removing the spurious timestamps from the process log simplifies
1272 TF/04 Improved ratelimit ACL condition.
1274 The /noupdate option has been deprecated in favour of /readonly which
1275 has clearer semantics. The /leaky, /strict, and /readonly update modes
1276 are mutually exclusive. The update mode is no longer included in the
1277 database key; it just determines when the database is updated. (This
1278 means that when you upgrade Exim will forget old rate measurements.)
1280 Exim now checks that the per_* options are used with an update mode that
1281 makes sense for the current ACL. For example, when Exim is processing a
1282 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1283 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1284 must specify per_mail/readonly. If you omit the update mode it defaults to
1285 /leaky where that makes sense (as before) or /readonly where required.
1287 The /noupdate option is now undocumented but still supported for
1288 backwards compatibility. It is equivalent to /readonly except that in
1289 ACLs where /readonly is required you may specify /leaky/noupdate or
1290 /strict/noupdate which are treated the same as /readonly.
1292 A useful new feature is the /count= option. This is a generalization
1293 of the per_byte option, so that you can measure the throughput of other
1294 aggregate values. For example, the per_byte option is now equivalent
1295 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1297 The per_rcpt option has been generalized using the /count= mechanism
1298 (though it's more complicated than the per_byte equivalence). When it is
1299 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1300 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1301 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1302 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1303 Note that using per_rcpt with a non-readonly update mode in more than
1304 one ACL will cause the recipients to be double-counted. (The per_mail
1305 and per_byte options don't have this problem.)
1307 The handling of very low rates has changed slightly. If the computed rate
1308 is less than the event's count (usually one) then this event is the first
1309 after a long gap. In this case the rate is set to the same as this event's
1310 count, so that the first message of a spam run is counted properly.
1312 The major new feature is a mechanism for counting the rate of unique
1313 events. The new per_addr option counts the number of different
1314 recipients that someone has sent messages to in the last time period. It
1315 behaves like per_rcpt if all the recipient addresses are different, but
1316 duplicate recipient addresses do not increase the measured rate. Like
1317 the /count= option this is a general mechanism, so the per_addr option
1318 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1319 example, measure the rate that a client uses different sender addresses
1320 with the options per_mail/unique=$sender_address. There are further
1321 details in the main documentation.
1323 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1325 TF/06 Removed a few PCRE remnants.
1327 TF/07 Automatically extract Exim's version number from tags in the git
1328 repository when doing development or release builds.
1330 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1331 Bugzilla 879. Patch from Paul Fisher.
1333 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1334 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1337 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1339 PP/05 Variable $av_failed, true if the AV scanner deferred.
1340 Bugzilla 1078. Patch from John Horne.
1342 PP/06 Stop make process more reliably on build failure.
1343 Bugzilla 1087. Patch from Heiko Schlittermann.
1345 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1346 Bugzilla 1089. Patch from Heiko Schlittermann.
1348 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1349 Bugzilla 1131. Patch from Holger Weiß.
1351 PP/09 Handle IPv6 addresses with SPF.
1352 Bugzilla 860. Patch from Wolfgang Breyha.
1354 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1356 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1359 PP/11 match_* no longer expand right-hand-side by default.
1360 New compile-time build option, EXPAND_LISTMATCH_RHS.
1361 New expansion conditions, "inlist", "inlisti".
1363 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1365 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1367 PP/14 fix log_write() format string regression from TF/03.
1368 Bugzilla 1152. Patch from Dmitry Isaikin.
1374 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1376 PP/02 Harmonised TLS library version reporting; only show if debugging.
1377 Layout now matches that introduced for other libraries in 4.74 PP/03.
1379 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1381 PP/04 New "dns_use_edns0" global option.
1383 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1386 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1387 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1389 TK/01 Updated PolarSSL code to 0.14.2.
1390 Bugzilla 1097. Patch from Andreas Metzler.
1392 PP/07 Catch divide-by-zero in ${eval:...}.
1393 Fixes bugzilla 1102.
1395 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1398 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1399 format-string attack -- SECURITY: remote arbitrary code execution.
1401 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1402 time unintentionally subject to list matching rules, letting the header
1403 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1404 arbitrary string expansion). This allowed for information disclosure.
1406 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1407 INT_MIN/-1 -- value coerced to INT_MAX.
1413 NM/01 Workaround for PCRE version dependency in version reporting
1416 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1417 This fixes portability to compilers other than gcc, notably
1418 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1420 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1421 makefiles for portability to HP-UX and POSIX correctness.
1423 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1424 Also via indented variable definition in the Makefile.
1425 (Debugging by Oliver Heesakkers).
1427 PP/02 Restore caching of spamd results with expanded spamd_address.
1428 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1430 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1431 Improves build reliability. Fix from: Frank Elsner
1433 NM/02 Fix wide character breakage in the rfc2047 coding
1434 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1436 NM/03 Allow underscore in dnslist lookups
1437 Fixes bug 1026. Patch from Graeme Fowler
1439 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1440 Code patches from Adam Ciarcinski of NetBSD.
1442 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1445 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1446 is logged, to avoid truncation. Patch from John Horne.
1448 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1449 Patch from Jakob Hirsch.
1451 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1452 SQL string expansion failure details.
1453 Patch from Andrey Oktyabrski.
1455 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1456 Patch from Simon Arlott.
1458 PP/09 New lookups functionality failed to compile on old gcc which rejects
1459 extern declarations in function scope.
1460 Patch from Oliver Fleischmann
1462 PP/10 Use sig_atomic_t for flags set from signal handlers.
1463 Check getgroups() return and improve debugging.
1464 Fixed developed for diagnosis in bug 927 (which turned out to be
1467 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1468 Patch from Mark Zealey.
1470 PP/12 Bugzilla 1056: Improved spamd server selection.
1471 Patch from Mark Zealey.
1473 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1474 Based on patch from Heiko Schlittermann.
1476 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1477 Patch from Uwe Doering, sign-off by Michael Haardt.
1479 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1480 variable declaration deep within a block. Bug and patch from
1483 PP/15 lookups-Makefile IRIX compatibility coercion.
1485 PP/16 Make DISABLE_DKIM build knob functional.
1487 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1488 Patch by Simon Arlott
1490 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1491 variable argument macros. Our copy now differs from upstream.
1497 TF/01 Failure to get a lock on a hints database can have serious
1498 consequences so log it to the panic log.
1500 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1501 controlled using the smtp_confirmation log selector.
1503 TF/03 Include the error message when we fail to unlink a spool file.
1505 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1506 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1507 for maintaining out-of-tree patches for some time.
1509 PP/01 Bugzilla 139: Documentation and portability issues.
1510 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1511 Handle per-OS dynamic-module compilation flags.
1513 PP/02 Let /dev/null have normal permissions.
1514 The 4.73 fixes were a little too stringent and complained about the
1515 permissions on /dev/null. Exempt it from some checks.
1516 Reported by Andreas M. Kirchwitz.
1518 PP/03 Report version information for many libraries, including
1519 Exim version information for dynamically loaded libraries. Created
1520 version.h, now support a version extension string for distributors
1521 who patch heavily. Dynamic module ABI change.
1523 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1524 privilege escalation vulnerability whereby the Exim run-time user
1525 can cause root to append content of the attacker's choosing to
1528 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1531 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1532 If dropping privileges for untrusted macros, we disabled normal logging
1533 on the basis that it would fail; for the Exim run-time user, this is not
1534 the case, and it resulted in successful deliveries going unlogged.
1535 Fixed. Reported by Andreas Metzler.
1541 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1542 only prepend for the Resent-* case. Fixes regression introduced in
1543 Exim 4.70 by NM/22 for Bugzilla 607.
1545 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1546 increasing numbers of administrators be bitten by this.
1548 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1550 PP/03 Bugzilla 994: added openssl_options main configuration option.
1552 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1554 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1556 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1558 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1559 without a peer certificate, leading to a segfault because of an
1560 assumption that peers always have certificates. Be a little more
1561 paranoid. Problem reported by Martin Tscholak.
1563 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1564 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1565 NB: ClamAV planning to remove STREAM in "middle of 2010".
1566 CL also introduces -bmalware, various -d+acl logging additions and
1567 more caution in buffer sizes.
1569 PP/09 Implemented reverse_ip expansion operator.
1571 PP/10 Bugzilla 937: provide a "debug" ACL control.
1573 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1575 PP/12 Bugzilla 973: Implement --version.
1577 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1579 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1581 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1583 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1584 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1585 ignore trailing whitespace.
1587 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1589 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1592 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1593 Notification from Dr Andrew Aitchison.
1595 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1596 ExtendedDetectionInfo response format.
1597 Notification from John Horne.
1599 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1602 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1603 XSL and documented dependency on system catalogs, with examples of how
1606 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1609 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1610 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1611 configuration file which is writeable by the Exim user or group.
1613 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1614 of configuration files to cover files specified with the -C option if
1615 they are going to be used with root privileges, not just the default
1618 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1619 option (effectively making it always true).
1621 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1622 files to be used while preserving root privileges.
1624 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1625 that rogue child processes cannot use them.
1627 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1628 run-time user, instead of root.
1630 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1631 Exim run-time user without dropping privileges.
1633 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1634 result string, instead of calling string_vformat() twice with the same
1637 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1638 for other users. Others should always drop root privileges if they use
1639 -C on the command line, even for a whitelisted configure file.
1641 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1643 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1649 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1650 $data_path, and $header_path variables; fixed documentation bugs and
1653 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1654 exipick to access non-standard spools, including the "frozen" queue
1657 NM/01 Bugzilla 965: Support mysql stored procedures.
1658 Patch from Alain Williams
1660 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1662 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1663 Patch from Andreas Metzler
1665 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1666 Patch from Kirill Miazine
1668 NM/05 Bugzilla 671: Added umask to procmail example.
1670 JJ/03 installed exipick 20100323.0, fixing doc bug
1672 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1673 directory. Notification and patch from Dan Rosenberg.
1675 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1677 TK/02 Improve log output when DKIM signing operation fails.
1679 MH/01 Treat the transport option dkim_domain as a colon separated
1680 list, not as a single string, and sign the message with each element,
1681 omitting multiple occurences of the same signer.
1683 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1684 Bugzilla 985, 986. Patch by Simon Arlott
1686 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1687 Patch by Simon Arlott
1689 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1690 MBX locking. Notification from Dan Rosenberg.
1696 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1698 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1700 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1702 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1704 NM/04 Bugzilla 915: Flag broken perl installation during build.
1710 TK/01 Added patch by Johannes Berg that expands the main option
1711 "spamd_address" if it starts with a dollar sign.
1713 TK/02 Write list of recipients to X-Envelope-Sender header when building
1714 the mbox-format spool file for content scanning (suggested by Jakob
1717 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1718 (http://www.dcc-servers.net/) support via dccifd. Activated by
1719 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1721 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1722 by Mark Daniel Reidel <mr@df.eu>.
1724 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1725 When building exim an external PCRE library is now needed -
1726 PCRE is a system library on the majority of modern systems.
1727 See entry on PCRE_LIBS in EDITME file.
1729 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1730 conversation. Added nologin parameter to request.
1731 Patch contributed by Kirill Miazine.
1733 TF/01 Do not log submission mode rewrites if they do not change the address.
1735 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1737 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1738 log files in place. Contributed by Roberto Lima.
1740 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1742 TF/03 Bugzilla 615: When checking the local_parts router precondition
1743 after a local_part_suffix or local_part_prefix option, Exim now
1744 does not use the address's named list lookup cache, since this
1745 contains cached lookups for the whole local part.
1747 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1748 Robert Millan. Documentation is in experimental-spec.txt.
1750 TF/04 Bugzilla 668: Fix parallel build (make -j).
1752 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1754 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1755 Patch provided by Jan Srzednicki.
1757 TF/05 Leading white space used to be stripped from $spam_report which
1758 wrecked the formatting. Now it is preserved.
1760 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1761 that they are available at delivery time.
1763 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1765 TF/08 TLS error reporting now respects the incoming_interface and
1766 incoming_port log selectors.
1768 TF/09 Produce a more useful error message if an SMTP transport's hosts
1769 setting expands to an empty string.
1771 NM/06 Bugzilla 744: EXPN did not work under TLS.
1772 Patch provided by Phil Pennock.
1774 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1775 Patch provided by Richard Godbee.
1777 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1778 acl_smtp_notquit, added index entry.
1780 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1781 Patch provided by Eugene Bujak.
1783 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1784 accept(). Patch provided by Maxim Dounin.
1786 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1787 Patch provided by Phil Pennock.
1789 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1791 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1792 Patch provided by Brad "anomie" Jorsch.
1794 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1795 Patch provided by Dean Brooks.
1797 TK/05 Add native DKIM support (does not depend on external libraries).
1799 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1800 Patch provided by Graeme Fowler.
1802 NM/16 Bugzilla 851: Documentation example syntax fix.
1804 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1806 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1809 NM/19 Bugzilla 745: TLS version reporting.
1810 Patch provided by Phil Pennock.
1812 NM/20 Bugzilla 167: bool: condition support.
1813 Patch provided by Phil Pennock.
1815 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1816 clients. Patch provided by Phil Pennock.
1818 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1819 Patch provided by Brad "anomie" Jorsch.
1821 NM/23 Bugzilla 687: Fix misparses in eximstats.
1822 Patch provided by Heiko Schlittermann.
1824 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1825 Patch provided by Heiko Schlittermann.
1827 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1828 plus update to original patch.
1830 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1832 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1833 Patch provided by David Brownlee.
1835 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1837 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1839 NM/30 Bugzilla 888: TLS documentation bugfixes.
1841 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1843 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1844 Unlike the original bugzilla I have changed all shell scripts in src tree.
1846 NM/33 Bugzilla 898: Transport filter timeout fix.
1847 Patch by Todd Rinaldo.
1849 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1850 Patch by Serge Demonchaux.
1852 NM/35 Bugzilla 39: Base64 decode bug fixes.
1853 Patch by Jakob Hirsch.
1855 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1857 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1859 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1861 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1867 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1868 ALT-N's libdkim that I have put here:
1869 http://duncanthrax.net/exim-experimental/
1871 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1872 were called 'true' and it seems that C99 defines that as a reserved
1873 keyword to be used with 'bool' variable types. That means you could
1874 not include C99-style headers which use bools without triggering
1875 build errors in sieve.c.
1877 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1878 as mailq or other aliases. Changed the --help handling significantly
1879 to do whats expected. exim_usage() emits usage/help information.
1881 SC/01 Added the -bylocaldomain option to eximstats.
1883 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1885 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1887 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1893 PH/01 Another patch from the Sieve maintainer.
1895 PH/02 When an IPv6 address is converted to a string for single-key lookup
1896 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1897 dots are used instead of colons so that keys in lsearch files need not
1898 contain colons. This was done some time before quoting was made available
1899 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1900 (notated using the quote facility) so as to distinguish them from IPv4
1901 keys. This meant that lookups for IP addresses in host lists did not work
1902 for iplsearch lookups.
1904 This has been fixed by arranging for IPv6 addresses to be expressed with
1905 colons if the lookup type is iplsearch. This is not incompatible, because
1906 previously such lookups could never work.
1908 The situation is now rather anomalous, since one *can* have colons in
1909 ordinary lsearch keys. However, making the change in all cases is
1910 incompatible and would probably break a number of configurations.
1912 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1915 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1916 conversion specification without a maximum field width, thereby enabling
1917 a rogue spamd server to cause a buffer overflow. While nobody in their
1918 right mind would setup Exim to query an untrusted spamd server, an
1919 attacker that gains access to a server running spamd could potentially
1920 exploit this vulnerability to run arbitrary code as the Exim user.
1922 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1923 $primary_hostname instead of what libspf2 thinks the hosts name is.
1925 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1926 a directory entry by the name of the lookup key. Previously, if a
1927 symlink pointed to a non-existing file or a file in a directory that
1928 Exim lacked permissions to read, a lookup for a key matching that
1929 symlink would fail. Now it is enough that a matching directory entry
1930 exists, symlink or not. (Bugzilla 503.)
1932 PH/03 The body_linecount and body_zerocount variables are now exported in the
1935 PH/04 Added the $dnslist_matched variable.
1937 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1938 This means they are set thereafter only if the connection becomes
1941 PH/06 Added the client_condition to authenticators so that some can be skipped
1942 by clients under certain conditions.
1944 PH/07 The error message for a badly-placed control=no_multiline_responses left
1945 "_responses" off the end of the name.
1947 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1949 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1950 (without spaces) instead of just copying the configuration text.
1952 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1954 PH/11 Added $max_received_linelength.
1956 PH/12 Added +ignore_defer and +include_defer to host lists.
1958 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1959 way in which PCRE > 7.0 is built.
1961 PH/14 Implemented queue_only_load_latch.
1963 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1964 MAIL command. The effect was to mangle the value on 64-bit systems.
1966 PH/16 Another patch from the Sieve maintainer.
1968 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1970 PH/18 If a system quota error occurred while trying to create the file for
1971 a maildir delivery, the message "Mailbox is full" was not appended to the
1972 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1973 only to a quota excession during the actual writing of the file.
1975 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1976 characters?) which causes problems in log lines. The DN values are now
1977 passed through string_printing() before being added to log lines.
1979 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1980 and InterBase are left for another time.)
1982 PH/21 Added message_body_newlines option.
1984 PH/22 Guard against possible overflow in moan_check_errorcopy().
1986 PH/23 POSIX allows open() to be a macro; guard against that.
1988 PH/24 If the recipient of an error message contained an @ in the local part
1989 (suitably quoted, of course), incorrect values were put in $domain and
1990 $local_part during the evaluation of errors_copy.
1996 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1997 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2000 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2001 issue a MAIL command.
2003 PH/02 In an ACL statement such as
2005 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2007 if a client was not listed at all, or was listed with a value other than
2008 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2009 the condition was not true (as it should be), so access was not denied.
2010 The bug was that the ! inversion was incorrectly passed on to the second
2011 item. This has been fixed.
2013 PH/03 Added additional dnslists conditions == and =& which are different from
2014 = and & when the dns lookup returns more than one IP address.
2016 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2017 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2019 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2020 FSYNC, which compiles an option called disable_fsync that allows for
2021 bypassing fsync(). The documentation is heavily laced with warnings.
2023 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2025 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2026 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2027 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2028 including adding "make clean"; (3) Added -fPIC when compiling the test
2029 dynamically loaded module, to get rid of a warning.
2031 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2032 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2033 = 0s. The bug is otherwise harmless.
2035 PH/07 There was a bug in the dovecot authenticator such that the value of
2036 $auth1 could be overwritten, and so not correctly preserved, after a
2037 successful authentication. This usually meant that the value preserved by
2038 the server_setid option was incorrect.
2040 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2042 PH/09 Installed PCRE release 7.0.
2044 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2045 run for batched SMTP input. It is now run at the start of every message
2046 in the batch. While fixing this I discovered that the process information
2047 (output by running exiwhat) was not always getting set for -bs and -bS
2048 input. This is fixed, and it now also says "batched" for BSMTP.
2050 PH/11 Added control=no_pipelining.
2052 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2053 patch, slightly modified), and move the expansion of helo_data till after
2054 the connection is made in the smtp transport (so it can use these
2057 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2059 PH/14 Added log_selector = +pid.
2061 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2063 PH/16 Add ${if forany and ${if forall.
2065 PH/17 Added dsn_from option to vary the From: line in DSNs.
2067 PH/18 Flush SMTP output before performing a callout, unless control =
2068 no_callout_flush is set.
2070 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2071 was true (the default) a successful delivery failed to delete the retry
2072 item, thus causing premature timeout of the address. The bug is now
2075 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2077 PH/21 Long custom messages for fakedefer and fakereject are now split up
2078 into multiline responses in the same way that messages for "deny" and
2079 other ACL rejections are.
2081 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2082 with slight modification.
2084 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2085 draft, changing the syntax and factoring some duplicate code".
2087 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2088 for deliveries of the second and subsequent messages over the same SMTP
2091 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2092 ${reduce, with only minor "tidies".
2094 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2096 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2097 expansion side effects.
2099 PH/27 When a message times out after an over-quota error from an Exim-imposed
2100 quota, the bounce message says "mailbox is full". This message was not
2101 being given when it was a system quota that was exceeded. It now should
2104 MH/03 Made $recipients available in local_scan(). local_scan() already has
2105 better access to the recipient list through recipients_list[], but
2106 $recipients can be useful in postmaster-provided expansion strings.
2108 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2109 in the case of a MAIL command with additional options following the
2110 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2111 were accidentally chopped off.
2113 PH/29 SMTP synchronization checks are implemented when a command is read -
2114 there is a check that no more input is waiting when there shouldn't be
2115 any. However, for some commands, a delay in an ACL can mean that it is
2116 some time before the response is written. In this time, more input might
2117 arrive, invalidly. So now there are extra checks after an ACL has run for
2118 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2119 pipelining has not been advertised.
2121 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2123 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2124 returned in the SMTP response when smtp_return_error_details was false.
2125 This has been fixed.
2127 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2128 instead of the C I/O that was originally supplied, because problems were
2129 reported on Solaris.
2131 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2132 Exim which did not show up earlier: it was assuming that a call to
2133 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2134 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2135 was a macro that became an assignment, so it seemed to work. This has
2136 changed to a proper function call with a void return, hence the compile
2137 error. Exim's code has been fixed.
2139 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2142 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2144 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2146 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2147 because it turns out to be too big or there is a timeout) there is no
2148 "Completed" line in the log. When some messages of this type were
2149 selected by exigrep, they were listed as "not completed". Others were
2150 picked up by some special patterns. I have improved the selection
2151 criteria to be more general.
2153 PH/38 The host_find_failed option in the manualroute router can now be set
2154 to "ignore", to completely ignore a host whose IP address cannot be
2155 found. If all hosts are ignored, the behaviour is controlled by the new
2156 host_all_ignored option.
2158 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2159 homing or because of multiple MX records with /mx) generated more than
2160 one IP address, and the following item turned out to be the local host,
2161 all the secondary addresses of the first item were incorrectly removed
2162 from the list, along with the local host and any following hosts (which
2163 is what is supposed to happen).
2165 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2166 whoever called Exim into the -H file. In the case of the daemon it was
2167 behaving confusingly. When first started, it used values for whoever
2168 started the daemon, but after a SIGHUP it used the Exim user (because it
2169 calls itself on a restart). I have changed the code so that it now always
2172 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2173 message are rejected with the same error (e.g. no authentication or bad
2174 sender address), and a DATA command is nevertheless sent (as can happen
2175 with PIPELINING or a stupid MUA), the error message that was given to the
2176 RCPT commands is included in the rejection of the DATA command. This is
2177 intended to be helpful for MUAs that show only the final error to their
2180 PH/42 Another patch from the Sieve maintainer.
2182 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2183 Eximstats - Fixed some broken HTML links and added missing column headers
2185 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2186 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2188 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2190 PH/43 Yet another patch from the Sieve maintainer.
2192 PH/44 I found a way to check for a TCP/IP connection going away before sending
2193 the response to the final '.' that terminates a message, but only in the
2194 case where the client has not sent further data following the '.'
2195 (unfortunately, this is allowed). However, in many cases there won't be
2196 any further data because there won't be any more messages to send. A call
2197 to select() can be used: if it shows that the input is "ready", there is
2198 either input waiting, or the socket has been closed. An attempt to read
2199 the next input character can distinguish the two cases. Previously, Exim
2200 would have sent an OK response which the client would never have see.
2201 This could lead to message repetition. This fix should cure that, at
2202 least in a lot of common cases.
2204 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2205 advertised in response to EHLO.
2211 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2212 fixed by 4.65/MH/01 (is this a record?) are fixed:
2214 (i) An empty string was always treated as zero by the numeric comparison
2215 operators. This behaviour has been restored.
2217 (ii) It is documented that the numeric comparison operators always treat
2218 their arguments as decimal numbers. This was broken in that numbers
2219 starting with 0 were being interpreted as octal.
2221 While fixing these problems I realized that there was another issue that
2222 hadn't been noticed. Values of message_size_limit (both the global option
2223 and the transport option) were treated as octal if they started with 0.
2224 The documentation was vague. These values are now always treated as
2225 decimal, and I will make that clear in the documentation.
2231 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2232 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2235 MH/01 Don't check that the operands of numeric comparison operators are
2236 integers when their expansion is in "skipping" mode (fixes bug
2237 introduced by 4.64-PH/07).
2239 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2240 child addresses, Exim now panics and dies. Previously, because the count
2241 is held in a short int, deliveries were likely to be lost. As such a
2242 large number of recipients for a single message is ridiculous
2243 (performance will be very, very poor), I have chosen to impose a limit
2244 rather than extend the field.
2250 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2251 leftover -K file (the existence of which was triggered by #402).
2252 While we were at it, introduced process PID as part of the -K
2253 filename. This should rule out race conditions when creating
2256 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2257 processing considerably. Previous code took too long for large mails,
2258 triggering a timeout which in turn triggers #401.
2260 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2261 in the DK code in transports.c. sendfile() is not really portable,
2262 hence the _LINUX specificness.
2264 TF/01 In the add_headers option to the mail command in an Exim filter,
2265 there was a bug that Exim would claim a syntax error in any
2266 header after the first one which had an odd number of characters
2269 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2270 callout verification, Exim cached a "reject" for the entire domain. This
2271 is correct for most verifications, but it is not correct for a recipient
2272 verification with use_sender or use_postmaster set, because in that case
2273 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2274 case of MAIL FROM:<> rejection from other early rejections (e.g.
2275 rejection of HELO). When verifying a recipient using a non-null MAIL
2276 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2277 Whatever the result of the callout, the value of the domain cache is
2278 left unchanged (for any other kind of callout, getting as far as trying
2279 RCPT means that the domain itself is ok).
2281 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2284 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2285 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2286 session, the daemon ignores ECONNECT errors and logs others; it now
2287 ignores EPIPE as well.
2289 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2290 (quoted-printable decoding).
2292 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2293 later the small subsequent patch to fix an introduced bug.
2295 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2297 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2299 PH/08 An error is now given if message_size_limit is specified negative.
2301 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2302 to be given (somewhat) arbitrary names.
2304 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2307 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2308 miscellaneous code fixes
2310 PH/10 Added the log_reject_target ACL modifier to specify where to log
2313 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2314 hostname. This is wrong, because it relates to the incoming message (and
2315 probably the interface on which it is arriving) and not to the outgoing
2316 callout (which could be using a different interface). This has been
2317 changed to use the value of the helo_data option from the smtp transport
2318 instead - this is what is used when a message is actually being sent. If
2319 there is no remote transport (possible with a router that sets up host
2320 addresses), $smtp_active_hostname is used.
2322 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2323 tweaks were necessary in order to get it to work (see also 21 below):
2324 (a) The code assumed that strncpy() returns a negative number on buffer
2325 overflow, which isn't the case. Replaced with Exim's string_format()
2327 (b) There were several signed/unsigned issues. I just did the minimum
2328 hacking in of casts. There is scope for a larger refactoring.
2329 (c) The code used strcasecmp() which is not a standard C function.
2330 Replaced with Exim's strcmpic() function.
2331 (d) The code set only $1; it now sets $auth1 as well.
2332 (e) A simple test gave the error "authentication client didn't specify
2333 service in request". It would seem that Dovecot has changed its
2334 interface. Fortunately there's a specification; I followed it and
2335 changed what the client sends and it appears to be working now.
2337 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2340 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2341 address A is aliased to B and C, where B exists and C does not. Without
2342 -v the output is "A verified" because verification stops after a
2343 successful redirection if more than one address is generated. However,
2344 with -v the child addresses are also verified. Exim was outputting "A
2345 failed to verify" and then showing the successful verification for C,
2346 with its parentage. It now outputs "B failed to verify", showing B's
2347 parentage before showing the successful verification of C.
2349 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2350 look up a TXT record in a specific list after matching in a combined
2353 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2354 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2355 they consult the DNS. I had assumed they would set it the way they
2356 wanted; and indeed my experiments on Linux seem to show that in some
2357 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2358 To be on the safe side, however, I have now made the interface to
2359 host_find_byname() similar to host_find_bydns(), with an argument
2360 containing the DNS resolver options. The host_find_byname() function now
2361 sets these options at its start, just as host_find_bydns() does. The smtp
2362 transport options dns_qualify_single and dns_search_parents are passed to
2363 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2364 of host_find_byname() use the default settings of RES_DEFNAMES
2365 (qualify_single) but not RES_DNSRCH (search_parents).
2367 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2368 spool_read_header() do less string testing, by means of a preliminary
2369 switch on the second character of optional "-foo" lines. (This is
2370 overdue, caused by the large number of possibilities that now exist.
2371 Originally there were few.) While I was there, I also converted the
2372 str(n)cmp tests so they don't re-test the leading "-" and the first
2373 character, in the hope this might squeeze out yet more improvement.
2375 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2376 flag allowing group syntax was set by the header_syntax check but not
2377 turned off, possible causing trouble later; (2) The flag was not being
2378 set at all for the header_verify test, causing "group"-style headers to
2379 be rejected. I have now set it in this case, and also caused header_
2380 verify to ignore an empty address taken from a group. While doing this, I
2381 came across some other cases where the code for allowing group syntax
2382 while scanning a header line wasn't quite right (mostly, not resetting
2383 the flag correctly in the right place). These bugs could have caused
2384 trouble for malformed header lines. I hope it is now all correct.
2386 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2387 with the "reply" argument non-NULL. The code, however (which originally
2388 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2389 but it didn't always do it. This confused somebody who was copying the
2390 code for some other use. I have removed all the tests.
2392 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2393 feature that was used to support insecure browsers during the U.S. crypto
2394 embargo. It requires special client support, and Exim is probably the
2395 only MTA that supported it -- and would never use it because real RSA is
2396 always available. This code has been removed, because it had the bad
2397 effect of slowing Exim down by computing (never used) parameters for the
2398 RSA_EXPORT functionality.
2400 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2401 authenticator to fail if there's a tab character in the incoming data
2402 (there should never be unless someone is messing about, as it's supposed
2403 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2404 if the connection is using TLS or if the remote IP is the same as the
2405 local IP, and the "valid-client-cert option" if a client certificate has
2408 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2409 authenticators. This can be used for authorization after authentication
2410 succeeds. (In the case of plaintext, it servers for both authentication
2413 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2414 if any retry times were supplied.
2416 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2417 connection, where there is no client IP address to check. In this
2418 situation, the verify now always succeeds.
2420 PH/25 Applied John Jetmore's -Mset patch.
2422 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2424 PH/27 In a string expansion for a processed (not raw) header when multiple
2425 headers of the same name were present, leading whitespace was being
2426 removed from all of them, but trailing whitespace was being removed only
2427 from the last one. Now trailing whitespace is removed from each header
2428 before concatenation. Completely empty headers in a concatenation (as
2429 before) are ignored.
2431 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2432 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2434 PH/29 [Removed. This was a change that I later backed out, and forgot to
2435 correct the ChangeLog entry (that I had efficiently created) before
2436 committing the later change.]
2438 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2439 address errors (4xx response to RCPT) over the same connection as other
2440 messages routed to the same hosts. Such deliveries are always "forced",
2441 so retry times are not inspected. This resulted in far too many retries
2442 for the affected addresses. The effect occurred only when there were more
2443 hosts than the hosts_max_try setting in the smtp transport when it had
2444 the 4xx errors. Those hosts that it had tried were not added to the list
2445 of hosts for which the message was waiting, so if all were tried, there
2446 was no problem. Two fixes have been applied:
2448 (i) If there are any address or message errors in an SMTP delivery, none
2449 of the hosts (tried or untried) are now added to the list of hosts
2450 for which the message is waiting, so the message should not be a
2451 candidate for sending over the same connection that was used for a
2452 successful delivery of some other message. This seems entirely
2453 reasonable: after all the message is NOT "waiting for some host".
2454 This is so "obvious" that I'm not sure why it wasn't done
2455 previously. Hope I haven't missed anything, but it can't do any
2456 harm, as the worst effect is to miss an optimization.
2458 (ii) If, despite (i), such a delivery is accidentally attempted, the
2459 routing retry time is respected, so at least it doesn't keep
2460 hammering the server.
2462 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2463 in ${readsocket because some servers need this prod.
2465 PH/32 Added some extra debug output when updating a wait-xxx database.
2467 PH/33 The hint "could be header name not terminated by colon", which has been
2468 given for certain expansion errors for a long time, was not being given
2469 for the ${if def:h_colon_omitted{... case.
2471 PH/34 The spec says: "With one important exception, whenever a domain list is
2472 being scanned, $domain contains the subject domain." There was at least
2473 one case where this was not true.
2475 PH/35 The error "getsockname() failed: connection reset by peer" was being
2476 written to the panic log as well as the main log, but it isn't really
2477 panic-worthy as it just means the connection died rather early on. I have
2478 removed the panic log writing for the ECONNRESET error when getsockname()
2481 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2482 runs only) independently of the message's sender address. This meant
2483 that, if the 4xx error was in fact related to the sender, a different
2484 message to the same recipient with a different sender could confuse
2485 things. In particular, this can happen when sending to a greylisting
2486 server, but other circumstances could also provoke similar problems.
2487 I have changed the default so that the retry time for these errors is now
2488 based a combination of the sender and recipient addresses. This change
2489 can be overridden by setting address_retry_include_sender=false in the
2492 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2493 remote server are returned as part of bounce messages. This was not
2494 happening for LMTP over a pipe (the lmtp transport), but now it is the
2495 same for both kinds of LMTP.
2497 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2498 in header lines that were in fact CNAMEs. This is no longer the case.
2500 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2501 and queue runs started by the daemon processed all messages. This has
2502 been fixed so that -R and -S can now usefully be given with -q<time>.
2504 PH/40 Import PCRE release 6.7 (fixes some bugs).
2506 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2508 PH/42 Give an error if -q is specified more than once.
2510 PH/43 Renamed the variables $interface_address and $interface_port as
2511 $received_ip_address and $received_port, to make it clear that these
2512 values apply to message reception, and not to the outgoing interface when
2513 a message is delivered. (The old names remain recognized, of course.)
2515 PH/44 There was no timeout on the connect() call when using a Unix domain
2516 socket in the ${readsocket expansion. There now is.
2518 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2519 be meaningful with "accept".
2521 SC/01 Eximstats V1.43
2522 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2524 SC/02 Eximstats V1.44
2525 Use a glob alias rather than an array ref in the generated
2526 parser. This improves both readability and performance.
2528 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2529 Collect SpamAssassin and rejection statistics.
2530 Don't display local sender or destination tables unless
2531 there is data to show.
2532 Added average volumes into the top table text output.
2534 SC/04 Eximstats V1.46
2535 Collect data on the number of addresses (recipients)
2536 as well as the number of messages.
2538 SC/05 Eximstats V1.47
2539 Added 'Message too big' to the list of mail rejection
2540 reasons (thanks to Marco Gaiarin).
2542 SC/06 Eximstats V1.48
2543 Mainlog lines which have GMT offsets and are too short to
2544 have a flag are now skipped.
2546 SC/07 Eximstats V1.49 (Alain Williams)
2547 Added the -emptyok flag.
2549 SC/08 Eximstats V1.50
2550 Fixes for obtaining the IP address from reject messages.
2552 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2553 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2554 whitespace changes from 4.64-PH/27
2556 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2559 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2560 are found, allow negative numbers in numeric criteria)
2562 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2564 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2565 to match changes made in 4.64-PH/43
2567 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2569 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2570 "domain" is actually an IP address, and force a failure. This locks out
2571 those revolvers/nameservers that support "A-for-A" lookups, in
2572 contravention of the specifications.
2574 PH/48 When a host name was looked up from an IP address, and the subsequent
2575 forward lookup of the name timed out, the host name was left in
2576 $sender_host_name, contrary to the specification.
2578 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2579 restricted to single-key lookups, Exim was not diagnosing an error if
2580 * or *@ was used with a query-style lookup.
2582 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2584 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2585 long ago, but noone interested enough thought of it. Let's just say that
2586 the "1.1" means that there are some new functions that weren't there at
2587 some point in the past.
2589 PH/51 Error processing for expansion failure of helo_data from an smtp
2590 transport during callout processing was broken.
2592 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2593 tested/used via the -bh/-bhc/-bs options.
2595 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2596 bug, fixed in subsequent PCRE releases).
2598 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2599 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2601 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2607 SC/01 Use a glob alias rather than an array ref in eximstats generated
2608 parser. This improves both readability and performance.
2610 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2611 Don't display local sender or destination tables in eximstats unless
2612 there is data to show.
2613 Added average volumes into the eximstats top table text output.
2615 SC/03 Collect data on the number of addresses (recipients) as well
2616 as the number of messages in eximstats.
2618 TF/01 Correct an error in the documentation for the redirect router. Exim
2619 does (usually) call initgroups() when daemonizing.
2621 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2622 with consistent privilege compared to when running as a daemon.
2624 TF/03 Note in the spec that $authenticated_id is not set for local
2625 submissions from trusted users.
2627 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2628 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2630 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2631 by adding some example configuration directives to the default
2632 configuration file. A little bit of work is required to uncomment the
2633 directives and define how usernames and passwords are checked, but
2634 there is now a framework to start from.
2636 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2637 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2638 without this. I don't know how relevant this is to other LDAP libraries.
2640 PH/02 Add the verb name to the "unknown ACL verb" error.
2642 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2644 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2646 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2647 directory not expanded when it should be if an expanded home directory
2648 was set for the address (which is overridden by the transport).
2650 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2653 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2654 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2655 because it is too late at that time, and has no effect.
2657 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2658 security issue with \' (bugzilla #107). I could not use the
2659 PQescapeStringConn() function, because it needs a PGconn value as one of
2662 PH/08 When testing addresses using -bt, indicate those final addresses that
2663 are duplicates that would not cause an additional delivery. At least one
2664 person was confused, thinking that -bt output corresponded to deliveries.
2665 (Suppressing duplicates isn't a good idea as you lose the information
2666 about possibly different redirections that led to the duplicates.)
2668 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2669 systems where poll() doesn't work, in particular OS X.
2671 PH/10 Added more information to debugging output for retry time not reached.
2673 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2674 operations in malware.c.
2676 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2679 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2680 syslog_duplication set false, log lines that would normally be written
2681 both the the main log and to the reject log were not written to syslog at
2684 PH/14 In the default configuration, change the use of "message" in ACL warn
2685 statements to "add_header".
2687 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2688 not followed by a command (e.g. "seen endif").
2690 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2691 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2694 PH/17 Added extra conditions to the default value of delay_warning_condition
2698 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2699 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2700 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2703 The Auto-Submitted: and various List- headers are standardised, whereas I
2704 don't think Precedence: ever was.
2706 PH/18 Refactored debugging code in route_finduser() to show more information,
2707 in particular, the error code if getpwnam() issues one.
2709 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2710 This is apparently needed in addition to the PH/07 change above to avoid
2711 any possible encoding problems.
2713 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2714 but not after initializing Perl.
2716 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2717 output them only if debugging. By default they are written stderr,
2718 apparently, which is not desirable.
2720 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2723 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2726 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2728 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2729 authenticated or an ident call has been made. Suppress the default
2730 values for $authenticated_id and $authenticated_sender (but permit -oMai
2731 and -oMas) when testing with -bh.
2733 PH/24 Re-jigged the order of the tests in the default configuration so that the
2734 tests for valid domains and recipients precede the DNS black list and CSA
2735 tests, on the grounds that those ones are more expensive.
2737 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2738 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2739 This bug exists in every version of Exim that I still have, right back to
2742 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2743 However, an attempt to turn on case-sensitivity in a regex key by
2744 including (?-i) didn't work because the subject string was already
2745 lowercased, and the effects were non-intuitive. It turns out that a
2746 one-line patch can be used to allow (?-i) to work as expected.
2752 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2753 other effects) broke the use of negated acl sub-conditions.
2755 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2758 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2759 "Deny" causes Exim to reject the incoming connection with a 554 error.
2760 Unfortunately, if there is a major crisis, such as a disk failure,
2761 tcp-wrappers gives "deny", whereas what one would like would be some
2762 kind of temporary error. A kludge has been added to help with this.
2763 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2764 554 error is used if errno is still zero or contains ENOENT (which occurs
2765 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2768 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2770 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2771 errors. Otherwise a message that provokes a temporary error (when other
2772 messages do not) can cause a whole host to time out.
2774 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2775 addresses were routed directly to files or pipes from a redirect router.
2776 File deliveries just didn't batch; pipe deliveries might have suffered
2779 PH/06 A failure to get a lock for a hints database would erroneously always say
2780 "Failed to get write lock", even when it was really a read lock.
2782 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2783 of 0600. This has been changed to use the value of the lockfile_mode
2784 option (which defaults to 0600).
2786 PH/08 Applied small patch from the Sieve maintainer.
2788 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2789 folder from quota calculations, a direct delivery into this folder messed
2790 up the contents of the maildirsize file. This was because the regex was
2791 used only to exclude .Trash (or whatever) when the size of the mailbox
2792 was calculated. There was no check that a delivery was happening into an
2793 excluded directory. This bug has been fixed by ignoring all quota
2794 processing for deliveries into excluded directories.
2796 PH/10 Added the maildirfolder_create_regex option to appendfile.
2802 PH/01 The code for finding all the local interface addresses on a FreeBSD
2803 system running IPv6 was broken. This may well have applied to all BSD
2804 systems, as well as to others that have similar system calls. The broken
2805 code found IPv4 interfaces correctly, but gave incorrect values for the
2806 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2807 that it would not match correctly against @[] and not recognize the IPv6
2810 PH/02 The ipliteral router was not recognizing addresses of the form user@
2811 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2813 PH/03 Added disable_ipv6.
2815 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2816 decoded form, because it is most often used to construct To: headers
2817 lines in autoreplies, and the decoded form may well be syntactically
2818 invalid. However, $reply_address has leading white space removed, and all
2819 newlines turned into spaces so that the autoreply transport does not
2822 PH/05 If group was specified without a user on a router, and no group or user
2823 was specified on a transport, the group from the router was ignored.
2825 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2826 for visible compile-time settings that can be used to change these
2827 numbers, for those that want even more. Backwards compatibility with old
2828 spool files has been maintained. However, going back to a previous Exim
2829 release will lost any variables that are in spool files.
2831 PH/07 Two small changes when running in the test harness: increase delay when
2832 passing a TCP/IP connection to a new process, in case the original
2833 process has to generate a bounce, and remove special handling of
2834 127.0.0.2 (sic), which is no longer necessary.
2836 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2837 be the same on different OS.
2839 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2842 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2843 whether --show-vars was specified or not
2845 JJ/02 exipick: Added support for new ACL variable spool format introduced
2848 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2849 syntactically invalid From: or Reply-to: line, and a filter used this to
2850 generate an autoreply, and therefore failed to obtain an address for the
2851 autoreply, Exim could try to deliver to a non-existent relative file
2852 name, causing unrelated and misleading errors. What now happens is that
2853 it logs this as a hard delivery error, but does not attempt to create a
2856 PH/11 The exinext utility has a -C option for testing purposes, but although
2857 the given file was scanned by exinext itself; it wasn't being passed on
2858 when Exim was called.
2860 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2861 an end-of-file indication when reading a command response.
2863 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2864 compiled. In many other places in Exim, IPv6 addresses are always
2865 recognized, so I have changed this. It also means that IPv4 domain
2866 literals of the form [IPV4:n.n.n.n] are now always recognized.
2868 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2869 used if the router is not running as root, for example, when verifying at
2870 ACL time, or when using -bh. The debugging output from this situation was
2871 non-existent - all you got was a failure to exec. I have made two
2874 (a) Failures to set uid/gid, the current directory, or a process leader
2875 in a subprocess such as that created by queryprogram now generate
2876 suitable debugging output when -d is set.
2878 (b) The queryprogram router detects when it is not running as root,
2879 outputs suitable debugging information if -d is set, and then runs
2880 the subprocess without attempting to change uid/gid.
2882 PH/15 Minor change to Makefile for building test_host (undocumented testing
2885 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2886 additional section of a DNS packet that returns MX or SRV records.
2887 Instead, it always explicitly searches for A/AAAA records. This avoids
2888 major problems that occur when a DNS server includes only records of one
2889 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2890 fixed another bug: if SRV records were looked up and the corresponding
2891 address records were *not* found in the additional section, the port
2892 values from the SRV records were lost.
2894 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2895 using the correct key (the original address) when searching the retry
2896 rules in order to find which one to use for generating the retry hint.
2898 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2899 adding the default one. Similarly, if it contains a Reply-To: header, the
2900 errors_reply_to option, if set, is not used.
2902 PH/19 When calculating a retry time, Exim used to measure the "time since
2903 failure" by looking at the "first failed" field in the retry record. Now
2904 it does not use this if it is later than than the arrival time of the
2905 message. Instead it uses the arrival time. This makes for better
2906 behaviour in cases where some deliveries succeed, thus re-setting the
2907 "first failed" field. An example is a quota failure for a huge message
2908 when small messages continue to be delivered. Without this change, the
2909 "time since failure" will always be short, possible causing more frequent
2910 delivery attempts for the huge message than are intended.
2911 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2913 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2914 $1, $2, $3) because the numerical variables can be reset during some
2915 expansion items (e.g. "match"), thereby losing the authentication data.
2917 PH/21 Make -bV show the size of off_t variables so that the test suite can
2918 decide whether to run tests for quotas > 2G.
2920 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2921 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2922 filecount value is greater than 2G or if a quota value is greater than 2G
2923 on a system where the size of off_t is not greater than 4, a panic error
2926 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2927 never match. The debug and -bh output now contains an explicit error
2928 message indicating a malformed IPv4 address or mask.
2930 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2931 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2932 PH/23 above applies.
2934 PH/25 Do not write to syslog when running in the test harness. The only
2935 occasion when this arises is a failure to open the main or panic logs
2936 (for which there is an explicit test).
2938 PH/26 Added the /no_tell option to "control=freeze".
2940 PH/27 If a host name lookup failed very early in a connection, for example, if
2941 the IP address matched host_lookup and the reverse lookup yielded a name
2942 that did not have a forward lookup, an error message of the form "no IP
2943 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2944 could be logged. Now it outputs the IP address instead of "NULL".
2946 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2947 allows the sender and the authenticated sender to be set when
2948 submitting a message from within Exim. Since child_open_exim() is
2949 documented for local_scan(), the new function should be too.
2951 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2952 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2953 results in an empty string is now treated as unset.
2955 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2957 PH/31 Added sender_verify_fail logging option.
2959 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2960 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2961 tidied the source and removed it altogether.
2963 PH/33 When a queue run was abandoned because the load average was too high, a
2964 log line was always written; now it is written only if the queue_run log
2965 selector is set. In addition, the log line for abandonment now contains
2966 information about the queue run such as the pid. This is always present
2967 in "start" and "stop" lines but was omitted from the "abandon" line.
2969 PH/34 Omit spaces between a header name and the colon in the error message that
2970 is given when verify = headers_syntax fails (if there are lots of them,
2971 the message gets confusing).
2973 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2974 names, as there are now some PTR records that contain slashes. This check
2975 is only to protect against broken name servers that fall over on strange
2976 characters, so the fact that it applies to all lookups doesn't matter.
2978 PH/36 Now that the new test suite is complete, we can remove some of the
2979 special code in Exim that was needed for the old test suite. For example,
2980 sorting DNS records because real resolvers return them in an arbitrary
2981 order. The new test suite's fake resolver always returns records in the
2984 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2985 bounces) except when queue_only is set, to avoid logging races between
2986 the different processes.
2988 PH/38 Panic-die if .include specifies a non-absolute path.
2990 PH/39 A tweak to the "H" retry rule from its user.
2992 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2993 a label. They prevented compilation on older perls.
2995 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2996 a warning to be raised on newish perls.
2998 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2999 on queue. Changes to match documented behaviour of showing count of
3000 messages matching specified criteria.
3002 PH/40 Changed the default ident timeout from 30s to 5s.
3004 PH/41 Added support for the use of login_cap features, on those BSD systems
3005 that have them, for controlling the resources used by pipe deliveries.
3007 PH/42 The content-scanning code uses fopen() to create files in which to put
3008 message data. Previously it was not paying any attention to the mode of
3009 the files. Exim runs with umask(0) because the rest of the code creates
3010 files with open(), and sets the required mode explicitly. Thus, these
3011 files were ending up world-writeable. This was not a big issue, because,
3012 being within the spool directory, they were not world-accessible. I have
3013 created a function called modefopen, which takes an additional mode
3014 argument. It sets umask(777), creates the file, chmods it to the required
3015 mode, then resets the umask. All the relevant calls to fopen() in the
3016 content scanning code have been changed to use this function.
3018 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3019 to 24 hours. This avoids potential overflow problems when processing G
3020 and H retry rules. I suspect nobody ever tinkers with this value.
3022 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3024 PH/45 When the plaintext authenticator is running as a client, the server's
3025 challenges are checked to ensure they are valid base64 strings. By
3026 default, the authentication attempt is cancelled if an invalid string is
3027 received. Setting client_ignore_invalid_base64 true ignores these errors.
3028 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3029 they are received. Thus, the responses can be made to depend on the
3030 challenges. If an invalid string is ignored, an empty string is placed in
3033 PH/46 Messages that are created by the autoreply transport now contains a
3034 References: header, in accordance with RFCs 2822 and 3834.
3036 PH/47 Added authenticated_sender_force to the smtp transport.
3038 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3040 PH/49 Installed latest patch from the Sieve maintainer.
3042 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3043 was also set, the appendfile transport was unnecessarily scanning a
3044 directory of message files (e.g. for maildir delivery) to find the count
3045 of files (along with the size), even though it did not need this
3046 information. It now does the scan only if it needs to find either the
3047 size of the count of files.
3049 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3051 PH/52 Two bugs concerned with error handling when the smtp transport is
3054 (i) Exim was not creating retry information for temporary errors given
3055 for individual recipients after the DATA command when the smtp transport
3056 was used in LMTP mode. This meant that they could be retried too
3057 frequently, and not timed out correctly.
3059 (ii) Exim was setting the flag that allows error details to be returned
3060 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3061 recipients that were returned after the DATA command.
3063 PH/53 This is related to PH/52, but is more general: for any failing address,
3064 when detailed error information was permitted to be returned to the
3065 sender, but the error was temporary, then after the final timeout, only
3066 "retry timeout exceeded" was returned. Now it returns the full error as
3067 well as "retry timeout exceeded".
3069 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3070 do this, and (what is worse) MTAs that accept it.
3072 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3073 will now be deprecated.
3075 PH/56 New os.c-cygwin from the Cygwin maintainer.
3077 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3078 formats (previously only available in exim formats via -bpr, -bpru,
3079 and -bpra. Now also available in native and exiqgrep formats)
3081 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3082 with very large, slow to parse queues
3084 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3086 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3088 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3089 responses to authentication challenges, though it was showing the
3090 challenges; (ii) I've removed the CR characters from the debug output for
3093 PH/58 Allow for the insertion of a newline as well as a space when a string
3094 is turned into more than one encoded-word during RFC 2047 encoding. The
3095 Sieve code now uses this.
3097 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3098 data_4xx, lost_connection, tls_required.
3100 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3101 message was being sent as an SMTP response.
3103 PH/61 Add -l and -k options to exicyclog.
3105 PH/62 When verifying, if an address was redirected to one new address, so that
3106 verification continued, and the new address failed or deferred after
3107 having set something in $address_data, the value of $address_data was not
3108 passed back to the ACL. This was different to the case when no
3109 redirection occurred. The value is now passed back in both cases.
3111 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3112 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3113 use login_cap.h, so on its own it isn't the distinguishing feature. The
3114 new name refers directly to the setclassresources() function.
3116 PH/65 Added configuration files for NetBSD3.
3118 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3120 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3121 is preferred over IPv4.
3123 PH/68 The bounce_return_message and bounce_return_body options were not being
3124 honoured for bounces generated during the reception of non-SMTP messages.
3125 In particular, this applied to messages rejected by the ACL. This bug has
3126 been fixed. However, if bounce_return_message is true and bounce_return_
3127 body is false, the headers that are returned for a non-SMTP message
3128 include only those that have been read before the error was detected.
3129 (In the case of an ACL rejection, they have all been read.)
3131 PH/69 The HTML version of the specification is now built in a directory called
3132 spec_html instead of spec.html, because the latter looks like a path with
3133 a MIME-type, and this confuses some software.
3135 PH/70 Catch two compiler warnings in sieve.c.
3137 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3138 function verify_get_ident() calls ip_connect() to connect a socket, but
3139 if the "connect()" function timed out, ip_connect() used to close the
3140 socket. However, verify_get_ident() also closes the socket later, and in
3141 between Exim writes to the log, which may get opened at this point. When
3142 the socket was closed in ip_connect(), the log could get the same file
3143 descriptor number as the socket. This naturally causes chaos. The fix is
3144 not to close the socket in ip_connect(); the socket should be closed by
3145 the function that creates it. There was only one place in the code where
3146 this was missing, in the iplookup router, which I don't think anybody now
3147 uses, but I've fixed it anyway.
3149 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3150 well as to direct DNS lookups. Otherwise the handling of names in host
3151 lists is inconsistent and therefore confusing.
3157 PH/01 Two changes to the default runtime configuration:
3159 (1) Move the checks for relay_from_hosts and authenticated clients from
3160 after to before the (commented out) DNS black list checks.
3162 (2) Add control=submission to the relay_from_hosts and authenticated
3163 clients checks, on the grounds that messages accepted by these
3164 statements are most likely to be submissions.
3166 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3168 (1) Generate an error if the third argument for the ${prvs expansion is
3171 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3174 (3) Reset the variables that are obtained from the first argument of
3175 ${prvscheck and used in the second argument before leaving the code,
3176 because their memory is reclaimed, so using them afterwards may do
3179 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3180 one (it's much easier than Tom thought :-).
3182 (5) Because of (4), we can now allow for the use of $prvscheck_result
3183 inside the third argument.
3185 PH/03 For some reason, the default setting of PATH when running a command from
3186 a pipe transport was just "/usr/bin". I have changed it to
3189 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3190 anything to be listed in the output from -bV.
3192 PH/05 When a filter generated an autoreply, the entire To: header line was
3193 quoted in the delivery log line, like this:
3195 => >A.N.Other <ano@some.domain> <original@ddress> ...
3197 This has been changed so that it extracts the operative address. There
3198 may be more than one such address. If so, they are comma-separated, like
3201 => >ano@some.domain,ona@other.domain <original@ddress> ...
3203 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3204 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3205 not being looked up in the rDNS to get a host name, Exim was showing the
3206 IP address twice in Received: lines, even though the IP addresses were
3207 identical. For example:
3209 Received: from [1.2.3.4] (helo=[1.2.3.4])
3211 However, if the real host name was known, it was omitting the HELO data
3212 if it matched the actual IP address. This has been tidied up so that it
3213 doesn't show the same IP address twice.
3215 PH/07 When both +timestamp and +memory debugging was on, the value given by
3216 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3217 called by the debug printing, thereby overwriting the timestamp buffer.
3218 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3220 PH/08 When the original message was included in an autoreply transport, it
3221 always said "this is a copy of the message, including all the headers",
3222 even if body_only or headers_only was set. It now gives an appropriate
3225 PH/09 Applied a patch from the Sieve maintainer which:
3227 o fixes some comments
3228 o adds the (disabled) notify extension core
3229 o adds some debug output for the result of if/elsif tests
3230 o points to the current vacation draft in the documentation
3231 and documents the missing references header update
3235 o fixes a bug in processing the envelope test (when testing
3236 multiple envelope elements, the last element determined the
3239 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3240 Electronic Mail") by including:
3242 Auto-submitted: auto-generated
3244 in the messages that it generates (bounce messages and others, such as
3245 warnings). In the case of bounce messages for non-SMTP messages, there was
3246 also a typo: it was using "Auto_submitted" (underscore instead of
3247 hyphen). Since every message generated by Exim is necessarily in response
3248 to another message, thes have all been changed to:
3250 Auto-Submitted: auto-replied
3252 in accordance with these statements in the RFC:
3254 The auto-replied keyword:
3256 - SHOULD be used on messages sent in direct response to another
3257 message by an automatic process,
3259 - MUST NOT be used on manually-generated messages,
3261 - MAY be used on Delivery Status Notifications (DSNs) and Message
3262 Disposition Notifications (MDNs),
3264 - MUST NOT be used on messages generated by automatic or periodic
3265 processes, except for messages which are automatic responses to
3268 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3269 to the default Received: header definition.
3271 PH/12 Added log selector acl_warn_skipped (default on).
3273 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3274 variables because (a) they are in the wrong storage pool and (b) even if
3275 they were copied, it wouldn't work properly because of the caching.
3277 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3278 checking when decoding. Apparently there are clients that generate
3279 overlong encoded strings. Why am I not surprised?
3281 PH/15 If the first argument of "${if match_address" was not empty, but did not
3282 contain an "@" character, Exim crashed. Now it writes a panic log message
3283 and treats the condition as false.
3285 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3287 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3288 extension "envelope-auth". The code is finished and in agreement with
3289 other implementations, but there is no documentation so far and in fact,
3290 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3291 not changing the active code.
3293 Print executed "if" and "elsif" statements when debugging is used. This
3294 helps a great deal to understand what a filter does.
3296 Document more things not specified clearly in RFC3028. I had all this
3297 sorted out, when out of a sudden new issues came to my mind. Oops."
3299 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3302 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3303 canonical form (as documented). However, after a host name lookup from
3304 the IP address, check_host() was doing a simple string comparison with
3305 addresses acquired from the DNS when checking that the found name did
3306 have the original IP as one of its addresses. Since any found IPv6
3307 addresses are likely to be in abbreviated form, the comparison could
3308 fail. Luckily, there already exists a function for doing the comparison
3309 by converting both addresses to binary, so now that is used instead of
3310 the text comparison.
3312 PH/20 There was another similar case to PH/19, when a complete host name was
3313 given in a host list; looking up its IP address could give an abbreviated
3314 form, whereas the current host's name might or might not be abbreviated.
3315 The same fix has been applied.
3321 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3322 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3325 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3326 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3328 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3330 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3331 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3332 Also, for "Auto-Submitted", treat anything other than "no" as
3333 non-personal, in accordance with RFC 3834. (Previously it treated
3334 anything starting "auto-" as non-personal.)
3336 TF/01 The control=submission/name=... option had a problem with syntax
3337 errors if the name included a slash character. The /name= option
3338 now slurps the rest of the string, so it can include any characters
3339 but it must come last in the list of options (after /sender_retain
3342 PH/05 Some modifications to the interface to the fake nameserver for the new
3350 TK/01 Added the "success_on_redirect" address verification option. See
3351 NewStuff for rationale and an example.
3353 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3355 PH/02 Patch to exigrep to allow it to work on syslog lines.
3357 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3358 fread() to skip over the body file's header line, because in Cygwin the
3359 header line is locked and is inaccessible.
3361 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3362 co-exist for some time) to make it clear that it is the Exim ID that is
3363 referenced, not the Message-ID: header line.
3365 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3366 string_format() function, because snprintf() does not exist on all
3369 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3370 ${stat: expansion item.
3372 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3373 protocol synchronization error", to keep the pedants happy.
3375 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3376 well as for IRIX systems, when gcc is being used. See the host.c source
3379 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3381 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3384 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3385 transport and to the smtp transport in LMTP mode.
3387 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3389 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3390 run a filter in a subprocess. This could lead to confusion in subsequent
3391 lookups in the parent process. There should also be a search_tidyup() at
3392 the end of the subprocess.
3394 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3395 only if the host matched helo_try_verify_hosts, which caused the
3396 verification to occur when the EHLO/HELO command was issued. The ACL just
3397 tested the remembered result. Now, if a previous verification attempt has
3398 not happened, "verify = helo" does it there and then.
3400 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3402 TK/03 Fix log output including CR from clamd.
3404 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3405 exist provoked a memory error which could cause a segfault.
3407 PH/15 Installed PCRE 6.2
3409 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3411 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3412 of the problem. Specifically, suggested +O2 rather than +O1 for the
3415 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3417 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3418 record, the port was not correctly logged when the outgoing_port log
3419 selector was set (it logged the transort's default port).
3421 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3422 fallback_hosts, and "hosts" in the smtp transport.
3424 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3425 host errors such as "Connection refused".
3427 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3428 authentication with radiusclient 0.4.9:
3430 - Error returned from rc_read_config was caught wrongly
3431 - Username/password not passed on to radius server due to wrong length.
3433 The presumption is that some radiusclient API changes for 4.51/PH/17
3434 were not taken care of correctly. The code is still untested by me (my
3435 Linux distribution still has 0.3.2 of radiusclient), but it was
3436 contributed by a Radius user.
3438 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3439 expanding the "port" option of the smtp transport.
3441 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3442 while reading a MIME header. Thanks to Tom Hughes for a patch.
3444 PH/24 Include config.h inside local_scan.h so that configuration settings are
3447 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3448 that in an ACL for RCPT (for example), you can examine exactly what was
3451 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3452 commands, but it was not correctly comparing the address with the actual
3453 client host address. Thus, it would show the EHLO address in Received:
3454 header lines when this was not necessary.
3456 PH/27 Added the % operator to ${eval:}.
3458 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3459 it should be ignoring failures (because with -C, for example, it has lost
3460 privilege). It wasn't ignoring creation failures other than "already
3463 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3466 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3467 because an input line was too long, either on its own, or by virtue of
3468 too many continuations, the temporary file was not being removed, and the
3469 return code was incorrect.
3471 PH/31 Missing "BOOL" in function definition in filtertest.c.
3473 PH/32 Applied Sieve patches from the maintainer.
3475 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3477 PH/33 Added "verify = not_blind".
3479 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3480 Local/Makefile (with some defaults set). These are used in built scripts
3481 such as exicyclog, but they have never been used in the exim_install
3482 script (though there are many overriding facilities there). I have
3483 arranged that the exim_install script now takes note of these two
3486 PH/35 Installed configuration files for Dragonfly.
3488 PH/36 When a locally submitted message by a trusted user did not contain a
3489 From: header, and the sender address was obtained from -f or from an SMTP
3490 MAIL command, and the trusted user did not use -F to supply a sender
3491 name, $originator_name was incorrectly used when constructing a From:
3492 header. Furthermore, $originator_name was used for submission mode
3493 messages from external hosts without From: headers in a similar way,
3494 which is clearly wrong.
3496 PH/37 Added control=suppress_local_fixups.
3498 PH/38 When log_selector = +received_sender was set, and the addition of the
3499 sender made the log line's construction buffer exactly full, or one byte
3500 less than full, an overflow happened when the terminating "\n" was
3503 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3504 when the result of a list match is failure because a DNS lookup failed.
3506 PH/40 RM_COMMAND is now used in the building process.
3508 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3509 the "build-* directories that it finds.
3511 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3512 address in a domain literal was a prefix of an interface address.
3514 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3515 when verifying a sender address, unless rewrite_headers is false.
3517 PH/44 Wrote a long comment about why errors_to addresses are verified as
3518 recipients, not senders.
3520 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3521 the ratelimit ACL was added.
3523 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3525 PH/46 Added extra information about PostgreSQL errors to the error string.
3527 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3528 avoiding the need to install special zones in a real server. This is
3529 backwards compatible; if it can't find the fake resolver, it drops back.
3530 Thus, both old and new test suites can be run.
3532 TF/02 Added util/ratelimit.pl
3534 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3535 clock is set back in time.
3537 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3538 Candler <B.Candler@pobox.com>.
3540 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3541 OK for addresses that are the result of redirections.
3543 PH/48 A number of further additions for the benefit of the new test suite,
3544 including a fake gethostbyname() that interfaces to the fake DNS resolver
3547 TF/06 The fix for widen_domains has also been applied to qualify_single and
3548 search_parents which are the other dnslookup options that can cause
3551 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3554 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3556 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3557 service" if it encountered an MX record with an empty target hostname.
3558 The message is now "an MX or SRV record indicated no SMTP service".
3560 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3561 if the DNS of the sending site is misconfigured. This is quite a
3562 common situation. This change restores the behaviour of treating a
3563 helo verification defer as a failure.
3565 PH/51 If self=fail was set on a router, the bounce message did not include the
3566 actual error message.
3572 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3574 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3575 command itself ended in error, the underlying message about the transport
3576 filter timeout was being overwritten with the pipe command error. Now the
3577 underlying error message should be appended to the second error message.
3579 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3581 PH/02 Reduce the amount of output that "make" produces by default. Full output
3582 can still be requested.
3584 PH/03 The warning log line about a condition test deferring for a "warn" verb
3585 was being output only once per connection, rather than after each
3586 occurrence (because it was using the same function as for successful
3587 "warn" verbs). This seems wrong, so I have changed it.
3589 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3590 it should not have, which might have caused a crash in the right
3591 circumstances, but probably never did.
3593 PH/04 Installed a modified version of Tony Finch's patch to make submission
3594 mode fix the return path as well as the Sender: header line, and to
3595 add a /name= option so that you can make the user's friendly name appear
3598 TF/03 Added the control = fakedefer ACL modifier.
3600 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3601 Mark Lowes for thorough testing.
3603 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3605 TK/03 Merged latest SRS patch from Miles Wilton.
3607 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3608 with the definition in sysexits.h (which is #included earlier).
3609 Fortunately, Exim does not actually use EX_OK. The code used to try to
3610 preserve the sysexits.h value, by assuming that macro definitions were
3611 scanned for macro replacements. I have been disabused of this notion,
3612 so now the code just undefines EX_OK before #including unistd.h.
3614 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3615 in the smtp transport. When a block could not be written in a single
3616 write() function, the timeout was being re-applied to each part-write.
3617 This seems wrong - if the receiver was accepting one byte at a time it
3618 would take for ever. The timeout is now adjusted when this happens. It
3619 doesn't have to be particularly precise.
3621 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3622 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3624 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3625 without a domain if the check to <postmaster@domain> fails.
3627 SC/01 Eximstats: added -xls and the ability to specify output files
3628 (patch written by Frank Heydlauf).
3630 SC/02 Eximstats: use FileHandles for outputting results.
3632 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3634 SC/04 Eximstats: fixed display of large numbers with -nvr option
3636 SC/05 Eximstats: fixed merging of reports with empty tables.
3638 SC/06 Eximstats: added the -include_original_destination flag
3640 SC/07 Eximstats: removed tabs and trailing whitespace.
3642 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3644 TK/06 MBOX spool code: Add real "From " MBOX separator line
3645 so the .eml file is really in mbox format (even though
3646 most programs do not really care). Patch from Alex Miller.
3648 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3649 The latter is generated from $received_to and is only set if the
3650 message has one envelope recipient. SA can use these headers,
3651 obviously out-of-the-box. Patch from Alex Miller.
3653 PH/08 The ${def test on a variable was returning false if the variable's
3654 value was "0", contrary to what the specification has always said!
3655 The result should be true unless the variable is empty.
3657 PH/09 The syntax error of a character other than { following "${if
3658 def:variable_name" (after optional whitespace) was not being diagnosed.
3659 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3660 accidental colon was present, for example, could give incorrect results.
3662 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3663 result is used (not including appendfile, where other changes are about
3666 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3667 This involved changing a lot of size variables from int to off_t. It
3668 should work with maildirs and everything.
3670 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3671 spamd dying while we are connected to it.
3673 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3676 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3677 function for BDB 4.3.
3679 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3681 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3682 the sources for PCRE have been re-arranged and more files are now
3685 PH/15 The code I had for printing potentially long long variables in PH/11
3686 above was not the best (it lost precision). The length of off_t variables
3687 is now inspected at build time, and an appropriate printing format (%ld
3688 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3689 to be "long long int" or "long int". This is needed for the internal
3690 formatting function string_vformat().
3692 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3693 the configuration file to be ":syslog", then the script "guesses" where
3694 the logs files are, rather than using the compiled in default. In our
3695 case the guess is not the same as the compiled default, so the script
3696 suddenly stopped working when I started to use syslog. The patch checks
3697 to see if log_file_path is "". If so, it attempts to read it from exim
3698 with no configuration file to get the compiled in version, before it
3699 falls back to the previous guessing code."
3701 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3702 implementing BATV in an Exim configuration. See NewStuff for the gory
3705 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3706 Makefile that are specific to HP-UX.
3708 PH/18 If the "use_postmaster" option was set for a recipient callout together
3709 with the "random" option, the postmaster address was used as the MAIL
3710 FROM address for the random test, but not for the subsequent recipient
3711 test. It is now used for both.
3713 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3714 patch removes a few documentation additions to RFC 3028, because the
3715 latest draft now contains them. It adds the new en;ascii-case comparator
3716 and a new error check for 8bit text in MIME parts. Comparator and
3717 require names are now matched exactly. I enabled the subaddress
3718 extension, but it is not well tested yet (read: it works for me)."
3720 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3721 rework some of the code of TK/09 above to avoid the hardwired use of
3722 "%lld" and "long long". Replaced the call to snprintf() with a call to
3725 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3726 records point to non-existent hosts", "retry timeout exceeded", and
3727 "retry time not reached for any host after a long failure period".
3729 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3730 experimental DomainKeys support:
3732 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3733 (2) On an error such as an illegally used "control", the wrong name for
3734 the control was given.
3736 These problems did NOT occur unless DomainKeys support was compiled.
3738 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3740 PH/24 Added ${if match_ip condition.
3742 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3743 databases so that it will be absolutely obvious if a crash occurs in the
3744 DB library. This is a regular occurrence (often caused by mis-matched
3747 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3748 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3749 picked up on a user's system that detects such things. There doesn't seem
3750 to be a gcc warning option for this - only an attribute that has to be
3751 put on the function's prototype. It seems that in Fedora Core 4 they have
3752 set this on a number of new functions. No doubt there will be more in due
3755 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3756 specify a transport. However, if an address that was verified by such a
3757 router was the subject of a callout, Exim crashed because it tried to
3758 read the rcpt_include_affixes from the non-existent transport. Now it
3759 just assumes that the setting of that option is false. This bug was
3760 introduced by 4.51/PH/31.
3762 PH/28 Changed -d+all to exclude +memory, because that information is very
3763 rarely of interest, but it makes the output a lot bigger. People tend to
3764 do -d+all out of habit.
3766 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3767 code in os-type was giving problems when libc.so lives in lib64, like on
3770 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3771 aren't the modern standard, and it seems that some systems' include files
3772 don't always have them. Exim was already checking for some of the newer
3773 ones like T_AAAA, and defining it itself. I've added checks for all the
3774 record types that Exim uses.
3776 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3777 not automatically generating a new one, as it is supposed to. This
3778 prevented TLS from working. If the file did exist, but contained invalid
3779 data, a new version was generated, as expected. It was only the case of a
3780 non-existent file that was broken.
3782 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3783 with a change in libdomainkeys > 0.64.
3785 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3786 from DNS. If the selector record carries the flag, it now has
3787 precedence over the domain-wide flag.
3789 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3791 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3792 the use of a port name that isn't defined in /etc/services) occurred, the
3793 message was deferred as in a normal delivery, and thus remained on the
3794 spool, instead of being failed because of the mua_wrapper setting. This
3795 is now fixed, and I tidied up some of the mua_wrapper messages at the
3798 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3799 the messages in a hash of arrays rather than using individual hashes.
3800 This is a bit cleaner and results in dramatic memory savings, albeit
3801 at a slight CPU cost.
3803 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3804 as requested by Marc Sherman.
3806 SC/10 Eximstats: added histograms for user specified patterns as requested
3809 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3811 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3812 fopen() in the content-scanning modules that did not already have it.
3818 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3819 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3821 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3823 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3825 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3826 to test Sieve filters that use "vacation".
3828 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3829 that changes the way the GnuTLS parameters are stored in the cache file.
3830 The new format can be generated externally. For backward compatibility,
3831 if the data in the cache doesn't make sense, Exim assumes it has read an
3832 old-format file, and it generates new data and writes a new file. This
3833 means that you can't go back to an older release without removing the
3836 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3837 work if there are any delivery delays because "one_time" forces the
3838 parent to be marked "delivered", so its unseen clone is never tried
3839 again. For this reason, Exim now forbids the simultaneous setting of
3842 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3843 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3844 entry if you want to know the details. The fix, however, neglected to
3845 consider the case when local delivery batching is involved. The test for
3846 "previously delivered" was not happening when checking to see if an
3847 address could be batched with a previous (undelivered) one; under
3848 certain circumstances this could lead to multiple deliveries to the same
3851 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3852 in its include files, and this causes problems building Exim.
3854 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3855 header_syntax) but Exim was just ignoring anything given after a slash.
3856 In particular, this caused confusion with an attempt to use "verify =
3857 reverse_host_lookup/defer_ok". An error is now given when options are
3858 supplied for verify items that do not have them. (Maybe reverse_host_
3859 lookup should have a defer_ok option, but that's a different point.)
3861 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3862 defined by RFC 821) to 2048, because there were problems with some AUTH
3863 commands, and RFC 1869 says the size should be increased for extended
3864 SMTP commands that take arguments.
3866 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3869 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3870 "unknown" error; now it says that the functionality isn't in the binary.
3872 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3873 an address' error message when a string expansion fails (syntax or
3874 whatever). Otherwise the password may appear in the log. Following change
3875 PH/42 below, there is no longer a chance of it appearing in a bounce
3878 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3880 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3881 of its addresses was ever tried. (Bugzilla bug #2.)
3883 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3884 the result incorrectly in the debug output. (It correctly added a newline
3885 to what was transported.)
3887 TF/01 Added $received_time.
3889 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3890 commented out examples of how to interface to a virus scanner and to
3891 SpamAssassin. Also added commented examples of av_scanner and
3892 spamd_address settings.
3894 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3895 and controls are allowed in which ACLs. There were a couple of minor
3896 errors. Some of the entries in the conditions table (which is a table of
3897 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3898 negation of where the condition IS allowed.
3900 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3902 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3903 header file does not have a version number, so I've had to invent a new
3904 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3905 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3906 radiusclient), but it was contributed by a Radius user.
3908 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3909 files or directories, for OpenSSL.
3911 PH/19 When an Exim process that is running as root has to create an Exim log
3912 file, it does so in a subprocess that runs as exim:exim so as to get the
3913 ownership right at creation (otherwise, other Exim processes might see
3914 the file with the wrong ownership). There was no test for failure of this
3915 fork() call, which would lead to the process getting stuck as it waited
3916 for a non-existent subprocess. Forks do occasionally fail when resources
3917 run out. I reviewed all the other calls to fork(); they all seem to check
3920 PH/20 When checking for unexpected SMTP input at connect time (before writing
3921 the banner), Exim was not dealing correctly with a non-positive return
3922 from the read() function. If the client had disconnected by this time,
3923 the result was a log entry for a synchronization error with an empty
3924 string after "input=" when read() returned zero. If read() returned -1
3925 (an event I could not check), uninitialized data bytes were printed.
3926 There were reports of junk text (parts of files, etc) appearing after
3929 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3931 PH/22 Added support for macro redefinition, and (re)definition in between
3932 driver and ACL definitions.
3934 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3935 forgetting to use the resulting value; it was using the unexpanded value.
3937 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3938 hadn't been configured. The fix is from Juergen Kreileder, who
3939 understands it better than I do:
3941 "Here's what I see happening with three configured cyrus_sasl
3942 authenticators configured (plain, login, cram-md5):
3944 On startup auth_cyrus_sasl_init() gets called for each of these.
3945 This means three calls to sasl_listmech() without a specified mech_list.
3946 => SASL tests which mechs of all available mechs actually work
3947 => three warnings about OTP not working
3948 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3950 With the patch, sasl_listmech() also gets called three times. But now
3951 SASL's mech_list option is set to the server_mech specified in the the
3952 authenticator. Or in other words, the answer from sasl_listmech()
3953 gets limited to just the mech you're testing for (which is different
3955 => the return list contains just 'plain' or 'login', 'cram-md5' or
3956 nothing depending on the value of ob->server_mech.
3958 I've just tested the patch: Authentication still works fine,
3959 unavailable mechs specified in the exim configuration are still
3960 caught, and the auth.log warnings about OTP are gone."
3962 PH/25 When debugging is enabled, the contents of the command line are added
3963 to the debugging output, even when log_selector=+arguments is not
3966 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3967 answer is "GNU", and only if the return is "GNU/something" is the answer
3970 PH/27 $acl_verify_message is now set immediately after the failure of a
3971 verification in an ACL, and so is available in subsequent modifiers. In
3972 particular, the message can be preserved by coding like this:
3974 warn !verify = sender
3975 set acl_m0 = $acl_verify_message
3977 Previously, $acl_verify_message was set only while expanding "message"
3978 and "log_message" when a very denied access.
3980 PH/28 Modified OS/os.c-Linux with
3982 -#ifndef OS_LOAD_AVERAGE
3983 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3985 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3986 nomenclature these days.)
3988 PH/29 Installed patch from the Sieve maintainer that adds the options
3989 sieve_useraddress and sieve_subaddress to the redirect router.
3991 PH/30 In these circumstances:
3992 . Two addresses routed to the same list of hosts;
3993 . First host does not offer TLS;
3994 . First host accepts first address;
3995 . First host gives temporary error to second address;
3996 . Second host offers TLS and a TLS session is established;
3997 . Second host accepts second address.
3998 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3999 and peerdn, if requested) that were in fact used only for the second
4002 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4003 attention to any local part prefix or suffix that was matched by the
4004 router that accepted the address. It now behaves in the same way as it
4005 does for delivery: the affixes are removed from the local part unless
4006 rcpt_include_affixes is set on the transport.
4008 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4009 timeout during the DATA phase of an incoming message.
4011 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4012 applied a patch sanctioned by the Sieve maintainer.
4014 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4015 the uid or gid is negative. A case of a negative gid caused this to be
4016 noticed. The fix allows for either to be negative.
4018 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4019 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4022 PH/36 The change PH/12 above was broken. Fixed it.
4024 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4025 the grounds that routing the same address twice would always produce the
4026 same answer. This might have been true once, but it is certainly no
4027 longer true now. Routing a child address may depend on the previous
4028 routing that produced that child. Some complicated redirection strategies
4029 went wrong when messages had multiple recipients, and made Exim's
4030 behaviour dependent on the order in which the addresses were given.
4032 I have moved the duplicate checking until after the routing is complete.
4033 Exim scans the addresses that are assigned to local and remote
4034 transports, and removes any duplicates. This means that more work will be
4035 done, as duplicates will always all be routed, but duplicates are
4036 presumably rare, so I don't expect this is of any significance.
4038 For deliveries to pipes, files, and autoreplies, the duplicate checking
4039 still happens during the routing process, since they are not going to be
4042 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4043 It corrects a timeout issue with spamd. This is Ian's comment: "The
4044 background is that sometimes spamd either never reads data from a
4045 connection it has accepted, or it never writes response data. The exiscan
4046 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4047 blindly assumes that writes won't block so it may never time out."
4049 PH/39 Allow G after quota size as well as K and M.
4051 PH/40 The value set for $authenticated_id in an authenticator may not contain
4052 binary zeroes or newlines because the value is written to log lines and
4053 to spool files. There was no check on this. Now the value is run through
4054 the string_printing() function so that such characters are converted to
4055 printable escape sequences.
4057 PH/41 $message_linecount is a new variable that contains the total number of
4058 lines in the message. Compare $body_linecount, which is the count for the
4061 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4062 bounce and delay warning messages, except in certain special cases, which
4065 (a) An SMTP error message from a remote host;
4066 (b) A message specified in a :fail: redirection;
4067 (c) A message specified in a "fail" command in a system filter;
4068 (d) A message specified in a FAIL return from the queryprogram router;
4069 (e) A message specified by the cannot_route_message router option.
4071 In these cases only, Exim does include the error details in bounce and
4072 warning messages. There are also a few cases where bland messages such
4073 as "unrouteable address" or "local delivery error" are given.
4075 PH/43 $value is now also set for the "else" part of a ${run expansion.
4077 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4078 being worked on, but at least Exim now implements the latest version to
4081 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4082 process to complete was treated as a delivery failure, a timeout while
4083 writing the message to the pipe was logged, but erroneously treated as a
4084 successful delivery. Such timeouts include transport filter timeouts. For
4085 consistency with the overall process timeout, these timeouts are now
4086 treated as errors, giving rise to delivery failures by default. However,
4087 there is now a new Boolean option for the pipe transport called
4088 timeout_defer, which, if set TRUE, converts the failures into defers for
4089 both kinds of timeout. A transport filter timeout is now identified in
4092 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4093 systems where "make" and "gmake" are different, calling "gmake" at top
4094 level broke things. I've arranged for the value of $(MAKE) to be passed
4095 from the Makefile to this script so that it can call the same version of
4099 A note about Exim versions 4.44 and 4.50
4100 ----------------------------------------
4102 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4103 changes of various kinds. As a consequence, a big documentation update was
4104 needed. This delayed the release for rather longer than seemed good, especially
4105 in the light of a couple of (minor) security issues. Therefore, the changes
4106 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4107 release. So 4.44 and 4.50 are in effect two different branches that both start
4110 I have left the 4.50 change log unchanged; it contains all the changes since
4111 4.43. The change log for 4.44 is below; many of its items are identical to
4112 those for 4.50. This seems to be the most sensible way to preserve the
4113 historical information.
4119 1. Minor wording change to the doc/README.SIEVE file.
4121 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4122 computation of the current number of files was incorrect.
4124 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4125 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4128 4. Give more explanation in the error message when the command for a transport
4129 filter fails to execute.
4131 5. There are several places where Exim runs a non-Exim command in a
4132 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4133 was being done only for the command run by the queryprogram router. It is
4134 now done for all such subprocesses. The other cases are: ${run, transport
4135 filters, and the commands run by the lmtp and pipe transports.
4137 6. Added CONFIGURE_GROUP build-time option.
4139 7. Some older OS have a limit of 256 on the maximum number of file
4140 descriptors. Exim was using setrlimit() to set 1000 as a large value
4141 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4142 systems. I've change it so that if it can't get 1000, it tries for 256.
4144 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4145 was an oversight, and furthermore, ever since the addition of extra
4146 controls (e.g. 4.43/32), the checks on when to allow different forms of
4147 "control" were broken. There should now be diagnostics for all cases when a
4148 control that does not make sense is encountered.
4150 9. Added the /retain_sender option to "control=submission".
4152 10. $recipients is now available in the predata ACL (oversight).
4154 11. Tidy the search cache before the fork to do a delivery from a message
4155 received from the command line. Otherwise the child will trigger a lookup
4156 failure and thereby defer the delivery if it tries to use (for example) a
4157 cached ldap connection that the parent has called unbind on.
4159 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4160 of $address_data from the recipient verification was clobbered by the
4161 sender verification.
4163 13. The value of address_data from a sender verification is now available in
4164 $sender_address_data in subsequent conditions in the ACL statement.
4166 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4168 15. Added a new option "connect=<time>" to callout options, to set a different
4171 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4172 was its contents. (It was OK if the option was not defined at all.)
4174 17. A "Completed" log line is now written for messages that are removed from
4175 the spool by the -Mrm option.
4177 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4178 information about exactly what failed.
4180 19. Added -dd to debug only the daemon process.
4182 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4183 handles timeouts, both on the server side and network timeouts. Renamed the
4184 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4186 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4187 It is now set to "smtps".
4189 22. $host_address is now set to the target address during the checking of
4190 ignore_target_hosts.
4192 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4193 being passed; this would have caused $sender_host_name to have been used if
4194 matching the list had actually called for a host name (not very likely,
4195 since this list is usually IP addresses). A host name is now passed as
4198 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4199 code that specifies a non-restarting handler (typically sigaction() in
4200 modern systems) in an attempt to fix a rare and obscure crash bug.
4202 25. Narrowed the window for a race in the daemon that could cause it to ignore
4203 SIGCHLD signals. This is not a major problem, because they are used only to
4204 wake it up if nothing else does.
4206 26. A malformed maildirsize file could cause Exim to calculate negative values
4207 for the mailbox size or file count. Odd effects could occur as a result.
4208 The maildirsize information is now recalculated if the size or filecount
4211 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4212 support for a long time. Removed HAVE_SYS_VFS_H.
4214 28. Installed the latest version of exipick from John Jetmore.
4216 29. In an address list, if the pattern was not a regular expression, an empty
4217 subject address (from a bounce message) matched only if the pattern was an
4218 empty string. Non-empty patterns were not even tested. This was the wrong
4219 because it is perfectly reasonable to use an empty address as part of a
4220 database query. An empty address is now tested by patterns that are
4221 lookups. However, all the other forms of pattern expect the subject to
4222 contain a local part and a domain, and therefore, for them, an empty
4223 address still always fails if the pattern is not itself empty.
4225 30. Exim went into a mad DNS loop when attempting to do a callout where the
4226 host was specified on an smtp transport, and looking it up yielded more
4227 than one IP address.
4229 31. Re-factored the code for checking spool and log partition space into a
4230 function that finds that data and another that does the check. The former
4231 is then used to implement four new variables: $spool_space, $log_space,
4232 $spool_inodes, and $log_inodes.
4234 32. The RFC2047 encoding function was originally intended for short strings
4235 such as real names; it was not keeping to the 75-character limit for
4236 encoded words that the RFC imposes. It now respects the limit, and
4237 generates multiple encoded words if necessary. To be on the safe side, I
4238 have increased the buffer size for the ${rfc2047: expansion operator from
4241 33. It is now permitted to omit both strings after an "if" condition; if the
4242 condition is true, the result is "true". As before, when the second string
4243 is omitted, a false condition yields an empty string. This makes it less
4244 cumbersome to write custom ACL and router conditions.
4246 34. Failure to deliver a bounce message always caused it to be frozen, even if
4247 there was an errors_to setting on the router. The errors_to setting is now
4250 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4251 canonical form (fully expanded) before being placed in
4252 $sender_host_address.
4254 36. The table in the code that translates DNS record types into text (T_A to
4255 "A" for instance) was missing entries for NS and CNAME. It is just possible
4256 that this could have caused confusion if both these types were looked up
4257 for the same domain, because the text type is used as part of Exim's
4258 per-process caching. But the chance of anyone hitting this buglet seems
4261 37. The dnsdb lookup has been extended in a number of ways.
4263 (1) There is a new type, "zns", which walks up the domain tree until it
4264 finds some nameserver records. It should be used with care.
4266 (2) There is a new type, "mxh", which is like "mx" except that it returns
4267 just the host names, not the priorities.
4269 (3) It is now possible to give a list of domains (or IP addresses) to be
4270 looked up. The behaviour when one of the lookups defers can be
4271 controlled by a keyword.
4273 (4) It is now possible to specify the separator character for use when
4274 multiple records are returned.
4276 38. The dnslists ACL condition has been extended: it is now possible to supply
4277 a list of IP addresses and/or domains to be looked up in a particular DNS
4280 39. Added log_selector=+queue_time_overall.
4282 40. When running the queue in the test harness, wait just a tad after forking a
4283 delivery process, to get repeatability of debugging output.
4285 41. Include certificate and key file names in error message when GnuTLS fails
4286 to set them up, because the GnuTLS error message doesn't include the name
4287 of the failing file when there is a problem reading it.
4289 42. Allow both -bf and -bF in the same test run.
4291 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4293 44. The "Exiscan patch" is now merged into the mainline Exim source.
4295 45. Sometimes the final signoff response after QUIT could fail to get
4296 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4297 < 0 before doing a fflush(). This bug looks as though it goes back to the
4298 introduction of TLS in release 3.20, but "sometimes" must have been rare
4299 because the tests only now provoked it.
4301 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4302 (this can affect the format of dates).
4304 47. exim_tidydb, when checking for the continued existence of a message for
4305 which it has found a message-specific retry record, was not finding
4306 messages that were in split spool directories. Consequently, it was
4307 deleting retry records that should have stayed in existence.
4309 48. Steve fixed some bugs in eximstats.
4311 49. The SPA authentication driver was not abandoning authentication and moving
4312 on to the next authenticator when an expansion was forced to fail,
4313 contradicting the general specification for all authenticators. Instead it
4314 was generating a temporary error. It now behaves as specified.
4316 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4317 (the order specifies the preference for clients). The order is now AES256,
4318 AES128, 3DES, ARCFOUR128.
4320 51. Small patch to Sieve code - explicitly set From: when generating an
4323 52. Exim crashed if a remote delivery caused a very long error message to be
4324 recorded - for instance if somebody sent an entire SpamAssassin report back
4325 as a large number of 550 error lines. This bug was coincidentally fixed by
4326 increasing the size of one of Exim's internal buffers (big_buffer) that
4327 happened as part of the Exiscan merge. However, to be on the safe side, I
4328 have made the code more robust (and fixed the comments that describe what
4331 53. Now that there can be additional text after "Completed" in log lines (if
4332 the queue_time_overall log selector is set), a one-byte patch to exigrep
4333 was needed to allow it to recognize "Completed" as not the last thing in
4336 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4337 patch that reportedly fixes this has been added. I am not expert enough to
4338 create a test for it. This is what the patch creator wrote:
4340 "I found a little strange behaviour of ldap code when working with
4341 Windows 2003 AD Domain, where users was placed in more than one
4342 Organization Units. When I tried to give exim partial DN, the exit code
4343 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4344 But simultaneously result of request was absolutely normal ldap result,
4345 so I produce this patch..."
4347 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4348 REFERENCE, so I have modified the code to exclude the patch when that macro
4351 55. Some experimental protocols are using DNS PTR records for new purposes. The
4352 keys for these records are domain names, not reversed IP addresses. The
4353 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4354 leaves it alone. Component reversal etc. now happens only for IP addresses.
4357 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4359 57. Double the size of the debug message buffer (to 2048) so that more of very
4360 long debug lines gets shown.
4362 58. The exicyclog utility now does better if the number of log files to keep
4363 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4365 59. Two changes related to the smtp_active_hostname option:
4367 (1) $smtp_active_hostname is now available as a variable.
4368 (2) The default for smtp_banner uses $smtp_active_hostname instead
4369 of $primary_hostname.
4371 60. The host_aton() function is supposed to be passed a string that is known
4372 to be a valid IP address. However, in the case of IPv6 addresses, it was
4373 not checking this. This is a hostage to fortune. Exim now panics and dies
4374 if the condition is not met. A case was found where this could be provoked
4375 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4376 components; fortuitously, this particular loophole had already been fixed
4377 by change 4.50/55 above.
4379 If there are any other similar loopholes, the new check in host_aton()
4380 itself should stop them being exploited. The report I received stated that
4381 data on the command line could provoke the exploit when Exim was running as
4382 exim, but did not say which command line option was involved. All I could
4383 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4384 running as the user.
4387 61. There was a buffer overflow vulnerability in the SPA authentication code
4388 (which came originally from the Samba project). I have added a test to the
4389 spa_base64_to_bits() function which I hope fixes it.
4392 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4393 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4395 63. The daemon start-up calls getloadavg() while still root for those OS that
4396 need the first call to be done as root, but it missed one case: when
4397 deliver_queue_load_max is set with deliver_drop_privilege. This is
4398 necessary for the benefit of the queue runner, because there is no re-exec
4399 when deliver_drop_privilege is set.
4401 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4402 This has been fixed.
4404 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4405 was in use, was not putting the data itself into the right store pool;
4406 consequently, it could be overwritten for a subsequent message in the same
4407 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4410 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4412 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4413 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4414 the function were treating the return as a boolean value, which happened to
4415 work because 0=false and not-0=true, but is not correct code.
4417 68. The host_aton() function was not handling scoped IPv6 addresses (those
4418 with, for example, "%eth0" on the end) correctly.
4420 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4421 negated items (that is, ~something) in unsigned ints. Some compilers
4422 apparently mutter when there is no cast.
4424 70. If an address verification called from an ACL failed, and did not produce a
4425 user-specific message (i.e. there was only a "system" message), nothing was
4426 put in $acl_verify_message. In this situation, it now puts the system
4429 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4430 session; change 4.31/43 added the unwanted input to the log line - except
4431 that it did not do this in the start of session case. It now does.
4433 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4434 This is wrong and can cause the other end to generate a synchronization
4435 error if it is another Exim or anything else that does the synchronization
4436 check. A QUIT command is no longer sent after a timeout.
4438 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4439 during host lookups.
4441 74. The defer_ok option of callout verification was not working if it was used
4442 when verifying addresses in header lines, that is, for this case:
4444 verify = header_sender/callout=defer_ok
4446 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4447 those file descriptors could be used for SMTP connections. If anything
4448 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4449 could be sent to the SMTP client, causing chaos. The daemon now opens
4450 stdin, stdout, and stderr to /dev/null when it puts itself into the
4453 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4454 log by default. The user can override this with suitable Perl magic.
4456 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4457 the log message when discard triggers, was not working for the DATA ACL or
4458 for the non-SMTP ACL.
4460 78. Error message wording change in sieve.c.
4462 79. If smtp_accept_max_per_host was set, the number of connections could be
4463 restricted to fewer than expected, because the daemon was trying to set up
4464 a new connection before checking whether the processes handling previous
4465 connections had finished. The check for completed processes is now done
4466 earlier. On busy systems, this bug wouldn't be noticed because something
4467 else would have woken the daemon, and it would have reaped the completed
4470 80. If a message was submitted locally by a user whose login name contained one
4471 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4472 It caused a spool format error. I have fixed the spool reading code. A
4473 related problem was that the "from" clause in the Received: line became
4474 illegal because of the space(s). It is now covered by ${quote_local_part.
4476 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4479 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4481 83. Patch from Sieve maintainer to fix unterminated string problem in
4482 "vacation" handling.
4484 84. Some minor changes to the Linux configuration files to help with other
4485 OS variants using glibc.
4487 85. One more patch for Sieve to update vacation handling to latest spec.
4490 ----------------------------------------------------
4491 See the note above about the 4.44 and 4.50 releases.
4492 ----------------------------------------------------
4498 1. Change 4.43/35 introduced a bug that caused file counts to be
4499 incorrectly computed when quota_filecount was set in an appendfile
4502 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4503 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4506 3. Give more explanation in the error message when the command for a transport
4507 filter fails to execute.
4509 4. There are several places where Exim runs a non-Exim command in a
4510 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4511 was being done only for the command run by the queryprogram router. It is
4512 now done for all such subprocesses. The other cases are: ${run, transport
4513 filters, and the commands run by the lmtp and pipe transports.
4515 5. Some older OS have a limit of 256 on the maximum number of file
4516 descriptors. Exim was using setrlimit() to set 1000 as a large value
4517 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4518 systems. I've change it so that if it can't get 1000, it tries for 256.
4520 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4521 was an oversight, and furthermore, ever since the addition of extra
4522 controls (e.g. 4.43/32), the checks on when to allow different forms of
4523 "control" were broken. There should now be diagnostics for all cases when a
4524 control that does not make sense is encountered.
4526 7. $recipients is now available in the predata ACL (oversight).
4528 8. Tidy the search cache before the fork to do a delivery from a message
4529 received from the command line. Otherwise the child will trigger a lookup
4530 failure and thereby defer the delivery if it tries to use (for example) a
4531 cached ldap connection that the parent has called unbind on.
4533 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4534 of $address_data from the recipient verification was clobbered by the
4535 sender verification.
4537 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4538 was its contents. (It was OK if the option was not defined at all.)
4540 11. A "Completed" log line is now written for messages that are removed from
4541 the spool by the -Mrm option.
4543 12. $host_address is now set to the target address during the checking of
4544 ignore_target_hosts.
4546 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4547 being passed; this would have caused $sender_host_name to have been used if
4548 matching the list had actually called for a host name (not very likely,
4549 since this list is usually IP addresses). A host name is now passed as
4552 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4553 code that specifies a non-restarting handler (typically sigaction() in
4554 modern systems) in an attempt to fix a rare and obscure crash bug.
4556 15. Narrowed the window for a race in the daemon that could cause it to ignore
4557 SIGCHLD signals. This is not a major problem, because they are used only to
4558 wake it up if nothing else does.
4560 16. A malformed maildirsize file could cause Exim to calculate negative values
4561 for the mailbox size or file count. Odd effects could occur as a result.
4562 The maildirsize information is now recalculated if the size or filecount
4565 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4566 support for a long time. Removed HAVE_SYS_VFS_H.
4568 18. Updated exipick to current release from John Jetmore.
4570 19. Allow an empty sender to be matched against a lookup in an address list.
4571 Previously the only cases considered were a regular expression, or an
4574 20. Exim went into a mad DNS lookup loop when doing a callout where the
4575 host was specified on the transport, if the DNS lookup yielded more than
4578 21. The RFC2047 encoding function was originally intended for short strings
4579 such as real names; it was not keeping to the 75-character limit for
4580 encoded words that the RFC imposes. It now respects the limit, and
4581 generates multiple encoded words if necessary. To be on the safe side, I
4582 have increased the buffer size for the ${rfc2047: expansion operator from
4585 22. Failure to deliver a bounce message always caused it to be frozen, even if
4586 there was an errors_to setting on the router. The errors_to setting is now
4589 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4590 canonical form (fully expanded) before being placed in
4591 $sender_host_address.
4593 24. Updated eximstats to version 1.33
4595 25. Include certificate and key file names in error message when GnuTLS fails
4596 to set them up, because the GnuTLS error message doesn't include the name
4597 of the failing file when there is a problem reading it.
4599 26. Expand error message when OpenSSL has problems setting up cert/key files.
4602 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4603 (this can affect the format of dates).
4605 28. exim_tidydb, when checking for the continued existence of a message for
4606 which it has found a message-specific retry record, was not finding
4607 messages that were in split spool directories. Consequently, it was
4608 deleting retry records that should have stayed in existence.
4610 29. eximstats updated to version 1.35
4611 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4612 1.35 - bugfix such that pie charts by volume are generated correctly
4614 30. The SPA authentication driver was not abandoning authentication and moving
4615 on to the next authenticator when an expansion was forced to fail,
4616 contradicting the general specification for all authenticators. Instead it
4617 was generating a temporary error. It now behaves as specified.
4619 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4620 (the order specifies the preference for clients). The order is now AES256,
4621 AES128, 3DES, ARCFOUR128.
4623 31. Small patch to Sieve code - explicitly set From: when generating an
4626 32. Exim crashed if a remote delivery caused a very long error message to be
4627 recorded - for instance if somebody sent an entire SpamAssassin report back
4628 as a large number of 550 error lines. This bug was coincidentally fixed by
4629 increasing the size of one of Exim's internal buffers (big_buffer) that
4630 happened as part of the Exiscan merge. However, to be on the safe side, I
4631 have made the code more robust (and fixed the comments that describe what
4634 33. Some experimental protocols are using DNS PTR records for new purposes. The
4635 keys for these records are domain names, not reversed IP addresses. The
4636 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4637 leaves it alone. Component reversal etc. now happens only for IP addresses.
4640 34. The host_aton() function is supposed to be passed a string that is known
4641 to be a valid IP address. However, in the case of IPv6 addresses, it was
4642 not checking this. This is a hostage to fortune. Exim now panics and dies
4643 if the condition is not met. A case was found where this could be provoked
4644 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4645 components; fortuitously, this particular loophole had already been fixed
4646 by change 4.50/55 or 4.44/33 above.
4648 If there are any other similar loopholes, the new check in host_aton()
4649 itself should stop them being exploited. The report I received stated that
4650 data on the command line could provoke the exploit when Exim was running as
4651 exim, but did not say which command line option was involved. All I could
4652 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4653 running as the user.
4656 35. There was a buffer overflow vulnerability in the SPA authentication code
4657 (which came originally from the Samba project). I have added a test to the
4658 spa_base64_to_bits() function which I hope fixes it.
4661 36. The daemon start-up calls getloadavg() while still root for those OS that
4662 need the first call to be done as root, but it missed one case: when
4663 deliver_queue_load_max is set with deliver_drop_privilege. This is
4664 necessary for the benefit of the queue runner, because there is no re-exec
4665 when deliver_drop_privilege is set.
4667 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4668 was in use, was not putting the data itself into the right store pool;
4669 consequently, it could be overwritten for a subsequent message in the same
4670 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4673 38. Sometimes the final signoff response after QUIT could fail to get
4674 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4675 < 0 before doing a fflush(). This bug looks as though it goes back to the
4676 introduction of TLS in release 3.20, but "sometimes" must have been rare
4677 because the tests only now provoked it.
4683 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4684 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4685 Now it returns an integer. A number of calls were still expecting a T/F
4686 return. Fortuitously, in all cases, the tests worked in OK situations,
4687 which is the norm. However, things would have gone wrong on any write
4688 failures on the smtp file descriptor. This function is used when sending
4689 messages over SMTP and also when doing verify callouts.
4691 2. When Exim is called to do synchronous delivery of a locally submitted
4692 message (the -odf or -odi options), it no longer closes stderr before doing
4695 3. Implemented the mua_wrapper option.
4697 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4699 5. Implemented the functions header_remove(), header_testname(),
4700 header_add_at_position(), and receive_remove_recipient(), and exported them
4703 6. If an ACL "warn" statement specified the addition of headers, Exim already
4704 inserted X-ACL-Warn: at the start if there was no header name. However, it
4705 was not making this test for the second and subsequent header lines if
4706 there were newlines in the string. This meant that an invalid header could
4707 be inserted if Exim was badly configured.
4709 7. Allow an ACL "warn" statement to add header lines at the start or after all
4710 the Received: headers, as well as at the end.
4712 8. Added the rcpt_4xx retry error code.
4714 9. Added postmaster_mailfrom=xxx to callout verification option.
4716 10. Added mailfrom=xxxx to the callout verification option, for verify=
4719 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4720 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4722 12. Inserted some casts to stop certain compilers warning when using pointer
4723 differences as field lengths or precisions in printf-type calls (mostly
4724 affecting debugging statements).
4726 13. Added optional readline() support for -be (dynamically loaded).
4728 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4729 same clock tick as a message's arrival, so that its received time was the
4730 same as the "first fail" time on the retry record, and that message
4731 remained on the queue past the ultimate address timeout, every queue runner
4732 would try a delivery (because it was past the ultimate address timeout) but
4733 after another failure, the ultimate address timeout, which should have then
4734 bounced the address, did not kick in. This was a "< instead of <=" error;
4735 in most cases the first failure would have been in the next clock tick
4736 after the received time, and all would be well.
4738 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4739 being recognized when the domain list was tested by the match_domain
4740 condition in an expansion string.
4742 16. Added the ${str2b64: operator.
4744 17. Exim was always calling setrlimit() to set a large limit for the number of
4745 processes, without checking whether the existing limit was already
4746 adequate. (It did check for the limit on file descriptors.) Furthermore,
4747 errors from getrlimit() and setrlimit() were being ignored. Now they are
4748 logged to the main and panic logs, but Exim does carry on, to try to do its
4749 job under whatever limits there are.
4751 18. Imported PCRE 5.0.
4753 19. Trivial typo in log message " temporarily refused connection" (the leading
4756 20. If the log selector return_path_on_delivery was set and an address was
4757 redirected to /dev/null, the delivery process crashed because it assumed
4758 that a return path would always be set for a "successful" delivery. In this
4759 case, the whole delivery is bypassed as an optimization, and therefore no
4762 21. Internal re-arrangement: the function for sending a challenge and reading
4763 a response while authentication was assuming a zero-terminated challenge
4764 string. It's now changed to take a pointer and a length, to allow for
4765 binary data in such strings.
4767 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4769 23. Exim was not respecting finduser_retries when seeking the login of the
4770 uid under which it was called; it was always trying 10 times. (The default
4771 setting of finduser_retries is zero.) Also, it was sleeping after the final
4772 failure, which is pointless.
4774 24. Implemented tls_on_connect_ports.
4776 25. Implemented acl_smtp_predata.
4778 26. If the domain in control=submission is set empty, Exim assumes that the
4779 authenticated id is a complete email address when it generates From: or
4780 Sender: header lines.
4782 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4783 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4784 chown and chgrp in /bin and hostname in /usr/bin.
4786 28. Exim was keeping the "process log" file open after each use, just as it
4787 does for the main log. This opens the possibility of it remaining open for
4788 long periods when the USR1 signal hits a daemon. Occasional processlog
4789 errors were reported, that could have been caused by this. Anyway, it seems
4790 much more sensible not to leave this file open at all, so that is what now
4793 29. The long-running daemon process does not normally write to the log once it
4794 has entered its main loop, and it closes the log before doing so. This is
4795 so that log files can straightforwardly be renamed and moved. However,
4796 there are a couple of unusual error situations where the daemon does write
4797 log entries, and I had neglected to close the log afterwards.
4799 30. The text of an SMTP error response that was received during a remote
4800 delivery was being truncated at 512 bytes. This is too short for some of
4801 the long messages that one sometimes sees. I've increased the limit to
4804 31. It is now possible to make retry rules that apply only when a message has a
4805 specific sender, in particular, an empty sender.
4807 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4808 it possible to be selective about when SMTP synchronization is enforced.
4810 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4812 32. Implemented hosts_connection_nolog.
4814 33. Added an ACL for QUIT.
4816 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4819 35. Added mailbox_size and mailbox_filecount to appendfile.
4821 36. Added control = no_multiline_responses to ACLs.
4823 37. There was a bug in the logic of the code that waits for the clock to tick
4824 in the case where the clock went backwards by a substantial amount such
4825 that the microsecond fraction of "now" was more than the microsecond
4826 fraction of "then" (but the whole seconds number was less).
4828 38. Added support for the libradius Radius client library this is found on
4829 FreeBSD (previously only the radiusclient library was supported).
4835 1. When certain lookups returned multiple values in the form name=value, the
4836 quoting of the values was not always being done properly. Specifically:
4837 (a) If the value started with a double quote, but contained no whitespace,
4839 (b) If the value contained whitespace other than a space character (i.e.
4840 tabs or newlines or carriage returns) it was not quoted.
4841 This fix has been applied to the mysql and pgsql lookups by writing a
4842 separate quoting function and calling it from the lookup code. The fix
4843 should probably also be applied to nisplus, ibase and oracle lookups, but
4844 since I cannot test any of those, I have not disturbed their existing code.
4846 2. A hit in the callout cache for a specific address caused a log line with no
4847 reason for rejecting RCPT. Now it says "Previous (cached) callout
4848 verification failure".
4850 3. There was an off-by-one bug in the queryprogram router. An over-long
4851 return line was truncated at 256 instead of 255 characters, thereby
4852 overflowing its buffer with the terminating zero. As well as fixing this, I
4853 have increased the buffer size to 1024 (and made a note to document this).
4855 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4856 when Exim is waiting for an SMTP response from a remote server, Exim
4857 restarts its select() call on the socket, thereby resetting its timeout.
4858 This is not a problem when such interrupts are rare. Somebody set up a cron
4859 job to run exiwhat every 2 minutes, which is less than the normal select()
4860 timeout (5 or 10 minutes). This meant that the select() timeout never
4861 kicked in because it was always reset. I have fixed this by comparing the
4862 time when an interrupt arrives with the time at the start of the first call
4863 to select(). If more time than the timeout has elapsed, the interrupt is
4864 treated as a timeout.
4866 5. Some internal re-factoring in preparation for the addition of Sieve
4867 extensions (by MH). In particular, the "personal" test is moved to a
4868 separate function, and given an option for scanning Cc: and Bcc: (which is
4869 not set for Exim filters).
4871 6. When Exim created an email address using the login of the caller as the
4872 local part (e.g. when creating a From: or Sender: header line), it was not
4873 quoting the local part when it contained special characters such as @.
4875 7. Installed new OpenBSD configuration files.
4877 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4878 try to make them clearer.
4880 9. Callout options, other than the timeout value, were being ignored when
4881 verifying sender addresses in header lines. For example, when using
4883 verify = header_sender/callout=no_cache
4885 the cache was (incorrectly) being used.
4887 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4888 only the Cygwin environment.
4890 11. When return_path_on_delivery was set as a log selector, if different remote
4891 addresses in the same message used different return paths and parallel
4892 remote delivery occurred, the wrong values would sometimes be logged.
4893 (Whenever a remote delivery process finished, the return path value from
4894 the most recently started remote delivery process was logged.)
4896 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4897 lines when AUTH and/or TLS are in use. This is the "received protocol"
4898 field. Exim used to use "asmtp" for authenticated SMTP, without any
4899 indication (in the protocol name) for TLS use. Now it follows the RFC and
4900 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4901 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4902 names appear in log lines as well as in Received: header lines.
4904 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4905 extensions, and comparison tests, and to fix some bugs.
4907 14. Changes to the "personal" filter test:
4909 (1) The test was buggy in that it was just doing the equivalent of
4910 "contains" tests on header lines. For example, if a user's address was
4911 anne@some.where, the "personal" test would incorrectly be true for
4913 To: susanne@some.where
4915 This test is now done by extracting each address from the header in turn,
4916 and checking the entire address. Other tests that are part of "personal"
4917 are now done using regular expressions (for example, to check local parts
4918 of addresses in From: header lines).
4920 (2) The list of non-personal local parts in From: addresses has been
4921 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4922 taken from the Sieve specification recommendations.
4924 (3) If the message contains any header line starting with "List-" it is
4925 treated as non-personal.
4927 (4) The test for "circular" in the Subject: header line has been removed
4928 because it now seems ill-conceived.
4930 15. Minor typos in src/EDITME comments corrected.
4932 16. Installed latest exipick from John Jetmore.
4934 17. If headers_add on a router specified a text string that was too long for
4935 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4936 of string_sprintf() is now avoided.
4938 18. $message_body_size was not set (it was always zero) when running the DATA
4939 ACL and the local_scan() function.
4941 19. For the "mail" command in an Exim filter, no default was being set for
4942 the once_repeat time, causing a random time value to be used if "once" was
4943 specified. (If the value happened to be <= 0, no repeat happened.) The
4944 default is now 0s, meaning "never repeat". The "vacation" command was OK
4945 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4946 (I found it when inspecting the code).
4948 20. There is now an overall timeout for performing a callout verification. It
4949 defaults to 4 times the callout timeout, which applies to individual SMTP
4950 commands during the callout. The overall timeout applies when there is more
4951 than one host that can be tried. The timeout is checked before trying the
4952 next host. This prevents very long delays if there are a large number of
4953 hosts and all are timing out (e.g. when the network connections are timing
4954 out). The value of the overall timeout can be changed by specifying an
4955 additional sub-option for "callout", called "maxwait". For example:
4957 verify = sender/callout=5s,maxwait=20s
4959 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4960 to the end before writing, but this should make it even safer).
4962 22. Exim was forgetting that it had advertised PIPELINING for the second and
4963 subsequent messages on an SMTP connection. It was also not resetting its
4964 memory on STARTTLS and an internal HELO.
4966 23. When Exim logs an SMTP synchronization error within a session, it now
4967 records whether PIPELINING has been advertised or not.
4969 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4970 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4971 rather than long int.
4973 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4975 26. Added the never_mail option to autoreply.
4981 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4982 crash if the getsockname() call failed; this can happen if a connection is
4983 closed very soon after it is established. The problem was simply in the
4984 order in which certain operations were done, causing Exim to try to write
4985 to the SMTP stream before it had set up the file descriptor. The bug has
4986 been fixed by making things happen in the correct order.
4992 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4993 before the connection was closed, thus losing the rejection response.
4995 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4996 some early Solaris releases, but causes trouble in current releases where
4997 socklen_t is defined.
4999 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5002 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5005 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5006 that releasing the top of it at the end releases what was used for sub-
5007 expansions (unless the block got too big). However, discard this block if
5008 the first thing is a variable or header, so that we can use its block when
5009 it is dynamic (useful for very large $message_headers, for example).
5011 6. Lookups now cache *every* query, not just the most recent. A new, separate
5012 store pool is used for this. It can be recovered when all lookup caches are
5013 flushed. Lookups now release memory at the end of their result strings.
5014 This has involved some general refactoring of the lookup sources.
5016 7. Some code has been added to the store_xxx() functions to reduce the amount
5017 of flapping under certain conditions.
5019 8. log_incoming_interface used to affect only the <= reception log lines. Now
5020 it causes the local interface and port to be added to several more SMTP log
5021 lines, for example "SMTP connection from", and rejection lines.
5023 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5025 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5027 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5028 could be overwritten at the end of the current message (or the start of a
5029 new message if it was set in a HELO ACL). The value is now preserved for
5030 the duration of the SMTP connection.
5032 12. If a transport had a headers_rewrite setting, and a matching header line
5033 contained an unqualified address, that address was qualified, even if it
5034 did not match any rewriting rules. The underlying bug was that the values
5035 of the flags that permit the existence of unqualified sender and recipient
5036 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5037 non-local messages, and by -bnq for local messages) were not being
5038 preserved with the message after it was received.
5040 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5041 "next input=" as part of the text comprising the host identity instead of
5042 the correct text. The code was using the same buffer for two different
5043 strings. However, depending on which order the printing function evaluated
5044 its arguments, the bug did not always show up. Under Linux, for example, my
5045 test suite worked just fine.
5047 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5048 doesn't work with some older versions of Perl. It has been changed to "my",
5049 which in any case is probably the better facility to use.
5051 15. A really picky compiler found some instances of statements for creating
5052 error messages that either had too many or two few arguments for the format
5055 16. The size of the buffer for calls to the DNS resolver has been increased
5056 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5057 for addresses that have a lot of PTR records. This alleviates a problem; it
5058 does not fully solve it.
5060 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5061 buffer now truncates the list and logs the incident, which is the same
5062 action as happens when Exim is looking up a host name and its aliases.
5063 Previously in this situation something unpredictable would happen;
5064 sometimes it was "internal error: store_reset failed".
5066 18. If a server dropped the connection unexpectedly when an Exim client was
5067 using GnuTLS and trying to read a response, the client delivery process
5068 crashed while trying to generate an error log message.
5070 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5071 string, for example:
5073 warn message = H1: something\nH2: something
5075 the text was added as a single header line from Exim's point of view
5076 though it ended up OK in the delivered message. However, searching for the
5077 second and subsequent header lines using $h_h2: did not work. This has been
5078 fixed. Similarly, if a system filter added multiple headers in this way,
5079 the routers could not see them.
5081 20. Expanded the error message when iplsearch is called with an invalid key to
5082 suggest using net-iplsearch in a host list.
5084 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5085 ACLs are no longer actually imposed (and a message to that effect is
5088 22. If a "gecos" field in a passwd entry contained escaped characters, in
5089 particular, if it contained a \" sequence, Exim got it wrong when building
5090 a From: or a Sender: header from that name. A second bug also caused
5091 incorrect handling when an unquoted " was present following a character
5092 that needed quoting.
5094 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5095 was not being matched caselessly.
5097 24. Arranged for all hyphens in the exim.8 source to be escaped with
5100 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5101 properly. Recipient callout cache records were still being keyed to include
5102 the sender, even when use_sender was set false. This led to far more
5103 callouts that were necessary. The sender is no longer included in the key
5104 when use_sender is false.
5106 26. Added "control = submission" modifier to ACLs.
5108 27. Added the ${base62d: operator to decode base 62 numbers.
5110 28. dnsdb lookups can now access SRV records.
5112 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5113 the configuration file.
5115 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5116 (-v) mode. This makes the output for a verbose queue run more intelligible.
5118 31. Added a use_postmaster feature to recipient callouts.
5120 32. Added the $body_zerocount variable, containing the number of binary zero
5121 bytes in the message body.
5123 33. The time of last modification of the "new" subdirectory is now used as the
5124 "mailbox time last read" when there is a quota error for a maildir
5127 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5129 35. Added +ignore_unknown as a special item in host lists.
5131 36. Code for decoding IPv6 addresses in host lists is now included, even if
5132 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5133 address was recognized as an IP address, but was then not correctly decoded
5134 into binary, causing unexpected and incorrect effects when compared with
5141 1. Very minor rewording of debugging text in manualroute to say "list of
5142 hosts" instead of "hostlist".
5144 2. If verify=header_syntax was set, and a header line with an unqualified
5145 address (no domain) and a large number of spaces between the end of the
5146 name and the colon was received, the reception process suffered a buffer
5147 overflow, and (when I tested it) crashed. This was caused by some obsolete
5148 code that should have been removed. The fix is to remove it!
5150 3. When running in the test harness, delay a bit after writing a bounce
5151 message to get a bit more predictability in the log output.
5153 4. Added a call to search_tidyup() just before forking a reception process. In
5154 theory, someone could use a lookup in the expansion of smtp_accept_max_
5155 per_host which, without the tidyup, could leave open a database connection.
5157 5. Added the variables $recipient_data and $sender_data which get set from a
5158 lookup success in an ACL "recipients" or "senders" condition, or a router
5159 "senders" option, similar to $domain_data and $local_part_data.
5161 6. Moved the writing of debug_print from before to after the "senders" test
5164 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5165 problems for message scanning, either using a data ACL, or using
5166 local_scan() because the Received: header was not generated till after they
5167 were called (in order to set the time as the time of reception completion).
5168 I have revised the way this works. The header is now generated after the
5169 body is received, but before the ACL or local_scan() are called. After they
5170 are run, the timestamp in the header is updated.
5176 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5177 before starting a queue runner without re-exec. This happened only when
5178 deliver_drop_privilege was set or when the Exim user was set to root. The
5179 effect of the bug was that timeouts during subsequent deliveries caused
5180 crashes instead of being properly handled. The handler is now left at its
5181 default (and expected) setting.
5183 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5184 message, again when deliver_drop_privilege is set or Exim is run as root.
5185 The bug described in (1) was not present in this case, but the tidying up
5186 of the other signals was missing. I have made the two cases consistent.
5188 3. The ignore_target_hosts setting on a manualroute router was being ignored
5189 for hosts that were looked up using the /MX notation.
5191 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5194 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5195 operated on the sender address. After changing the $sender_address to <>
5196 for the sender address verify, Exim was re-instated it as the original
5197 (before rewriting) address, but remembering that it had rewritten it, so it
5198 wasn't rewriting it again. This bug also had the effect of breaking the
5199 sender address verification caching when the sender address was rewritten.
5201 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5202 This has been changed so that if the ip literal address matches
5203 ignore_target_hosts, the router declines.
5205 7. Added expansion conditions match_domain, match_address, and match_local_
5206 part (NOT match_host).
5208 8. The placeholder for the Received: header didn't have a length field set.
5210 9. Added code to Exim itself and to exim_lock to test for a specific race
5211 condition that could lead to file corruption when using MBX delivery. The
5212 issue is with the lockfile that is created in /tmp. If this file is removed
5213 after a process has opened it but before that process has acquired a lock,
5214 there is the potential for a second process to recreate the file and also
5215 acquire a lock. This could lead to two Exim processes writing to the file
5216 at the same time. The added code performs the same test as UW imapd; it
5217 checks after acquiring the lock that its file descriptor still refers to
5218 the same named file.
5220 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5221 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5224 11. Added the smtp_active_hostname option. If used, this will typically be made
5225 to depend on the incoming interface address. Because $interface_address is
5226 not set up until the daemon has forked a reception process, error responses
5227 that can happen earlier (such as "too many connections") no longer contain
5230 12. If an expansion in a condition on a "warn" statement fails because a lookup
5231 defers, the "warn" statement is abandoned, and the next ACL statement is
5232 processed. Previously this caused the whole ACL to be aborted.
5234 13. Added the iplsearch lookup type.
5236 14. Added ident_timeout as a log selector.
5238 15. Added tls_certificate_verified as a log selector.
5240 16. Added a global option tls_require_ciphers (compare the smtp transport
5241 option of the same name). This controls incoming TLS connections.
5243 17. I finally figured out how to make tls_require_ciphers do a similar thing
5244 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5245 before starting the TLS session.
5247 18. Tabs are now shown as \t in -bP output.
5249 19. If the log selector return_path_on_delivery was set, Exim crashed when
5250 bouncing a message because it had too many Received: header lines.
5252 20. If two routers both had headers_remove settings, and the first one included
5253 a superfluous trailing colon, the final name in the first list and the
5254 first name in the second list were incorrectly joined into one item (with a
5255 colon in the middle).
5261 1. Added -C and -D options to the exinext utility, mainly to make it easier
5262 to include in the automated testing, but these could be helpful when
5263 multiple configurations are in use.
5265 2. The exinext utility was not formatting the output nicely when there was
5266 an alternate port involved in the retry record key, nor when there was a
5267 message id as well (for retries that were specific to a specific message
5268 and a specific host). It was also confused by IPv6 addresses, because of
5269 the additional colons they contain. I have fixed the IPv4 problem, and
5270 patched it up to do a reasonable job for IPv6.
5272 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5273 delivery, the log line now contains "pipelined" if PIPELINING was used.
5275 4. An SMTP transport process used to panic and die if the bind() call to set
5276 an explicit outgoing interface failed. This has been changed; it is now
5277 treated in the same way as a connect() failure.
5279 5. A reference to $sender_host_name in the part of a conditional expansion
5280 that was being skipped was still causing a DNS lookup. This no longer
5283 6. The def: expansion condition was not recognizing references to header lines
5284 that used bh_ and bheader_.
5286 7. Added the _cache feature to named lists.
5288 8. The code for checking quota_filecount in the appendfile transport was
5289 allowing one more file than it should have been.
5291 9. For compatibility with Sendmail, the command line option
5299 and sets the incoming protocol and host name (for trusted callers). The
5300 host name and its colon can be omitted when only the protocol is to be set.
5301 Note the Exim already has two private options, -pd and -ps, that refer to
5302 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5303 "s", but I don't think that's a major issue.
5305 10. A number of refactoring changes to the code, none of which should affect
5308 (a) The number of logging options was getting close to filling up the
5309 32-bit word that was used as a bit map. I have split them into two classes:
5310 those that are passed in the argument to log_write(), and those that are
5311 only ever tested independently outside of that function. These are now in
5312 separate 32-bit words, so there is plenty of room for expansion again.
5313 There is no change in the user interface or the logging behaviour.
5315 (b) When building, for example, log lines, the code previously used a
5316 macro that called string_cat() twice, in order to add two strings. This is
5317 not really sufficiently general. Furthermore, there was one instance where
5318 it was actually wrong because one of the argument was used twice, and in
5319 one call a function was used. (As it happened, calling the function twice
5320 did not affect the overall behaviour.) The macro has been replaced by a
5321 function that can join an arbitrary number of extra strings onto a growing
5324 (c) The code for expansion conditions now uses a table and a binary chop
5325 instead of a serial search (which was left over from when there were very
5326 few conditions). Also, it now recognizes conditions like "pam" even when
5327 the relevant support is not compiled in: a suitably worded error message is
5328 given if an attempt is made to use such a condition.
5330 11. Added ${time_interval:xxxxx}.
5332 12. A bug was causing one of the ddress fields not to be passed back correctly
5333 from remote delivery subprocesses. The field in question was not being
5334 subsequently used, so this caused to problems in practice.
5336 13. Added new log selectors queue_time and deliver_time.
5338 14. Might have fixed a bug in maildirsizefile handling that threw up
5339 "unexpected character" debug warnings, and recalculated the data
5340 unnecessarily. In any case, I expanded the warning message to give more
5343 15. Added the message "Restricted characters in address" to the statements in
5344 the default ACL that block characters like @ and % in local parts.
5346 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5347 Three changes have been made:
5349 (a) There was a serious bug; a negative response to MAIL caused the whole
5350 recipient domain to be cached as invalid, thereby blocking all messages
5351 to all local parts at the same domain, from all senders. This bug has
5352 been fixed. The domain is no longer cached after a negative response to
5353 MAIL if the sender used is not empty.
5355 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5358 (c) A new callout option, "use_sender" has been added for people who want
5359 the modified behaviour.
5365 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5368 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5369 indeed breaks things for older releases.
5371 3. Added additional logging to the case where there is a problem reading data
5372 from a filter that is running in a subprocess using a pipe, in order to
5373 try to track down a specific problem.
5375 4. Testing facility fudge: when running in the test harness and attempting
5376 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5377 getting "No route to host". Convert this to a timeout.
5379 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5382 6. Some OS don't have socklen_t but use size_t instead. This affects the
5383 fifth argument of getsockopt() amongst other things. This is now
5384 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5385 can be set for individual OS. I have set it for SunOS5, OSF1, and
5386 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5387 some earlier ones do not.
5389 7. Change 4.30/15 was not doing the test caselessly.
5391 8. The standard form for an IPv6 address literal was being rejected by address
5392 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5393 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5394 this, as well as the form without the "IPv6" on the front (but only when
5395 address literals are enabled, of course).
5397 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5399 10. Exim crashed if a message with an empty sender address specified by -f
5400 encountered a router with an errors_to setting. This could be provoked only
5401 by a command such as
5405 where an empty string was supplied; "<>" did not hit this bug.
5407 11. Installed PCRE release 4.5.
5409 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5410 remained set. It is now erased.
5412 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5413 times from message ids (which are base 36 rather than the normal 62).
5415 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5416 were being counted as actual protocol errors, and logged if the log
5417 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5418 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5419 and DATA following a set of rejected RCPTs do not count as protocol errors.
5420 In other words, Exim assumes they were pipelined, though this may not
5421 actually be the case. Of course, in all cases the client gets an
5422 appropriate error code.
5424 15. If a lookup fails in an ACL condition, a message about the failure may
5425 be available; it is used if testing the ACL cannot continue, because most
5426 such messages specify what the cause of the deferral is. However, some
5427 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5428 that caused an old message to be retained and used if a later statement
5429 caused a defer, replacing the real cause of the deferral.
5431 16. If an IP address had so many PTR records that the DNS lookup buffer
5432 was not large enough to hold them, Exim could crash while trying to process
5433 the truncated data. It now detects and logs this case.
5435 17. Further to 4.21/58, another change has been made: if (and only if) the
5436 first line of a message (the first header line) ends with CRLF, a bare LF
5437 in a subsequent header line has a space inserted after it, so as not to
5438 terminate the header.
5440 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5441 unnecessarily, used atoi() instead of strtol(), and didn't check the
5442 termination when getting file sizes from file names by regex.
5444 19. Completely re-implemented the support for maildirsize files, in the light
5445 of a number of problems with the previous contributed implementation
5446 (4.30/29). In particular:
5448 . If the quota is zero, the maildirsize file is maintained, but no quota is
5451 . If the maildir directory does not exist, it is created before any attempt
5452 to write a maildirsize file.
5454 . The quota value in the file is just a cache; if the quota is changed in
5455 the transport, the new value overrides.
5457 . A regular expression is available for excluding directories from the
5460 20. The autoreply transport checks the characters in options that define the
5461 message's headers; it allows continued headers, but it was checking with
5462 isspace() after an embedded newline instead of explicitly looking for a
5465 21. If all the "regular" hosts to which an address was routed had passed their
5466 expiry times, and had not reached their retry times, the address was
5467 bounced, even if fallback hosts were defined. Now Exim should go on to try
5470 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5471 equivalent code in the SMTP transport. Some hosts send humungous responses
5472 to HELO/EHLO, more than 1024 it seems.
5474 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5475 gives compiler warnings in some environments. I've now done it "properly",
5478 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5479 (because of problems with the built-in one) was declared to return uschar *
5480 instead of char *, causing compiler failure.
5482 25. Fixed a file descriptor leak when processing alias/forward files.
5484 26. Fixed a minor format string issue in dbfn.c.
5486 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5488 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5489 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5492 29. When a sender address is verified, it is cached, to save repeating the test
5493 when there is more than one recipient in a message. However, when the
5494 verification involves a callout, it is possible for different callout
5495 options to be set for different recipients. It is too complicated to keep
5496 track of this in the cache, so now Exim always runs a verification when a
5497 callout is required, relying on the callout cache for the optimization.
5498 The overhead is duplication of the address routing, but this should not be
5501 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5502 to be verified with callout=postmaster, and the main callout worked but the
5503 postmaster check failed, the verification correctly failed. However, if a
5504 subsequent RCPT command asked for sender verification *without* the
5505 postmaster check, incorrect caching caused this verification also to fail,
5508 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5509 it was not caching the DNS options (qualify_single, search_parents) that
5510 were used when the lookup failed. A subsequent lookup with different
5511 options therefore always gave the same answer, though there were cases
5512 where it should not have. (Example: a "domains = !$mx_any" option on a
5513 dnslookup router: the "domains" option is always processed without any
5514 widening, but the router might have qualify_single set.) Now Exim uses the
5515 cached value only when the same options are set.
5517 32. Added John Jetmore's "exipick" utility to the distribution.
5519 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5520 than a timeout (e.g. a certificate is required, and is not provided), an
5521 Exim server now closes the connection immediately. Previously it waited for
5522 the client to close - but if the client is SSL, it seems that they each
5523 wait for each other, leading to a delay before one of them times out.
5525 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5526 maintained 0.8.x compatibility because I don't think many are using it, and
5527 it is clearly obsolete.
5529 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5532 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5533 client certificate was expired. A simple patch fixes this, though I don't
5534 understand the full logic of why the verify callback is called multiple
5537 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5538 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5539 which causes problems with some clients (such as the Certicom SSL Plus
5540 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5541 disables the coutermeasure allowing Eudora to connect."
5543 38. Exim was not checking that a write() to a log file succeeded. This could
5544 lead to Bad Things if a log got too big, in particular if it hit a file
5545 size limit. Exim now panics and dies if it cannot write to a log file, just
5546 as it does if it cannot open a log file.
5548 39. Modified OS/Makefile-Linux so that it now contains
5550 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5552 The two -D definitions ensure that Exim is compiled with large file
5553 support, which makes it possible to handle log files that are bigger than
5556 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5557 instance) a domain was checked against a named list that involved a lookup,
5558 causing $domain_data to be set, then another domain was checked against the
5559 same list, then the first domain was re-checked, the value of $domain_data
5560 after the final check could be wrong. In particular, if the second check
5561 failed, it could be set empty. This bug probably also applied to
5564 41. The strip_trailing_dot option was not being applied to the address given
5565 with the -f command-line option.
5567 42. The code for reading a message's header from the spool was incrementing
5568 $received_count, but never initializing it. This meant that the value was
5569 incorrect (doubled) while delivering a message in the same process in which
5570 it was received. In the most common configuration of Exim, this never
5571 happens - a fresh exec is done - but it can happen when
5572 deliver_drop_privilege is set.
5574 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5575 it now includes up to 150 characters of the unexpected data in the log
5578 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5579 and building data strings. The size of both of these buffers was 10 000
5580 bytes - far larger than anybody would *ever* want, thought I. Needless to
5581 say, somebody hit the limit. I have increased the maximum line length to
5582 20 000 and the maximum data length of concatenated lines to 100 000. I have
5583 also fixed two bugs, because there was no checking on these buffers. Tsk,
5584 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5585 buffer is too small.
5587 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5588 lsearch lookups. Now it does.
5590 46. When parsing a route_list item in a manualroute router, a fixed-length
5591 buffer was used for the list of hosts. I made this 1024 bytes long,
5592 thinking that nobody would ever have a list of hosts that long. Wrong.
5593 Somebody had a whole pile of complicated expansion conditions, and the
5594 string was silently truncated, leading to an expansion error. It turns out
5595 that it is easier to change to an unlimited length (owing to other changes
5596 that have happened since this code was originally written) than to build
5597 structure for giving a limitation error. The length of the item that
5598 expands into the list of hosts is now unlimited.
5600 47. The lsearch lookup could not handle data where the length of text line was
5601 more than 4095 characters. Such lines were truncated, leading to shortened
5602 data being returned. It should now handle lines of any length.
5604 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5605 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5608 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5609 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5610 no longer happens in comment lines. A list of replacements is now placed
5611 at the head of all of the source files, except those whose only change is
5612 to replace PERL_COMMAND in the very first #! line.
5614 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5615 messages on the queue, with a bottom-up merge sort, using code contributed
5616 by Michael Haardt. This should make operations like -bp somewhat faster on
5617 large queues. It won't affect queue runners, except when queue_run_in_order
5620 51. Installed eximstats 1.31 in the distribution.
5622 52. Added support for SRV lookups to the dnslookup router.
5624 53. If an ACL referred to $message_body or $message_body_end, the value was not
5625 reset for any messages that followed in the same SMTP session.
5627 54. The store-handling optimization for building very long strings was not
5628 differentiating between the different store pools. I don't think this
5629 actually made any difference in practice, but I've tidied it.
5631 55. While running the routers to verify a sender address, $sender_address
5632 was still set to the sender address. This is wrong, because when routing to
5633 send a bounce to the sender, it would be empty. Therefore, I have changed
5634 it so that, while verifying a sender address, $sender_address is set to <>.
5635 (There is no change to what happens when verifying a recipient address.)
5637 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5638 target A or AAAA records (if not already returned) without resetting the
5639 qualify_single or search_parents options of the DNS resolver. These are
5640 inappropriate in this case because the targets of MX and SRV records must
5641 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5642 target that, when qualified, matched something in the local domain. These
5643 two options are now turned off when doing these lookups.
5645 57. It seems that at least some releases of Reiserfs (which does not have the
5646 concept of a fixed number of inodes) returns zero and not -1 for the
5647 number of available inodes. This interacted badly with check_spool_inodes,
5648 which assumed that -1 was the "no such thing" setting. What I have done is
5649 to check that the total number of inodes is greater than zero before doing
5650 the test of how many are available.
5652 58. When a "warn" ACL statement has a log_message modifier, the message is
5653 remembered, and not repeated. This is to avoid a lot of repetition when a
5654 message has many recipients that cause the same warning to be written.
5655 However, Exim was preserving the list of already written lines for an
5656 entire SMTP session, which doesn't seem right. The memory is now reset if a
5657 new message is started.
5659 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5660 debugging output unless log_rewrite was also set.
5662 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5663 of (int)(handle) when we know that handle contains (void *)(-1).
5665 61. The Exim daemon panic-logs an error return when it closes the incoming
5666 connection. However "connection reset by peer" seems to be common, and
5667 isn't really an error worthy of noting specially, so that particular error
5670 62. When Exim is trying to find all the local interfaces, it used to panic and
5671 die if the ioctl to get the interface flags failed. However, it seems that
5672 on at least one OS (Solaris 9) it is possible to have an interface that is
5673 included in the list of interfaces, but for which you get a failure error
5674 for this call. This happens when the interface is not "plumbed" into a
5675 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5676 failure of the "get flags" call assumes that the interface is down.
5678 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5679 makes life easier for people who are doing arithmetic on fields extracted
5680 from dates, where you often get leading zeros that should not be
5681 interpreted as octal.
5683 64. Added qualify_domain to the redirect router, to override the global
5686 65. If a pathologically long header line contained very many addresses (the
5687 report of this problem mentioned 10 000) and each of them was rewritten,
5688 Exim could use up a very large amount of memory. (It kept on making new
5689 copies of the header line as it rewrote, and never released the old ones.)
5690 At the expense of a bit more processing, the header rewriting function has
5691 been changed so that it no longer eats memory in this way.
5693 66. The generation of the Received: header has been moved from the time that a
5694 message starts to be received, to the time that it finishes. The timestamp
5695 in the Received: header should now be very close to that of the <= log
5696 line. There are two side-effects of this change:
5698 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5699 logged header lines no longer include the local Received: line, because
5700 it has not yet been created. The same applies to a copy of the message
5701 that is returned to a non-SMTP sender when a message is rejected.
5703 (b) When a filter file is tested using -bf, no additional Received: header
5704 is added to the test message. After some thought, I decided that this
5707 This change does not affect the value of $received_for. It is still set
5708 after address rewriting, but before local_scan() is called.
5710 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5712 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5713 gave an unhelpful panic error message, and a defer error. I have managed to
5714 change this behaviour so that it now rejects any supplied certificate,
5715 which seems right, as the list of acceptable certificates is empty.
5717 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5718 gave an unhelpful defer error. I have not managed to make this reject any
5719 supplied certificates, but the error message it gives is "no certificate
5720 supplied", which is not helpful.
5722 70. exigrep's output now also includes lines that are not associated with any
5723 message, but which match the given pattern. Implemented by a patch from
5724 Martin Sluka, which also tidied up the Perl a bit.
5726 71. Recipient callout verification, like sender verification, was using <> in
5727 the MAIL FROM command. This isn't really the right thing, since the actual
5728 sender may affect whether the remote host accepts the recipient or not. I
5729 have changed it to use the actual sender in the callout; this means that
5730 the cache record is now keyed on a recipient/sender pair, not just the
5731 recipient address. There doesn't seem to be a real danger of callout loops,
5732 since a callout by the remote host to check the sender would use <>.
5733 [SEE ABOVE: changed after hitting problems.]
5735 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5736 temporary errors. However, in the case of such a code being given after
5737 the end of a data transmission (i.e. after ".") Exim was failing to write
5738 a retry record for the message. (Yes, there was some broken host that was
5739 actually sending 8xx at this point.)
5741 73. An unknown lookup type in a host list could cause Exim to panic-die when
5742 the list was checked. (An example that provoked this was putting <; in the
5743 middle of a list instead of at the start.) If this happened during a DATA
5744 ACL check, a -D file could be left lying around. This kind of configuration
5745 error no longer causes Exim to die; instead it causes a defer error. The
5746 incident is still logged to the main and panic logs.
5748 74. Buglet left over from Exim 3 conversion. The message "too many messages
5749 in one connection" was written to the rejectlog but not the mainlog, except
5750 when address rewriting (yes!) was being logged.
5752 75. Added write_rejectlog option.
5754 76. When a system filter was run not as root (that is, when system_filter_user
5755 was set), the values of the $n variables were not being returned to the
5756 main process; thus, they were not subsequently available in the $sn
5759 77. Added +return_path_on_delivery log selector.
5761 78. A connection timeout was being treated differently from recipients deferred
5762 when testing hosts_max_try with a message that was older than the host's
5763 retry timeout. (The host should not be counted, thus allowing all hosts to
5764 be tried at least once before bouncing.) This may have been the cause of an
5765 occasionally reported bug whereby a message would remain on the queue
5766 longer than the retry timeout, but would be bounced if a delivery was
5767 forced. I say "may" because I never totally pinned down the problem;
5768 setting up timeout/retry tests is difficult. See also the next item.
5770 79. The ultimate address timeout was not being applied to errors that involved
5771 a combination of host plus message (for example, a timeout on a MAIL
5772 command). When an address resolved to a number of possible hosts, and they
5773 were not all tried for each delivery (e.g. because of hosts_max_try), a
5774 message could remain on the queue longer than the retry timeout.
5776 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5779 81. Fixed an obscure SMTP outgoing bug which required at least the following
5780 conditions: (a) there was another message waiting for the same server;
5781 (b) the server returned 5xx to all RCPT commands in the first message so
5782 that the message was not completed; (c) the server dropped the connection
5783 or gave a negative response to the RSET that Exim sends to abort the
5784 transaction. The observed case was a dropped connection after DATA that had
5785 been sent in pipelining mode. That is, the server had advertised PIPELINING
5786 but was not implementing it correctly. The effect of the bug was incorrect
5787 behaviour, such as trying another host, and this could lead to a crash.
5793 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5794 and daemon.c were passed as pointers to ints; they should have been
5795 pointers to socklen_t variables (which are typically unsigned ints).
5797 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5800 3. Fixed a really odd bug that affected only the testing scheme; patching a
5801 certain fixed string in the binary changed the value of another string that
5802 happened to be identical to the end of the original first string.
5804 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5805 name", it returns that address as the IP address. On some operating
5806 systems (e.g. Solaris), it also passes back the IP address string as the
5807 "host name". However, on others (e.g. Linux), it passes back an empty
5808 string. Exim wasn't checking for this, and was changing the host name to an
5809 empty string, assuming it had been canonicalized.
5811 5. Although rare, it is permitted to have more than one PTR record for a given
5812 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5813 all the names associated with an address, because they do in Solaris.
5814 However, it seems that they do not in Linux for data that comes from the
5815 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5816 I found this out when I moved to a new Linux workstation and tried to run
5817 the Exim test suite.
5819 To get round this problem I have changed the code so that it now does its
5820 own call to the DNS to look up PTR records when searching for a host name.
5821 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5822 addresses that are only in /etc/hosts are still found.
5824 This behaviour is, however, controlled by an option called host_lookup_
5825 order, which defaults to "bydns:byaddr". If people want to use the other
5826 order, or indeed, just use one or the other means of lookup, they can
5827 specify it in this variable.
5829 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5830 some operating systems, this comes back from gethostbyaddr() as an empty
5831 string, and this is what Exim used to test for. However, it seems that in
5832 other systems, "." is yielded. Exim now tests for this case too.
5834 7. The values of check_spool_space and check_log_space are now held internally
5835 as a number of kilobytes instead of an absolute number of bytes. If a
5836 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5837 kilobyte. This means that much larger values can be stored.
5839 8. Exim monitor: an attempt to get the action menu when not actually pointing
5840 at a message produces an empty menu entitled "No message selected". This
5841 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5842 no entries in it ("Shell widget menu has zero width and/or height"). So I
5843 have added a single, blank menu entry in this case.
5845 9. Added ${quote_local_part.
5847 10. MIME decoding is now applied to the contents of Subject: header lines when
5850 11. Now that a reference to $sender_host_address automatically causes a reverse
5851 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5852 host lookup before query-style lookups in lists that might use this
5853 variable. This has therefore been abolished, and the "net-" prefix is no
5854 longer necessary for query-style lookups.
5856 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5857 have been a typo for LFLAGS, so it has been changed.
5859 13. The install script calls Exim with "-C /dev/null" in order to find the
5860 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5861 to be output. However, since Exim outputs its version number before the
5862 error, it didn't break the script. It just looked ugly. I fixed this by
5863 always allowing "-C /dev/null" if the caller is root.
5865 14. Ignore overlarge ACL variable number when reading spool file - insurance
5866 against a later release with more variables having written the file.
5868 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5869 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5870 this, as well as the form without the "IPv6" on the front.
5872 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5873 OS/Makefile-Darwin file.
5875 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5876 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5878 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5879 variable. However, some OS use ints for inodes. I've added cast to long int
5880 to get rid of the compiler warning.
5882 19. I had forgotten to lock out "/../" in configuration file names when
5883 ALT_CONFIG_PREFIX was set.
5885 20. Routers used for verification do not need to specify transports. However,
5886 if such a router generated a host list, and callout was configured, Exim
5887 crashed, because it could not find a port number from the (non-existent)
5888 transport. It now assumes port 25 in this circumstance.
5890 21. Added the -t option to exigrep.
5892 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5893 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5894 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5895 them is set, LOOKUP_LSEARCH is forced.
5897 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5899 24. Added sender and host information to the "rejected by local_scan()" log
5900 line; previously there was no indication of these.
5902 25. Added .include_if_exists.
5904 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5905 Linux. It turns out that not all file systems support this. Apparently some
5906 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5907 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5908 not supported on the file descriptor, is now ignored when Exim is trying to
5909 sync a directory. This applies only to Linux.
5911 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5913 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5914 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5917 29. Added support for maildirsize files from supplied patch (modified a bit).
5919 30. The use of :fail: followed by an empty string could lead Exim to respond to
5920 sender verification failures with (e.g.):
5922 550 Verification failed for <xxx>
5923 550 Sender verify failed
5925 where the first response line was missing the '-' that indicates it is not
5926 the final line of the response.
5928 31. The loop for finding the name of the user that called Exim had a hardwired
5929 limit of 10; it now uses the value of finduser_retries, which is used for
5930 all other user lookups.
5932 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5935 33. Exim was neglecting to zero errno before one call of strtol() when
5936 expanding a string and expecting an integer value. On some systems this
5937 resulted in spurious "integer overflow" errors. Also, it was casting the
5938 result into an int without checking.
5940 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5941 did not work. The code looks as if it has *never* worked, though it appears
5942 to have been documented since at least release 1.62. I have made it work.
5944 35. The "timeout_DNS" error in retry rules, also documented since at least
5945 1.62, also never worked. As it isn't clear exactly what this means, and
5946 clearly it isn't a major issue, I have abolished the feature by treating it
5947 as "timeout", and writing a warning to the main and panic logs.
5949 36. The display of retry rules for -brt wasn't always showing the error code
5952 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5953 timeout_connect_A, timeout_connect_MX.
5955 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5956 to the empty sender.
5958 39. The daemon was not analysing the content of -oX till after it had closed
5959 stderr and disconnected from the controlling terminal. This meant that any
5960 syntax errors were only noted on the panic log, and the return code from
5961 the command was 0. By re-arranging the code a little, I've made the
5962 decoding happen first, so such errors now appear on stderr, and the return
5963 code is 1. However, the actual setting up of the sockets still happens in
5964 the disconnected process, so errors there are still only recorded on the
5967 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5968 connections (as happens on some IP stacks) was logged at start up time as
5969 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5970 it from "IPv6 and IPv4", which means that two separate sockets are being
5973 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5974 says whether AF_INET or AF_INET6 was passed as an argument.
5976 42. Exiwhat output was messed up when time zones were included in log
5979 43. Exiwhat now gives more information about the daemon's listening ports,
5980 and whether -tls-on-connect was used.
5982 44. The "port" option of the smtp transport is now expanded.
5984 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5985 silently ignored. Now an error message is written to the main and panic
5988 46. There's a new ACL modifier called "logwrite" which writes to a log file
5989 as soon as it is encountered.
5991 47. Added $local_user_uid and $local_user_gid at routing time.
5993 48. Exim crashed when trying to verify a sender address that was being
5996 49. Exim was recognizing only a space character after ".include". It now also
5997 recognizes a tab character.
5999 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6000 extracting the relevant information from the specification. The man page no
6001 longer contains scrambled data for the -d option, and I've added a section
6002 at the front about calling Exim under different names.
6004 51. Added "extra_headers" argument to the "mail" command in filter files.
6006 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6009 53. Installed eximstats 1.29.
6011 54. Added transport_filter_timeout as a generic transport option.
6013 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6014 Cc: header lines. This was required by RFC 822, but it not required by RFC
6017 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6018 incoming messages that did not have them. Now it does so only if the
6019 message originates locally, that is, if there is no associated remote host
6020 address. When Resent- header lines are present, this applies to the Resent-
6021 lines rather than the non-Resent- lines.
6023 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6024 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6026 58. Messages for configuration errors now include the name of the main
6027 configuration file - useful now that there may be more than one file in a
6028 list (.included file names were always shown).
6030 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6031 for those rare installations that do not start the daemon as root or run it
6032 setuid root. I've cut out the call to initgroups() if the daemon is not
6035 60. The Exim user and group can now be bound into the binary as text strings
6036 that are looked up at the start of Exim's processing.
6038 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6040 62. Added $mailstore_basename variable.
6042 63. Installed patch to sieve.c from Michael Haardt.
6044 64. When Exim failed to open the panic log after failing to open the main log,
6045 the original message it was trying to log was written to stderr and debug
6046 output, but if they were not available (the usual case in production), it
6047 was lost. Now it is written to syslog before the two lines that record the
6048 failures to open the logs.
6050 65. Users' Exim filters run in subprocesses under the user's uid. It is
6051 possible for a "deliver" command or an alias in a "personal" command to
6052 provoke an address rewrite. If logging of address rewriting is configured,
6053 this fails because the process is not running as root or exim. There may be
6054 a better way of dealing with this, but for the moment (because 4.30 needs
6055 to be released), I have disabled address rewrite logging when running a
6056 filter in a non-root, non-exim process.
6062 1. The buildconfig auxiliary program wasn't quoting the value set for
6063 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6064 not defined. This bug was masked in 4.22 by the effect that was fixed in
6067 2. Some messages that were rejected after a message id was allocated were
6068 shown as "incomplete" by exigrep. It no longer does this for messages that
6069 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6071 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6072 have allow_domain_literals set, the ID did not get logged in the <= line.
6073 Domain literals are now always recognized in Message-ID: header lines.
6075 4. The first argument for a ${extract expansion item is the key name or field
6076 number. Leading and trailing spaces in this item were not being ignored,
6077 causing some misleading effects.
6079 5. When deliver_drop_privilege was set, single queue runner processes started
6080 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6081 same command in the process it spins off) were not dropping privilege.
6083 6. When the daemon running as "exim" started a queue runner, it always
6084 re-executed Exim in the spun-off process. This is a waste of effort when
6085 deliver_drop_privilege is set. The new process now just calls the
6086 queue-runner function directly.
6092 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6095 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6096 ignored. Though the use of -oP was forcing the writing of a pid file, it
6097 was always written to the default place.
6099 3. If the message "no IP address found for host xxxx" is generated during
6100 incoming verification, it is now followed by identification of the incoming
6101 connection (so you can more easily find what provoked it).
6103 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6105 5. Added some features to "harden" Exim a bit more against certain attacks:
6107 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6108 be put in Local/Makefile. This is like the never_users runtime option,
6109 but it cannot be overridden. The default setting is "root".
6111 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6112 prefix string with which any file named in a -C command line option
6115 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6116 is retained for -C and -D only if the caller of Exim is root. Without
6117 it, the exim user may also use -C and -D and retain privilege.
6119 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6120 command line option is disabled.
6122 6. Macro names set by the -D option must start with an upper case letter, just
6123 like macro names defined in the configuration file.
6125 7. Added "dereference=" facility to LDAP.
6127 8. Two instances of the typo "uknown" in the source files are fixed.
6129 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6130 the Configure-Makefile script screwed up while processing it.
6132 10. Incorporated PCRE 4.4.
6134 11. The SMTP synchronization check was not operating right at the start of an
6135 SMTP session. For example, it could not catch a HELO sent before the client
6136 waited for the greeting. There is now a check for outstanding input at the
6137 point when the greeting is written. Because of the duplex, asynchronous
6138 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6139 way, but not yet received, when the check is performed.
6141 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6142 on TCP/IP sockets, because this apparently causes some broken clients to
6145 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6146 unchanged) from the Cygwin maintainer.
6148 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6149 was supported instead of testing for mbx. Effectively a typo.
6151 15. The spa authenticator server code was not checking that the input it
6152 received was valid base64.
6154 16. The debug output line for the "set" modifier in ACLs was not showing the
6155 name of the variable that was being set.
6157 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6159 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6160 lookup on its own. Something else was needed to trigger the lookup. For
6161 example, a match in host_lookup or the need for a host name in a host list.
6162 Now, if $sender_host_name is referenced and the host name has not yet been
6163 looked up, a lookup is performed. If the lookup fails, the variable remains
6164 empty, and $host_lookup_failed is set to "1".
6166 19. Added "eqi" as a case-independent comparison operator.
6168 20. The saslauthd authentication condition could segfault if neither service
6169 nor realm was specified.
6171 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6172 error message that was logged was misleading, and incoming SMTP
6173 connections were dropped. The message is now more accurate, and temporary
6174 errors are given to SMTP connections.
6176 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6177 (except RSET) with a 421 error, until QUIT is received. However, it was
6178 failing to send a response to QUIT.
6180 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6181 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6182 ACL is now run after the helo_try_verify_hosts code.
6184 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6185 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6186 case-independent, so other case variants are also recognized). Apparently
6187 some systems use these upper case variants.
6189 25. If more than two messages were waiting for the same host, and a transport
6190 filter was specified for the transport, Exim sent two messages over the
6191 same TCP/IP connection, and then failed with "socket operation on non-
6192 socket" when it tried to send the third.
6194 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6196 27. The extern definition of crypt16() in expand.c was not being excluded when
6197 the OS had its own crypt16() function.
6199 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6200 as a preferred synonym for return_size_limit, both as an option and as an
6203 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6205 30. Changed the default configuration ACL to relax the local part checking rule
6206 for addresses that are not in any local domains. For these addresses,
6207 slashes and pipe symbols are allowed within local parts, but the sequence
6208 /../ is explicitly forbidden.
6210 31. SPA server authentication was not clearing the challenge buffer before
6213 32. log_message in a "warn" ACL statement was writing to the reject log as
6214 well as to the main log, which contradicts the documentation and doesn't
6215 seem right (because no rejection is happening). So I have stopped it.
6217 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6218 However, I am unable to do any testing of this.
6220 34. Fixed an infelicity in the appendfile transport. When checking directories
6221 for a mailbox, to see if any needed to be created, it was accidentally
6222 using path names with one or more superfluous leading slashes; tracing
6223 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6225 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6226 contents are added to the log line that is written for every discarded
6227 recipient. (Previously a log_message setting was ignored.)
6229 36. The ${quote: operator now quotes the string if it is empty.
6231 37. The install script runs exim in order to find its version number. If for
6232 some reason other than non-existence or emptiness, which it checks, it
6233 could not run './exim', it was installing it with an empty version number,
6234 i.e. as "exim-". This error state is now caught, and the installation is
6237 38. An argument was missing from the function that creates an error message
6238 when Exim fails to connect to the socket for saslauthd authentication.
6239 This could cause Exim to crash, or give a corrupted message.
6241 39. Added isip, isip4, and isip6 to ${if conditions.
6243 40. The ACL variables $acl_xx are now saved with the message, and can be
6244 accessed later in routers, transports, and filters.
6246 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6247 strings in the file are not string-expanded.
6249 42. If a MAIL command specified a SIZE value that was too large to fit into an
6250 int variable, the check against message_size_limit failed. Such values are
6251 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6252 day this will have to be increased, but I don't think I want to be around
6253 when emails are that large.
6260 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6261 iconv() is not standard in FreeBSD.
6263 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6264 IPv6 enabled. The observed symptom was a segmentation fault on return from
6265 the function os_common_find_running_interfaces() in src/os.c.
6267 3. In the check_special_case() function in daemon.c I had used "errno" as an
6268 argument name, which causes warnings on some systems. This was basically a
6269 typo, since it was named "eno" in the comments!
6271 4. The code that waits for the clock to tick (at a resolution of some fraction
6272 of a second) so as to ensure message-id uniqueness was always waiting for
6273 at least one whole tick, when it could have waited for less. [This is
6274 almost certainly not relevant at current processor speeds, where it is
6275 unlikely to ever wait at all. But we try to future-proof.]
6277 5. The function that sleeps for a time interval that includes fractions of a
6278 second contained a race. It did not block SIGALRM between setting the
6279 timer, and suspending (a couple of lines later). If the interval was short
6280 and the sigsuspend() was delayed until after it had expired, the suspension
6281 never ended. On busy systems this could lead to processes getting stuck for
6284 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6285 process, before it forks any delivery processes. The open lookup caching
6286 mechanism meant that the open file or database connection was passed into
6287 the delivery process. The problem was that delivery processes always tidy
6288 up cached lookup data. This could cause a problem for the next delivery
6289 process started by the queue runner, because the external queue runner
6290 process does not know about the closure. So the next delivery process
6291 still has data in the lookup cache. In the case of a file lookup, there was
6292 no problem because closing a file descriptor in a subprocess doesn't affect
6293 the parent. However, if the lookup was caching a connection to a database,
6294 the connection was closed, and the second delivery process was likely to
6295 see errors such as "PGSQL: query failed: server closed the connection
6296 unexpectedly". The problem has been fixed by closing all cached lookups
6297 in a queue runner before running a delivery process.
6299 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6300 seem to have the "const" qualifier which it has on other OS. I've
6303 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6304 *of the same type* (client or server) with the same public name that an
6305 error should be diagnosed.
6307 9. When Exim looked up a host name for an IP address, but failed to find the
6308 original IP address when looking up the host name (a safety check), it
6309 output the message "<ip address> does not match any IP for NULL", which was
6310 confusing, to say the least. The bug was that the host name should have
6311 appeared instead of "NULL".
6313 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6314 user that is built into the binary, and the -C or -D options is used, root
6315 privilege is dropped before the configuration file is read. In addition,
6316 logging is switched to stderr instead of the normal log files. If the
6317 configuration then re-defines the Exim user, the unprivileged environment
6318 is probably not what is expected, so Exim logs a panic warning message (but
6321 However, if deliver_drop_privilege is set, the unprivileged state may well
6322 be exactly what is intended, so the warning has been cut out in that case,
6323 and Exim is allowed to try to write to its normal log files.
6329 1. smtp_return_error_details was not giving details for temporary sender
6330 or receiver verification errors.
6332 2. Diagnose a configuration error if two authenticators have the same public
6335 3. Exim used not to create the message log file for a message until the first
6336 delivery attempt. This could be confusing when incoming messages were held
6337 for policy or load reasons. The message log file is now created at the time
6338 the message is received, and an initial "Received" line is written to it.
6340 4. The automatically generated man page for command line options had a minor
6341 bug that caused no ill effects; however, a more serious problem was that
6342 the procedure for building the man page automatically didn't always
6343 operate. Consequently, release 4.20 contains an out-of-date version. This
6344 shouldn't happen again.
6346 5. When building Exim with embedded Perl support, the script that builds the
6347 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6348 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6350 6. The freeze_tell option was not being used for messages that were frozen on
6351 arrival, either by an ACL or by local_scan().
6353 7. Added the smtp_incomplete_transaction log selector.
6355 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6356 was accepting AUTH without a new EHLO.
6358 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6359 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6362 10. Logging of TCP/IP connections (when configured) now happens in the main
6363 daemon process instead of the child process, so that the TCP/IP connection
6364 count is more accurate (but it can never be perfect).
6366 11. The use of "drop" in a nested ACL was not being handled correctly in the
6367 outer ACL. Now, if condition failure induced by the nested "drop" causes
6368 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6369 or "require"), the connection is dropped.
6371 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6372 that yield "discard" can now be used with an "accept" or a "discard" verb,
6373 but an error is generated for any others (because I can't see a useful way
6374 to define what should happen).
6376 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6377 are now processed in the same way as lines in the Exim configuration file.
6378 In particular, continuation lines are supported.
6380 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6382 15. Added -ti meaning -t -i.
6384 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6385 domains, and warn by logging if others are found.
6387 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6388 returned from GIFCONF when Exim is trying to find the list of interfaces on
6389 a host. The code in os.c has been modified to copy each ifreq to an aligned
6390 structure in all cases.
6392 Also, in some cases, the returned ifreq's were being copied to a 'struct
6393 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6394 means the last couple of bytes of an IPv6 address could be chopped if the
6395 ifreq contained only a normal sockaddr (14 bytes storage).
6397 18. Named domain lists were not supported in the hosts_treat_as_local option.
6398 An entry such as +xxxx was not recognized, and was treated as a literal
6401 19. Ensure that header lines added by a DATA ACL are included in the reject log
6402 if the ACL subsequently rejects the message.
6404 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6405 MD5 (which is deprecated).
6407 21. When testing a filter file using -bf, Exim was writing a message when it
6408 took the sender from a "From " line in the message, but it was not doing so
6409 when it took $return_path from a Return-Path: header line. It now does.
6411 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6412 with a valid header line field name (a series of printing characters
6413 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6415 23. Changed "disc" in the source to "disk" to conform to the documentation and
6416 the book and for uniformity.
6418 24. Ignore Sendmail's -Ooption=value command line item.
6420 25. When execve() failed while trying to run a command in a pipe transport,
6421 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6422 could be confused with a return value of 69 from the command itself. This
6423 has been changed to 127, the value the shell returns if it is asked to run
6424 a non-existent command. The wording for the related log line suggests a
6425 non-existent command as the problem.
6427 26. If received_header_text expands to an empty string, do not add a Received:
6428 header line to the message. (Well, it adds a token one on the spool, but
6429 marks it "old" so that it doesn't get used or transmitted.)
6431 27. Installed eximstats 1.28 (addition of -nt option).
6433 28. There was no check for failure on the call to getsockname() in the daemon
6434 code. This can fail if there is a shortage of resources on the system, with
6435 ENOMEM, for example. A temporary error is now given on failure.
6437 29. Contrary to the C standard, it seems that in some environments, the
6438 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6439 program. Exim now does this explicitly; it affects the formatting of
6440 timestamps using strftime().
6442 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6443 complaints. I've now initialized all the variables, to avoid this.
6445 32. Header lines added by a system filter were not being "seen" during
6446 transport-time rewrites.
6448 33. The info_callback() function passed to OpenSSL is set up with type void
6449 (*)(SSL *, int, int), as described somewhere. However, when calling the
6450 function (actually a macro) that sets it up, the type void(*)() is
6451 expected. I've put in a cast to prevent warnings from picky compilers.
6453 34. If a DNS black list lookup found a CNAME record, but there were no A
6454 records associated with the domain it pointed at, Exim crashed.
6456 35. If a DNS black list lookup returned more than one A record, Exim ignored
6457 all but the first. It now scans all returned addresses if a particular IP
6458 value is being sought. In this situation, the contents of the
6459 $dnslist_value variable are a list of all the addresses, separated by a
6462 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6463 to accept a host name and all its aliases if the forward lookup for any of
6464 them yielded the IP address of the incoming connection. Now it accepts only
6465 those names whose forward lookup yields the correct IP address. Any other
6466 names are discarded. This closes a loophole whereby a rogue DNS
6467 administrator could create reverse DNS records to break through a
6468 wildcarded host restriction in an ACL.
6470 37. If a user filter or a system filter that ran in a subprocess used any of
6471 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6472 the wrong values were passed to the pipe command ($thisaddress had the
6473 value of $0, $0 had the value of $1, etc). This bug was introduced by
6474 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6476 38. Improved the line breaking for long SMTP error messages from ACLs.
6477 Previously, if there was no break point between 40 and 75 characters, Exim
6478 left the rest of the message alone. Two changes have been made: (a) I've
6479 reduced the minimum length to 35 characters; (b) if it can't find a break
6480 point between 35 and 75 characters, it looks ahead and uses the first one
6481 that it finds. This may give the occasional overlong line, but at least the
6482 remaining text gets split now.
6484 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6485 file descriptors might be low, and that setting 1000 would always raise it.
6486 It turns out that in some environments, the limit is already over 1000 and
6487 that lowering it causes trouble. So now Exim takes care not to decrease it.
6489 40. When delivering a message, the value of $return_path is set to $sender_
6490 address at the start of routing (routers may change the value). By an
6491 oversight, this default was not being set up when an address was tested by
6492 -bt or -bv, which affected the outcome if any router or filter referred to
6495 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6496 log only when "message" or "log_message" are set. However, if one of the
6497 conditions was an address verification, or a call to a nested ACL, the
6498 messages generated by the underlying test were being passed through. This
6499 no longer happens. The underlying message is available in $acl_verify_
6500 message for both "message" and "log_message" expansions, so it can be
6501 passed through if needed.
6503 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6504 new expansion $bh_ to give the encoded byte string without charset
6505 translation. Translation happens only if iconv() is available; HAVE_ICONV
6506 indicates this at build time. HEADERS_CHARSET gives the charset to
6507 translate to; headers_charset can change it in the configuration, and
6508 "headers charset" can change it in an individual filter file.
6510 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6511 that creates RFC 2047 encoded "words" labels them as that charset instead
6512 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6513 expansion operator; (ii) when Exim creates a From: line for a local
6514 message; (iii) when a header line is rewritten to include a "phrase" part.
6516 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6517 buggy, causing it to skip the first lines of messages whose message ID
6518 ended in 'D'. This would not have bitten before Exim release 4.14, because
6519 message IDs were unlikely to end in 'D' before then. The effect was to have
6520 incorrect size information for certain domains.
6522 45. #include "config.h" was missing at the start of the crypt16.c module. This
6523 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6526 46. If there was a timeout during a "random" callout check, Exim treated it as
6527 a failure of the random address, and carried on sending RSET and the real
6528 address. If the delay was just some slowness somewhere, the response to the
6529 original RCPT would be taken as a response to RSET and so on, causing
6530 mayhem of various kinds.
6532 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6533 when I implemented it. It didn't allow for the fact that some option values
6534 may legitimately be negative (e.g. size_addition), and it didn't even do
6535 the right test for positive values.
6537 48. Domain names in DNS records are case-independent. Exim always looks them up
6538 in lower case. Some resolvers return domain names in exactly the case they
6539 appear in the zone file, that is, they may contain uppercase letters. Not
6540 all resolvers do this - some return always lower case. Exim was treating a
6541 change of case by a resolver as a change of domain, similar to a widening
6542 of a domain abbreviation. This triggered its re-routing code and so it was
6543 trying to route what was effectively the same domain again. This normally
6544 caused routing to fail (because the router wouldn't handle the domain
6545 twice). Now Exim checks for this case specially, and just changes the
6546 casing of the domain that it ultimately uses when it transmits the message
6549 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6552 50. If a filter generated a file delivery with a non-absolute name (possible if
6553 no home directory exists for the router), the forbid_file option was not
6556 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6557 the existing equality matching.
6559 52. Exim was using ints instead of ino_t variables in some places where it was
6560 dealing with inode numbers.
6562 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6563 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6564 called TMPDIR, and if it finds it is different, it changes its value.
6566 54. The smtp_printf() function is now made available to local_scan() so
6567 additional output lines can be written before returning. There is also an
6568 smtp_fflush() function to enable the detection of a dropped connection.
6569 The variables smtp_input and smtp_batched_input are exported to
6572 55. Changed the default runtime configuration: the message "Unknown user"
6573 has been removed from the ACL, and instead placed on the localuser router,
6574 using the cannot_route_message feature. This means that any verification
6575 failures that generate their own messages won't get overridden. Similarly,
6576 the "Unrouteable address" message that was in the ACL for unverifiable
6577 relay addresses has also been removed.
6579 56. Added hosts_avoid_esmtp to the smtp transport.
6581 57. The exicyclog script was not checking for the esoteric option
6582 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6583 will work only if exicyclog is run under the appropriate euid.
6585 58. Following a discussion on the list, the rules by which Exim recognises line
6586 endings on incoming messages have been changed. The -dropcr and drop_cr
6587 options are now no-ops, retained only for backwards compatibility. The
6588 following line terminators are recognized: LF CRLF CR. However, special
6589 processing applies to CR:
6591 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6592 nor a local message in the state where . is a terminator.
6594 (ii) If a bare CR is encountered in a header line, an extra space is added
6595 after the line terminator so as not to end the header. The reasoning
6596 behind this is that bare CRs in header lines are most likely either
6597 to be mistakes, or people trying to play silly games.
6599 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6600 was being incorrectly given as 18 bytes larger than it should have been.
6601 This is a VOB (very old bug).
6603 60. This may never have affected anything current, but just in case it has:
6604 When the local host is found other than at the start of a list of hosts,
6605 the local host, those with the same MX, and any that follow, are discarded.
6606 When the list in question was part of a longer list of hosts, the following
6607 hosts (not currently being processed) were also being discarded. This no
6608 longer happens. I'm not sure if this situation could ever has previously
6611 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6614 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6615 header. This is something that is recommended in a new Internet Draft, and
6616 is something that is documented as being done by Sendmail. There are two
6617 possible values. For messages generated by the autoreply transport, Exim
6620 Auto-Submitted: auto-replied
6622 whereas for all other generated messages (e.g. bounces) it adds
6624 Auto-Submitted: auto-generated
6626 63. The "personal" condition in filters now includes a test for the
6627 Auto-Submitted: header. If it contains the string "auto-" the message it
6628 not considered personal.
6630 64. Added rcpt_include_affixes as a generic transport option.
6632 65. Added queue_only_override (default true).
6634 66. Added the syslog_duplication option.
6636 67. If what should have been the first header line of a message consisted of
6637 a space followed by a colon, Exim was mis-interpreting it as a header line.
6638 It isn't of course - it is syntactically invalid and should therefore be
6639 treated as the start of the message body. The misbehaviour could have
6640 caused a number of strange effects, including loss of data in subsequent
6641 header lines, and spool format errors.
6643 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6644 client host had authenticated. This control can now be exercised by an ACL
6645 for more flexibility.
6647 69. By default, callouts do not happen when testing with -bh. There is now a
6648 variant, -bhc, which does actually run the callout code, including
6649 consulting and updating the callout cache.
6651 70. Added support for saslauthd authentication, courtesy of Alexander
6654 71. If statvfs() failed on the spool or log directories while checking their
6655 size for availability, Exim confusingly gave the error "space shortage".
6656 Furthermore, in debugging mode it crashed with a floating point exception.
6657 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6658 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6659 serious problem, Exim now writes to the main and panic logs when this
6660 happens, with details of the failure. It then refuses to accept the
6661 incoming message, giving the message "spool directory problem" or "log
6662 directory problem" with a 421 code for SMTP messages.
6664 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6665 0, 1, and 2 exist, because some OS complain for execs without them (see
6666 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6667 descriptors. However, the code omitted to check that the open succeeded,
6668 causing mysterious errors if for some reason the permissions on /dev/null
6669 got screwed. Now Exim writes a message to the main and panic logs, and
6670 bombs out if it can't open /dev/null.
6672 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6673 interact so that it is all more flexible. It is supposed to remain
6674 backwards compatible. Also added extra_local_interfaces.
6676 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6677 to bomb out with an assertion failure - to the client this appears as a
6678 connection drop. This problem occurs in the part of the code that was taken
6679 from the Samba project. Fortunately, the assertion is in a very simple
6680 function, so I have fixed this by reproducing the function inline in the
6681 one place where it is called, and arranging for authentication to fail
6682 instead of killing the process with assert().
6684 75. The SPA client code was not working when the server requested OEM rather
6685 than Unicode encoding.
6687 76. Added code to make require_files with a specific uid setting more usable in
6688 the case where statting the file as root fails - usually a non-root-mounted
6689 NFS file system. When this happens and the failure is EACCES, Exim now
6690 forks a subprocess and does the per-uid checking as the relevant uid.
6692 77. Added process_log_path.
6694 78. If log_file_path was not explicitly set, a setting of check_log_space or
6695 check_log_inodes was ignored.
6697 79. If a space check for the spool or log partitions fails, the incident is now
6698 logged. Of course, in the latter case the data may get lost...
6700 80. Added the %p formatting code to string_format() so that it can be used to
6701 print addresses in debug_print(). Adjusted all the address printing in the
6702 debugging in store.c to use %p rather than %d.
6704 81. There was a concern that a line of code in smtp_in.c could overflow a
6705 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6706 initially expressed, the concern was not well-founded, because trailing
6707 spaces are removed early. However, if the trailing spaces were followed by
6708 a NULL, they did not get removed, so the overflow was possible. Two fixes
6711 (a) I re-wrote the offending code in a cleaner fashion.
6712 (b) If an incoming SMTP command contains a NULL character, it is rejected
6715 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6716 runs initgroups(), so that if the Exim user is in any additional groups,
6717 they will be used during message reception.
6723 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob