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.
52 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
53 supports it and a size is available (ie. the sending peer gave us one).
55 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
56 years of being deprecated). The replacements are the ACLs
57 acl_smtp_mime and acl_not_smtp_mime.
59 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
60 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
61 or even an in-clear connection were permitted. Now, if the host lookup
62 was dnssec and dane was requested then the host is only used if the
63 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
64 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
65 if one fails this test.
66 This means that a poorly-configured remote DNS will make it incommunicado;
67 but it protects against a DNS-interception attack on it.
69 JH/04 Bug 1810: make continued-use of an open smtp transport connection
70 non-noisy when a race steals the message being considered.
72 JH/05 If main configuration option tls_certificate is unset, generate a
73 self-signed certificate for inbound TLS connections.
75 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
76 in rewrites and routers.
78 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
79 and logged a warning sing 4.83; now they are a configuration file error.
81 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
82 (lacking @domain). Apply the same qualification processing as RCPT.
84 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
86 JH/10 Support ${sha256:} applied to a string (as well as the previous
89 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
90 a cutthrough deliver is pending, as we always want to make a connection.
91 This also avoids re-routing the message when later placing the cutthrough
92 connection after a verify cache hit.
93 Do not update it with the verify result either.
95 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
96 when routing results in more than one destination address.
98 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
99 signing (which inhibits the cutthrough capability). Previously only
100 the presence of an option was tested; now an expansion evaluating as
101 empty is permissible (obviously it should depend only on data available
102 when the cutthrough connection is made).
104 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
105 the relevant preceding SMTP command did not note the pipelining mode.
107 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
108 Previously they were not counted.
110 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
111 as one having no matching records. Previously we deferred the message
112 that needed the lookup.
114 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
115 distinguished as "(=".
117 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
118 for missing MX records. Previously it only worked for missing A records.
120 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
122 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
123 after the data-go-ahead and data-ack. Patch from Jason Betts.
125 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
126 even for a "none" policy. Patch from Tony Meyer.
128 JH/22 Fix continued use of a connection for further deliveries. If a port was
129 specified by a router, it must also match for the delivery to be
132 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
133 When one of the recipients of a message was unsuitable for the connection
134 (has no matching addresses), we lost track of needing to mark it
135 deferred. As a result mail would be lost.
137 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
139 JH/25 Decoding ACL controls is now done using a binary search; the source code
140 takes up less space and should be simpler to maintain. Merge the ACL
141 condition decode tables also, with similar effect.
143 JH/26 Fix problem with one_time used on a redirect router which returned the
144 parent address unchanged. A retry would see the parent address marked as
145 delivered, so not attempt the (identical) child. As a result mail would
148 JH/27 Fix a possible security hole, wherein a process operating with the Exim
149 UID can gain a root shell. Credit to http://www.halfdog.net/ for
150 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
153 JH/28 Enable {spool,log} filesystem space and inode checks as default.
154 Main config options check_{log,spool}_{inodes,space} are now
155 100 inodes, 10MB unless set otherwise in the configuration.
157 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
158 Previously it only applied to the main-section connection policy
161 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
163 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
164 by me. Added RFC7919 DH primes as an alternative.
166 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
167 are not in the system include path.
169 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
170 GnuTLS, when a session startup failed (eg because the client disconnected)
171 Exim did stdio operations after fclose. This was exposed by a recent
172 change which nulled out the file handle after the fclose.
174 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
175 signed directly by the cert-signing cert, rather than an intermediate
176 OCSP-signing cert. This is the model used by LetsEncrypt.
178 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
180 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
181 an incoming connection.
183 HS/02 Bug 1802: Do not half-close the connection after sending a request
186 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
187 fallback to "prime256v1".
189 JH/34 SECURITY: Use proper copy of DATA command in error message.
190 Could leak key material. Remotely exploitable. CVE-2016-9963.
196 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
197 and 3.4.4 - once the server is enabled to respond to an OCSP request
198 it does even when not requested, resulting in a stapling non-aware
199 client dropping the TLS connection.
201 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
202 support variable-length bit vectors. No functional change.
204 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
205 The I= interface field on outgoing lines is now after the H= remote
206 host field, same as incoming lines. There is a separate
207 outgoing_interface log selector which allows you to disable the
210 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
211 If not running log_selector +smtp_connection the mainlog would be held
212 open indefinitely after a "too many connections" event, including to a
213 deleted file after a log rotate. Leave the per net connection logging
214 leaving it open for efficiency as that will be quickly detected by the
215 check on the next write.
217 HS/01 Bug 1671: Fix post transport crash.
218 Processing the wait-<transport> messages could crash the delivery
219 process if the message IDs didn't exist for some reason. When
220 using 'split_spool_directory=yes' the construction of the spool
221 file name failed already, exposing the same netto behaviour.
223 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
224 mime_regex ACL conditions.
226 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
227 to DSN fail messages (bounces): remote IP, remote greeting, remote response
228 to HELO, local diagnostic string.
230 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
231 log line to debug. Even when configured with a tls authenticator many
232 client connections are expected to not authenticate in this way, so
233 an authenticate fail is not an error.
235 HS/02 Add the Exim version string to the process info. This way exiwhat
236 gives some more detail about the running daemon.
238 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
239 matter for fast-change records such as DNSBLs.
241 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
242 retry record, even if constant. There may be multiple transports with
243 different interface settings and the retry behaviour needs to be kept
246 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
248 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
250 JH/10 Bug 840: fix log_defer_output option of pipe transport
252 JH/11 Bug 830: use same host for all RCPTS of a message, even under
253 hosts_randomize. This matters a lot when combined with mua_wrapper.
255 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
256 ${quote_pgsql:<string>} operator.
258 JH/13 Bug 1708: avoid misaligned access in cached lookup.
260 JH/14 Change header file name for freeradius-client. Relevant if compiling
261 with Radius support; from the Gentoo tree and checked under Fedora.
263 JH/15 Bug 1712: Introduce $prdr_requested flag variable
265 JH/16 Bug 1714: Permit an empty string as expansion result for transport
266 option transport_filter, meaning no filtering.
268 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
270 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
271 defaults to "*" (all hosts). The variable is now available when not built
272 with TLS, default unset, mainly to enable keeping the testsuite sane.
273 If a server certificate is not supplied (via tls_certificate) an error is
274 logged, and clients will find TLS connections fail on startup. Presumably
275 they will retry in-clear.
276 Packagers of Exim are strongly encouraged to create a server certificate
277 at installation time.
279 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
280 with the $config_file variable.
282 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
283 in transport context, after the attempt, and per-recipient. The latter type
284 is per host attempted. The event data is the error message, and the errno
285 information encodes the lookup type (A vs. MX) used for the (first) host,
286 and the trailing two digits of the smtp 4xx response.
288 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
289 to write to mainlog (or rejectlog, paniclog) in the window between file
290 creation and permissions/ownership being changed. Particularly affects
291 installations where exicyclog is run as root, rather than exim user;
292 result is that the running daemon panics and dies.
294 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
296 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
297 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
298 "pri" and "weight". Note that the previous implicit priority given by the
299 list order is no longer honoured.
301 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
304 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
305 by defining SUPPORT_SOCKS.
307 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
308 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
309 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
310 variables are renamed to proxy_{local,external}_{address,port}.
312 JH/27 Move Internationalisation support from Experimental to mainline, enabled
313 for a build by defining SUPPORT_I18N
315 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
316 of the query string, and make ${quote_redis:} do that quoting.
318 JH/29 Move Events support from Experimental to mainline, enabled by default
319 and removable for a build by defining DISABLE_EVENT.
321 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
323 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
324 cached by the daemon.
326 JH/32 Move Redis support from Experimental to mainline, enabled for a build
327 by defining LOOKUP_REDIS. The libhiredis library is required.
329 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
330 keys are given for lookup.
332 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
333 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
334 only supported when built with TLS support. The PolarSSL SHA routines
335 are still used when the TLS library is too old for convenient support.
337 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
338 openssl_options), for security. OpenSSL forces this from version 1.1.0
339 server-side so match that on older versions.
341 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
342 allocation for $value could be released as the expansion processing
343 concluded, but leaving the global pointer active for it.
345 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
346 and to use the domains and local_parts ACL conditions.
348 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
349 incorrectly not doubled on cutthrough transmission, hence seen as a
350 body-termination at the receiving system - resulting in truncated mails.
351 Commonly the sender saw a TCP-level error, and retransmitted the message
352 via the normal store-and-forward channel. This could result in duplicates
353 received - but deduplicating mailstores were liable to retain only the
354 initial truncated version.
356 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
358 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
360 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
361 we're in there, support oversigning also; bug 1309.
363 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
365 HS/04 Add support for keep_environment and add_environment options.
367 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
368 either intentional arithmetic overflow during PRNG, or testing config-
371 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
372 delivery resulted in actual delivery. Cancel cutthrough before DATA
375 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
376 rejects a recipient: pass the reject to the originator.
378 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
379 Many were false-positives and ignorable, but it's worth fixing the
382 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
383 for the new environment-manipulation done at startup. Move the routines
384 from being local to tls.c to being global via the os.c file.
386 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
387 an extract embedded as result-arg for a map, the first arg for extract
388 is unavailable so we cannot tell if this is a numbered or keyed
389 extraction. Accept either.
395 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
398 JH/02 The smtp transport option "multi_domain" is now expanded.
400 JH/03 The smtp transport now requests PRDR by default, if the server offers
403 JH/04 Certificate name checking on server certificates, when exim is a client,
404 is now done by default. The transport option tls_verify_cert_hostnames
405 can be used to disable this per-host. The build option
406 EXPERIMENTAL_CERTNAMES is withdrawn.
408 JH/05 The value of the tls_verify_certificates smtp transport and main options
409 default to the word "system" to access the system default CA bundle.
410 For GnuTLS, only version 3.0.20 or later.
412 JH/06 Verification of the server certificate for a TLS connection is now tried
413 (but not required) by default. The verification status is now logged by
414 default, for both outbound TLS and client-certificate supplying inbound
417 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
420 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
421 Status Notification (bounce) messages are now MIME format per RFC 3464.
422 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
423 under the control of the dsn_advertise_hosts option, and routers may
424 have a dsn_lasthop option.
426 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
427 default, modifiable by a malware= option. The list separator for
428 the options can now be changed in the usual way. Bug 68.
430 JH/10 The smtp_receive_timeout main option is now expanded before use.
432 JH/11 The incoming_interface log option now also enables logging of the
433 local interface on delivery outgoing connections.
435 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
436 if the interface and destination host and port all match.
438 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
441 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
442 Patch from Andrew Lewis.
444 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
445 now supports optional time-restrictions, weighting, and priority
446 modifiers per server. Patch originally by <rommer@active.by>.
448 JH/16 The spamd_address main option now supports a mixed list of local
449 and remote servers. Remote servers can be IPv6 addresses, and
450 specify a port-range.
452 JH/17 Bug 68: The spamd_address main option now supports an optional
453 timeout value per server.
455 JH/18 Bug 1581: Router and transport options headers_add/remove can
456 now have the list separator specified.
458 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
461 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
464 JH/21 Support for the A6 type of dns record is withdrawn.
466 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
467 rather than the verbs used.
469 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
470 from 255 to 1024 chars.
472 JH/24 Verification callouts now attempt to use TLS by default.
474 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
475 are generic router options now. The defaults didn't change.
477 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
478 Original patch from Alexander Shikoff, worked over by JH.
480 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
481 files if ZCAT_COMMAND is not executable.
483 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
485 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
487 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
488 Normally benign, it bites when the pair was led to by a CNAME;
489 modern usage is to not canonicalize the domain to a CNAME target
490 (and we were inconsistent anyway for A-only vs AAAA+A).
492 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
494 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
495 when evaluating $sender_host_dnssec.
497 JH/31 Check the HELO verification lookup for DNSSEC, adding new
498 $sender_helo_dnssec variable.
500 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
502 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
504 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
506 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
507 documented as working, but never had. Support all but $spam_report.
509 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
510 added for tls authenticator.
512 HS/03 Add perl_taintmode main config option
518 TL/01 When running the test suite, the README says that variables such as
519 no_msglog_check are global and can be placed anywhere in a specific
520 test's script, however it was observed that placement needed to be near
521 the beginning for it to behave that way. Changed the runtest perl
522 script to read through the entire script once to detect and set these
523 variables, reset to the beginning of the script, and then run through
524 the script parsing/test process like normal.
526 TL/02 The BSD's have an arc4random API. One of the functions to induce
527 adding randomness was arc4random_stir(), but it has been removed in
528 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
529 function when detected.
531 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
532 cause callback expansion.
534 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
535 syntax errors in an expansion can be treated as a string instead of
536 logging or causing an error, due to the internal use of bool_lax
537 instead of bool when processing it.
539 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
540 server certificates when making smtp deliveries.
542 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
544 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
546 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
548 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
549 Merged patch from Sebastian Wiedenroth.
551 JH/05 Fix results-pipe from transport process. Several recipients, combined
552 with certificate use, exposed issues where response data items split
553 over buffer boundaries were not parsed properly. This eventually
554 resulted in duplicates being sent. This issue only became common enough
555 to notice due to the introduction of connection certificate information,
556 the item size being so much larger. Found and fixed by Wolfgang Breyha.
558 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
559 size buffer was used, resulting in syntax errors when an expansion
562 JH/07 Add support for directories of certificates when compiled with a GnuTLS
563 version 3.3.6 or later.
565 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
566 is EXPERIMENTAL_EVENT, the main-configuration and transport options
567 both become "event_action", the variables become $event_name, $event_data
568 and $event_defer_errno. There is a new variable $verify_mode, usable in
569 routers, transports and related events. The tls:cert event is now also
570 raised for inbound connections, if the main configuration event_action
573 TL/06 In test suite, disable OCSP for old versions of openssl which contained
574 early OCSP support, but no stapling (appears to be less than 1.0.0).
576 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
577 server certificate names available under the smtp transport option
578 "tls_verify_cert_hostname" now do not permit multi-component wildcard
581 JH/10 Time-related extraction expansions from certificates now use the main
582 option "timezone" setting for output formatting, and are consistent
583 between OpenSSL and GnuTLS compilations. Bug 1541.
585 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
586 encoded parameter in the incoming message. Bug 1558.
588 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
589 include certificate info, eximon was claiming there were spoolfile
592 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
594 JH/14 Log delivery-related information more consistently, using the sequence
595 "H=<name> [<ip>]" wherever possible.
597 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
598 are problematic for Debian distribution, omit them from the release
601 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
603 JH/16 Fix string representation of time values on 64bit time_t architectures.
606 JH/17 Fix a null-indirection in certextract expansions when a nondefault
607 output list separator was used.
612 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
613 checkers that were complaining about end of non-void function with no
616 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
617 This was a regression introduced in 4.83 by another bugfix.
619 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
621 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
622 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
628 TF/01 Correctly close the server side of TLS when forking for delivery.
630 When a message was received over SMTP with TLS, Exim failed to clear up
631 the incoming connection properly after forking off the child process to
632 deliver the message. In some situations the subsequent outgoing
633 delivery connection happened to have the same fd number as the incoming
634 connection previously had. Exim would try to use TLS and fail, logging
635 a "Bad file descriptor" error.
637 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
638 utilities have not been installed.
640 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
641 temporary space as the ACL may create new global variables.
643 TL/01 LDAP support uses per connection or global context settings, depending
644 upon the detected version of the libraries at build time.
646 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
647 to extract and use the src ip:port in logging and expansions as if it
648 were a direct connection from the outside internet. PPv2 support was
649 updated based on HAProxy spec change in May 2014.
651 JH/02 Add ${listextract {number}{list}{success}{fail}}.
653 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
654 Properly escape header and check for NULL return.
656 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
659 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
661 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
662 characters in header names, implemented as a verify condition.
663 Contributed by Michael Fischer v. Mollard.
665 TL/05 Rename SPF condition results err_perm and err_temp to standardized
666 results permerror and temperror. Previous values are deprecated but
667 still accepted. In a future release, err_perm and err_temp will be
668 completely removed, which will be a backward incompatibility if the
669 ACL tests for either of these two old results. Patch contributed by
670 user bes-internal on the mailing list.
672 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
674 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
675 selectors, in both main and reject logs.
677 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
680 JH/07 Add malware type "sock" for talking to simple daemon.
682 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
684 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
685 routers/transports under cutthrough routing.
687 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
688 numbers. Touch up "bool" conditional to keep the same definition.
690 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
692 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
694 JH/12 Expand items in router/transport headers_add or headers_remove lists
695 individually rather than the list as a whole. Bug 1452.
697 Required for reasonable handling of multiple headers_ options when
698 they may be empty; requires that headers_remove items with embedded
699 colons must have them doubled (or the list-separator changed).
701 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
702 view the policy declared in the DMARC record. Currently, $dmarc_status
703 is a combined value of both the record presence and the result of the
706 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
708 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
709 dnslookup router and the smtp transport (applying to the forward
712 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
713 of ldap servers used for a specific lookup. Patch provided by Heiko
716 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
717 New variable $lookup_dnssec_authenticated for observability.
719 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
720 Patch submitted by Lars Timman.
722 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
724 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
725 Requires trusted mode and valid format message id, aborts otherwise.
726 Patch contributed by Heiko Schlichting.
728 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
729 certextract with support for various fields. Bug 1358.
731 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
732 is requested by default, modifiable by smtp transport option
735 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
736 operate on certificate variables to give certificate fingerprints
737 Also new ${sha256:cert_variable}.
739 JH/23 The PRDR feature is moved from being Experimental into the mainline.
741 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
742 Christian Aistleitner.
744 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
746 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
747 file. Patch from Wolfgang Breyha.
749 JH/25 Expand the coverage of the delivery $host and $host_address to
750 client authenticators run in verify callout. Bug 1476.
752 JH/26 Port service names are now accepted for tls_on_connect_ports, to
753 align with daemon_smtp_ports. Bug 72.
755 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
756 support and error reporting did not work properly.
758 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
759 and is readable. Patch from Andrew Colin Kissa.
761 TL/14 Enhance documentation of ${run expansion and how it parses the
762 commandline after expansion, particularly in the case when an
763 unquoted variable expansion results in an empty value.
765 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
767 PP/02 Fix internal collision of T_APL on systems which support RFC3123
768 by renaming away from it. Addresses GH issue 15, reported by
771 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
773 TL/15 SECURITY: prevent double expansion in math comparison functions
774 (can expand unsanitized data). Not remotely exploitable.
781 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
783 PP/02 Make -n do something, by making it not do something.
784 When combined with -bP, the name of an option is not output.
786 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
789 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
790 $sender_host_name and config options to manage this, and basic check
793 PP/05 DSCP support for outbound connections and control modifier for inbound.
795 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
796 (Only plugin which currently uses this is kerberos4, which nobody should
797 be using, but we should make it available and other future plugins might
798 conceivably use it, even though it would break NAT; stuff *should* be
799 using channel bindings instead).
801 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
802 name; added for Sendmail compatibility; requires admin caller.
803 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
804 just ignore it); requires trusted caller.
805 Also parse but ignore: -Ac -Am -X<logfile>
808 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
810 TL/02 Add +smtp_confirmation as a default logging option.
812 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
813 Patch by Magnus Holmgren from 2007-02-20.
815 TL/04 Bugzilla 1281 - Spec typo.
816 Bugzilla 1283 - Spec typo.
817 Bugzilla 1290 - Spec grammar fixes.
819 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
821 TL/06 Add Experimental DMARC support using libopendmarc libraries.
823 TL/07 Fix an out of order global option causing a segfault. Reported to dev
824 mailing list by by Dmitry Isaikin.
826 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
828 JH/02 Support "G" suffix to numbers in ${if comparisons.
830 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
832 NM/01 Bugzilla 1197 - Spec typo
833 Bugzilla 1196 - Spec examples corrections
835 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
837 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
838 gnutls_enable_pkcs11, but renamed to more accurately indicate its
841 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
842 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
844 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
845 "acl {{name}{arg}...}", and optional args on acl condition
848 JH/05 Permit multiple router/transport headers_add/remove lines.
850 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
852 JH/07 Avoid using a waiting database for a single-message-only transport.
853 Performance patch from Paul Fisher. Bugzilla 1262.
855 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
858 JH/09 Add $headers_added variable, with content from use of ACL modifier
859 add_header (but not yet added to the message). Bugzilla 199.
861 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
862 Pulled from Bugzilla 817 by Wolfgang Breyha.
864 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
866 (nb: this is the same fix as in Exim 4.80.1)
868 JH/11 Add A= logging on delivery lines, and a client_set_id option on
871 JH/12 Add optional authenticated_sender logging to A= and a log_selector
874 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
876 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
877 advertise SMTP AUTH mechanism to us, instead of a generic
878 protocol violation error. Also, make Exim more robust to bad
879 data from the Dovecot auth socket.
881 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
883 When a queue runner is handling a message, Exim first routes the
884 recipient addresses, during which it prunes them based on the retry
885 hints database. After that it attempts to deliver the message to
886 any remaining recipients. It then updates the hints database using
889 So if a recipient address works intermittently, it can get repeatedly
890 deferred at routing time. The retry hints record remains fresh so the
891 address never reaches the final cutoff time.
893 This is a fairly common occurrence when a user is bumping up against
894 their storage quota. Exim had some logic in its local delivery code
895 to deal with this. However it did not apply to per-recipient defers
896 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
898 This change adds a proper retry rule check during routing so that the
899 final cutoff time is checked against the message's age. We only do
900 this check if there is an address retry record and there is not a
901 domain retry record; this implies that previous attempts to handle
902 the address had the retry_use_local_parts option turned on. We use
903 this as an approximation for the destination being like a local
904 delivery, as in LMTP.
906 I suspect this new check makes the old local delivery cutoff check
907 redundant, but I have not verified this so I left the code in place.
909 TF/02 Correct gecos expansion when From: is a prefix of the username.
911 Test 0254 submits a message to Exim with the header
915 When I ran the test suite under the user fanf2, Exim expanded
916 the header to contain my full name, whereas it should have added
917 a Resent-Sender: header. It erroneously treats any prefix of the
918 username as equal to the username.
920 This change corrects that bug.
922 GF/01 DCC debug and logging tidyup
923 Error conditions log to paniclog rather than rejectlog.
924 Debug lines prefixed by "DCC: " to remove any ambiguity.
926 TF/03 Avoid unnecessary rebuilds of lookup-related code.
928 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
929 Bug spotted by Jeremy Harris; was flawed since initial commit.
930 Would have resulted in OCSP responses post-SNI triggering an Exim
931 NULL dereference and crash.
933 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
935 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
936 Bug detection, analysis and fix by Samuel Thibault.
937 Bugzilla 1331, Debian bug #698092.
939 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
941 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
942 Server implementation by Todd Lyons, client by JH.
943 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
944 config variable "prdr_enable" controls whether the server
945 advertises the facility. If the client requests PRDR a new
946 acl_data_smtp_prdr ACL is called once for each recipient, after
947 the body content is received and before the acl_smtp_data ACL.
948 The client is controlled by both of: a hosts_try_prdr option
949 on the smtp transport, and the server advertisement.
950 Default client logging of deliveries and rejections involving
951 PRDR are flagged with the string "PRDR".
953 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
954 fclose(). Diagnosis by Todd Lyons.
956 PP/17 Update configure.default to handle IPv6 localhost better.
957 Patch by Alain Williams (plus minor tweaks).
960 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
961 This is now consistent with GnuTLS, and is now documented: the
962 previous undocumented portable approach to treating the option as
963 unset was to force an expansion failure. That still works, and
964 an empty string is now equivalent.
966 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
967 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
968 not performing validation itself.
970 PP/20 Added force_command boolean option to pipe transport.
971 Patch from Nick Koston, of cPanel Inc.
973 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
976 TF/04 Added udpsend ACL modifier and hexquote expansion operator
978 PP/21 Fix eximon continuous updating with timestamped log-files.
979 Broken in a format-string cleanup in 4.80, missed when I repaired the
980 other false fix of the same issue.
981 Report and fix from Heiko Schlichting.
984 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
985 Report from Prashanth Katuri.
987 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
988 It's SecureTransport, so affects any MacOS clients which use the
989 system-integrated TLS libraries, including email clients.
991 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
992 using a MIME ACL for non-SMTP local injection.
993 Report and assistance in diagnosis by Warren Baker.
995 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
997 JH/16 Fix comparisons for 64b. Bugzilla 1385.
999 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1000 last id that failed so it may be referenced in subsequent ACL's.
1002 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1005 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1006 ldap library initialization, allowing self-signed CA's to be
1007 used. Also properly sets require_cert option later in code by
1008 using NULL (global ldap config) instead of ldap handle (per
1009 session). Bug diagnosis and testing by alxgomz.
1011 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1012 the src/util/ subdirectory.
1014 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1015 renamed to Transport Post Delivery Action by Jeremy Harris, as
1018 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1019 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1020 redis_servers = needs to be configured which will be used by the redis
1021 lookup. Patch from Warren Baker, of The Packet Hub.
1023 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1025 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1026 hostname or reverse DNS when processing a host list. Used suggestions
1027 from multiple comments on this bug.
1029 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1031 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1032 Missed a few lines, added it to make the runtest require no keyboard
1035 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1036 contains upper case chars. Make router use caseful_local_part.
1038 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1039 support when GnuTLS has been built with p11-kit.
1045 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1047 This, or similar/improved, will also be change PP/11 of 4.82.
1053 PP/01 Handle short writes when writing local log-files.
1054 In practice, only affects FreeBSD (8 onwards).
1055 Bugzilla 1053, with thanks to Dmitry Isaikin.
1057 NM/01 Bugzilla 949 - Documentation tweak
1059 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1062 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1064 PP/02 Implemented gsasl authenticator.
1066 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1068 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1069 `pkg-config foo` for cflags/libs.
1071 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1072 with rest of GSASL and with heimdal_gssapi.
1074 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1075 `pkg-config foo` for cflags/libs for the TLS implementation.
1077 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1078 properties get this fed in as external SSF. A number of robustness
1079 and debugging improvements to the cyrus_sasl authenticator.
1081 PP/08 cyrus_sasl server now expands the server_realm option.
1083 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1084 Patch by Jeremy Harris.
1086 PP/10 Added dbmjz lookup type.
1088 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1090 PP/12 MAIL args handles TAB as well as SP, for better interop with
1091 non-compliant senders.
1092 Analysis and variant patch by Todd Lyons.
1094 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1095 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1096 Patch from Dirk Mueller <dmueller@suse.com>
1098 PP/13 tls_peerdn now print-escaped for spool files.
1099 Observed some $tls_peerdn in wild which contained \n, which resulted
1100 in spool file corruption.
1102 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1103 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1104 or write after TLS renegotiation, which otherwise led to messages
1107 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1108 as a tracking header (ie: a signed header comes before the signature).
1109 Patch from Wolfgang Breyha.
1111 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1112 comma-sep list; embedded commas doubled.
1114 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1116 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1118 Report and patch from Dmitry Banschikov.
1120 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1121 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1122 failures. We appear to now support TLS1.1+ with Exim.
1124 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1125 lets Exim select keys and certificates based upon TLS SNI from client.
1126 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1127 before an outbound SMTP session. New log_selector, +tls_sni.
1129 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1130 NULL dereference. Report and patch from Alun Jones.
1132 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1133 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1134 Not seeing resolver debug output on NetBSD, but suspect this is a
1135 resolver implementation change.
1137 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1138 Left warnings. Added "eximon gdb" invocation mode.
1140 PP/21 Defaulting "accept_8bitmime" to true, not false.
1142 PP/22 Added -bw for inetd wait mode support.
1144 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1145 locate the relevant includes and libraries. Made this the default.
1147 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1148 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1150 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1151 This may cause build issues on older platforms.
1153 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1154 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1155 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1156 Added SNI support via GnuTLS too.
1157 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1159 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1161 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1162 Applied second patch from Janne, implementing suggestion to default
1163 multiple-strings-in-record handling to match SPF spec.
1165 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1167 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1168 read-only, out of scope).
1169 Patch from Wolfgang Breyha, report from Stuart Northfield.
1171 PP/29 Fix three issues highlighted by clang analyser static analysis.
1172 Only crash-plausible issue would require the Cambridge-specific
1173 iplookup router and a misconfiguration.
1174 Report from Marcin Mirosław.
1176 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1178 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1179 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1180 As part of this, removing so much warning spew let me fix some minor
1181 real issues in debug logging.
1183 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1184 assignment on my part. Fixed.
1186 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1187 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1188 Janne Snabb (who went above and beyond: thank you).
1190 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1191 string otherwise requires a connection and a bunch more work and it's
1192 relatively easy to get wrong. Should also expose TLS library linkage
1195 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1196 64-bit ${eval} (JH/03).
1198 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1199 GNU libc to support some of the 64-bit stuff, should not lead to
1200 conflicts. Defined before os.h is pulled in, so if a given platform
1201 needs to override this, it can.
1203 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1204 protection layer was required, which is not implemented.
1205 Bugzilla 1254, patch from Wolfgang Breyha.
1207 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1208 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1209 tls_dhparam take prime identifiers. Also unbreak combination of
1210 OpenSSL+DH_params+TLSSNI.
1212 PP/39 Disable SSLv2 by default in OpenSSL support.
1218 PP/01 Solaris build fix for Oracle's LDAP libraries.
1219 Bugzilla 1109, patch from Stephen Usher.
1221 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1223 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1226 TF/02 Fix a couple more cases where we did not log the error message
1227 when unlink() failed. See also change 4.74-TF/03.
1229 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1230 lock up or crash if it happened to be inside a call to libc when it
1231 got a SIGUSR1 from exiwhat.
1233 The SIGUSR1 handler appends the current process status to the process
1234 log which is later printed by exiwhat. It used to use the general
1235 purpose logging code to do this, but several functions it calls are
1236 not safe for signals.
1238 The new output code in the SIGUSR1 handler is specific to the process
1239 log, and simple enough that it's easy to inspect for signal safety.
1240 Removing some special cases also simplifies the general logging code.
1241 Removing the spurious timestamps from the process log simplifies
1244 TF/04 Improved ratelimit ACL condition.
1246 The /noupdate option has been deprecated in favour of /readonly which
1247 has clearer semantics. The /leaky, /strict, and /readonly update modes
1248 are mutually exclusive. The update mode is no longer included in the
1249 database key; it just determines when the database is updated. (This
1250 means that when you upgrade Exim will forget old rate measurements.)
1252 Exim now checks that the per_* options are used with an update mode that
1253 makes sense for the current ACL. For example, when Exim is processing a
1254 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1255 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1256 must specify per_mail/readonly. If you omit the update mode it defaults to
1257 /leaky where that makes sense (as before) or /readonly where required.
1259 The /noupdate option is now undocumented but still supported for
1260 backwards compatibility. It is equivalent to /readonly except that in
1261 ACLs where /readonly is required you may specify /leaky/noupdate or
1262 /strict/noupdate which are treated the same as /readonly.
1264 A useful new feature is the /count= option. This is a generalization
1265 of the per_byte option, so that you can measure the throughput of other
1266 aggregate values. For example, the per_byte option is now equivalent
1267 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1269 The per_rcpt option has been generalized using the /count= mechanism
1270 (though it's more complicated than the per_byte equivalence). When it is
1271 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1272 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1273 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1274 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1275 Note that using per_rcpt with a non-readonly update mode in more than
1276 one ACL will cause the recipients to be double-counted. (The per_mail
1277 and per_byte options don't have this problem.)
1279 The handling of very low rates has changed slightly. If the computed rate
1280 is less than the event's count (usually one) then this event is the first
1281 after a long gap. In this case the rate is set to the same as this event's
1282 count, so that the first message of a spam run is counted properly.
1284 The major new feature is a mechanism for counting the rate of unique
1285 events. The new per_addr option counts the number of different
1286 recipients that someone has sent messages to in the last time period. It
1287 behaves like per_rcpt if all the recipient addresses are different, but
1288 duplicate recipient addresses do not increase the measured rate. Like
1289 the /count= option this is a general mechanism, so the per_addr option
1290 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1291 example, measure the rate that a client uses different sender addresses
1292 with the options per_mail/unique=$sender_address. There are further
1293 details in the main documentation.
1295 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1297 TF/06 Removed a few PCRE remnants.
1299 TF/07 Automatically extract Exim's version number from tags in the git
1300 repository when doing development or release builds.
1302 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1303 Bugzilla 879. Patch from Paul Fisher.
1305 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1306 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1309 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1311 PP/05 Variable $av_failed, true if the AV scanner deferred.
1312 Bugzilla 1078. Patch from John Horne.
1314 PP/06 Stop make process more reliably on build failure.
1315 Bugzilla 1087. Patch from Heiko Schlittermann.
1317 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1318 Bugzilla 1089. Patch from Heiko Schlittermann.
1320 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1321 Bugzilla 1131. Patch from Holger Weiß.
1323 PP/09 Handle IPv6 addresses with SPF.
1324 Bugzilla 860. Patch from Wolfgang Breyha.
1326 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1328 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1331 PP/11 match_* no longer expand right-hand-side by default.
1332 New compile-time build option, EXPAND_LISTMATCH_RHS.
1333 New expansion conditions, "inlist", "inlisti".
1335 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1337 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1339 PP/14 fix log_write() format string regression from TF/03.
1340 Bugzilla 1152. Patch from Dmitry Isaikin.
1346 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1348 PP/02 Harmonised TLS library version reporting; only show if debugging.
1349 Layout now matches that introduced for other libraries in 4.74 PP/03.
1351 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1353 PP/04 New "dns_use_edns0" global option.
1355 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1358 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1359 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1361 TK/01 Updated PolarSSL code to 0.14.2.
1362 Bugzilla 1097. Patch from Andreas Metzler.
1364 PP/07 Catch divide-by-zero in ${eval:...}.
1365 Fixes bugzilla 1102.
1367 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1370 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1371 format-string attack -- SECURITY: remote arbitrary code execution.
1373 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1374 time unintentionally subject to list matching rules, letting the header
1375 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1376 arbitrary string expansion). This allowed for information disclosure.
1378 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1379 INT_MIN/-1 -- value coerced to INT_MAX.
1385 NM/01 Workaround for PCRE version dependency in version reporting
1388 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1389 This fixes portability to compilers other than gcc, notably
1390 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1392 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1393 makefiles for portability to HP-UX and POSIX correctness.
1395 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1396 Also via indented variable definition in the Makefile.
1397 (Debugging by Oliver Heesakkers).
1399 PP/02 Restore caching of spamd results with expanded spamd_address.
1400 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1402 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1403 Improves build reliability. Fix from: Frank Elsner
1405 NM/02 Fix wide character breakage in the rfc2047 coding
1406 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1408 NM/03 Allow underscore in dnslist lookups
1409 Fixes bug 1026. Patch from Graeme Fowler
1411 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1412 Code patches from Adam Ciarcinski of NetBSD.
1414 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1417 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1418 is logged, to avoid truncation. Patch from John Horne.
1420 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1421 Patch from Jakob Hirsch.
1423 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1424 SQL string expansion failure details.
1425 Patch from Andrey Oktyabrski.
1427 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1428 Patch from Simon Arlott.
1430 PP/09 New lookups functionality failed to compile on old gcc which rejects
1431 extern declarations in function scope.
1432 Patch from Oliver Fleischmann
1434 PP/10 Use sig_atomic_t for flags set from signal handlers.
1435 Check getgroups() return and improve debugging.
1436 Fixed developed for diagnosis in bug 927 (which turned out to be
1439 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1440 Patch from Mark Zealey.
1442 PP/12 Bugzilla 1056: Improved spamd server selection.
1443 Patch from Mark Zealey.
1445 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1446 Based on patch from Heiko Schlittermann.
1448 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1449 Patch from Uwe Doering, sign-off by Michael Haardt.
1451 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1452 variable declaration deep within a block. Bug and patch from
1455 PP/15 lookups-Makefile IRIX compatibility coercion.
1457 PP/16 Make DISABLE_DKIM build knob functional.
1459 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1460 Patch by Simon Arlott
1462 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1463 variable argument macros. Our copy now differs from upstream.
1469 TF/01 Failure to get a lock on a hints database can have serious
1470 consequences so log it to the panic log.
1472 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1473 controlled using the smtp_confirmation log selector.
1475 TF/03 Include the error message when we fail to unlink a spool file.
1477 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1478 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1479 for maintaining out-of-tree patches for some time.
1481 PP/01 Bugzilla 139: Documentation and portability issues.
1482 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1483 Handle per-OS dynamic-module compilation flags.
1485 PP/02 Let /dev/null have normal permissions.
1486 The 4.73 fixes were a little too stringent and complained about the
1487 permissions on /dev/null. Exempt it from some checks.
1488 Reported by Andreas M. Kirchwitz.
1490 PP/03 Report version information for many libraries, including
1491 Exim version information for dynamically loaded libraries. Created
1492 version.h, now support a version extension string for distributors
1493 who patch heavily. Dynamic module ABI change.
1495 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1496 privilege escalation vulnerability whereby the Exim run-time user
1497 can cause root to append content of the attacker's choosing to
1500 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1503 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1504 If dropping privileges for untrusted macros, we disabled normal logging
1505 on the basis that it would fail; for the Exim run-time user, this is not
1506 the case, and it resulted in successful deliveries going unlogged.
1507 Fixed. Reported by Andreas Metzler.
1513 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1514 only prepend for the Resent-* case. Fixes regression introduced in
1515 Exim 4.70 by NM/22 for Bugzilla 607.
1517 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1518 increasing numbers of administrators be bitten by this.
1520 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1522 PP/03 Bugzilla 994: added openssl_options main configuration option.
1524 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1526 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1528 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1530 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1531 without a peer certificate, leading to a segfault because of an
1532 assumption that peers always have certificates. Be a little more
1533 paranoid. Problem reported by Martin Tscholak.
1535 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1536 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1537 NB: ClamAV planning to remove STREAM in "middle of 2010".
1538 CL also introduces -bmalware, various -d+acl logging additions and
1539 more caution in buffer sizes.
1541 PP/09 Implemented reverse_ip expansion operator.
1543 PP/10 Bugzilla 937: provide a "debug" ACL control.
1545 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1547 PP/12 Bugzilla 973: Implement --version.
1549 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1551 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1553 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1555 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1556 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1557 ignore trailing whitespace.
1559 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1561 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1564 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1565 Notification from Dr Andrew Aitchison.
1567 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1568 ExtendedDetectionInfo response format.
1569 Notification from John Horne.
1571 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1574 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1575 XSL and documented dependency on system catalogs, with examples of how
1578 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1581 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1582 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1583 configuration file which is writeable by the Exim user or group.
1585 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1586 of configuration files to cover files specified with the -C option if
1587 they are going to be used with root privileges, not just the default
1590 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1591 option (effectively making it always true).
1593 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1594 files to be used while preserving root privileges.
1596 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1597 that rogue child processes cannot use them.
1599 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1600 run-time user, instead of root.
1602 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1603 Exim run-time user without dropping privileges.
1605 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1606 result string, instead of calling string_vformat() twice with the same
1609 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1610 for other users. Others should always drop root privileges if they use
1611 -C on the command line, even for a whitelisted configure file.
1613 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1615 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1621 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1622 $data_path, and $header_path variables; fixed documentation bugs and
1625 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1626 exipick to access non-standard spools, including the "frozen" queue
1629 NM/01 Bugzilla 965: Support mysql stored procedures.
1630 Patch from Alain Williams
1632 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1634 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1635 Patch from Andreas Metzler
1637 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1638 Patch from Kirill Miazine
1640 NM/05 Bugzilla 671: Added umask to procmail example.
1642 JJ/03 installed exipick 20100323.0, fixing doc bug
1644 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1645 directory. Notification and patch from Dan Rosenberg.
1647 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1649 TK/02 Improve log output when DKIM signing operation fails.
1651 MH/01 Treat the transport option dkim_domain as a colon separated
1652 list, not as a single string, and sign the message with each element,
1653 omitting multiple occurences of the same signer.
1655 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1656 Bugzilla 985, 986. Patch by Simon Arlott
1658 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1659 Patch by Simon Arlott
1661 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1662 MBX locking. Notification from Dan Rosenberg.
1668 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1670 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1672 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1674 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1676 NM/04 Bugzilla 915: Flag broken perl installation during build.
1682 TK/01 Added patch by Johannes Berg that expands the main option
1683 "spamd_address" if it starts with a dollar sign.
1685 TK/02 Write list of recipients to X-Envelope-Sender header when building
1686 the mbox-format spool file for content scanning (suggested by Jakob
1689 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1690 (http://www.dcc-servers.net/) support via dccifd. Activated by
1691 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1693 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1694 by Mark Daniel Reidel <mr@df.eu>.
1696 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1697 When building exim an external PCRE library is now needed -
1698 PCRE is a system library on the majority of modern systems.
1699 See entry on PCRE_LIBS in EDITME file.
1701 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1702 conversation. Added nologin parameter to request.
1703 Patch contributed by Kirill Miazine.
1705 TF/01 Do not log submission mode rewrites if they do not change the address.
1707 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1709 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1710 log files in place. Contributed by Roberto Lima.
1712 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1714 TF/03 Bugzilla 615: When checking the local_parts router precondition
1715 after a local_part_suffix or local_part_prefix option, Exim now
1716 does not use the address's named list lookup cache, since this
1717 contains cached lookups for the whole local part.
1719 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1720 Robert Millan. Documentation is in experimental-spec.txt.
1722 TF/04 Bugzilla 668: Fix parallel build (make -j).
1724 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1726 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1727 Patch provided by Jan Srzednicki.
1729 TF/05 Leading white space used to be stripped from $spam_report which
1730 wrecked the formatting. Now it is preserved.
1732 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1733 that they are available at delivery time.
1735 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1737 TF/08 TLS error reporting now respects the incoming_interface and
1738 incoming_port log selectors.
1740 TF/09 Produce a more useful error message if an SMTP transport's hosts
1741 setting expands to an empty string.
1743 NM/06 Bugzilla 744: EXPN did not work under TLS.
1744 Patch provided by Phil Pennock.
1746 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1747 Patch provided by Richard Godbee.
1749 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1750 acl_smtp_notquit, added index entry.
1752 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1753 Patch provided by Eugene Bujak.
1755 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1756 accept(). Patch provided by Maxim Dounin.
1758 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1759 Patch provided by Phil Pennock.
1761 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1763 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1764 Patch provided by Brad "anomie" Jorsch.
1766 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1767 Patch provided by Dean Brooks.
1769 TK/05 Add native DKIM support (does not depend on external libraries).
1771 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1772 Patch provided by Graeme Fowler.
1774 NM/16 Bugzilla 851: Documentation example syntax fix.
1776 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1778 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1781 NM/19 Bugzilla 745: TLS version reporting.
1782 Patch provided by Phil Pennock.
1784 NM/20 Bugzilla 167: bool: condition support.
1785 Patch provided by Phil Pennock.
1787 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1788 clients. Patch provided by Phil Pennock.
1790 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1791 Patch provided by Brad "anomie" Jorsch.
1793 NM/23 Bugzilla 687: Fix misparses in eximstats.
1794 Patch provided by Heiko Schlittermann.
1796 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1797 Patch provided by Heiko Schlittermann.
1799 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1800 plus update to original patch.
1802 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1804 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1805 Patch provided by David Brownlee.
1807 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1809 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1811 NM/30 Bugzilla 888: TLS documentation bugfixes.
1813 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1815 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1816 Unlike the original bugzilla I have changed all shell scripts in src tree.
1818 NM/33 Bugzilla 898: Transport filter timeout fix.
1819 Patch by Todd Rinaldo.
1821 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1822 Patch by Serge Demonchaux.
1824 NM/35 Bugzilla 39: Base64 decode bug fixes.
1825 Patch by Jakob Hirsch.
1827 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1829 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1831 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1833 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1839 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1840 ALT-N's libdkim that I have put here:
1841 http://duncanthrax.net/exim-experimental/
1843 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1844 were called 'true' and it seems that C99 defines that as a reserved
1845 keyword to be used with 'bool' variable types. That means you could
1846 not include C99-style headers which use bools without triggering
1847 build errors in sieve.c.
1849 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1850 as mailq or other aliases. Changed the --help handling significantly
1851 to do whats expected. exim_usage() emits usage/help information.
1853 SC/01 Added the -bylocaldomain option to eximstats.
1855 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1857 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1859 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1865 PH/01 Another patch from the Sieve maintainer.
1867 PH/02 When an IPv6 address is converted to a string for single-key lookup
1868 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1869 dots are used instead of colons so that keys in lsearch files need not
1870 contain colons. This was done some time before quoting was made available
1871 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1872 (notated using the quote facility) so as to distinguish them from IPv4
1873 keys. This meant that lookups for IP addresses in host lists did not work
1874 for iplsearch lookups.
1876 This has been fixed by arranging for IPv6 addresses to be expressed with
1877 colons if the lookup type is iplsearch. This is not incompatible, because
1878 previously such lookups could never work.
1880 The situation is now rather anomalous, since one *can* have colons in
1881 ordinary lsearch keys. However, making the change in all cases is
1882 incompatible and would probably break a number of configurations.
1884 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1887 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1888 conversion specification without a maximum field width, thereby enabling
1889 a rogue spamd server to cause a buffer overflow. While nobody in their
1890 right mind would setup Exim to query an untrusted spamd server, an
1891 attacker that gains access to a server running spamd could potentially
1892 exploit this vulnerability to run arbitrary code as the Exim user.
1894 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1895 $primary_hostname instead of what libspf2 thinks the hosts name is.
1897 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1898 a directory entry by the name of the lookup key. Previously, if a
1899 symlink pointed to a non-existing file or a file in a directory that
1900 Exim lacked permissions to read, a lookup for a key matching that
1901 symlink would fail. Now it is enough that a matching directory entry
1902 exists, symlink or not. (Bugzilla 503.)
1904 PH/03 The body_linecount and body_zerocount variables are now exported in the
1907 PH/04 Added the $dnslist_matched variable.
1909 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1910 This means they are set thereafter only if the connection becomes
1913 PH/06 Added the client_condition to authenticators so that some can be skipped
1914 by clients under certain conditions.
1916 PH/07 The error message for a badly-placed control=no_multiline_responses left
1917 "_responses" off the end of the name.
1919 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1921 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1922 (without spaces) instead of just copying the configuration text.
1924 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1926 PH/11 Added $max_received_linelength.
1928 PH/12 Added +ignore_defer and +include_defer to host lists.
1930 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1931 way in which PCRE > 7.0 is built.
1933 PH/14 Implemented queue_only_load_latch.
1935 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1936 MAIL command. The effect was to mangle the value on 64-bit systems.
1938 PH/16 Another patch from the Sieve maintainer.
1940 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1942 PH/18 If a system quota error occurred while trying to create the file for
1943 a maildir delivery, the message "Mailbox is full" was not appended to the
1944 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1945 only to a quota excession during the actual writing of the file.
1947 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1948 characters?) which causes problems in log lines. The DN values are now
1949 passed through string_printing() before being added to log lines.
1951 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1952 and InterBase are left for another time.)
1954 PH/21 Added message_body_newlines option.
1956 PH/22 Guard against possible overflow in moan_check_errorcopy().
1958 PH/23 POSIX allows open() to be a macro; guard against that.
1960 PH/24 If the recipient of an error message contained an @ in the local part
1961 (suitably quoted, of course), incorrect values were put in $domain and
1962 $local_part during the evaluation of errors_copy.
1968 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1969 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1972 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1973 issue a MAIL command.
1975 PH/02 In an ACL statement such as
1977 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1979 if a client was not listed at all, or was listed with a value other than
1980 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1981 the condition was not true (as it should be), so access was not denied.
1982 The bug was that the ! inversion was incorrectly passed on to the second
1983 item. This has been fixed.
1985 PH/03 Added additional dnslists conditions == and =& which are different from
1986 = and & when the dns lookup returns more than one IP address.
1988 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1989 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1991 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1992 FSYNC, which compiles an option called disable_fsync that allows for
1993 bypassing fsync(). The documentation is heavily laced with warnings.
1995 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1997 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1998 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1999 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2000 including adding "make clean"; (3) Added -fPIC when compiling the test
2001 dynamically loaded module, to get rid of a warning.
2003 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2004 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2005 = 0s. The bug is otherwise harmless.
2007 PH/07 There was a bug in the dovecot authenticator such that the value of
2008 $auth1 could be overwritten, and so not correctly preserved, after a
2009 successful authentication. This usually meant that the value preserved by
2010 the server_setid option was incorrect.
2012 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2014 PH/09 Installed PCRE release 7.0.
2016 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2017 run for batched SMTP input. It is now run at the start of every message
2018 in the batch. While fixing this I discovered that the process information
2019 (output by running exiwhat) was not always getting set for -bs and -bS
2020 input. This is fixed, and it now also says "batched" for BSMTP.
2022 PH/11 Added control=no_pipelining.
2024 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2025 patch, slightly modified), and move the expansion of helo_data till after
2026 the connection is made in the smtp transport (so it can use these
2029 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2031 PH/14 Added log_selector = +pid.
2033 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2035 PH/16 Add ${if forany and ${if forall.
2037 PH/17 Added dsn_from option to vary the From: line in DSNs.
2039 PH/18 Flush SMTP output before performing a callout, unless control =
2040 no_callout_flush is set.
2042 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2043 was true (the default) a successful delivery failed to delete the retry
2044 item, thus causing premature timeout of the address. The bug is now
2047 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2049 PH/21 Long custom messages for fakedefer and fakereject are now split up
2050 into multiline responses in the same way that messages for "deny" and
2051 other ACL rejections are.
2053 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2054 with slight modification.
2056 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2057 draft, changing the syntax and factoring some duplicate code".
2059 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2060 for deliveries of the second and subsequent messages over the same SMTP
2063 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2064 ${reduce, with only minor "tidies".
2066 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2068 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2069 expansion side effects.
2071 PH/27 When a message times out after an over-quota error from an Exim-imposed
2072 quota, the bounce message says "mailbox is full". This message was not
2073 being given when it was a system quota that was exceeded. It now should
2076 MH/03 Made $recipients available in local_scan(). local_scan() already has
2077 better access to the recipient list through recipients_list[], but
2078 $recipients can be useful in postmaster-provided expansion strings.
2080 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2081 in the case of a MAIL command with additional options following the
2082 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2083 were accidentally chopped off.
2085 PH/29 SMTP synchronization checks are implemented when a command is read -
2086 there is a check that no more input is waiting when there shouldn't be
2087 any. However, for some commands, a delay in an ACL can mean that it is
2088 some time before the response is written. In this time, more input might
2089 arrive, invalidly. So now there are extra checks after an ACL has run for
2090 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2091 pipelining has not been advertised.
2093 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2095 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2096 returned in the SMTP response when smtp_return_error_details was false.
2097 This has been fixed.
2099 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2100 instead of the C I/O that was originally supplied, because problems were
2101 reported on Solaris.
2103 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2104 Exim which did not show up earlier: it was assuming that a call to
2105 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2106 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2107 was a macro that became an assignment, so it seemed to work. This has
2108 changed to a proper function call with a void return, hence the compile
2109 error. Exim's code has been fixed.
2111 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2114 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2116 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2118 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2119 because it turns out to be too big or there is a timeout) there is no
2120 "Completed" line in the log. When some messages of this type were
2121 selected by exigrep, they were listed as "not completed". Others were
2122 picked up by some special patterns. I have improved the selection
2123 criteria to be more general.
2125 PH/38 The host_find_failed option in the manualroute router can now be set
2126 to "ignore", to completely ignore a host whose IP address cannot be
2127 found. If all hosts are ignored, the behaviour is controlled by the new
2128 host_all_ignored option.
2130 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2131 homing or because of multiple MX records with /mx) generated more than
2132 one IP address, and the following item turned out to be the local host,
2133 all the secondary addresses of the first item were incorrectly removed
2134 from the list, along with the local host and any following hosts (which
2135 is what is supposed to happen).
2137 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2138 whoever called Exim into the -H file. In the case of the daemon it was
2139 behaving confusingly. When first started, it used values for whoever
2140 started the daemon, but after a SIGHUP it used the Exim user (because it
2141 calls itself on a restart). I have changed the code so that it now always
2144 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2145 message are rejected with the same error (e.g. no authentication or bad
2146 sender address), and a DATA command is nevertheless sent (as can happen
2147 with PIPELINING or a stupid MUA), the error message that was given to the
2148 RCPT commands is included in the rejection of the DATA command. This is
2149 intended to be helpful for MUAs that show only the final error to their
2152 PH/42 Another patch from the Sieve maintainer.
2154 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2155 Eximstats - Fixed some broken HTML links and added missing column headers
2157 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2158 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2160 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2162 PH/43 Yet another patch from the Sieve maintainer.
2164 PH/44 I found a way to check for a TCP/IP connection going away before sending
2165 the response to the final '.' that terminates a message, but only in the
2166 case where the client has not sent further data following the '.'
2167 (unfortunately, this is allowed). However, in many cases there won't be
2168 any further data because there won't be any more messages to send. A call
2169 to select() can be used: if it shows that the input is "ready", there is
2170 either input waiting, or the socket has been closed. An attempt to read
2171 the next input character can distinguish the two cases. Previously, Exim
2172 would have sent an OK response which the client would never have see.
2173 This could lead to message repetition. This fix should cure that, at
2174 least in a lot of common cases.
2176 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2177 advertised in response to EHLO.
2183 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2184 fixed by 4.65/MH/01 (is this a record?) are fixed:
2186 (i) An empty string was always treated as zero by the numeric comparison
2187 operators. This behaviour has been restored.
2189 (ii) It is documented that the numeric comparison operators always treat
2190 their arguments as decimal numbers. This was broken in that numbers
2191 starting with 0 were being interpreted as octal.
2193 While fixing these problems I realized that there was another issue that
2194 hadn't been noticed. Values of message_size_limit (both the global option
2195 and the transport option) were treated as octal if they started with 0.
2196 The documentation was vague. These values are now always treated as
2197 decimal, and I will make that clear in the documentation.
2203 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2204 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2207 MH/01 Don't check that the operands of numeric comparison operators are
2208 integers when their expansion is in "skipping" mode (fixes bug
2209 introduced by 4.64-PH/07).
2211 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2212 child addresses, Exim now panics and dies. Previously, because the count
2213 is held in a short int, deliveries were likely to be lost. As such a
2214 large number of recipients for a single message is ridiculous
2215 (performance will be very, very poor), I have chosen to impose a limit
2216 rather than extend the field.
2222 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2223 leftover -K file (the existence of which was triggered by #402).
2224 While we were at it, introduced process PID as part of the -K
2225 filename. This should rule out race conditions when creating
2228 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2229 processing considerably. Previous code took too long for large mails,
2230 triggering a timeout which in turn triggers #401.
2232 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2233 in the DK code in transports.c. sendfile() is not really portable,
2234 hence the _LINUX specificness.
2236 TF/01 In the add_headers option to the mail command in an Exim filter,
2237 there was a bug that Exim would claim a syntax error in any
2238 header after the first one which had an odd number of characters
2241 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2242 callout verification, Exim cached a "reject" for the entire domain. This
2243 is correct for most verifications, but it is not correct for a recipient
2244 verification with use_sender or use_postmaster set, because in that case
2245 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2246 case of MAIL FROM:<> rejection from other early rejections (e.g.
2247 rejection of HELO). When verifying a recipient using a non-null MAIL
2248 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2249 Whatever the result of the callout, the value of the domain cache is
2250 left unchanged (for any other kind of callout, getting as far as trying
2251 RCPT means that the domain itself is ok).
2253 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2256 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2257 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2258 session, the daemon ignores ECONNECT errors and logs others; it now
2259 ignores EPIPE as well.
2261 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2262 (quoted-printable decoding).
2264 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2265 later the small subsequent patch to fix an introduced bug.
2267 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2269 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2271 PH/08 An error is now given if message_size_limit is specified negative.
2273 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2274 to be given (somewhat) arbitrary names.
2276 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2279 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2280 miscellaneous code fixes
2282 PH/10 Added the log_reject_target ACL modifier to specify where to log
2285 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2286 hostname. This is wrong, because it relates to the incoming message (and
2287 probably the interface on which it is arriving) and not to the outgoing
2288 callout (which could be using a different interface). This has been
2289 changed to use the value of the helo_data option from the smtp transport
2290 instead - this is what is used when a message is actually being sent. If
2291 there is no remote transport (possible with a router that sets up host
2292 addresses), $smtp_active_hostname is used.
2294 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2295 tweaks were necessary in order to get it to work (see also 21 below):
2296 (a) The code assumed that strncpy() returns a negative number on buffer
2297 overflow, which isn't the case. Replaced with Exim's string_format()
2299 (b) There were several signed/unsigned issues. I just did the minimum
2300 hacking in of casts. There is scope for a larger refactoring.
2301 (c) The code used strcasecmp() which is not a standard C function.
2302 Replaced with Exim's strcmpic() function.
2303 (d) The code set only $1; it now sets $auth1 as well.
2304 (e) A simple test gave the error "authentication client didn't specify
2305 service in request". It would seem that Dovecot has changed its
2306 interface. Fortunately there's a specification; I followed it and
2307 changed what the client sends and it appears to be working now.
2309 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2312 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2313 address A is aliased to B and C, where B exists and C does not. Without
2314 -v the output is "A verified" because verification stops after a
2315 successful redirection if more than one address is generated. However,
2316 with -v the child addresses are also verified. Exim was outputting "A
2317 failed to verify" and then showing the successful verification for C,
2318 with its parentage. It now outputs "B failed to verify", showing B's
2319 parentage before showing the successful verification of C.
2321 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2322 look up a TXT record in a specific list after matching in a combined
2325 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2326 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2327 they consult the DNS. I had assumed they would set it the way they
2328 wanted; and indeed my experiments on Linux seem to show that in some
2329 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2330 To be on the safe side, however, I have now made the interface to
2331 host_find_byname() similar to host_find_bydns(), with an argument
2332 containing the DNS resolver options. The host_find_byname() function now
2333 sets these options at its start, just as host_find_bydns() does. The smtp
2334 transport options dns_qualify_single and dns_search_parents are passed to
2335 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2336 of host_find_byname() use the default settings of RES_DEFNAMES
2337 (qualify_single) but not RES_DNSRCH (search_parents).
2339 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2340 spool_read_header() do less string testing, by means of a preliminary
2341 switch on the second character of optional "-foo" lines. (This is
2342 overdue, caused by the large number of possibilities that now exist.
2343 Originally there were few.) While I was there, I also converted the
2344 str(n)cmp tests so they don't re-test the leading "-" and the first
2345 character, in the hope this might squeeze out yet more improvement.
2347 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2348 flag allowing group syntax was set by the header_syntax check but not
2349 turned off, possible causing trouble later; (2) The flag was not being
2350 set at all for the header_verify test, causing "group"-style headers to
2351 be rejected. I have now set it in this case, and also caused header_
2352 verify to ignore an empty address taken from a group. While doing this, I
2353 came across some other cases where the code for allowing group syntax
2354 while scanning a header line wasn't quite right (mostly, not resetting
2355 the flag correctly in the right place). These bugs could have caused
2356 trouble for malformed header lines. I hope it is now all correct.
2358 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2359 with the "reply" argument non-NULL. The code, however (which originally
2360 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2361 but it didn't always do it. This confused somebody who was copying the
2362 code for some other use. I have removed all the tests.
2364 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2365 feature that was used to support insecure browsers during the U.S. crypto
2366 embargo. It requires special client support, and Exim is probably the
2367 only MTA that supported it -- and would never use it because real RSA is
2368 always available. This code has been removed, because it had the bad
2369 effect of slowing Exim down by computing (never used) parameters for the
2370 RSA_EXPORT functionality.
2372 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2373 authenticator to fail if there's a tab character in the incoming data
2374 (there should never be unless someone is messing about, as it's supposed
2375 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2376 if the connection is using TLS or if the remote IP is the same as the
2377 local IP, and the "valid-client-cert option" if a client certificate has
2380 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2381 authenticators. This can be used for authorization after authentication
2382 succeeds. (In the case of plaintext, it servers for both authentication
2385 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2386 if any retry times were supplied.
2388 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2389 connection, where there is no client IP address to check. In this
2390 situation, the verify now always succeeds.
2392 PH/25 Applied John Jetmore's -Mset patch.
2394 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2396 PH/27 In a string expansion for a processed (not raw) header when multiple
2397 headers of the same name were present, leading whitespace was being
2398 removed from all of them, but trailing whitespace was being removed only
2399 from the last one. Now trailing whitespace is removed from each header
2400 before concatenation. Completely empty headers in a concatenation (as
2401 before) are ignored.
2403 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2404 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2406 PH/29 [Removed. This was a change that I later backed out, and forgot to
2407 correct the ChangeLog entry (that I had efficiently created) before
2408 committing the later change.]
2410 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2411 address errors (4xx response to RCPT) over the same connection as other
2412 messages routed to the same hosts. Such deliveries are always "forced",
2413 so retry times are not inspected. This resulted in far too many retries
2414 for the affected addresses. The effect occurred only when there were more
2415 hosts than the hosts_max_try setting in the smtp transport when it had
2416 the 4xx errors. Those hosts that it had tried were not added to the list
2417 of hosts for which the message was waiting, so if all were tried, there
2418 was no problem. Two fixes have been applied:
2420 (i) If there are any address or message errors in an SMTP delivery, none
2421 of the hosts (tried or untried) are now added to the list of hosts
2422 for which the message is waiting, so the message should not be a
2423 candidate for sending over the same connection that was used for a
2424 successful delivery of some other message. This seems entirely
2425 reasonable: after all the message is NOT "waiting for some host".
2426 This is so "obvious" that I'm not sure why it wasn't done
2427 previously. Hope I haven't missed anything, but it can't do any
2428 harm, as the worst effect is to miss an optimization.
2430 (ii) If, despite (i), such a delivery is accidentally attempted, the
2431 routing retry time is respected, so at least it doesn't keep
2432 hammering the server.
2434 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2435 in ${readsocket because some servers need this prod.
2437 PH/32 Added some extra debug output when updating a wait-xxx database.
2439 PH/33 The hint "could be header name not terminated by colon", which has been
2440 given for certain expansion errors for a long time, was not being given
2441 for the ${if def:h_colon_omitted{... case.
2443 PH/34 The spec says: "With one important exception, whenever a domain list is
2444 being scanned, $domain contains the subject domain." There was at least
2445 one case where this was not true.
2447 PH/35 The error "getsockname() failed: connection reset by peer" was being
2448 written to the panic log as well as the main log, but it isn't really
2449 panic-worthy as it just means the connection died rather early on. I have
2450 removed the panic log writing for the ECONNRESET error when getsockname()
2453 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2454 runs only) independently of the message's sender address. This meant
2455 that, if the 4xx error was in fact related to the sender, a different
2456 message to the same recipient with a different sender could confuse
2457 things. In particular, this can happen when sending to a greylisting
2458 server, but other circumstances could also provoke similar problems.
2459 I have changed the default so that the retry time for these errors is now
2460 based a combination of the sender and recipient addresses. This change
2461 can be overridden by setting address_retry_include_sender=false in the
2464 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2465 remote server are returned as part of bounce messages. This was not
2466 happening for LMTP over a pipe (the lmtp transport), but now it is the
2467 same for both kinds of LMTP.
2469 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2470 in header lines that were in fact CNAMEs. This is no longer the case.
2472 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2473 and queue runs started by the daemon processed all messages. This has
2474 been fixed so that -R and -S can now usefully be given with -q<time>.
2476 PH/40 Import PCRE release 6.7 (fixes some bugs).
2478 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2480 PH/42 Give an error if -q is specified more than once.
2482 PH/43 Renamed the variables $interface_address and $interface_port as
2483 $received_ip_address and $received_port, to make it clear that these
2484 values apply to message reception, and not to the outgoing interface when
2485 a message is delivered. (The old names remain recognized, of course.)
2487 PH/44 There was no timeout on the connect() call when using a Unix domain
2488 socket in the ${readsocket expansion. There now is.
2490 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2491 be meaningful with "accept".
2493 SC/01 Eximstats V1.43
2494 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2496 SC/02 Eximstats V1.44
2497 Use a glob alias rather than an array ref in the generated
2498 parser. This improves both readability and performance.
2500 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2501 Collect SpamAssassin and rejection statistics.
2502 Don't display local sender or destination tables unless
2503 there is data to show.
2504 Added average volumes into the top table text output.
2506 SC/04 Eximstats V1.46
2507 Collect data on the number of addresses (recipients)
2508 as well as the number of messages.
2510 SC/05 Eximstats V1.47
2511 Added 'Message too big' to the list of mail rejection
2512 reasons (thanks to Marco Gaiarin).
2514 SC/06 Eximstats V1.48
2515 Mainlog lines which have GMT offsets and are too short to
2516 have a flag are now skipped.
2518 SC/07 Eximstats V1.49 (Alain Williams)
2519 Added the -emptyok flag.
2521 SC/08 Eximstats V1.50
2522 Fixes for obtaining the IP address from reject messages.
2524 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2525 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2526 whitespace changes from 4.64-PH/27
2528 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2531 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2532 are found, allow negative numbers in numeric criteria)
2534 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2536 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2537 to match changes made in 4.64-PH/43
2539 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2541 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2542 "domain" is actually an IP address, and force a failure. This locks out
2543 those revolvers/nameservers that support "A-for-A" lookups, in
2544 contravention of the specifications.
2546 PH/48 When a host name was looked up from an IP address, and the subsequent
2547 forward lookup of the name timed out, the host name was left in
2548 $sender_host_name, contrary to the specification.
2550 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2551 restricted to single-key lookups, Exim was not diagnosing an error if
2552 * or *@ was used with a query-style lookup.
2554 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2556 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2557 long ago, but noone interested enough thought of it. Let's just say that
2558 the "1.1" means that there are some new functions that weren't there at
2559 some point in the past.
2561 PH/51 Error processing for expansion failure of helo_data from an smtp
2562 transport during callout processing was broken.
2564 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2565 tested/used via the -bh/-bhc/-bs options.
2567 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2568 bug, fixed in subsequent PCRE releases).
2570 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2571 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2573 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2579 SC/01 Use a glob alias rather than an array ref in eximstats generated
2580 parser. This improves both readability and performance.
2582 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2583 Don't display local sender or destination tables in eximstats unless
2584 there is data to show.
2585 Added average volumes into the eximstats top table text output.
2587 SC/03 Collect data on the number of addresses (recipients) as well
2588 as the number of messages in eximstats.
2590 TF/01 Correct an error in the documentation for the redirect router. Exim
2591 does (usually) call initgroups() when daemonizing.
2593 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2594 with consistent privilege compared to when running as a daemon.
2596 TF/03 Note in the spec that $authenticated_id is not set for local
2597 submissions from trusted users.
2599 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2600 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2602 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2603 by adding some example configuration directives to the default
2604 configuration file. A little bit of work is required to uncomment the
2605 directives and define how usernames and passwords are checked, but
2606 there is now a framework to start from.
2608 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2609 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2610 without this. I don't know how relevant this is to other LDAP libraries.
2612 PH/02 Add the verb name to the "unknown ACL verb" error.
2614 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2616 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2618 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2619 directory not expanded when it should be if an expanded home directory
2620 was set for the address (which is overridden by the transport).
2622 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2625 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2626 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2627 because it is too late at that time, and has no effect.
2629 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2630 security issue with \' (bugzilla #107). I could not use the
2631 PQescapeStringConn() function, because it needs a PGconn value as one of
2634 PH/08 When testing addresses using -bt, indicate those final addresses that
2635 are duplicates that would not cause an additional delivery. At least one
2636 person was confused, thinking that -bt output corresponded to deliveries.
2637 (Suppressing duplicates isn't a good idea as you lose the information
2638 about possibly different redirections that led to the duplicates.)
2640 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2641 systems where poll() doesn't work, in particular OS X.
2643 PH/10 Added more information to debugging output for retry time not reached.
2645 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2646 operations in malware.c.
2648 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2651 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2652 syslog_duplication set false, log lines that would normally be written
2653 both the the main log and to the reject log were not written to syslog at
2656 PH/14 In the default configuration, change the use of "message" in ACL warn
2657 statements to "add_header".
2659 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2660 not followed by a command (e.g. "seen endif").
2662 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2663 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2666 PH/17 Added extra conditions to the default value of delay_warning_condition
2670 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2671 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2672 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2675 The Auto-Submitted: and various List- headers are standardised, whereas I
2676 don't think Precedence: ever was.
2678 PH/18 Refactored debugging code in route_finduser() to show more information,
2679 in particular, the error code if getpwnam() issues one.
2681 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2682 This is apparently needed in addition to the PH/07 change above to avoid
2683 any possible encoding problems.
2685 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2686 but not after initializing Perl.
2688 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2689 output them only if debugging. By default they are written stderr,
2690 apparently, which is not desirable.
2692 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2695 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2698 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2700 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2701 authenticated or an ident call has been made. Suppress the default
2702 values for $authenticated_id and $authenticated_sender (but permit -oMai
2703 and -oMas) when testing with -bh.
2705 PH/24 Re-jigged the order of the tests in the default configuration so that the
2706 tests for valid domains and recipients precede the DNS black list and CSA
2707 tests, on the grounds that those ones are more expensive.
2709 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2710 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2711 This bug exists in every version of Exim that I still have, right back to
2714 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2715 However, an attempt to turn on case-sensitivity in a regex key by
2716 including (?-i) didn't work because the subject string was already
2717 lowercased, and the effects were non-intuitive. It turns out that a
2718 one-line patch can be used to allow (?-i) to work as expected.
2724 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2725 other effects) broke the use of negated acl sub-conditions.
2727 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2730 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2731 "Deny" causes Exim to reject the incoming connection with a 554 error.
2732 Unfortunately, if there is a major crisis, such as a disk failure,
2733 tcp-wrappers gives "deny", whereas what one would like would be some
2734 kind of temporary error. A kludge has been added to help with this.
2735 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2736 554 error is used if errno is still zero or contains ENOENT (which occurs
2737 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2740 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2742 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2743 errors. Otherwise a message that provokes a temporary error (when other
2744 messages do not) can cause a whole host to time out.
2746 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2747 addresses were routed directly to files or pipes from a redirect router.
2748 File deliveries just didn't batch; pipe deliveries might have suffered
2751 PH/06 A failure to get a lock for a hints database would erroneously always say
2752 "Failed to get write lock", even when it was really a read lock.
2754 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2755 of 0600. This has been changed to use the value of the lockfile_mode
2756 option (which defaults to 0600).
2758 PH/08 Applied small patch from the Sieve maintainer.
2760 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2761 folder from quota calculations, a direct delivery into this folder messed
2762 up the contents of the maildirsize file. This was because the regex was
2763 used only to exclude .Trash (or whatever) when the size of the mailbox
2764 was calculated. There was no check that a delivery was happening into an
2765 excluded directory. This bug has been fixed by ignoring all quota
2766 processing for deliveries into excluded directories.
2768 PH/10 Added the maildirfolder_create_regex option to appendfile.
2774 PH/01 The code for finding all the local interface addresses on a FreeBSD
2775 system running IPv6 was broken. This may well have applied to all BSD
2776 systems, as well as to others that have similar system calls. The broken
2777 code found IPv4 interfaces correctly, but gave incorrect values for the
2778 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2779 that it would not match correctly against @[] and not recognize the IPv6
2782 PH/02 The ipliteral router was not recognizing addresses of the form user@
2783 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2785 PH/03 Added disable_ipv6.
2787 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2788 decoded form, because it is most often used to construct To: headers
2789 lines in autoreplies, and the decoded form may well be syntactically
2790 invalid. However, $reply_address has leading white space removed, and all
2791 newlines turned into spaces so that the autoreply transport does not
2794 PH/05 If group was specified without a user on a router, and no group or user
2795 was specified on a transport, the group from the router was ignored.
2797 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2798 for visible compile-time settings that can be used to change these
2799 numbers, for those that want even more. Backwards compatibility with old
2800 spool files has been maintained. However, going back to a previous Exim
2801 release will lost any variables that are in spool files.
2803 PH/07 Two small changes when running in the test harness: increase delay when
2804 passing a TCP/IP connection to a new process, in case the original
2805 process has to generate a bounce, and remove special handling of
2806 127.0.0.2 (sic), which is no longer necessary.
2808 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2809 be the same on different OS.
2811 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2814 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2815 whether --show-vars was specified or not
2817 JJ/02 exipick: Added support for new ACL variable spool format introduced
2820 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2821 syntactically invalid From: or Reply-to: line, and a filter used this to
2822 generate an autoreply, and therefore failed to obtain an address for the
2823 autoreply, Exim could try to deliver to a non-existent relative file
2824 name, causing unrelated and misleading errors. What now happens is that
2825 it logs this as a hard delivery error, but does not attempt to create a
2828 PH/11 The exinext utility has a -C option for testing purposes, but although
2829 the given file was scanned by exinext itself; it wasn't being passed on
2830 when Exim was called.
2832 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2833 an end-of-file indication when reading a command response.
2835 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2836 compiled. In many other places in Exim, IPv6 addresses are always
2837 recognized, so I have changed this. It also means that IPv4 domain
2838 literals of the form [IPV4:n.n.n.n] are now always recognized.
2840 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2841 used if the router is not running as root, for example, when verifying at
2842 ACL time, or when using -bh. The debugging output from this situation was
2843 non-existent - all you got was a failure to exec. I have made two
2846 (a) Failures to set uid/gid, the current directory, or a process leader
2847 in a subprocess such as that created by queryprogram now generate
2848 suitable debugging output when -d is set.
2850 (b) The queryprogram router detects when it is not running as root,
2851 outputs suitable debugging information if -d is set, and then runs
2852 the subprocess without attempting to change uid/gid.
2854 PH/15 Minor change to Makefile for building test_host (undocumented testing
2857 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2858 additional section of a DNS packet that returns MX or SRV records.
2859 Instead, it always explicitly searches for A/AAAA records. This avoids
2860 major problems that occur when a DNS server includes only records of one
2861 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2862 fixed another bug: if SRV records were looked up and the corresponding
2863 address records were *not* found in the additional section, the port
2864 values from the SRV records were lost.
2866 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2867 using the correct key (the original address) when searching the retry
2868 rules in order to find which one to use for generating the retry hint.
2870 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2871 adding the default one. Similarly, if it contains a Reply-To: header, the
2872 errors_reply_to option, if set, is not used.
2874 PH/19 When calculating a retry time, Exim used to measure the "time since
2875 failure" by looking at the "first failed" field in the retry record. Now
2876 it does not use this if it is later than than the arrival time of the
2877 message. Instead it uses the arrival time. This makes for better
2878 behaviour in cases where some deliveries succeed, thus re-setting the
2879 "first failed" field. An example is a quota failure for a huge message
2880 when small messages continue to be delivered. Without this change, the
2881 "time since failure" will always be short, possible causing more frequent
2882 delivery attempts for the huge message than are intended.
2883 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2885 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2886 $1, $2, $3) because the numerical variables can be reset during some
2887 expansion items (e.g. "match"), thereby losing the authentication data.
2889 PH/21 Make -bV show the size of off_t variables so that the test suite can
2890 decide whether to run tests for quotas > 2G.
2892 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2893 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2894 filecount value is greater than 2G or if a quota value is greater than 2G
2895 on a system where the size of off_t is not greater than 4, a panic error
2898 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2899 never match. The debug and -bh output now contains an explicit error
2900 message indicating a malformed IPv4 address or mask.
2902 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2903 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2904 PH/23 above applies.
2906 PH/25 Do not write to syslog when running in the test harness. The only
2907 occasion when this arises is a failure to open the main or panic logs
2908 (for which there is an explicit test).
2910 PH/26 Added the /no_tell option to "control=freeze".
2912 PH/27 If a host name lookup failed very early in a connection, for example, if
2913 the IP address matched host_lookup and the reverse lookup yielded a name
2914 that did not have a forward lookup, an error message of the form "no IP
2915 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2916 could be logged. Now it outputs the IP address instead of "NULL".
2918 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2919 allows the sender and the authenticated sender to be set when
2920 submitting a message from within Exim. Since child_open_exim() is
2921 documented for local_scan(), the new function should be too.
2923 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2924 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2925 results in an empty string is now treated as unset.
2927 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2929 PH/31 Added sender_verify_fail logging option.
2931 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2932 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2933 tidied the source and removed it altogether.
2935 PH/33 When a queue run was abandoned because the load average was too high, a
2936 log line was always written; now it is written only if the queue_run log
2937 selector is set. In addition, the log line for abandonment now contains
2938 information about the queue run such as the pid. This is always present
2939 in "start" and "stop" lines but was omitted from the "abandon" line.
2941 PH/34 Omit spaces between a header name and the colon in the error message that
2942 is given when verify = headers_syntax fails (if there are lots of them,
2943 the message gets confusing).
2945 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2946 names, as there are now some PTR records that contain slashes. This check
2947 is only to protect against broken name servers that fall over on strange
2948 characters, so the fact that it applies to all lookups doesn't matter.
2950 PH/36 Now that the new test suite is complete, we can remove some of the
2951 special code in Exim that was needed for the old test suite. For example,
2952 sorting DNS records because real resolvers return them in an arbitrary
2953 order. The new test suite's fake resolver always returns records in the
2956 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2957 bounces) except when queue_only is set, to avoid logging races between
2958 the different processes.
2960 PH/38 Panic-die if .include specifies a non-absolute path.
2962 PH/39 A tweak to the "H" retry rule from its user.
2964 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2965 a label. They prevented compilation on older perls.
2967 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2968 a warning to be raised on newish perls.
2970 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2971 on queue. Changes to match documented behaviour of showing count of
2972 messages matching specified criteria.
2974 PH/40 Changed the default ident timeout from 30s to 5s.
2976 PH/41 Added support for the use of login_cap features, on those BSD systems
2977 that have them, for controlling the resources used by pipe deliveries.
2979 PH/42 The content-scanning code uses fopen() to create files in which to put
2980 message data. Previously it was not paying any attention to the mode of
2981 the files. Exim runs with umask(0) because the rest of the code creates
2982 files with open(), and sets the required mode explicitly. Thus, these
2983 files were ending up world-writeable. This was not a big issue, because,
2984 being within the spool directory, they were not world-accessible. I have
2985 created a function called modefopen, which takes an additional mode
2986 argument. It sets umask(777), creates the file, chmods it to the required
2987 mode, then resets the umask. All the relevant calls to fopen() in the
2988 content scanning code have been changed to use this function.
2990 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2991 to 24 hours. This avoids potential overflow problems when processing G
2992 and H retry rules. I suspect nobody ever tinkers with this value.
2994 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2996 PH/45 When the plaintext authenticator is running as a client, the server's
2997 challenges are checked to ensure they are valid base64 strings. By
2998 default, the authentication attempt is cancelled if an invalid string is
2999 received. Setting client_ignore_invalid_base64 true ignores these errors.
3000 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3001 they are received. Thus, the responses can be made to depend on the
3002 challenges. If an invalid string is ignored, an empty string is placed in
3005 PH/46 Messages that are created by the autoreply transport now contains a
3006 References: header, in accordance with RFCs 2822 and 3834.
3008 PH/47 Added authenticated_sender_force to the smtp transport.
3010 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3012 PH/49 Installed latest patch from the Sieve maintainer.
3014 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3015 was also set, the appendfile transport was unnecessarily scanning a
3016 directory of message files (e.g. for maildir delivery) to find the count
3017 of files (along with the size), even though it did not need this
3018 information. It now does the scan only if it needs to find either the
3019 size of the count of files.
3021 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3023 PH/52 Two bugs concerned with error handling when the smtp transport is
3026 (i) Exim was not creating retry information for temporary errors given
3027 for individual recipients after the DATA command when the smtp transport
3028 was used in LMTP mode. This meant that they could be retried too
3029 frequently, and not timed out correctly.
3031 (ii) Exim was setting the flag that allows error details to be returned
3032 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3033 recipients that were returned after the DATA command.
3035 PH/53 This is related to PH/52, but is more general: for any failing address,
3036 when detailed error information was permitted to be returned to the
3037 sender, but the error was temporary, then after the final timeout, only
3038 "retry timeout exceeded" was returned. Now it returns the full error as
3039 well as "retry timeout exceeded".
3041 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3042 do this, and (what is worse) MTAs that accept it.
3044 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3045 will now be deprecated.
3047 PH/56 New os.c-cygwin from the Cygwin maintainer.
3049 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3050 formats (previously only available in exim formats via -bpr, -bpru,
3051 and -bpra. Now also available in native and exiqgrep formats)
3053 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3054 with very large, slow to parse queues
3056 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3058 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3060 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3061 responses to authentication challenges, though it was showing the
3062 challenges; (ii) I've removed the CR characters from the debug output for
3065 PH/58 Allow for the insertion of a newline as well as a space when a string
3066 is turned into more than one encoded-word during RFC 2047 encoding. The
3067 Sieve code now uses this.
3069 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3070 data_4xx, lost_connection, tls_required.
3072 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3073 message was being sent as an SMTP response.
3075 PH/61 Add -l and -k options to exicyclog.
3077 PH/62 When verifying, if an address was redirected to one new address, so that
3078 verification continued, and the new address failed or deferred after
3079 having set something in $address_data, the value of $address_data was not
3080 passed back to the ACL. This was different to the case when no
3081 redirection occurred. The value is now passed back in both cases.
3083 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3084 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3085 use login_cap.h, so on its own it isn't the distinguishing feature. The
3086 new name refers directly to the setclassresources() function.
3088 PH/65 Added configuration files for NetBSD3.
3090 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3092 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3093 is preferred over IPv4.
3095 PH/68 The bounce_return_message and bounce_return_body options were not being
3096 honoured for bounces generated during the reception of non-SMTP messages.
3097 In particular, this applied to messages rejected by the ACL. This bug has
3098 been fixed. However, if bounce_return_message is true and bounce_return_
3099 body is false, the headers that are returned for a non-SMTP message
3100 include only those that have been read before the error was detected.
3101 (In the case of an ACL rejection, they have all been read.)
3103 PH/69 The HTML version of the specification is now built in a directory called
3104 spec_html instead of spec.html, because the latter looks like a path with
3105 a MIME-type, and this confuses some software.
3107 PH/70 Catch two compiler warnings in sieve.c.
3109 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3110 function verify_get_ident() calls ip_connect() to connect a socket, but
3111 if the "connect()" function timed out, ip_connect() used to close the
3112 socket. However, verify_get_ident() also closes the socket later, and in
3113 between Exim writes to the log, which may get opened at this point. When
3114 the socket was closed in ip_connect(), the log could get the same file
3115 descriptor number as the socket. This naturally causes chaos. The fix is
3116 not to close the socket in ip_connect(); the socket should be closed by
3117 the function that creates it. There was only one place in the code where
3118 this was missing, in the iplookup router, which I don't think anybody now
3119 uses, but I've fixed it anyway.
3121 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3122 well as to direct DNS lookups. Otherwise the handling of names in host
3123 lists is inconsistent and therefore confusing.
3129 PH/01 Two changes to the default runtime configuration:
3131 (1) Move the checks for relay_from_hosts and authenticated clients from
3132 after to before the (commented out) DNS black list checks.
3134 (2) Add control=submission to the relay_from_hosts and authenticated
3135 clients checks, on the grounds that messages accepted by these
3136 statements are most likely to be submissions.
3138 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3140 (1) Generate an error if the third argument for the ${prvs expansion is
3143 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3146 (3) Reset the variables that are obtained from the first argument of
3147 ${prvscheck and used in the second argument before leaving the code,
3148 because their memory is reclaimed, so using them afterwards may do
3151 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3152 one (it's much easier than Tom thought :-).
3154 (5) Because of (4), we can now allow for the use of $prvscheck_result
3155 inside the third argument.
3157 PH/03 For some reason, the default setting of PATH when running a command from
3158 a pipe transport was just "/usr/bin". I have changed it to
3161 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3162 anything to be listed in the output from -bV.
3164 PH/05 When a filter generated an autoreply, the entire To: header line was
3165 quoted in the delivery log line, like this:
3167 => >A.N.Other <ano@some.domain> <original@ddress> ...
3169 This has been changed so that it extracts the operative address. There
3170 may be more than one such address. If so, they are comma-separated, like
3173 => >ano@some.domain,ona@other.domain <original@ddress> ...
3175 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3176 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3177 not being looked up in the rDNS to get a host name, Exim was showing the
3178 IP address twice in Received: lines, even though the IP addresses were
3179 identical. For example:
3181 Received: from [1.2.3.4] (helo=[1.2.3.4])
3183 However, if the real host name was known, it was omitting the HELO data
3184 if it matched the actual IP address. This has been tidied up so that it
3185 doesn't show the same IP address twice.
3187 PH/07 When both +timestamp and +memory debugging was on, the value given by
3188 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3189 called by the debug printing, thereby overwriting the timestamp buffer.
3190 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3192 PH/08 When the original message was included in an autoreply transport, it
3193 always said "this is a copy of the message, including all the headers",
3194 even if body_only or headers_only was set. It now gives an appropriate
3197 PH/09 Applied a patch from the Sieve maintainer which:
3199 o fixes some comments
3200 o adds the (disabled) notify extension core
3201 o adds some debug output for the result of if/elsif tests
3202 o points to the current vacation draft in the documentation
3203 and documents the missing references header update
3207 o fixes a bug in processing the envelope test (when testing
3208 multiple envelope elements, the last element determined the
3211 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3212 Electronic Mail") by including:
3214 Auto-submitted: auto-generated
3216 in the messages that it generates (bounce messages and others, such as
3217 warnings). In the case of bounce messages for non-SMTP messages, there was
3218 also a typo: it was using "Auto_submitted" (underscore instead of
3219 hyphen). Since every message generated by Exim is necessarily in response
3220 to another message, thes have all been changed to:
3222 Auto-Submitted: auto-replied
3224 in accordance with these statements in the RFC:
3226 The auto-replied keyword:
3228 - SHOULD be used on messages sent in direct response to another
3229 message by an automatic process,
3231 - MUST NOT be used on manually-generated messages,
3233 - MAY be used on Delivery Status Notifications (DSNs) and Message
3234 Disposition Notifications (MDNs),
3236 - MUST NOT be used on messages generated by automatic or periodic
3237 processes, except for messages which are automatic responses to
3240 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3241 to the default Received: header definition.
3243 PH/12 Added log selector acl_warn_skipped (default on).
3245 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3246 variables because (a) they are in the wrong storage pool and (b) even if
3247 they were copied, it wouldn't work properly because of the caching.
3249 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3250 checking when decoding. Apparently there are clients that generate
3251 overlong encoded strings. Why am I not surprised?
3253 PH/15 If the first argument of "${if match_address" was not empty, but did not
3254 contain an "@" character, Exim crashed. Now it writes a panic log message
3255 and treats the condition as false.
3257 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3259 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3260 extension "envelope-auth". The code is finished and in agreement with
3261 other implementations, but there is no documentation so far and in fact,
3262 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3263 not changing the active code.
3265 Print executed "if" and "elsif" statements when debugging is used. This
3266 helps a great deal to understand what a filter does.
3268 Document more things not specified clearly in RFC3028. I had all this
3269 sorted out, when out of a sudden new issues came to my mind. Oops."
3271 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3274 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3275 canonical form (as documented). However, after a host name lookup from
3276 the IP address, check_host() was doing a simple string comparison with
3277 addresses acquired from the DNS when checking that the found name did
3278 have the original IP as one of its addresses. Since any found IPv6
3279 addresses are likely to be in abbreviated form, the comparison could
3280 fail. Luckily, there already exists a function for doing the comparison
3281 by converting both addresses to binary, so now that is used instead of
3282 the text comparison.
3284 PH/20 There was another similar case to PH/19, when a complete host name was
3285 given in a host list; looking up its IP address could give an abbreviated
3286 form, whereas the current host's name might or might not be abbreviated.
3287 The same fix has been applied.
3293 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3294 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3297 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3298 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3300 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3302 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3303 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3304 Also, for "Auto-Submitted", treat anything other than "no" as
3305 non-personal, in accordance with RFC 3834. (Previously it treated
3306 anything starting "auto-" as non-personal.)
3308 TF/01 The control=submission/name=... option had a problem with syntax
3309 errors if the name included a slash character. The /name= option
3310 now slurps the rest of the string, so it can include any characters
3311 but it must come last in the list of options (after /sender_retain
3314 PH/05 Some modifications to the interface to the fake nameserver for the new
3322 TK/01 Added the "success_on_redirect" address verification option. See
3323 NewStuff for rationale and an example.
3325 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3327 PH/02 Patch to exigrep to allow it to work on syslog lines.
3329 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3330 fread() to skip over the body file's header line, because in Cygwin the
3331 header line is locked and is inaccessible.
3333 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3334 co-exist for some time) to make it clear that it is the Exim ID that is
3335 referenced, not the Message-ID: header line.
3337 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3338 string_format() function, because snprintf() does not exist on all
3341 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3342 ${stat: expansion item.
3344 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3345 protocol synchronization error", to keep the pedants happy.
3347 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3348 well as for IRIX systems, when gcc is being used. See the host.c source
3351 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3353 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3356 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3357 transport and to the smtp transport in LMTP mode.
3359 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3361 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3362 run a filter in a subprocess. This could lead to confusion in subsequent
3363 lookups in the parent process. There should also be a search_tidyup() at
3364 the end of the subprocess.
3366 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3367 only if the host matched helo_try_verify_hosts, which caused the
3368 verification to occur when the EHLO/HELO command was issued. The ACL just
3369 tested the remembered result. Now, if a previous verification attempt has
3370 not happened, "verify = helo" does it there and then.
3372 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3374 TK/03 Fix log output including CR from clamd.
3376 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3377 exist provoked a memory error which could cause a segfault.
3379 PH/15 Installed PCRE 6.2
3381 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3383 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3384 of the problem. Specifically, suggested +O2 rather than +O1 for the
3387 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3389 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3390 record, the port was not correctly logged when the outgoing_port log
3391 selector was set (it logged the transort's default port).
3393 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3394 fallback_hosts, and "hosts" in the smtp transport.
3396 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3397 host errors such as "Connection refused".
3399 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3400 authentication with radiusclient 0.4.9:
3402 - Error returned from rc_read_config was caught wrongly
3403 - Username/password not passed on to radius server due to wrong length.
3405 The presumption is that some radiusclient API changes for 4.51/PH/17
3406 were not taken care of correctly. The code is still untested by me (my
3407 Linux distribution still has 0.3.2 of radiusclient), but it was
3408 contributed by a Radius user.
3410 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3411 expanding the "port" option of the smtp transport.
3413 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3414 while reading a MIME header. Thanks to Tom Hughes for a patch.
3416 PH/24 Include config.h inside local_scan.h so that configuration settings are
3419 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3420 that in an ACL for RCPT (for example), you can examine exactly what was
3423 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3424 commands, but it was not correctly comparing the address with the actual
3425 client host address. Thus, it would show the EHLO address in Received:
3426 header lines when this was not necessary.
3428 PH/27 Added the % operator to ${eval:}.
3430 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3431 it should be ignoring failures (because with -C, for example, it has lost
3432 privilege). It wasn't ignoring creation failures other than "already
3435 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3438 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3439 because an input line was too long, either on its own, or by virtue of
3440 too many continuations, the temporary file was not being removed, and the
3441 return code was incorrect.
3443 PH/31 Missing "BOOL" in function definition in filtertest.c.
3445 PH/32 Applied Sieve patches from the maintainer.
3447 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3449 PH/33 Added "verify = not_blind".
3451 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3452 Local/Makefile (with some defaults set). These are used in built scripts
3453 such as exicyclog, but they have never been used in the exim_install
3454 script (though there are many overriding facilities there). I have
3455 arranged that the exim_install script now takes note of these two
3458 PH/35 Installed configuration files for Dragonfly.
3460 PH/36 When a locally submitted message by a trusted user did not contain a
3461 From: header, and the sender address was obtained from -f or from an SMTP
3462 MAIL command, and the trusted user did not use -F to supply a sender
3463 name, $originator_name was incorrectly used when constructing a From:
3464 header. Furthermore, $originator_name was used for submission mode
3465 messages from external hosts without From: headers in a similar way,
3466 which is clearly wrong.
3468 PH/37 Added control=suppress_local_fixups.
3470 PH/38 When log_selector = +received_sender was set, and the addition of the
3471 sender made the log line's construction buffer exactly full, or one byte
3472 less than full, an overflow happened when the terminating "\n" was
3475 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3476 when the result of a list match is failure because a DNS lookup failed.
3478 PH/40 RM_COMMAND is now used in the building process.
3480 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3481 the "build-* directories that it finds.
3483 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3484 address in a domain literal was a prefix of an interface address.
3486 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3487 when verifying a sender address, unless rewrite_headers is false.
3489 PH/44 Wrote a long comment about why errors_to addresses are verified as
3490 recipients, not senders.
3492 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3493 the ratelimit ACL was added.
3495 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3497 PH/46 Added extra information about PostgreSQL errors to the error string.
3499 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3500 avoiding the need to install special zones in a real server. This is
3501 backwards compatible; if it can't find the fake resolver, it drops back.
3502 Thus, both old and new test suites can be run.
3504 TF/02 Added util/ratelimit.pl
3506 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3507 clock is set back in time.
3509 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3510 Candler <B.Candler@pobox.com>.
3512 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3513 OK for addresses that are the result of redirections.
3515 PH/48 A number of further additions for the benefit of the new test suite,
3516 including a fake gethostbyname() that interfaces to the fake DNS resolver
3519 TF/06 The fix for widen_domains has also been applied to qualify_single and
3520 search_parents which are the other dnslookup options that can cause
3523 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3526 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3528 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3529 service" if it encountered an MX record with an empty target hostname.
3530 The message is now "an MX or SRV record indicated no SMTP service".
3532 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3533 if the DNS of the sending site is misconfigured. This is quite a
3534 common situation. This change restores the behaviour of treating a
3535 helo verification defer as a failure.
3537 PH/51 If self=fail was set on a router, the bounce message did not include the
3538 actual error message.
3544 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3546 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3547 command itself ended in error, the underlying message about the transport
3548 filter timeout was being overwritten with the pipe command error. Now the
3549 underlying error message should be appended to the second error message.
3551 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3553 PH/02 Reduce the amount of output that "make" produces by default. Full output
3554 can still be requested.
3556 PH/03 The warning log line about a condition test deferring for a "warn" verb
3557 was being output only once per connection, rather than after each
3558 occurrence (because it was using the same function as for successful
3559 "warn" verbs). This seems wrong, so I have changed it.
3561 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3562 it should not have, which might have caused a crash in the right
3563 circumstances, but probably never did.
3565 PH/04 Installed a modified version of Tony Finch's patch to make submission
3566 mode fix the return path as well as the Sender: header line, and to
3567 add a /name= option so that you can make the user's friendly name appear
3570 TF/03 Added the control = fakedefer ACL modifier.
3572 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3573 Mark Lowes for thorough testing.
3575 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3577 TK/03 Merged latest SRS patch from Miles Wilton.
3579 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3580 with the definition in sysexits.h (which is #included earlier).
3581 Fortunately, Exim does not actually use EX_OK. The code used to try to
3582 preserve the sysexits.h value, by assuming that macro definitions were
3583 scanned for macro replacements. I have been disabused of this notion,
3584 so now the code just undefines EX_OK before #including unistd.h.
3586 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3587 in the smtp transport. When a block could not be written in a single
3588 write() function, the timeout was being re-applied to each part-write.
3589 This seems wrong - if the receiver was accepting one byte at a time it
3590 would take for ever. The timeout is now adjusted when this happens. It
3591 doesn't have to be particularly precise.
3593 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3594 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3596 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3597 without a domain if the check to <postmaster@domain> fails.
3599 SC/01 Eximstats: added -xls and the ability to specify output files
3600 (patch written by Frank Heydlauf).
3602 SC/02 Eximstats: use FileHandles for outputting results.
3604 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3606 SC/04 Eximstats: fixed display of large numbers with -nvr option
3608 SC/05 Eximstats: fixed merging of reports with empty tables.
3610 SC/06 Eximstats: added the -include_original_destination flag
3612 SC/07 Eximstats: removed tabs and trailing whitespace.
3614 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3616 TK/06 MBOX spool code: Add real "From " MBOX separator line
3617 so the .eml file is really in mbox format (even though
3618 most programs do not really care). Patch from Alex Miller.
3620 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3621 The latter is generated from $received_to and is only set if the
3622 message has one envelope recipient. SA can use these headers,
3623 obviously out-of-the-box. Patch from Alex Miller.
3625 PH/08 The ${def test on a variable was returning false if the variable's
3626 value was "0", contrary to what the specification has always said!
3627 The result should be true unless the variable is empty.
3629 PH/09 The syntax error of a character other than { following "${if
3630 def:variable_name" (after optional whitespace) was not being diagnosed.
3631 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3632 accidental colon was present, for example, could give incorrect results.
3634 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3635 result is used (not including appendfile, where other changes are about
3638 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3639 This involved changing a lot of size variables from int to off_t. It
3640 should work with maildirs and everything.
3642 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3643 spamd dying while we are connected to it.
3645 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3648 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3649 function for BDB 4.3.
3651 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3653 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3654 the sources for PCRE have been re-arranged and more files are now
3657 PH/15 The code I had for printing potentially long long variables in PH/11
3658 above was not the best (it lost precision). The length of off_t variables
3659 is now inspected at build time, and an appropriate printing format (%ld
3660 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3661 to be "long long int" or "long int". This is needed for the internal
3662 formatting function string_vformat().
3664 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3665 the configuration file to be ":syslog", then the script "guesses" where
3666 the logs files are, rather than using the compiled in default. In our
3667 case the guess is not the same as the compiled default, so the script
3668 suddenly stopped working when I started to use syslog. The patch checks
3669 to see if log_file_path is "". If so, it attempts to read it from exim
3670 with no configuration file to get the compiled in version, before it
3671 falls back to the previous guessing code."
3673 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3674 implementing BATV in an Exim configuration. See NewStuff for the gory
3677 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3678 Makefile that are specific to HP-UX.
3680 PH/18 If the "use_postmaster" option was set for a recipient callout together
3681 with the "random" option, the postmaster address was used as the MAIL
3682 FROM address for the random test, but not for the subsequent recipient
3683 test. It is now used for both.
3685 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3686 patch removes a few documentation additions to RFC 3028, because the
3687 latest draft now contains them. It adds the new en;ascii-case comparator
3688 and a new error check for 8bit text in MIME parts. Comparator and
3689 require names are now matched exactly. I enabled the subaddress
3690 extension, but it is not well tested yet (read: it works for me)."
3692 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3693 rework some of the code of TK/09 above to avoid the hardwired use of
3694 "%lld" and "long long". Replaced the call to snprintf() with a call to
3697 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3698 records point to non-existent hosts", "retry timeout exceeded", and
3699 "retry time not reached for any host after a long failure period".
3701 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3702 experimental DomainKeys support:
3704 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3705 (2) On an error such as an illegally used "control", the wrong name for
3706 the control was given.
3708 These problems did NOT occur unless DomainKeys support was compiled.
3710 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3712 PH/24 Added ${if match_ip condition.
3714 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3715 databases so that it will be absolutely obvious if a crash occurs in the
3716 DB library. This is a regular occurrence (often caused by mis-matched
3719 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3720 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3721 picked up on a user's system that detects such things. There doesn't seem
3722 to be a gcc warning option for this - only an attribute that has to be
3723 put on the function's prototype. It seems that in Fedora Core 4 they have
3724 set this on a number of new functions. No doubt there will be more in due
3727 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3728 specify a transport. However, if an address that was verified by such a
3729 router was the subject of a callout, Exim crashed because it tried to
3730 read the rcpt_include_affixes from the non-existent transport. Now it
3731 just assumes that the setting of that option is false. This bug was
3732 introduced by 4.51/PH/31.
3734 PH/28 Changed -d+all to exclude +memory, because that information is very
3735 rarely of interest, but it makes the output a lot bigger. People tend to
3736 do -d+all out of habit.
3738 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3739 code in os-type was giving problems when libc.so lives in lib64, like on
3742 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3743 aren't the modern standard, and it seems that some systems' include files
3744 don't always have them. Exim was already checking for some of the newer
3745 ones like T_AAAA, and defining it itself. I've added checks for all the
3746 record types that Exim uses.
3748 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3749 not automatically generating a new one, as it is supposed to. This
3750 prevented TLS from working. If the file did exist, but contained invalid
3751 data, a new version was generated, as expected. It was only the case of a
3752 non-existent file that was broken.
3754 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3755 with a change in libdomainkeys > 0.64.
3757 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3758 from DNS. If the selector record carries the flag, it now has
3759 precedence over the domain-wide flag.
3761 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3763 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3764 the use of a port name that isn't defined in /etc/services) occurred, the
3765 message was deferred as in a normal delivery, and thus remained on the
3766 spool, instead of being failed because of the mua_wrapper setting. This
3767 is now fixed, and I tidied up some of the mua_wrapper messages at the
3770 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3771 the messages in a hash of arrays rather than using individual hashes.
3772 This is a bit cleaner and results in dramatic memory savings, albeit
3773 at a slight CPU cost.
3775 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3776 as requested by Marc Sherman.
3778 SC/10 Eximstats: added histograms for user specified patterns as requested
3781 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3783 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3784 fopen() in the content-scanning modules that did not already have it.
3790 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3791 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3793 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3795 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3797 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3798 to test Sieve filters that use "vacation".
3800 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3801 that changes the way the GnuTLS parameters are stored in the cache file.
3802 The new format can be generated externally. For backward compatibility,
3803 if the data in the cache doesn't make sense, Exim assumes it has read an
3804 old-format file, and it generates new data and writes a new file. This
3805 means that you can't go back to an older release without removing the
3808 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3809 work if there are any delivery delays because "one_time" forces the
3810 parent to be marked "delivered", so its unseen clone is never tried
3811 again. For this reason, Exim now forbids the simultaneous setting of
3814 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3815 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3816 entry if you want to know the details. The fix, however, neglected to
3817 consider the case when local delivery batching is involved. The test for
3818 "previously delivered" was not happening when checking to see if an
3819 address could be batched with a previous (undelivered) one; under
3820 certain circumstances this could lead to multiple deliveries to the same
3823 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3824 in its include files, and this causes problems building Exim.
3826 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3827 header_syntax) but Exim was just ignoring anything given after a slash.
3828 In particular, this caused confusion with an attempt to use "verify =
3829 reverse_host_lookup/defer_ok". An error is now given when options are
3830 supplied for verify items that do not have them. (Maybe reverse_host_
3831 lookup should have a defer_ok option, but that's a different point.)
3833 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3834 defined by RFC 821) to 2048, because there were problems with some AUTH
3835 commands, and RFC 1869 says the size should be increased for extended
3836 SMTP commands that take arguments.
3838 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3841 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3842 "unknown" error; now it says that the functionality isn't in the binary.
3844 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3845 an address' error message when a string expansion fails (syntax or
3846 whatever). Otherwise the password may appear in the log. Following change
3847 PH/42 below, there is no longer a chance of it appearing in a bounce
3850 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3852 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3853 of its addresses was ever tried. (Bugzilla bug #2.)
3855 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3856 the result incorrectly in the debug output. (It correctly added a newline
3857 to what was transported.)
3859 TF/01 Added $received_time.
3861 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3862 commented out examples of how to interface to a virus scanner and to
3863 SpamAssassin. Also added commented examples of av_scanner and
3864 spamd_address settings.
3866 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3867 and controls are allowed in which ACLs. There were a couple of minor
3868 errors. Some of the entries in the conditions table (which is a table of
3869 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3870 negation of where the condition IS allowed.
3872 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3874 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3875 header file does not have a version number, so I've had to invent a new
3876 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3877 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3878 radiusclient), but it was contributed by a Radius user.
3880 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3881 files or directories, for OpenSSL.
3883 PH/19 When an Exim process that is running as root has to create an Exim log
3884 file, it does so in a subprocess that runs as exim:exim so as to get the
3885 ownership right at creation (otherwise, other Exim processes might see
3886 the file with the wrong ownership). There was no test for failure of this
3887 fork() call, which would lead to the process getting stuck as it waited
3888 for a non-existent subprocess. Forks do occasionally fail when resources
3889 run out. I reviewed all the other calls to fork(); they all seem to check
3892 PH/20 When checking for unexpected SMTP input at connect time (before writing
3893 the banner), Exim was not dealing correctly with a non-positive return
3894 from the read() function. If the client had disconnected by this time,
3895 the result was a log entry for a synchronization error with an empty
3896 string after "input=" when read() returned zero. If read() returned -1
3897 (an event I could not check), uninitialized data bytes were printed.
3898 There were reports of junk text (parts of files, etc) appearing after
3901 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3903 PH/22 Added support for macro redefinition, and (re)definition in between
3904 driver and ACL definitions.
3906 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3907 forgetting to use the resulting value; it was using the unexpanded value.
3909 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3910 hadn't been configured. The fix is from Juergen Kreileder, who
3911 understands it better than I do:
3913 "Here's what I see happening with three configured cyrus_sasl
3914 authenticators configured (plain, login, cram-md5):
3916 On startup auth_cyrus_sasl_init() gets called for each of these.
3917 This means three calls to sasl_listmech() without a specified mech_list.
3918 => SASL tests which mechs of all available mechs actually work
3919 => three warnings about OTP not working
3920 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3922 With the patch, sasl_listmech() also gets called three times. But now
3923 SASL's mech_list option is set to the server_mech specified in the the
3924 authenticator. Or in other words, the answer from sasl_listmech()
3925 gets limited to just the mech you're testing for (which is different
3927 => the return list contains just 'plain' or 'login', 'cram-md5' or
3928 nothing depending on the value of ob->server_mech.
3930 I've just tested the patch: Authentication still works fine,
3931 unavailable mechs specified in the exim configuration are still
3932 caught, and the auth.log warnings about OTP are gone."
3934 PH/25 When debugging is enabled, the contents of the command line are added
3935 to the debugging output, even when log_selector=+arguments is not
3938 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3939 answer is "GNU", and only if the return is "GNU/something" is the answer
3942 PH/27 $acl_verify_message is now set immediately after the failure of a
3943 verification in an ACL, and so is available in subsequent modifiers. In
3944 particular, the message can be preserved by coding like this:
3946 warn !verify = sender
3947 set acl_m0 = $acl_verify_message
3949 Previously, $acl_verify_message was set only while expanding "message"
3950 and "log_message" when a very denied access.
3952 PH/28 Modified OS/os.c-Linux with
3954 -#ifndef OS_LOAD_AVERAGE
3955 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3957 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3958 nomenclature these days.)
3960 PH/29 Installed patch from the Sieve maintainer that adds the options
3961 sieve_useraddress and sieve_subaddress to the redirect router.
3963 PH/30 In these circumstances:
3964 . Two addresses routed to the same list of hosts;
3965 . First host does not offer TLS;
3966 . First host accepts first address;
3967 . First host gives temporary error to second address;
3968 . Second host offers TLS and a TLS session is established;
3969 . Second host accepts second address.
3970 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3971 and peerdn, if requested) that were in fact used only for the second
3974 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3975 attention to any local part prefix or suffix that was matched by the
3976 router that accepted the address. It now behaves in the same way as it
3977 does for delivery: the affixes are removed from the local part unless
3978 rcpt_include_affixes is set on the transport.
3980 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3981 timeout during the DATA phase of an incoming message.
3983 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3984 applied a patch sanctioned by the Sieve maintainer.
3986 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3987 the uid or gid is negative. A case of a negative gid caused this to be
3988 noticed. The fix allows for either to be negative.
3990 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3991 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3994 PH/36 The change PH/12 above was broken. Fixed it.
3996 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3997 the grounds that routing the same address twice would always produce the
3998 same answer. This might have been true once, but it is certainly no
3999 longer true now. Routing a child address may depend on the previous
4000 routing that produced that child. Some complicated redirection strategies
4001 went wrong when messages had multiple recipients, and made Exim's
4002 behaviour dependent on the order in which the addresses were given.
4004 I have moved the duplicate checking until after the routing is complete.
4005 Exim scans the addresses that are assigned to local and remote
4006 transports, and removes any duplicates. This means that more work will be
4007 done, as duplicates will always all be routed, but duplicates are
4008 presumably rare, so I don't expect this is of any significance.
4010 For deliveries to pipes, files, and autoreplies, the duplicate checking
4011 still happens during the routing process, since they are not going to be
4014 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4015 It corrects a timeout issue with spamd. This is Ian's comment: "The
4016 background is that sometimes spamd either never reads data from a
4017 connection it has accepted, or it never writes response data. The exiscan
4018 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4019 blindly assumes that writes won't block so it may never time out."
4021 PH/39 Allow G after quota size as well as K and M.
4023 PH/40 The value set for $authenticated_id in an authenticator may not contain
4024 binary zeroes or newlines because the value is written to log lines and
4025 to spool files. There was no check on this. Now the value is run through
4026 the string_printing() function so that such characters are converted to
4027 printable escape sequences.
4029 PH/41 $message_linecount is a new variable that contains the total number of
4030 lines in the message. Compare $body_linecount, which is the count for the
4033 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4034 bounce and delay warning messages, except in certain special cases, which
4037 (a) An SMTP error message from a remote host;
4038 (b) A message specified in a :fail: redirection;
4039 (c) A message specified in a "fail" command in a system filter;
4040 (d) A message specified in a FAIL return from the queryprogram router;
4041 (e) A message specified by the cannot_route_message router option.
4043 In these cases only, Exim does include the error details in bounce and
4044 warning messages. There are also a few cases where bland messages such
4045 as "unrouteable address" or "local delivery error" are given.
4047 PH/43 $value is now also set for the "else" part of a ${run expansion.
4049 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4050 being worked on, but at least Exim now implements the latest version to
4053 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4054 process to complete was treated as a delivery failure, a timeout while
4055 writing the message to the pipe was logged, but erroneously treated as a
4056 successful delivery. Such timeouts include transport filter timeouts. For
4057 consistency with the overall process timeout, these timeouts are now
4058 treated as errors, giving rise to delivery failures by default. However,
4059 there is now a new Boolean option for the pipe transport called
4060 timeout_defer, which, if set TRUE, converts the failures into defers for
4061 both kinds of timeout. A transport filter timeout is now identified in
4064 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4065 systems where "make" and "gmake" are different, calling "gmake" at top
4066 level broke things. I've arranged for the value of $(MAKE) to be passed
4067 from the Makefile to this script so that it can call the same version of
4071 A note about Exim versions 4.44 and 4.50
4072 ----------------------------------------
4074 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4075 changes of various kinds. As a consequence, a big documentation update was
4076 needed. This delayed the release for rather longer than seemed good, especially
4077 in the light of a couple of (minor) security issues. Therefore, the changes
4078 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4079 release. So 4.44 and 4.50 are in effect two different branches that both start
4082 I have left the 4.50 change log unchanged; it contains all the changes since
4083 4.43. The change log for 4.44 is below; many of its items are identical to
4084 those for 4.50. This seems to be the most sensible way to preserve the
4085 historical information.
4091 1. Minor wording change to the doc/README.SIEVE file.
4093 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4094 computation of the current number of files was incorrect.
4096 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4097 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4100 4. Give more explanation in the error message when the command for a transport
4101 filter fails to execute.
4103 5. There are several places where Exim runs a non-Exim command in a
4104 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4105 was being done only for the command run by the queryprogram router. It is
4106 now done for all such subprocesses. The other cases are: ${run, transport
4107 filters, and the commands run by the lmtp and pipe transports.
4109 6. Added CONFIGURE_GROUP build-time option.
4111 7. Some older OS have a limit of 256 on the maximum number of file
4112 descriptors. Exim was using setrlimit() to set 1000 as a large value
4113 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4114 systems. I've change it so that if it can't get 1000, it tries for 256.
4116 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4117 was an oversight, and furthermore, ever since the addition of extra
4118 controls (e.g. 4.43/32), the checks on when to allow different forms of
4119 "control" were broken. There should now be diagnostics for all cases when a
4120 control that does not make sense is encountered.
4122 9. Added the /retain_sender option to "control=submission".
4124 10. $recipients is now available in the predata ACL (oversight).
4126 11. Tidy the search cache before the fork to do a delivery from a message
4127 received from the command line. Otherwise the child will trigger a lookup
4128 failure and thereby defer the delivery if it tries to use (for example) a
4129 cached ldap connection that the parent has called unbind on.
4131 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4132 of $address_data from the recipient verification was clobbered by the
4133 sender verification.
4135 13. The value of address_data from a sender verification is now available in
4136 $sender_address_data in subsequent conditions in the ACL statement.
4138 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4140 15. Added a new option "connect=<time>" to callout options, to set a different
4143 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4144 was its contents. (It was OK if the option was not defined at all.)
4146 17. A "Completed" log line is now written for messages that are removed from
4147 the spool by the -Mrm option.
4149 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4150 information about exactly what failed.
4152 19. Added -dd to debug only the daemon process.
4154 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4155 handles timeouts, both on the server side and network timeouts. Renamed the
4156 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4158 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4159 It is now set to "smtps".
4161 22. $host_address is now set to the target address during the checking of
4162 ignore_target_hosts.
4164 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4165 being passed; this would have caused $sender_host_name to have been used if
4166 matching the list had actually called for a host name (not very likely,
4167 since this list is usually IP addresses). A host name is now passed as
4170 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4171 code that specifies a non-restarting handler (typically sigaction() in
4172 modern systems) in an attempt to fix a rare and obscure crash bug.
4174 25. Narrowed the window for a race in the daemon that could cause it to ignore
4175 SIGCHLD signals. This is not a major problem, because they are used only to
4176 wake it up if nothing else does.
4178 26. A malformed maildirsize file could cause Exim to calculate negative values
4179 for the mailbox size or file count. Odd effects could occur as a result.
4180 The maildirsize information is now recalculated if the size or filecount
4183 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4184 support for a long time. Removed HAVE_SYS_VFS_H.
4186 28. Installed the latest version of exipick from John Jetmore.
4188 29. In an address list, if the pattern was not a regular expression, an empty
4189 subject address (from a bounce message) matched only if the pattern was an
4190 empty string. Non-empty patterns were not even tested. This was the wrong
4191 because it is perfectly reasonable to use an empty address as part of a
4192 database query. An empty address is now tested by patterns that are
4193 lookups. However, all the other forms of pattern expect the subject to
4194 contain a local part and a domain, and therefore, for them, an empty
4195 address still always fails if the pattern is not itself empty.
4197 30. Exim went into a mad DNS loop when attempting to do a callout where the
4198 host was specified on an smtp transport, and looking it up yielded more
4199 than one IP address.
4201 31. Re-factored the code for checking spool and log partition space into a
4202 function that finds that data and another that does the check. The former
4203 is then used to implement four new variables: $spool_space, $log_space,
4204 $spool_inodes, and $log_inodes.
4206 32. The RFC2047 encoding function was originally intended for short strings
4207 such as real names; it was not keeping to the 75-character limit for
4208 encoded words that the RFC imposes. It now respects the limit, and
4209 generates multiple encoded words if necessary. To be on the safe side, I
4210 have increased the buffer size for the ${rfc2047: expansion operator from
4213 33. It is now permitted to omit both strings after an "if" condition; if the
4214 condition is true, the result is "true". As before, when the second string
4215 is omitted, a false condition yields an empty string. This makes it less
4216 cumbersome to write custom ACL and router conditions.
4218 34. Failure to deliver a bounce message always caused it to be frozen, even if
4219 there was an errors_to setting on the router. The errors_to setting is now
4222 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4223 canonical form (fully expanded) before being placed in
4224 $sender_host_address.
4226 36. The table in the code that translates DNS record types into text (T_A to
4227 "A" for instance) was missing entries for NS and CNAME. It is just possible
4228 that this could have caused confusion if both these types were looked up
4229 for the same domain, because the text type is used as part of Exim's
4230 per-process caching. But the chance of anyone hitting this buglet seems
4233 37. The dnsdb lookup has been extended in a number of ways.
4235 (1) There is a new type, "zns", which walks up the domain tree until it
4236 finds some nameserver records. It should be used with care.
4238 (2) There is a new type, "mxh", which is like "mx" except that it returns
4239 just the host names, not the priorities.
4241 (3) It is now possible to give a list of domains (or IP addresses) to be
4242 looked up. The behaviour when one of the lookups defers can be
4243 controlled by a keyword.
4245 (4) It is now possible to specify the separator character for use when
4246 multiple records are returned.
4248 38. The dnslists ACL condition has been extended: it is now possible to supply
4249 a list of IP addresses and/or domains to be looked up in a particular DNS
4252 39. Added log_selector=+queue_time_overall.
4254 40. When running the queue in the test harness, wait just a tad after forking a
4255 delivery process, to get repeatability of debugging output.
4257 41. Include certificate and key file names in error message when GnuTLS fails
4258 to set them up, because the GnuTLS error message doesn't include the name
4259 of the failing file when there is a problem reading it.
4261 42. Allow both -bf and -bF in the same test run.
4263 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4265 44. The "Exiscan patch" is now merged into the mainline Exim source.
4267 45. Sometimes the final signoff response after QUIT could fail to get
4268 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4269 < 0 before doing a fflush(). This bug looks as though it goes back to the
4270 introduction of TLS in release 3.20, but "sometimes" must have been rare
4271 because the tests only now provoked it.
4273 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4274 (this can affect the format of dates).
4276 47. exim_tidydb, when checking for the continued existence of a message for
4277 which it has found a message-specific retry record, was not finding
4278 messages that were in split spool directories. Consequently, it was
4279 deleting retry records that should have stayed in existence.
4281 48. Steve fixed some bugs in eximstats.
4283 49. The SPA authentication driver was not abandoning authentication and moving
4284 on to the next authenticator when an expansion was forced to fail,
4285 contradicting the general specification for all authenticators. Instead it
4286 was generating a temporary error. It now behaves as specified.
4288 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4289 (the order specifies the preference for clients). The order is now AES256,
4290 AES128, 3DES, ARCFOUR128.
4292 51. Small patch to Sieve code - explicitly set From: when generating an
4295 52. Exim crashed if a remote delivery caused a very long error message to be
4296 recorded - for instance if somebody sent an entire SpamAssassin report back
4297 as a large number of 550 error lines. This bug was coincidentally fixed by
4298 increasing the size of one of Exim's internal buffers (big_buffer) that
4299 happened as part of the Exiscan merge. However, to be on the safe side, I
4300 have made the code more robust (and fixed the comments that describe what
4303 53. Now that there can be additional text after "Completed" in log lines (if
4304 the queue_time_overall log selector is set), a one-byte patch to exigrep
4305 was needed to allow it to recognize "Completed" as not the last thing in
4308 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4309 patch that reportedly fixes this has been added. I am not expert enough to
4310 create a test for it. This is what the patch creator wrote:
4312 "I found a little strange behaviour of ldap code when working with
4313 Windows 2003 AD Domain, where users was placed in more than one
4314 Organization Units. When I tried to give exim partial DN, the exit code
4315 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4316 But simultaneously result of request was absolutely normal ldap result,
4317 so I produce this patch..."
4319 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4320 REFERENCE, so I have modified the code to exclude the patch when that macro
4323 55. Some experimental protocols are using DNS PTR records for new purposes. The
4324 keys for these records are domain names, not reversed IP addresses. The
4325 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4326 leaves it alone. Component reversal etc. now happens only for IP addresses.
4329 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4331 57. Double the size of the debug message buffer (to 2048) so that more of very
4332 long debug lines gets shown.
4334 58. The exicyclog utility now does better if the number of log files to keep
4335 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4337 59. Two changes related to the smtp_active_hostname option:
4339 (1) $smtp_active_hostname is now available as a variable.
4340 (2) The default for smtp_banner uses $smtp_active_hostname instead
4341 of $primary_hostname.
4343 60. The host_aton() function is supposed to be passed a string that is known
4344 to be a valid IP address. However, in the case of IPv6 addresses, it was
4345 not checking this. This is a hostage to fortune. Exim now panics and dies
4346 if the condition is not met. A case was found where this could be provoked
4347 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4348 components; fortuitously, this particular loophole had already been fixed
4349 by change 4.50/55 above.
4351 If there are any other similar loopholes, the new check in host_aton()
4352 itself should stop them being exploited. The report I received stated that
4353 data on the command line could provoke the exploit when Exim was running as
4354 exim, but did not say which command line option was involved. All I could
4355 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4356 running as the user.
4359 61. There was a buffer overflow vulnerability in the SPA authentication code
4360 (which came originally from the Samba project). I have added a test to the
4361 spa_base64_to_bits() function which I hope fixes it.
4364 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4365 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4367 63. The daemon start-up calls getloadavg() while still root for those OS that
4368 need the first call to be done as root, but it missed one case: when
4369 deliver_queue_load_max is set with deliver_drop_privilege. This is
4370 necessary for the benefit of the queue runner, because there is no re-exec
4371 when deliver_drop_privilege is set.
4373 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4374 This has been fixed.
4376 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4377 was in use, was not putting the data itself into the right store pool;
4378 consequently, it could be overwritten for a subsequent message in the same
4379 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4382 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4384 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4385 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4386 the function were treating the return as a boolean value, which happened to
4387 work because 0=false and not-0=true, but is not correct code.
4389 68. The host_aton() function was not handling scoped IPv6 addresses (those
4390 with, for example, "%eth0" on the end) correctly.
4392 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4393 negated items (that is, ~something) in unsigned ints. Some compilers
4394 apparently mutter when there is no cast.
4396 70. If an address verification called from an ACL failed, and did not produce a
4397 user-specific message (i.e. there was only a "system" message), nothing was
4398 put in $acl_verify_message. In this situation, it now puts the system
4401 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4402 session; change 4.31/43 added the unwanted input to the log line - except
4403 that it did not do this in the start of session case. It now does.
4405 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4406 This is wrong and can cause the other end to generate a synchronization
4407 error if it is another Exim or anything else that does the synchronization
4408 check. A QUIT command is no longer sent after a timeout.
4410 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4411 during host lookups.
4413 74. The defer_ok option of callout verification was not working if it was used
4414 when verifying addresses in header lines, that is, for this case:
4416 verify = header_sender/callout=defer_ok
4418 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4419 those file descriptors could be used for SMTP connections. If anything
4420 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4421 could be sent to the SMTP client, causing chaos. The daemon now opens
4422 stdin, stdout, and stderr to /dev/null when it puts itself into the
4425 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4426 log by default. The user can override this with suitable Perl magic.
4428 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4429 the log message when discard triggers, was not working for the DATA ACL or
4430 for the non-SMTP ACL.
4432 78. Error message wording change in sieve.c.
4434 79. If smtp_accept_max_per_host was set, the number of connections could be
4435 restricted to fewer than expected, because the daemon was trying to set up
4436 a new connection before checking whether the processes handling previous
4437 connections had finished. The check for completed processes is now done
4438 earlier. On busy systems, this bug wouldn't be noticed because something
4439 else would have woken the daemon, and it would have reaped the completed
4442 80. If a message was submitted locally by a user whose login name contained one
4443 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4444 It caused a spool format error. I have fixed the spool reading code. A
4445 related problem was that the "from" clause in the Received: line became
4446 illegal because of the space(s). It is now covered by ${quote_local_part.
4448 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4451 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4453 83. Patch from Sieve maintainer to fix unterminated string problem in
4454 "vacation" handling.
4456 84. Some minor changes to the Linux configuration files to help with other
4457 OS variants using glibc.
4459 85. One more patch for Sieve to update vacation handling to latest spec.
4462 ----------------------------------------------------
4463 See the note above about the 4.44 and 4.50 releases.
4464 ----------------------------------------------------
4470 1. Change 4.43/35 introduced a bug that caused file counts to be
4471 incorrectly computed when quota_filecount was set in an appendfile
4474 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4475 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4478 3. Give more explanation in the error message when the command for a transport
4479 filter fails to execute.
4481 4. There are several places where Exim runs a non-Exim command in a
4482 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4483 was being done only for the command run by the queryprogram router. It is
4484 now done for all such subprocesses. The other cases are: ${run, transport
4485 filters, and the commands run by the lmtp and pipe transports.
4487 5. Some older OS have a limit of 256 on the maximum number of file
4488 descriptors. Exim was using setrlimit() to set 1000 as a large value
4489 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4490 systems. I've change it so that if it can't get 1000, it tries for 256.
4492 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4493 was an oversight, and furthermore, ever since the addition of extra
4494 controls (e.g. 4.43/32), the checks on when to allow different forms of
4495 "control" were broken. There should now be diagnostics for all cases when a
4496 control that does not make sense is encountered.
4498 7. $recipients is now available in the predata ACL (oversight).
4500 8. Tidy the search cache before the fork to do a delivery from a message
4501 received from the command line. Otherwise the child will trigger a lookup
4502 failure and thereby defer the delivery if it tries to use (for example) a
4503 cached ldap connection that the parent has called unbind on.
4505 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4506 of $address_data from the recipient verification was clobbered by the
4507 sender verification.
4509 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4510 was its contents. (It was OK if the option was not defined at all.)
4512 11. A "Completed" log line is now written for messages that are removed from
4513 the spool by the -Mrm option.
4515 12. $host_address is now set to the target address during the checking of
4516 ignore_target_hosts.
4518 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4519 being passed; this would have caused $sender_host_name to have been used if
4520 matching the list had actually called for a host name (not very likely,
4521 since this list is usually IP addresses). A host name is now passed as
4524 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4525 code that specifies a non-restarting handler (typically sigaction() in
4526 modern systems) in an attempt to fix a rare and obscure crash bug.
4528 15. Narrowed the window for a race in the daemon that could cause it to ignore
4529 SIGCHLD signals. This is not a major problem, because they are used only to
4530 wake it up if nothing else does.
4532 16. A malformed maildirsize file could cause Exim to calculate negative values
4533 for the mailbox size or file count. Odd effects could occur as a result.
4534 The maildirsize information is now recalculated if the size or filecount
4537 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4538 support for a long time. Removed HAVE_SYS_VFS_H.
4540 18. Updated exipick to current release from John Jetmore.
4542 19. Allow an empty sender to be matched against a lookup in an address list.
4543 Previously the only cases considered were a regular expression, or an
4546 20. Exim went into a mad DNS lookup loop when doing a callout where the
4547 host was specified on the transport, if the DNS lookup yielded more than
4550 21. The RFC2047 encoding function was originally intended for short strings
4551 such as real names; it was not keeping to the 75-character limit for
4552 encoded words that the RFC imposes. It now respects the limit, and
4553 generates multiple encoded words if necessary. To be on the safe side, I
4554 have increased the buffer size for the ${rfc2047: expansion operator from
4557 22. Failure to deliver a bounce message always caused it to be frozen, even if
4558 there was an errors_to setting on the router. The errors_to setting is now
4561 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4562 canonical form (fully expanded) before being placed in
4563 $sender_host_address.
4565 24. Updated eximstats to version 1.33
4567 25. Include certificate and key file names in error message when GnuTLS fails
4568 to set them up, because the GnuTLS error message doesn't include the name
4569 of the failing file when there is a problem reading it.
4571 26. Expand error message when OpenSSL has problems setting up cert/key files.
4574 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4575 (this can affect the format of dates).
4577 28. exim_tidydb, when checking for the continued existence of a message for
4578 which it has found a message-specific retry record, was not finding
4579 messages that were in split spool directories. Consequently, it was
4580 deleting retry records that should have stayed in existence.
4582 29. eximstats updated to version 1.35
4583 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4584 1.35 - bugfix such that pie charts by volume are generated correctly
4586 30. The SPA authentication driver was not abandoning authentication and moving
4587 on to the next authenticator when an expansion was forced to fail,
4588 contradicting the general specification for all authenticators. Instead it
4589 was generating a temporary error. It now behaves as specified.
4591 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4592 (the order specifies the preference for clients). The order is now AES256,
4593 AES128, 3DES, ARCFOUR128.
4595 31. Small patch to Sieve code - explicitly set From: when generating an
4598 32. Exim crashed if a remote delivery caused a very long error message to be
4599 recorded - for instance if somebody sent an entire SpamAssassin report back
4600 as a large number of 550 error lines. This bug was coincidentally fixed by
4601 increasing the size of one of Exim's internal buffers (big_buffer) that
4602 happened as part of the Exiscan merge. However, to be on the safe side, I
4603 have made the code more robust (and fixed the comments that describe what
4606 33. Some experimental protocols are using DNS PTR records for new purposes. The
4607 keys for these records are domain names, not reversed IP addresses. The
4608 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4609 leaves it alone. Component reversal etc. now happens only for IP addresses.
4612 34. The host_aton() function is supposed to be passed a string that is known
4613 to be a valid IP address. However, in the case of IPv6 addresses, it was
4614 not checking this. This is a hostage to fortune. Exim now panics and dies
4615 if the condition is not met. A case was found where this could be provoked
4616 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4617 components; fortuitously, this particular loophole had already been fixed
4618 by change 4.50/55 or 4.44/33 above.
4620 If there are any other similar loopholes, the new check in host_aton()
4621 itself should stop them being exploited. The report I received stated that
4622 data on the command line could provoke the exploit when Exim was running as
4623 exim, but did not say which command line option was involved. All I could
4624 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4625 running as the user.
4628 35. There was a buffer overflow vulnerability in the SPA authentication code
4629 (which came originally from the Samba project). I have added a test to the
4630 spa_base64_to_bits() function which I hope fixes it.
4633 36. The daemon start-up calls getloadavg() while still root for those OS that
4634 need the first call to be done as root, but it missed one case: when
4635 deliver_queue_load_max is set with deliver_drop_privilege. This is
4636 necessary for the benefit of the queue runner, because there is no re-exec
4637 when deliver_drop_privilege is set.
4639 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4640 was in use, was not putting the data itself into the right store pool;
4641 consequently, it could be overwritten for a subsequent message in the same
4642 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4645 38. Sometimes the final signoff response after QUIT could fail to get
4646 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4647 < 0 before doing a fflush(). This bug looks as though it goes back to the
4648 introduction of TLS in release 3.20, but "sometimes" must have been rare
4649 because the tests only now provoked it.
4655 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4656 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4657 Now it returns an integer. A number of calls were still expecting a T/F
4658 return. Fortuitously, in all cases, the tests worked in OK situations,
4659 which is the norm. However, things would have gone wrong on any write
4660 failures on the smtp file descriptor. This function is used when sending
4661 messages over SMTP and also when doing verify callouts.
4663 2. When Exim is called to do synchronous delivery of a locally submitted
4664 message (the -odf or -odi options), it no longer closes stderr before doing
4667 3. Implemented the mua_wrapper option.
4669 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4671 5. Implemented the functions header_remove(), header_testname(),
4672 header_add_at_position(), and receive_remove_recipient(), and exported them
4675 6. If an ACL "warn" statement specified the addition of headers, Exim already
4676 inserted X-ACL-Warn: at the start if there was no header name. However, it
4677 was not making this test for the second and subsequent header lines if
4678 there were newlines in the string. This meant that an invalid header could
4679 be inserted if Exim was badly configured.
4681 7. Allow an ACL "warn" statement to add header lines at the start or after all
4682 the Received: headers, as well as at the end.
4684 8. Added the rcpt_4xx retry error code.
4686 9. Added postmaster_mailfrom=xxx to callout verification option.
4688 10. Added mailfrom=xxxx to the callout verification option, for verify=
4691 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4692 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4694 12. Inserted some casts to stop certain compilers warning when using pointer
4695 differences as field lengths or precisions in printf-type calls (mostly
4696 affecting debugging statements).
4698 13. Added optional readline() support for -be (dynamically loaded).
4700 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4701 same clock tick as a message's arrival, so that its received time was the
4702 same as the "first fail" time on the retry record, and that message
4703 remained on the queue past the ultimate address timeout, every queue runner
4704 would try a delivery (because it was past the ultimate address timeout) but
4705 after another failure, the ultimate address timeout, which should have then
4706 bounced the address, did not kick in. This was a "< instead of <=" error;
4707 in most cases the first failure would have been in the next clock tick
4708 after the received time, and all would be well.
4710 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4711 being recognized when the domain list was tested by the match_domain
4712 condition in an expansion string.
4714 16. Added the ${str2b64: operator.
4716 17. Exim was always calling setrlimit() to set a large limit for the number of
4717 processes, without checking whether the existing limit was already
4718 adequate. (It did check for the limit on file descriptors.) Furthermore,
4719 errors from getrlimit() and setrlimit() were being ignored. Now they are
4720 logged to the main and panic logs, but Exim does carry on, to try to do its
4721 job under whatever limits there are.
4723 18. Imported PCRE 5.0.
4725 19. Trivial typo in log message " temporarily refused connection" (the leading
4728 20. If the log selector return_path_on_delivery was set and an address was
4729 redirected to /dev/null, the delivery process crashed because it assumed
4730 that a return path would always be set for a "successful" delivery. In this
4731 case, the whole delivery is bypassed as an optimization, and therefore no
4734 21. Internal re-arrangement: the function for sending a challenge and reading
4735 a response while authentication was assuming a zero-terminated challenge
4736 string. It's now changed to take a pointer and a length, to allow for
4737 binary data in such strings.
4739 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4741 23. Exim was not respecting finduser_retries when seeking the login of the
4742 uid under which it was called; it was always trying 10 times. (The default
4743 setting of finduser_retries is zero.) Also, it was sleeping after the final
4744 failure, which is pointless.
4746 24. Implemented tls_on_connect_ports.
4748 25. Implemented acl_smtp_predata.
4750 26. If the domain in control=submission is set empty, Exim assumes that the
4751 authenticated id is a complete email address when it generates From: or
4752 Sender: header lines.
4754 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4755 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4756 chown and chgrp in /bin and hostname in /usr/bin.
4758 28. Exim was keeping the "process log" file open after each use, just as it
4759 does for the main log. This opens the possibility of it remaining open for
4760 long periods when the USR1 signal hits a daemon. Occasional processlog
4761 errors were reported, that could have been caused by this. Anyway, it seems
4762 much more sensible not to leave this file open at all, so that is what now
4765 29. The long-running daemon process does not normally write to the log once it
4766 has entered its main loop, and it closes the log before doing so. This is
4767 so that log files can straightforwardly be renamed and moved. However,
4768 there are a couple of unusual error situations where the daemon does write
4769 log entries, and I had neglected to close the log afterwards.
4771 30. The text of an SMTP error response that was received during a remote
4772 delivery was being truncated at 512 bytes. This is too short for some of
4773 the long messages that one sometimes sees. I've increased the limit to
4776 31. It is now possible to make retry rules that apply only when a message has a
4777 specific sender, in particular, an empty sender.
4779 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4780 it possible to be selective about when SMTP synchronization is enforced.
4782 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4784 32. Implemented hosts_connection_nolog.
4786 33. Added an ACL for QUIT.
4788 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4791 35. Added mailbox_size and mailbox_filecount to appendfile.
4793 36. Added control = no_multiline_responses to ACLs.
4795 37. There was a bug in the logic of the code that waits for the clock to tick
4796 in the case where the clock went backwards by a substantial amount such
4797 that the microsecond fraction of "now" was more than the microsecond
4798 fraction of "then" (but the whole seconds number was less).
4800 38. Added support for the libradius Radius client library this is found on
4801 FreeBSD (previously only the radiusclient library was supported).
4807 1. When certain lookups returned multiple values in the form name=value, the
4808 quoting of the values was not always being done properly. Specifically:
4809 (a) If the value started with a double quote, but contained no whitespace,
4811 (b) If the value contained whitespace other than a space character (i.e.
4812 tabs or newlines or carriage returns) it was not quoted.
4813 This fix has been applied to the mysql and pgsql lookups by writing a
4814 separate quoting function and calling it from the lookup code. The fix
4815 should probably also be applied to nisplus, ibase and oracle lookups, but
4816 since I cannot test any of those, I have not disturbed their existing code.
4818 2. A hit in the callout cache for a specific address caused a log line with no
4819 reason for rejecting RCPT. Now it says "Previous (cached) callout
4820 verification failure".
4822 3. There was an off-by-one bug in the queryprogram router. An over-long
4823 return line was truncated at 256 instead of 255 characters, thereby
4824 overflowing its buffer with the terminating zero. As well as fixing this, I
4825 have increased the buffer size to 1024 (and made a note to document this).
4827 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4828 when Exim is waiting for an SMTP response from a remote server, Exim
4829 restarts its select() call on the socket, thereby resetting its timeout.
4830 This is not a problem when such interrupts are rare. Somebody set up a cron
4831 job to run exiwhat every 2 minutes, which is less than the normal select()
4832 timeout (5 or 10 minutes). This meant that the select() timeout never
4833 kicked in because it was always reset. I have fixed this by comparing the
4834 time when an interrupt arrives with the time at the start of the first call
4835 to select(). If more time than the timeout has elapsed, the interrupt is
4836 treated as a timeout.
4838 5. Some internal re-factoring in preparation for the addition of Sieve
4839 extensions (by MH). In particular, the "personal" test is moved to a
4840 separate function, and given an option for scanning Cc: and Bcc: (which is
4841 not set for Exim filters).
4843 6. When Exim created an email address using the login of the caller as the
4844 local part (e.g. when creating a From: or Sender: header line), it was not
4845 quoting the local part when it contained special characters such as @.
4847 7. Installed new OpenBSD configuration files.
4849 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4850 try to make them clearer.
4852 9. Callout options, other than the timeout value, were being ignored when
4853 verifying sender addresses in header lines. For example, when using
4855 verify = header_sender/callout=no_cache
4857 the cache was (incorrectly) being used.
4859 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4860 only the Cygwin environment.
4862 11. When return_path_on_delivery was set as a log selector, if different remote
4863 addresses in the same message used different return paths and parallel
4864 remote delivery occurred, the wrong values would sometimes be logged.
4865 (Whenever a remote delivery process finished, the return path value from
4866 the most recently started remote delivery process was logged.)
4868 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4869 lines when AUTH and/or TLS are in use. This is the "received protocol"
4870 field. Exim used to use "asmtp" for authenticated SMTP, without any
4871 indication (in the protocol name) for TLS use. Now it follows the RFC and
4872 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4873 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4874 names appear in log lines as well as in Received: header lines.
4876 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4877 extensions, and comparison tests, and to fix some bugs.
4879 14. Changes to the "personal" filter test:
4881 (1) The test was buggy in that it was just doing the equivalent of
4882 "contains" tests on header lines. For example, if a user's address was
4883 anne@some.where, the "personal" test would incorrectly be true for
4885 To: susanne@some.where
4887 This test is now done by extracting each address from the header in turn,
4888 and checking the entire address. Other tests that are part of "personal"
4889 are now done using regular expressions (for example, to check local parts
4890 of addresses in From: header lines).
4892 (2) The list of non-personal local parts in From: addresses has been
4893 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4894 taken from the Sieve specification recommendations.
4896 (3) If the message contains any header line starting with "List-" it is
4897 treated as non-personal.
4899 (4) The test for "circular" in the Subject: header line has been removed
4900 because it now seems ill-conceived.
4902 15. Minor typos in src/EDITME comments corrected.
4904 16. Installed latest exipick from John Jetmore.
4906 17. If headers_add on a router specified a text string that was too long for
4907 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4908 of string_sprintf() is now avoided.
4910 18. $message_body_size was not set (it was always zero) when running the DATA
4911 ACL and the local_scan() function.
4913 19. For the "mail" command in an Exim filter, no default was being set for
4914 the once_repeat time, causing a random time value to be used if "once" was
4915 specified. (If the value happened to be <= 0, no repeat happened.) The
4916 default is now 0s, meaning "never repeat". The "vacation" command was OK
4917 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4918 (I found it when inspecting the code).
4920 20. There is now an overall timeout for performing a callout verification. It
4921 defaults to 4 times the callout timeout, which applies to individual SMTP
4922 commands during the callout. The overall timeout applies when there is more
4923 than one host that can be tried. The timeout is checked before trying the
4924 next host. This prevents very long delays if there are a large number of
4925 hosts and all are timing out (e.g. when the network connections are timing
4926 out). The value of the overall timeout can be changed by specifying an
4927 additional sub-option for "callout", called "maxwait". For example:
4929 verify = sender/callout=5s,maxwait=20s
4931 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4932 to the end before writing, but this should make it even safer).
4934 22. Exim was forgetting that it had advertised PIPELINING for the second and
4935 subsequent messages on an SMTP connection. It was also not resetting its
4936 memory on STARTTLS and an internal HELO.
4938 23. When Exim logs an SMTP synchronization error within a session, it now
4939 records whether PIPELINING has been advertised or not.
4941 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4942 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4943 rather than long int.
4945 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4947 26. Added the never_mail option to autoreply.
4953 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4954 crash if the getsockname() call failed; this can happen if a connection is
4955 closed very soon after it is established. The problem was simply in the
4956 order in which certain operations were done, causing Exim to try to write
4957 to the SMTP stream before it had set up the file descriptor. The bug has
4958 been fixed by making things happen in the correct order.
4964 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4965 before the connection was closed, thus losing the rejection response.
4967 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4968 some early Solaris releases, but causes trouble in current releases where
4969 socklen_t is defined.
4971 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4974 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4977 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4978 that releasing the top of it at the end releases what was used for sub-
4979 expansions (unless the block got too big). However, discard this block if
4980 the first thing is a variable or header, so that we can use its block when
4981 it is dynamic (useful for very large $message_headers, for example).
4983 6. Lookups now cache *every* query, not just the most recent. A new, separate
4984 store pool is used for this. It can be recovered when all lookup caches are
4985 flushed. Lookups now release memory at the end of their result strings.
4986 This has involved some general refactoring of the lookup sources.
4988 7. Some code has been added to the store_xxx() functions to reduce the amount
4989 of flapping under certain conditions.
4991 8. log_incoming_interface used to affect only the <= reception log lines. Now
4992 it causes the local interface and port to be added to several more SMTP log
4993 lines, for example "SMTP connection from", and rejection lines.
4995 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4997 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4999 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5000 could be overwritten at the end of the current message (or the start of a
5001 new message if it was set in a HELO ACL). The value is now preserved for
5002 the duration of the SMTP connection.
5004 12. If a transport had a headers_rewrite setting, and a matching header line
5005 contained an unqualified address, that address was qualified, even if it
5006 did not match any rewriting rules. The underlying bug was that the values
5007 of the flags that permit the existence of unqualified sender and recipient
5008 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5009 non-local messages, and by -bnq for local messages) were not being
5010 preserved with the message after it was received.
5012 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5013 "next input=" as part of the text comprising the host identity instead of
5014 the correct text. The code was using the same buffer for two different
5015 strings. However, depending on which order the printing function evaluated
5016 its arguments, the bug did not always show up. Under Linux, for example, my
5017 test suite worked just fine.
5019 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5020 doesn't work with some older versions of Perl. It has been changed to "my",
5021 which in any case is probably the better facility to use.
5023 15. A really picky compiler found some instances of statements for creating
5024 error messages that either had too many or two few arguments for the format
5027 16. The size of the buffer for calls to the DNS resolver has been increased
5028 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5029 for addresses that have a lot of PTR records. This alleviates a problem; it
5030 does not fully solve it.
5032 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5033 buffer now truncates the list and logs the incident, which is the same
5034 action as happens when Exim is looking up a host name and its aliases.
5035 Previously in this situation something unpredictable would happen;
5036 sometimes it was "internal error: store_reset failed".
5038 18. If a server dropped the connection unexpectedly when an Exim client was
5039 using GnuTLS and trying to read a response, the client delivery process
5040 crashed while trying to generate an error log message.
5042 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5043 string, for example:
5045 warn message = H1: something\nH2: something
5047 the text was added as a single header line from Exim's point of view
5048 though it ended up OK in the delivered message. However, searching for the
5049 second and subsequent header lines using $h_h2: did not work. This has been
5050 fixed. Similarly, if a system filter added multiple headers in this way,
5051 the routers could not see them.
5053 20. Expanded the error message when iplsearch is called with an invalid key to
5054 suggest using net-iplsearch in a host list.
5056 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5057 ACLs are no longer actually imposed (and a message to that effect is
5060 22. If a "gecos" field in a passwd entry contained escaped characters, in
5061 particular, if it contained a \" sequence, Exim got it wrong when building
5062 a From: or a Sender: header from that name. A second bug also caused
5063 incorrect handling when an unquoted " was present following a character
5064 that needed quoting.
5066 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5067 was not being matched caselessly.
5069 24. Arranged for all hyphens in the exim.8 source to be escaped with
5072 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5073 properly. Recipient callout cache records were still being keyed to include
5074 the sender, even when use_sender was set false. This led to far more
5075 callouts that were necessary. The sender is no longer included in the key
5076 when use_sender is false.
5078 26. Added "control = submission" modifier to ACLs.
5080 27. Added the ${base62d: operator to decode base 62 numbers.
5082 28. dnsdb lookups can now access SRV records.
5084 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5085 the configuration file.
5087 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5088 (-v) mode. This makes the output for a verbose queue run more intelligible.
5090 31. Added a use_postmaster feature to recipient callouts.
5092 32. Added the $body_zerocount variable, containing the number of binary zero
5093 bytes in the message body.
5095 33. The time of last modification of the "new" subdirectory is now used as the
5096 "mailbox time last read" when there is a quota error for a maildir
5099 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5101 35. Added +ignore_unknown as a special item in host lists.
5103 36. Code for decoding IPv6 addresses in host lists is now included, even if
5104 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5105 address was recognized as an IP address, but was then not correctly decoded
5106 into binary, causing unexpected and incorrect effects when compared with
5113 1. Very minor rewording of debugging text in manualroute to say "list of
5114 hosts" instead of "hostlist".
5116 2. If verify=header_syntax was set, and a header line with an unqualified
5117 address (no domain) and a large number of spaces between the end of the
5118 name and the colon was received, the reception process suffered a buffer
5119 overflow, and (when I tested it) crashed. This was caused by some obsolete
5120 code that should have been removed. The fix is to remove it!
5122 3. When running in the test harness, delay a bit after writing a bounce
5123 message to get a bit more predictability in the log output.
5125 4. Added a call to search_tidyup() just before forking a reception process. In
5126 theory, someone could use a lookup in the expansion of smtp_accept_max_
5127 per_host which, without the tidyup, could leave open a database connection.
5129 5. Added the variables $recipient_data and $sender_data which get set from a
5130 lookup success in an ACL "recipients" or "senders" condition, or a router
5131 "senders" option, similar to $domain_data and $local_part_data.
5133 6. Moved the writing of debug_print from before to after the "senders" test
5136 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5137 problems for message scanning, either using a data ACL, or using
5138 local_scan() because the Received: header was not generated till after they
5139 were called (in order to set the time as the time of reception completion).
5140 I have revised the way this works. The header is now generated after the
5141 body is received, but before the ACL or local_scan() are called. After they
5142 are run, the timestamp in the header is updated.
5148 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5149 before starting a queue runner without re-exec. This happened only when
5150 deliver_drop_privilege was set or when the Exim user was set to root. The
5151 effect of the bug was that timeouts during subsequent deliveries caused
5152 crashes instead of being properly handled. The handler is now left at its
5153 default (and expected) setting.
5155 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5156 message, again when deliver_drop_privilege is set or Exim is run as root.
5157 The bug described in (1) was not present in this case, but the tidying up
5158 of the other signals was missing. I have made the two cases consistent.
5160 3. The ignore_target_hosts setting on a manualroute router was being ignored
5161 for hosts that were looked up using the /MX notation.
5163 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5166 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5167 operated on the sender address. After changing the $sender_address to <>
5168 for the sender address verify, Exim was re-instated it as the original
5169 (before rewriting) address, but remembering that it had rewritten it, so it
5170 wasn't rewriting it again. This bug also had the effect of breaking the
5171 sender address verification caching when the sender address was rewritten.
5173 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5174 This has been changed so that if the ip literal address matches
5175 ignore_target_hosts, the router declines.
5177 7. Added expansion conditions match_domain, match_address, and match_local_
5178 part (NOT match_host).
5180 8. The placeholder for the Received: header didn't have a length field set.
5182 9. Added code to Exim itself and to exim_lock to test for a specific race
5183 condition that could lead to file corruption when using MBX delivery. The
5184 issue is with the lockfile that is created in /tmp. If this file is removed
5185 after a process has opened it but before that process has acquired a lock,
5186 there is the potential for a second process to recreate the file and also
5187 acquire a lock. This could lead to two Exim processes writing to the file
5188 at the same time. The added code performs the same test as UW imapd; it
5189 checks after acquiring the lock that its file descriptor still refers to
5190 the same named file.
5192 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5193 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5196 11. Added the smtp_active_hostname option. If used, this will typically be made
5197 to depend on the incoming interface address. Because $interface_address is
5198 not set up until the daemon has forked a reception process, error responses
5199 that can happen earlier (such as "too many connections") no longer contain
5202 12. If an expansion in a condition on a "warn" statement fails because a lookup
5203 defers, the "warn" statement is abandoned, and the next ACL statement is
5204 processed. Previously this caused the whole ACL to be aborted.
5206 13. Added the iplsearch lookup type.
5208 14. Added ident_timeout as a log selector.
5210 15. Added tls_certificate_verified as a log selector.
5212 16. Added a global option tls_require_ciphers (compare the smtp transport
5213 option of the same name). This controls incoming TLS connections.
5215 17. I finally figured out how to make tls_require_ciphers do a similar thing
5216 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5217 before starting the TLS session.
5219 18. Tabs are now shown as \t in -bP output.
5221 19. If the log selector return_path_on_delivery was set, Exim crashed when
5222 bouncing a message because it had too many Received: header lines.
5224 20. If two routers both had headers_remove settings, and the first one included
5225 a superfluous trailing colon, the final name in the first list and the
5226 first name in the second list were incorrectly joined into one item (with a
5227 colon in the middle).
5233 1. Added -C and -D options to the exinext utility, mainly to make it easier
5234 to include in the automated testing, but these could be helpful when
5235 multiple configurations are in use.
5237 2. The exinext utility was not formatting the output nicely when there was
5238 an alternate port involved in the retry record key, nor when there was a
5239 message id as well (for retries that were specific to a specific message
5240 and a specific host). It was also confused by IPv6 addresses, because of
5241 the additional colons they contain. I have fixed the IPv4 problem, and
5242 patched it up to do a reasonable job for IPv6.
5244 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5245 delivery, the log line now contains "pipelined" if PIPELINING was used.
5247 4. An SMTP transport process used to panic and die if the bind() call to set
5248 an explicit outgoing interface failed. This has been changed; it is now
5249 treated in the same way as a connect() failure.
5251 5. A reference to $sender_host_name in the part of a conditional expansion
5252 that was being skipped was still causing a DNS lookup. This no longer
5255 6. The def: expansion condition was not recognizing references to header lines
5256 that used bh_ and bheader_.
5258 7. Added the _cache feature to named lists.
5260 8. The code for checking quota_filecount in the appendfile transport was
5261 allowing one more file than it should have been.
5263 9. For compatibility with Sendmail, the command line option
5271 and sets the incoming protocol and host name (for trusted callers). The
5272 host name and its colon can be omitted when only the protocol is to be set.
5273 Note the Exim already has two private options, -pd and -ps, that refer to
5274 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5275 "s", but I don't think that's a major issue.
5277 10. A number of refactoring changes to the code, none of which should affect
5280 (a) The number of logging options was getting close to filling up the
5281 32-bit word that was used as a bit map. I have split them into two classes:
5282 those that are passed in the argument to log_write(), and those that are
5283 only ever tested independently outside of that function. These are now in
5284 separate 32-bit words, so there is plenty of room for expansion again.
5285 There is no change in the user interface or the logging behaviour.
5287 (b) When building, for example, log lines, the code previously used a
5288 macro that called string_cat() twice, in order to add two strings. This is
5289 not really sufficiently general. Furthermore, there was one instance where
5290 it was actually wrong because one of the argument was used twice, and in
5291 one call a function was used. (As it happened, calling the function twice
5292 did not affect the overall behaviour.) The macro has been replaced by a
5293 function that can join an arbitrary number of extra strings onto a growing
5296 (c) The code for expansion conditions now uses a table and a binary chop
5297 instead of a serial search (which was left over from when there were very
5298 few conditions). Also, it now recognizes conditions like "pam" even when
5299 the relevant support is not compiled in: a suitably worded error message is
5300 given if an attempt is made to use such a condition.
5302 11. Added ${time_interval:xxxxx}.
5304 12. A bug was causing one of the ddress fields not to be passed back correctly
5305 from remote delivery subprocesses. The field in question was not being
5306 subsequently used, so this caused to problems in practice.
5308 13. Added new log selectors queue_time and deliver_time.
5310 14. Might have fixed a bug in maildirsizefile handling that threw up
5311 "unexpected character" debug warnings, and recalculated the data
5312 unnecessarily. In any case, I expanded the warning message to give more
5315 15. Added the message "Restricted characters in address" to the statements in
5316 the default ACL that block characters like @ and % in local parts.
5318 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5319 Three changes have been made:
5321 (a) There was a serious bug; a negative response to MAIL caused the whole
5322 recipient domain to be cached as invalid, thereby blocking all messages
5323 to all local parts at the same domain, from all senders. This bug has
5324 been fixed. The domain is no longer cached after a negative response to
5325 MAIL if the sender used is not empty.
5327 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5330 (c) A new callout option, "use_sender" has been added for people who want
5331 the modified behaviour.
5337 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5340 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5341 indeed breaks things for older releases.
5343 3. Added additional logging to the case where there is a problem reading data
5344 from a filter that is running in a subprocess using a pipe, in order to
5345 try to track down a specific problem.
5347 4. Testing facility fudge: when running in the test harness and attempting
5348 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5349 getting "No route to host". Convert this to a timeout.
5351 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5354 6. Some OS don't have socklen_t but use size_t instead. This affects the
5355 fifth argument of getsockopt() amongst other things. This is now
5356 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5357 can be set for individual OS. I have set it for SunOS5, OSF1, and
5358 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5359 some earlier ones do not.
5361 7. Change 4.30/15 was not doing the test caselessly.
5363 8. The standard form for an IPv6 address literal was being rejected by address
5364 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5365 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5366 this, as well as the form without the "IPv6" on the front (but only when
5367 address literals are enabled, of course).
5369 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5371 10. Exim crashed if a message with an empty sender address specified by -f
5372 encountered a router with an errors_to setting. This could be provoked only
5373 by a command such as
5377 where an empty string was supplied; "<>" did not hit this bug.
5379 11. Installed PCRE release 4.5.
5381 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5382 remained set. It is now erased.
5384 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5385 times from message ids (which are base 36 rather than the normal 62).
5387 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5388 were being counted as actual protocol errors, and logged if the log
5389 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5390 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5391 and DATA following a set of rejected RCPTs do not count as protocol errors.
5392 In other words, Exim assumes they were pipelined, though this may not
5393 actually be the case. Of course, in all cases the client gets an
5394 appropriate error code.
5396 15. If a lookup fails in an ACL condition, a message about the failure may
5397 be available; it is used if testing the ACL cannot continue, because most
5398 such messages specify what the cause of the deferral is. However, some
5399 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5400 that caused an old message to be retained and used if a later statement
5401 caused a defer, replacing the real cause of the deferral.
5403 16. If an IP address had so many PTR records that the DNS lookup buffer
5404 was not large enough to hold them, Exim could crash while trying to process
5405 the truncated data. It now detects and logs this case.
5407 17. Further to 4.21/58, another change has been made: if (and only if) the
5408 first line of a message (the first header line) ends with CRLF, a bare LF
5409 in a subsequent header line has a space inserted after it, so as not to
5410 terminate the header.
5412 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5413 unnecessarily, used atoi() instead of strtol(), and didn't check the
5414 termination when getting file sizes from file names by regex.
5416 19. Completely re-implemented the support for maildirsize files, in the light
5417 of a number of problems with the previous contributed implementation
5418 (4.30/29). In particular:
5420 . If the quota is zero, the maildirsize file is maintained, but no quota is
5423 . If the maildir directory does not exist, it is created before any attempt
5424 to write a maildirsize file.
5426 . The quota value in the file is just a cache; if the quota is changed in
5427 the transport, the new value overrides.
5429 . A regular expression is available for excluding directories from the
5432 20. The autoreply transport checks the characters in options that define the
5433 message's headers; it allows continued headers, but it was checking with
5434 isspace() after an embedded newline instead of explicitly looking for a
5437 21. If all the "regular" hosts to which an address was routed had passed their
5438 expiry times, and had not reached their retry times, the address was
5439 bounced, even if fallback hosts were defined. Now Exim should go on to try
5442 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5443 equivalent code in the SMTP transport. Some hosts send humungous responses
5444 to HELO/EHLO, more than 1024 it seems.
5446 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5447 gives compiler warnings in some environments. I've now done it "properly",
5450 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5451 (because of problems with the built-in one) was declared to return uschar *
5452 instead of char *, causing compiler failure.
5454 25. Fixed a file descriptor leak when processing alias/forward files.
5456 26. Fixed a minor format string issue in dbfn.c.
5458 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5460 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5461 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5464 29. When a sender address is verified, it is cached, to save repeating the test
5465 when there is more than one recipient in a message. However, when the
5466 verification involves a callout, it is possible for different callout
5467 options to be set for different recipients. It is too complicated to keep
5468 track of this in the cache, so now Exim always runs a verification when a
5469 callout is required, relying on the callout cache for the optimization.
5470 The overhead is duplication of the address routing, but this should not be
5473 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5474 to be verified with callout=postmaster, and the main callout worked but the
5475 postmaster check failed, the verification correctly failed. However, if a
5476 subsequent RCPT command asked for sender verification *without* the
5477 postmaster check, incorrect caching caused this verification also to fail,
5480 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5481 it was not caching the DNS options (qualify_single, search_parents) that
5482 were used when the lookup failed. A subsequent lookup with different
5483 options therefore always gave the same answer, though there were cases
5484 where it should not have. (Example: a "domains = !$mx_any" option on a
5485 dnslookup router: the "domains" option is always processed without any
5486 widening, but the router might have qualify_single set.) Now Exim uses the
5487 cached value only when the same options are set.
5489 32. Added John Jetmore's "exipick" utility to the distribution.
5491 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5492 than a timeout (e.g. a certificate is required, and is not provided), an
5493 Exim server now closes the connection immediately. Previously it waited for
5494 the client to close - but if the client is SSL, it seems that they each
5495 wait for each other, leading to a delay before one of them times out.
5497 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5498 maintained 0.8.x compatibility because I don't think many are using it, and
5499 it is clearly obsolete.
5501 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5504 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5505 client certificate was expired. A simple patch fixes this, though I don't
5506 understand the full logic of why the verify callback is called multiple
5509 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5510 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5511 which causes problems with some clients (such as the Certicom SSL Plus
5512 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5513 disables the coutermeasure allowing Eudora to connect."
5515 38. Exim was not checking that a write() to a log file succeeded. This could
5516 lead to Bad Things if a log got too big, in particular if it hit a file
5517 size limit. Exim now panics and dies if it cannot write to a log file, just
5518 as it does if it cannot open a log file.
5520 39. Modified OS/Makefile-Linux so that it now contains
5522 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5524 The two -D definitions ensure that Exim is compiled with large file
5525 support, which makes it possible to handle log files that are bigger than
5528 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5529 instance) a domain was checked against a named list that involved a lookup,
5530 causing $domain_data to be set, then another domain was checked against the
5531 same list, then the first domain was re-checked, the value of $domain_data
5532 after the final check could be wrong. In particular, if the second check
5533 failed, it could be set empty. This bug probably also applied to
5536 41. The strip_trailing_dot option was not being applied to the address given
5537 with the -f command-line option.
5539 42. The code for reading a message's header from the spool was incrementing
5540 $received_count, but never initializing it. This meant that the value was
5541 incorrect (doubled) while delivering a message in the same process in which
5542 it was received. In the most common configuration of Exim, this never
5543 happens - a fresh exec is done - but it can happen when
5544 deliver_drop_privilege is set.
5546 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5547 it now includes up to 150 characters of the unexpected data in the log
5550 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5551 and building data strings. The size of both of these buffers was 10 000
5552 bytes - far larger than anybody would *ever* want, thought I. Needless to
5553 say, somebody hit the limit. I have increased the maximum line length to
5554 20 000 and the maximum data length of concatenated lines to 100 000. I have
5555 also fixed two bugs, because there was no checking on these buffers. Tsk,
5556 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5557 buffer is too small.
5559 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5560 lsearch lookups. Now it does.
5562 46. When parsing a route_list item in a manualroute router, a fixed-length
5563 buffer was used for the list of hosts. I made this 1024 bytes long,
5564 thinking that nobody would ever have a list of hosts that long. Wrong.
5565 Somebody had a whole pile of complicated expansion conditions, and the
5566 string was silently truncated, leading to an expansion error. It turns out
5567 that it is easier to change to an unlimited length (owing to other changes
5568 that have happened since this code was originally written) than to build
5569 structure for giving a limitation error. The length of the item that
5570 expands into the list of hosts is now unlimited.
5572 47. The lsearch lookup could not handle data where the length of text line was
5573 more than 4095 characters. Such lines were truncated, leading to shortened
5574 data being returned. It should now handle lines of any length.
5576 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5577 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5580 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5581 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5582 no longer happens in comment lines. A list of replacements is now placed
5583 at the head of all of the source files, except those whose only change is
5584 to replace PERL_COMMAND in the very first #! line.
5586 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5587 messages on the queue, with a bottom-up merge sort, using code contributed
5588 by Michael Haardt. This should make operations like -bp somewhat faster on
5589 large queues. It won't affect queue runners, except when queue_run_in_order
5592 51. Installed eximstats 1.31 in the distribution.
5594 52. Added support for SRV lookups to the dnslookup router.
5596 53. If an ACL referred to $message_body or $message_body_end, the value was not
5597 reset for any messages that followed in the same SMTP session.
5599 54. The store-handling optimization for building very long strings was not
5600 differentiating between the different store pools. I don't think this
5601 actually made any difference in practice, but I've tidied it.
5603 55. While running the routers to verify a sender address, $sender_address
5604 was still set to the sender address. This is wrong, because when routing to
5605 send a bounce to the sender, it would be empty. Therefore, I have changed
5606 it so that, while verifying a sender address, $sender_address is set to <>.
5607 (There is no change to what happens when verifying a recipient address.)
5609 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5610 target A or AAAA records (if not already returned) without resetting the
5611 qualify_single or search_parents options of the DNS resolver. These are
5612 inappropriate in this case because the targets of MX and SRV records must
5613 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5614 target that, when qualified, matched something in the local domain. These
5615 two options are now turned off when doing these lookups.
5617 57. It seems that at least some releases of Reiserfs (which does not have the
5618 concept of a fixed number of inodes) returns zero and not -1 for the
5619 number of available inodes. This interacted badly with check_spool_inodes,
5620 which assumed that -1 was the "no such thing" setting. What I have done is
5621 to check that the total number of inodes is greater than zero before doing
5622 the test of how many are available.
5624 58. When a "warn" ACL statement has a log_message modifier, the message is
5625 remembered, and not repeated. This is to avoid a lot of repetition when a
5626 message has many recipients that cause the same warning to be written.
5627 However, Exim was preserving the list of already written lines for an
5628 entire SMTP session, which doesn't seem right. The memory is now reset if a
5629 new message is started.
5631 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5632 debugging output unless log_rewrite was also set.
5634 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5635 of (int)(handle) when we know that handle contains (void *)(-1).
5637 61. The Exim daemon panic-logs an error return when it closes the incoming
5638 connection. However "connection reset by peer" seems to be common, and
5639 isn't really an error worthy of noting specially, so that particular error
5642 62. When Exim is trying to find all the local interfaces, it used to panic and
5643 die if the ioctl to get the interface flags failed. However, it seems that
5644 on at least one OS (Solaris 9) it is possible to have an interface that is
5645 included in the list of interfaces, but for which you get a failure error
5646 for this call. This happens when the interface is not "plumbed" into a
5647 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5648 failure of the "get flags" call assumes that the interface is down.
5650 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5651 makes life easier for people who are doing arithmetic on fields extracted
5652 from dates, where you often get leading zeros that should not be
5653 interpreted as octal.
5655 64. Added qualify_domain to the redirect router, to override the global
5658 65. If a pathologically long header line contained very many addresses (the
5659 report of this problem mentioned 10 000) and each of them was rewritten,
5660 Exim could use up a very large amount of memory. (It kept on making new
5661 copies of the header line as it rewrote, and never released the old ones.)
5662 At the expense of a bit more processing, the header rewriting function has
5663 been changed so that it no longer eats memory in this way.
5665 66. The generation of the Received: header has been moved from the time that a
5666 message starts to be received, to the time that it finishes. The timestamp
5667 in the Received: header should now be very close to that of the <= log
5668 line. There are two side-effects of this change:
5670 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5671 logged header lines no longer include the local Received: line, because
5672 it has not yet been created. The same applies to a copy of the message
5673 that is returned to a non-SMTP sender when a message is rejected.
5675 (b) When a filter file is tested using -bf, no additional Received: header
5676 is added to the test message. After some thought, I decided that this
5679 This change does not affect the value of $received_for. It is still set
5680 after address rewriting, but before local_scan() is called.
5682 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5684 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5685 gave an unhelpful panic error message, and a defer error. I have managed to
5686 change this behaviour so that it now rejects any supplied certificate,
5687 which seems right, as the list of acceptable certificates is empty.
5689 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5690 gave an unhelpful defer error. I have not managed to make this reject any
5691 supplied certificates, but the error message it gives is "no certificate
5692 supplied", which is not helpful.
5694 70. exigrep's output now also includes lines that are not associated with any
5695 message, but which match the given pattern. Implemented by a patch from
5696 Martin Sluka, which also tidied up the Perl a bit.
5698 71. Recipient callout verification, like sender verification, was using <> in
5699 the MAIL FROM command. This isn't really the right thing, since the actual
5700 sender may affect whether the remote host accepts the recipient or not. I
5701 have changed it to use the actual sender in the callout; this means that
5702 the cache record is now keyed on a recipient/sender pair, not just the
5703 recipient address. There doesn't seem to be a real danger of callout loops,
5704 since a callout by the remote host to check the sender would use <>.
5705 [SEE ABOVE: changed after hitting problems.]
5707 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5708 temporary errors. However, in the case of such a code being given after
5709 the end of a data transmission (i.e. after ".") Exim was failing to write
5710 a retry record for the message. (Yes, there was some broken host that was
5711 actually sending 8xx at this point.)
5713 73. An unknown lookup type in a host list could cause Exim to panic-die when
5714 the list was checked. (An example that provoked this was putting <; in the
5715 middle of a list instead of at the start.) If this happened during a DATA
5716 ACL check, a -D file could be left lying around. This kind of configuration
5717 error no longer causes Exim to die; instead it causes a defer error. The
5718 incident is still logged to the main and panic logs.
5720 74. Buglet left over from Exim 3 conversion. The message "too many messages
5721 in one connection" was written to the rejectlog but not the mainlog, except
5722 when address rewriting (yes!) was being logged.
5724 75. Added write_rejectlog option.
5726 76. When a system filter was run not as root (that is, when system_filter_user
5727 was set), the values of the $n variables were not being returned to the
5728 main process; thus, they were not subsequently available in the $sn
5731 77. Added +return_path_on_delivery log selector.
5733 78. A connection timeout was being treated differently from recipients deferred
5734 when testing hosts_max_try with a message that was older than the host's
5735 retry timeout. (The host should not be counted, thus allowing all hosts to
5736 be tried at least once before bouncing.) This may have been the cause of an
5737 occasionally reported bug whereby a message would remain on the queue
5738 longer than the retry timeout, but would be bounced if a delivery was
5739 forced. I say "may" because I never totally pinned down the problem;
5740 setting up timeout/retry tests is difficult. See also the next item.
5742 79. The ultimate address timeout was not being applied to errors that involved
5743 a combination of host plus message (for example, a timeout on a MAIL
5744 command). When an address resolved to a number of possible hosts, and they
5745 were not all tried for each delivery (e.g. because of hosts_max_try), a
5746 message could remain on the queue longer than the retry timeout.
5748 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5751 81. Fixed an obscure SMTP outgoing bug which required at least the following
5752 conditions: (a) there was another message waiting for the same server;
5753 (b) the server returned 5xx to all RCPT commands in the first message so
5754 that the message was not completed; (c) the server dropped the connection
5755 or gave a negative response to the RSET that Exim sends to abort the
5756 transaction. The observed case was a dropped connection after DATA that had
5757 been sent in pipelining mode. That is, the server had advertised PIPELINING
5758 but was not implementing it correctly. The effect of the bug was incorrect
5759 behaviour, such as trying another host, and this could lead to a crash.
5765 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5766 and daemon.c were passed as pointers to ints; they should have been
5767 pointers to socklen_t variables (which are typically unsigned ints).
5769 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5772 3. Fixed a really odd bug that affected only the testing scheme; patching a
5773 certain fixed string in the binary changed the value of another string that
5774 happened to be identical to the end of the original first string.
5776 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5777 name", it returns that address as the IP address. On some operating
5778 systems (e.g. Solaris), it also passes back the IP address string as the
5779 "host name". However, on others (e.g. Linux), it passes back an empty
5780 string. Exim wasn't checking for this, and was changing the host name to an
5781 empty string, assuming it had been canonicalized.
5783 5. Although rare, it is permitted to have more than one PTR record for a given
5784 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5785 all the names associated with an address, because they do in Solaris.
5786 However, it seems that they do not in Linux for data that comes from the
5787 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5788 I found this out when I moved to a new Linux workstation and tried to run
5789 the Exim test suite.
5791 To get round this problem I have changed the code so that it now does its
5792 own call to the DNS to look up PTR records when searching for a host name.
5793 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5794 addresses that are only in /etc/hosts are still found.
5796 This behaviour is, however, controlled by an option called host_lookup_
5797 order, which defaults to "bydns:byaddr". If people want to use the other
5798 order, or indeed, just use one or the other means of lookup, they can
5799 specify it in this variable.
5801 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5802 some operating systems, this comes back from gethostbyaddr() as an empty
5803 string, and this is what Exim used to test for. However, it seems that in
5804 other systems, "." is yielded. Exim now tests for this case too.
5806 7. The values of check_spool_space and check_log_space are now held internally
5807 as a number of kilobytes instead of an absolute number of bytes. If a
5808 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5809 kilobyte. This means that much larger values can be stored.
5811 8. Exim monitor: an attempt to get the action menu when not actually pointing
5812 at a message produces an empty menu entitled "No message selected". This
5813 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5814 no entries in it ("Shell widget menu has zero width and/or height"). So I
5815 have added a single, blank menu entry in this case.
5817 9. Added ${quote_local_part.
5819 10. MIME decoding is now applied to the contents of Subject: header lines when
5822 11. Now that a reference to $sender_host_address automatically causes a reverse
5823 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5824 host lookup before query-style lookups in lists that might use this
5825 variable. This has therefore been abolished, and the "net-" prefix is no
5826 longer necessary for query-style lookups.
5828 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5829 have been a typo for LFLAGS, so it has been changed.
5831 13. The install script calls Exim with "-C /dev/null" in order to find the
5832 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5833 to be output. However, since Exim outputs its version number before the
5834 error, it didn't break the script. It just looked ugly. I fixed this by
5835 always allowing "-C /dev/null" if the caller is root.
5837 14. Ignore overlarge ACL variable number when reading spool file - insurance
5838 against a later release with more variables having written the file.
5840 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5841 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5842 this, as well as the form without the "IPv6" on the front.
5844 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5845 OS/Makefile-Darwin file.
5847 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5848 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5850 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5851 variable. However, some OS use ints for inodes. I've added cast to long int
5852 to get rid of the compiler warning.
5854 19. I had forgotten to lock out "/../" in configuration file names when
5855 ALT_CONFIG_PREFIX was set.
5857 20. Routers used for verification do not need to specify transports. However,
5858 if such a router generated a host list, and callout was configured, Exim
5859 crashed, because it could not find a port number from the (non-existent)
5860 transport. It now assumes port 25 in this circumstance.
5862 21. Added the -t option to exigrep.
5864 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5865 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5866 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5867 them is set, LOOKUP_LSEARCH is forced.
5869 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5871 24. Added sender and host information to the "rejected by local_scan()" log
5872 line; previously there was no indication of these.
5874 25. Added .include_if_exists.
5876 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5877 Linux. It turns out that not all file systems support this. Apparently some
5878 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5879 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5880 not supported on the file descriptor, is now ignored when Exim is trying to
5881 sync a directory. This applies only to Linux.
5883 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5885 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5886 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5889 29. Added support for maildirsize files from supplied patch (modified a bit).
5891 30. The use of :fail: followed by an empty string could lead Exim to respond to
5892 sender verification failures with (e.g.):
5894 550 Verification failed for <xxx>
5895 550 Sender verify failed
5897 where the first response line was missing the '-' that indicates it is not
5898 the final line of the response.
5900 31. The loop for finding the name of the user that called Exim had a hardwired
5901 limit of 10; it now uses the value of finduser_retries, which is used for
5902 all other user lookups.
5904 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5907 33. Exim was neglecting to zero errno before one call of strtol() when
5908 expanding a string and expecting an integer value. On some systems this
5909 resulted in spurious "integer overflow" errors. Also, it was casting the
5910 result into an int without checking.
5912 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5913 did not work. The code looks as if it has *never* worked, though it appears
5914 to have been documented since at least release 1.62. I have made it work.
5916 35. The "timeout_DNS" error in retry rules, also documented since at least
5917 1.62, also never worked. As it isn't clear exactly what this means, and
5918 clearly it isn't a major issue, I have abolished the feature by treating it
5919 as "timeout", and writing a warning to the main and panic logs.
5921 36. The display of retry rules for -brt wasn't always showing the error code
5924 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5925 timeout_connect_A, timeout_connect_MX.
5927 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5928 to the empty sender.
5930 39. The daemon was not analysing the content of -oX till after it had closed
5931 stderr and disconnected from the controlling terminal. This meant that any
5932 syntax errors were only noted on the panic log, and the return code from
5933 the command was 0. By re-arranging the code a little, I've made the
5934 decoding happen first, so such errors now appear on stderr, and the return
5935 code is 1. However, the actual setting up of the sockets still happens in
5936 the disconnected process, so errors there are still only recorded on the
5939 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5940 connections (as happens on some IP stacks) was logged at start up time as
5941 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5942 it from "IPv6 and IPv4", which means that two separate sockets are being
5945 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5946 says whether AF_INET or AF_INET6 was passed as an argument.
5948 42. Exiwhat output was messed up when time zones were included in log
5951 43. Exiwhat now gives more information about the daemon's listening ports,
5952 and whether -tls-on-connect was used.
5954 44. The "port" option of the smtp transport is now expanded.
5956 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5957 silently ignored. Now an error message is written to the main and panic
5960 46. There's a new ACL modifier called "logwrite" which writes to a log file
5961 as soon as it is encountered.
5963 47. Added $local_user_uid and $local_user_gid at routing time.
5965 48. Exim crashed when trying to verify a sender address that was being
5968 49. Exim was recognizing only a space character after ".include". It now also
5969 recognizes a tab character.
5971 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5972 extracting the relevant information from the specification. The man page no
5973 longer contains scrambled data for the -d option, and I've added a section
5974 at the front about calling Exim under different names.
5976 51. Added "extra_headers" argument to the "mail" command in filter files.
5978 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5981 53. Installed eximstats 1.29.
5983 54. Added transport_filter_timeout as a generic transport option.
5985 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5986 Cc: header lines. This was required by RFC 822, but it not required by RFC
5989 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5990 incoming messages that did not have them. Now it does so only if the
5991 message originates locally, that is, if there is no associated remote host
5992 address. When Resent- header lines are present, this applies to the Resent-
5993 lines rather than the non-Resent- lines.
5995 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5996 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5998 58. Messages for configuration errors now include the name of the main
5999 configuration file - useful now that there may be more than one file in a
6000 list (.included file names were always shown).
6002 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6003 for those rare installations that do not start the daemon as root or run it
6004 setuid root. I've cut out the call to initgroups() if the daemon is not
6007 60. The Exim user and group can now be bound into the binary as text strings
6008 that are looked up at the start of Exim's processing.
6010 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6012 62. Added $mailstore_basename variable.
6014 63. Installed patch to sieve.c from Michael Haardt.
6016 64. When Exim failed to open the panic log after failing to open the main log,
6017 the original message it was trying to log was written to stderr and debug
6018 output, but if they were not available (the usual case in production), it
6019 was lost. Now it is written to syslog before the two lines that record the
6020 failures to open the logs.
6022 65. Users' Exim filters run in subprocesses under the user's uid. It is
6023 possible for a "deliver" command or an alias in a "personal" command to
6024 provoke an address rewrite. If logging of address rewriting is configured,
6025 this fails because the process is not running as root or exim. There may be
6026 a better way of dealing with this, but for the moment (because 4.30 needs
6027 to be released), I have disabled address rewrite logging when running a
6028 filter in a non-root, non-exim process.
6034 1. The buildconfig auxiliary program wasn't quoting the value set for
6035 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6036 not defined. This bug was masked in 4.22 by the effect that was fixed in
6039 2. Some messages that were rejected after a message id was allocated were
6040 shown as "incomplete" by exigrep. It no longer does this for messages that
6041 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6043 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6044 have allow_domain_literals set, the ID did not get logged in the <= line.
6045 Domain literals are now always recognized in Message-ID: header lines.
6047 4. The first argument for a ${extract expansion item is the key name or field
6048 number. Leading and trailing spaces in this item were not being ignored,
6049 causing some misleading effects.
6051 5. When deliver_drop_privilege was set, single queue runner processes started
6052 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6053 same command in the process it spins off) were not dropping privilege.
6055 6. When the daemon running as "exim" started a queue runner, it always
6056 re-executed Exim in the spun-off process. This is a waste of effort when
6057 deliver_drop_privilege is set. The new process now just calls the
6058 queue-runner function directly.
6064 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6067 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6068 ignored. Though the use of -oP was forcing the writing of a pid file, it
6069 was always written to the default place.
6071 3. If the message "no IP address found for host xxxx" is generated during
6072 incoming verification, it is now followed by identification of the incoming
6073 connection (so you can more easily find what provoked it).
6075 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6077 5. Added some features to "harden" Exim a bit more against certain attacks:
6079 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6080 be put in Local/Makefile. This is like the never_users runtime option,
6081 but it cannot be overridden. The default setting is "root".
6083 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6084 prefix string with which any file named in a -C command line option
6087 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6088 is retained for -C and -D only if the caller of Exim is root. Without
6089 it, the exim user may also use -C and -D and retain privilege.
6091 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6092 command line option is disabled.
6094 6. Macro names set by the -D option must start with an upper case letter, just
6095 like macro names defined in the configuration file.
6097 7. Added "dereference=" facility to LDAP.
6099 8. Two instances of the typo "uknown" in the source files are fixed.
6101 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6102 the Configure-Makefile script screwed up while processing it.
6104 10. Incorporated PCRE 4.4.
6106 11. The SMTP synchronization check was not operating right at the start of an
6107 SMTP session. For example, it could not catch a HELO sent before the client
6108 waited for the greeting. There is now a check for outstanding input at the
6109 point when the greeting is written. Because of the duplex, asynchronous
6110 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6111 way, but not yet received, when the check is performed.
6113 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6114 on TCP/IP sockets, because this apparently causes some broken clients to
6117 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6118 unchanged) from the Cygwin maintainer.
6120 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6121 was supported instead of testing for mbx. Effectively a typo.
6123 15. The spa authenticator server code was not checking that the input it
6124 received was valid base64.
6126 16. The debug output line for the "set" modifier in ACLs was not showing the
6127 name of the variable that was being set.
6129 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6131 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6132 lookup on its own. Something else was needed to trigger the lookup. For
6133 example, a match in host_lookup or the need for a host name in a host list.
6134 Now, if $sender_host_name is referenced and the host name has not yet been
6135 looked up, a lookup is performed. If the lookup fails, the variable remains
6136 empty, and $host_lookup_failed is set to "1".
6138 19. Added "eqi" as a case-independent comparison operator.
6140 20. The saslauthd authentication condition could segfault if neither service
6141 nor realm was specified.
6143 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6144 error message that was logged was misleading, and incoming SMTP
6145 connections were dropped. The message is now more accurate, and temporary
6146 errors are given to SMTP connections.
6148 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6149 (except RSET) with a 421 error, until QUIT is received. However, it was
6150 failing to send a response to QUIT.
6152 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6153 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6154 ACL is now run after the helo_try_verify_hosts code.
6156 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6157 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6158 case-independent, so other case variants are also recognized). Apparently
6159 some systems use these upper case variants.
6161 25. If more than two messages were waiting for the same host, and a transport
6162 filter was specified for the transport, Exim sent two messages over the
6163 same TCP/IP connection, and then failed with "socket operation on non-
6164 socket" when it tried to send the third.
6166 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6168 27. The extern definition of crypt16() in expand.c was not being excluded when
6169 the OS had its own crypt16() function.
6171 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6172 as a preferred synonym for return_size_limit, both as an option and as an
6175 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6177 30. Changed the default configuration ACL to relax the local part checking rule
6178 for addresses that are not in any local domains. For these addresses,
6179 slashes and pipe symbols are allowed within local parts, but the sequence
6180 /../ is explicitly forbidden.
6182 31. SPA server authentication was not clearing the challenge buffer before
6185 32. log_message in a "warn" ACL statement was writing to the reject log as
6186 well as to the main log, which contradicts the documentation and doesn't
6187 seem right (because no rejection is happening). So I have stopped it.
6189 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6190 However, I am unable to do any testing of this.
6192 34. Fixed an infelicity in the appendfile transport. When checking directories
6193 for a mailbox, to see if any needed to be created, it was accidentally
6194 using path names with one or more superfluous leading slashes; tracing
6195 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6197 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6198 contents are added to the log line that is written for every discarded
6199 recipient. (Previously a log_message setting was ignored.)
6201 36. The ${quote: operator now quotes the string if it is empty.
6203 37. The install script runs exim in order to find its version number. If for
6204 some reason other than non-existence or emptiness, which it checks, it
6205 could not run './exim', it was installing it with an empty version number,
6206 i.e. as "exim-". This error state is now caught, and the installation is
6209 38. An argument was missing from the function that creates an error message
6210 when Exim fails to connect to the socket for saslauthd authentication.
6211 This could cause Exim to crash, or give a corrupted message.
6213 39. Added isip, isip4, and isip6 to ${if conditions.
6215 40. The ACL variables $acl_xx are now saved with the message, and can be
6216 accessed later in routers, transports, and filters.
6218 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6219 strings in the file are not string-expanded.
6221 42. If a MAIL command specified a SIZE value that was too large to fit into an
6222 int variable, the check against message_size_limit failed. Such values are
6223 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6224 day this will have to be increased, but I don't think I want to be around
6225 when emails are that large.
6232 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6233 iconv() is not standard in FreeBSD.
6235 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6236 IPv6 enabled. The observed symptom was a segmentation fault on return from
6237 the function os_common_find_running_interfaces() in src/os.c.
6239 3. In the check_special_case() function in daemon.c I had used "errno" as an
6240 argument name, which causes warnings on some systems. This was basically a
6241 typo, since it was named "eno" in the comments!
6243 4. The code that waits for the clock to tick (at a resolution of some fraction
6244 of a second) so as to ensure message-id uniqueness was always waiting for
6245 at least one whole tick, when it could have waited for less. [This is
6246 almost certainly not relevant at current processor speeds, where it is
6247 unlikely to ever wait at all. But we try to future-proof.]
6249 5. The function that sleeps for a time interval that includes fractions of a
6250 second contained a race. It did not block SIGALRM between setting the
6251 timer, and suspending (a couple of lines later). If the interval was short
6252 and the sigsuspend() was delayed until after it had expired, the suspension
6253 never ended. On busy systems this could lead to processes getting stuck for
6256 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6257 process, before it forks any delivery processes. The open lookup caching
6258 mechanism meant that the open file or database connection was passed into
6259 the delivery process. The problem was that delivery processes always tidy
6260 up cached lookup data. This could cause a problem for the next delivery
6261 process started by the queue runner, because the external queue runner
6262 process does not know about the closure. So the next delivery process
6263 still has data in the lookup cache. In the case of a file lookup, there was
6264 no problem because closing a file descriptor in a subprocess doesn't affect
6265 the parent. However, if the lookup was caching a connection to a database,
6266 the connection was closed, and the second delivery process was likely to
6267 see errors such as "PGSQL: query failed: server closed the connection
6268 unexpectedly". The problem has been fixed by closing all cached lookups
6269 in a queue runner before running a delivery process.
6271 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6272 seem to have the "const" qualifier which it has on other OS. I've
6275 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6276 *of the same type* (client or server) with the same public name that an
6277 error should be diagnosed.
6279 9. When Exim looked up a host name for an IP address, but failed to find the
6280 original IP address when looking up the host name (a safety check), it
6281 output the message "<ip address> does not match any IP for NULL", which was
6282 confusing, to say the least. The bug was that the host name should have
6283 appeared instead of "NULL".
6285 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6286 user that is built into the binary, and the -C or -D options is used, root
6287 privilege is dropped before the configuration file is read. In addition,
6288 logging is switched to stderr instead of the normal log files. If the
6289 configuration then re-defines the Exim user, the unprivileged environment
6290 is probably not what is expected, so Exim logs a panic warning message (but
6293 However, if deliver_drop_privilege is set, the unprivileged state may well
6294 be exactly what is intended, so the warning has been cut out in that case,
6295 and Exim is allowed to try to write to its normal log files.
6301 1. smtp_return_error_details was not giving details for temporary sender
6302 or receiver verification errors.
6304 2. Diagnose a configuration error if two authenticators have the same public
6307 3. Exim used not to create the message log file for a message until the first
6308 delivery attempt. This could be confusing when incoming messages were held
6309 for policy or load reasons. The message log file is now created at the time
6310 the message is received, and an initial "Received" line is written to it.
6312 4. The automatically generated man page for command line options had a minor
6313 bug that caused no ill effects; however, a more serious problem was that
6314 the procedure for building the man page automatically didn't always
6315 operate. Consequently, release 4.20 contains an out-of-date version. This
6316 shouldn't happen again.
6318 5. When building Exim with embedded Perl support, the script that builds the
6319 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6320 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6322 6. The freeze_tell option was not being used for messages that were frozen on
6323 arrival, either by an ACL or by local_scan().
6325 7. Added the smtp_incomplete_transaction log selector.
6327 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6328 was accepting AUTH without a new EHLO.
6330 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6331 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6334 10. Logging of TCP/IP connections (when configured) now happens in the main
6335 daemon process instead of the child process, so that the TCP/IP connection
6336 count is more accurate (but it can never be perfect).
6338 11. The use of "drop" in a nested ACL was not being handled correctly in the
6339 outer ACL. Now, if condition failure induced by the nested "drop" causes
6340 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6341 or "require"), the connection is dropped.
6343 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6344 that yield "discard" can now be used with an "accept" or a "discard" verb,
6345 but an error is generated for any others (because I can't see a useful way
6346 to define what should happen).
6348 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6349 are now processed in the same way as lines in the Exim configuration file.
6350 In particular, continuation lines are supported.
6352 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6354 15. Added -ti meaning -t -i.
6356 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6357 domains, and warn by logging if others are found.
6359 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6360 returned from GIFCONF when Exim is trying to find the list of interfaces on
6361 a host. The code in os.c has been modified to copy each ifreq to an aligned
6362 structure in all cases.
6364 Also, in some cases, the returned ifreq's were being copied to a 'struct
6365 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6366 means the last couple of bytes of an IPv6 address could be chopped if the
6367 ifreq contained only a normal sockaddr (14 bytes storage).
6369 18. Named domain lists were not supported in the hosts_treat_as_local option.
6370 An entry such as +xxxx was not recognized, and was treated as a literal
6373 19. Ensure that header lines added by a DATA ACL are included in the reject log
6374 if the ACL subsequently rejects the message.
6376 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6377 MD5 (which is deprecated).
6379 21. When testing a filter file using -bf, Exim was writing a message when it
6380 took the sender from a "From " line in the message, but it was not doing so
6381 when it took $return_path from a Return-Path: header line. It now does.
6383 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6384 with a valid header line field name (a series of printing characters
6385 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6387 23. Changed "disc" in the source to "disk" to conform to the documentation and
6388 the book and for uniformity.
6390 24. Ignore Sendmail's -Ooption=value command line item.
6392 25. When execve() failed while trying to run a command in a pipe transport,
6393 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6394 could be confused with a return value of 69 from the command itself. This
6395 has been changed to 127, the value the shell returns if it is asked to run
6396 a non-existent command. The wording for the related log line suggests a
6397 non-existent command as the problem.
6399 26. If received_header_text expands to an empty string, do not add a Received:
6400 header line to the message. (Well, it adds a token one on the spool, but
6401 marks it "old" so that it doesn't get used or transmitted.)
6403 27. Installed eximstats 1.28 (addition of -nt option).
6405 28. There was no check for failure on the call to getsockname() in the daemon
6406 code. This can fail if there is a shortage of resources on the system, with
6407 ENOMEM, for example. A temporary error is now given on failure.
6409 29. Contrary to the C standard, it seems that in some environments, the
6410 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6411 program. Exim now does this explicitly; it affects the formatting of
6412 timestamps using strftime().
6414 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6415 complaints. I've now initialized all the variables, to avoid this.
6417 32. Header lines added by a system filter were not being "seen" during
6418 transport-time rewrites.
6420 33. The info_callback() function passed to OpenSSL is set up with type void
6421 (*)(SSL *, int, int), as described somewhere. However, when calling the
6422 function (actually a macro) that sets it up, the type void(*)() is
6423 expected. I've put in a cast to prevent warnings from picky compilers.
6425 34. If a DNS black list lookup found a CNAME record, but there were no A
6426 records associated with the domain it pointed at, Exim crashed.
6428 35. If a DNS black list lookup returned more than one A record, Exim ignored
6429 all but the first. It now scans all returned addresses if a particular IP
6430 value is being sought. In this situation, the contents of the
6431 $dnslist_value variable are a list of all the addresses, separated by a
6434 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6435 to accept a host name and all its aliases if the forward lookup for any of
6436 them yielded the IP address of the incoming connection. Now it accepts only
6437 those names whose forward lookup yields the correct IP address. Any other
6438 names are discarded. This closes a loophole whereby a rogue DNS
6439 administrator could create reverse DNS records to break through a
6440 wildcarded host restriction in an ACL.
6442 37. If a user filter or a system filter that ran in a subprocess used any of
6443 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6444 the wrong values were passed to the pipe command ($thisaddress had the
6445 value of $0, $0 had the value of $1, etc). This bug was introduced by
6446 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6448 38. Improved the line breaking for long SMTP error messages from ACLs.
6449 Previously, if there was no break point between 40 and 75 characters, Exim
6450 left the rest of the message alone. Two changes have been made: (a) I've
6451 reduced the minimum length to 35 characters; (b) if it can't find a break
6452 point between 35 and 75 characters, it looks ahead and uses the first one
6453 that it finds. This may give the occasional overlong line, but at least the
6454 remaining text gets split now.
6456 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6457 file descriptors might be low, and that setting 1000 would always raise it.
6458 It turns out that in some environments, the limit is already over 1000 and
6459 that lowering it causes trouble. So now Exim takes care not to decrease it.
6461 40. When delivering a message, the value of $return_path is set to $sender_
6462 address at the start of routing (routers may change the value). By an
6463 oversight, this default was not being set up when an address was tested by
6464 -bt or -bv, which affected the outcome if any router or filter referred to
6467 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6468 log only when "message" or "log_message" are set. However, if one of the
6469 conditions was an address verification, or a call to a nested ACL, the
6470 messages generated by the underlying test were being passed through. This
6471 no longer happens. The underlying message is available in $acl_verify_
6472 message for both "message" and "log_message" expansions, so it can be
6473 passed through if needed.
6475 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6476 new expansion $bh_ to give the encoded byte string without charset
6477 translation. Translation happens only if iconv() is available; HAVE_ICONV
6478 indicates this at build time. HEADERS_CHARSET gives the charset to
6479 translate to; headers_charset can change it in the configuration, and
6480 "headers charset" can change it in an individual filter file.
6482 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6483 that creates RFC 2047 encoded "words" labels them as that charset instead
6484 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6485 expansion operator; (ii) when Exim creates a From: line for a local
6486 message; (iii) when a header line is rewritten to include a "phrase" part.
6488 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6489 buggy, causing it to skip the first lines of messages whose message ID
6490 ended in 'D'. This would not have bitten before Exim release 4.14, because
6491 message IDs were unlikely to end in 'D' before then. The effect was to have
6492 incorrect size information for certain domains.
6494 45. #include "config.h" was missing at the start of the crypt16.c module. This
6495 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6498 46. If there was a timeout during a "random" callout check, Exim treated it as
6499 a failure of the random address, and carried on sending RSET and the real
6500 address. If the delay was just some slowness somewhere, the response to the
6501 original RCPT would be taken as a response to RSET and so on, causing
6502 mayhem of various kinds.
6504 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6505 when I implemented it. It didn't allow for the fact that some option values
6506 may legitimately be negative (e.g. size_addition), and it didn't even do
6507 the right test for positive values.
6509 48. Domain names in DNS records are case-independent. Exim always looks them up
6510 in lower case. Some resolvers return domain names in exactly the case they
6511 appear in the zone file, that is, they may contain uppercase letters. Not
6512 all resolvers do this - some return always lower case. Exim was treating a
6513 change of case by a resolver as a change of domain, similar to a widening
6514 of a domain abbreviation. This triggered its re-routing code and so it was
6515 trying to route what was effectively the same domain again. This normally
6516 caused routing to fail (because the router wouldn't handle the domain
6517 twice). Now Exim checks for this case specially, and just changes the
6518 casing of the domain that it ultimately uses when it transmits the message
6521 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6524 50. If a filter generated a file delivery with a non-absolute name (possible if
6525 no home directory exists for the router), the forbid_file option was not
6528 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6529 the existing equality matching.
6531 52. Exim was using ints instead of ino_t variables in some places where it was
6532 dealing with inode numbers.
6534 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6535 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6536 called TMPDIR, and if it finds it is different, it changes its value.
6538 54. The smtp_printf() function is now made available to local_scan() so
6539 additional output lines can be written before returning. There is also an
6540 smtp_fflush() function to enable the detection of a dropped connection.
6541 The variables smtp_input and smtp_batched_input are exported to
6544 55. Changed the default runtime configuration: the message "Unknown user"
6545 has been removed from the ACL, and instead placed on the localuser router,
6546 using the cannot_route_message feature. This means that any verification
6547 failures that generate their own messages won't get overridden. Similarly,
6548 the "Unrouteable address" message that was in the ACL for unverifiable
6549 relay addresses has also been removed.
6551 56. Added hosts_avoid_esmtp to the smtp transport.
6553 57. The exicyclog script was not checking for the esoteric option
6554 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6555 will work only if exicyclog is run under the appropriate euid.
6557 58. Following a discussion on the list, the rules by which Exim recognises line
6558 endings on incoming messages have been changed. The -dropcr and drop_cr
6559 options are now no-ops, retained only for backwards compatibility. The
6560 following line terminators are recognized: LF CRLF CR. However, special
6561 processing applies to CR:
6563 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6564 nor a local message in the state where . is a terminator.
6566 (ii) If a bare CR is encountered in a header line, an extra space is added
6567 after the line terminator so as not to end the header. The reasoning
6568 behind this is that bare CRs in header lines are most likely either
6569 to be mistakes, or people trying to play silly games.
6571 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6572 was being incorrectly given as 18 bytes larger than it should have been.
6573 This is a VOB (very old bug).
6575 60. This may never have affected anything current, but just in case it has:
6576 When the local host is found other than at the start of a list of hosts,
6577 the local host, those with the same MX, and any that follow, are discarded.
6578 When the list in question was part of a longer list of hosts, the following
6579 hosts (not currently being processed) were also being discarded. This no
6580 longer happens. I'm not sure if this situation could ever has previously
6583 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6586 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6587 header. This is something that is recommended in a new Internet Draft, and
6588 is something that is documented as being done by Sendmail. There are two
6589 possible values. For messages generated by the autoreply transport, Exim
6592 Auto-Submitted: auto-replied
6594 whereas for all other generated messages (e.g. bounces) it adds
6596 Auto-Submitted: auto-generated
6598 63. The "personal" condition in filters now includes a test for the
6599 Auto-Submitted: header. If it contains the string "auto-" the message it
6600 not considered personal.
6602 64. Added rcpt_include_affixes as a generic transport option.
6604 65. Added queue_only_override (default true).
6606 66. Added the syslog_duplication option.
6608 67. If what should have been the first header line of a message consisted of
6609 a space followed by a colon, Exim was mis-interpreting it as a header line.
6610 It isn't of course - it is syntactically invalid and should therefore be
6611 treated as the start of the message body. The misbehaviour could have
6612 caused a number of strange effects, including loss of data in subsequent
6613 header lines, and spool format errors.
6615 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6616 client host had authenticated. This control can now be exercised by an ACL
6617 for more flexibility.
6619 69. By default, callouts do not happen when testing with -bh. There is now a
6620 variant, -bhc, which does actually run the callout code, including
6621 consulting and updating the callout cache.
6623 70. Added support for saslauthd authentication, courtesy of Alexander
6626 71. If statvfs() failed on the spool or log directories while checking their
6627 size for availability, Exim confusingly gave the error "space shortage".
6628 Furthermore, in debugging mode it crashed with a floating point exception.
6629 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6630 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6631 serious problem, Exim now writes to the main and panic logs when this
6632 happens, with details of the failure. It then refuses to accept the
6633 incoming message, giving the message "spool directory problem" or "log
6634 directory problem" with a 421 code for SMTP messages.
6636 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6637 0, 1, and 2 exist, because some OS complain for execs without them (see
6638 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6639 descriptors. However, the code omitted to check that the open succeeded,
6640 causing mysterious errors if for some reason the permissions on /dev/null
6641 got screwed. Now Exim writes a message to the main and panic logs, and
6642 bombs out if it can't open /dev/null.
6644 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6645 interact so that it is all more flexible. It is supposed to remain
6646 backwards compatible. Also added extra_local_interfaces.
6648 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6649 to bomb out with an assertion failure - to the client this appears as a
6650 connection drop. This problem occurs in the part of the code that was taken
6651 from the Samba project. Fortunately, the assertion is in a very simple
6652 function, so I have fixed this by reproducing the function inline in the
6653 one place where it is called, and arranging for authentication to fail
6654 instead of killing the process with assert().
6656 75. The SPA client code was not working when the server requested OEM rather
6657 than Unicode encoding.
6659 76. Added code to make require_files with a specific uid setting more usable in
6660 the case where statting the file as root fails - usually a non-root-mounted
6661 NFS file system. When this happens and the failure is EACCES, Exim now
6662 forks a subprocess and does the per-uid checking as the relevant uid.
6664 77. Added process_log_path.
6666 78. If log_file_path was not explicitly set, a setting of check_log_space or
6667 check_log_inodes was ignored.
6669 79. If a space check for the spool or log partitions fails, the incident is now
6670 logged. Of course, in the latter case the data may get lost...
6672 80. Added the %p formatting code to string_format() so that it can be used to
6673 print addresses in debug_print(). Adjusted all the address printing in the
6674 debugging in store.c to use %p rather than %d.
6676 81. There was a concern that a line of code in smtp_in.c could overflow a
6677 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6678 initially expressed, the concern was not well-founded, because trailing
6679 spaces are removed early. However, if the trailing spaces were followed by
6680 a NULL, they did not get removed, so the overflow was possible. Two fixes
6683 (a) I re-wrote the offending code in a cleaner fashion.
6684 (b) If an incoming SMTP command contains a NULL character, it is rejected
6687 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6688 runs initgroups(), so that if the Exim user is in any additional groups,
6689 they will be used during message reception.
6695 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob