1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
12 than -2003 did; needs libidn2 in addition to libidn.
14 JH/02 The path option on a pipe transport is now expanded before use.
16 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
17 Patch provided by "Björn", documentation fix added too.
19 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
20 missing a wire-to-host endian conversion.
22 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
23 close after a BDAT command line could be taken as a following command,
24 giving a synch failure. Fix by only checking for synch immediately
25 before acknowledging the chunk.
27 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
28 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
29 macro. Patches provided by Josh Soref.
31 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
32 Previously we did not; the RFC seems ambiguous and VRFY is not listed
33 by IANA as a service extension. However, John Klensin suggests that we
36 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
37 the dkim code may be unix-mode line endings rather than smtp wire-format
38 CRLF, so prepend a CR to any bare LF.
40 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
41 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
43 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
44 the first were themselves being wrongly included in the feed into dkim
45 processing; with most chunk sizes in use this resulted in an incorrect
46 body hash calculated value.
48 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
49 DKIM signature block, for verification. Although advised against by
50 standards it is specifically not ruled illegal.
52 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
54 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
55 missing a body hash (the bh= tag).
57 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
58 It seems that HAProxy sends the Proxy Protocol information in clear and
59 only then does a TLS startup, so do the same.
61 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
62 TCP connections (such as for Spamd) unless the daemon successfully set
63 Fast Open mode on its listening sockets. This fixes breakage seen on
64 too-old kernels or those not configured for Fast Open, at the cost of
65 requiring both directions being enabled for TFO, and TFO never being used
66 by non-daemon-related Exim processes.
68 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
69 endings, at least on the first header line. Try to canonify any that get
70 past that check, despite the cost.
72 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
73 now limited to an arbitrary five deep, while parsing addresses with the
74 strip_excess_angle_brackets option enabled.
76 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
77 instead leave the unprompted TLS handshake in socket buffer for the
78 TLS library to consume.
80 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
82 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
88 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
89 supports it and a size is available (ie. the sending peer gave us one).
91 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
92 years of being deprecated). The replacements are the ACLs
93 acl_smtp_mime and acl_not_smtp_mime.
95 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
96 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
97 or even an in-clear connection were permitted. Now, if the host lookup
98 was dnssec and dane was requested then the host is only used if the
99 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
100 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
101 if one fails this test.
102 This means that a poorly-configured remote DNS will make it incommunicado;
103 but it protects against a DNS-interception attack on it.
105 JH/04 Bug 1810: make continued-use of an open smtp transport connection
106 non-noisy when a race steals the message being considered.
108 JH/05 If main configuration option tls_certificate is unset, generate a
109 self-signed certificate for inbound TLS connections.
111 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
112 in rewrites and routers.
114 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
115 and logged a warning sing 4.83; now they are a configuration file error.
117 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
118 (lacking @domain). Apply the same qualification processing as RCPT.
120 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
122 JH/10 Support ${sha256:} applied to a string (as well as the previous
125 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
126 a cutthrough deliver is pending, as we always want to make a connection.
127 This also avoids re-routing the message when later placing the cutthrough
128 connection after a verify cache hit.
129 Do not update it with the verify result either.
131 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
132 when routing results in more than one destination address.
134 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
135 signing (which inhibits the cutthrough capability). Previously only
136 the presence of an option was tested; now an expansion evaluating as
137 empty is permissible (obviously it should depend only on data available
138 when the cutthrough connection is made).
140 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
141 the relevant preceding SMTP command did not note the pipelining mode.
143 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
144 Previously they were not counted.
146 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
147 as one having no matching records. Previously we deferred the message
148 that needed the lookup.
150 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
151 distinguished as "(=".
153 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
154 for missing MX records. Previously it only worked for missing A records.
156 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
158 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
159 after the data-go-ahead and data-ack. Patch from Jason Betts.
161 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
162 even for a "none" policy. Patch from Tony Meyer.
164 JH/22 Fix continued use of a connection for further deliveries. If a port was
165 specified by a router, it must also match for the delivery to be
168 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
169 When one of the recipients of a message was unsuitable for the connection
170 (has no matching addresses), we lost track of needing to mark it
171 deferred. As a result mail would be lost.
173 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
175 JH/25 Decoding ACL controls is now done using a binary search; the source code
176 takes up less space and should be simpler to maintain. Merge the ACL
177 condition decode tables also, with similar effect.
179 JH/26 Fix problem with one_time used on a redirect router which returned the
180 parent address unchanged. A retry would see the parent address marked as
181 delivered, so not attempt the (identical) child. As a result mail would
184 JH/27 Fix a possible security hole, wherein a process operating with the Exim
185 UID can gain a root shell. Credit to http://www.halfdog.net/ for
186 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
189 JH/28 Enable {spool,log} filesystem space and inode checks as default.
190 Main config options check_{log,spool}_{inodes,space} are now
191 100 inodes, 10MB unless set otherwise in the configuration.
193 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
194 Previously it only applied to the main-section connection policy
197 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
199 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
200 by me. Added RFC7919 DH primes as an alternative.
202 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
203 are not in the system include path.
205 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
206 GnuTLS, when a session startup failed (eg because the client disconnected)
207 Exim did stdio operations after fclose. This was exposed by a recent
208 change which nulled out the file handle after the fclose.
210 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
211 signed directly by the cert-signing cert, rather than an intermediate
212 OCSP-signing cert. This is the model used by LetsEncrypt.
214 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
216 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
217 an incoming connection.
219 HS/02 Bug 1802: Do not half-close the connection after sending a request
222 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
223 fallback to "prime256v1".
225 JH/34 SECURITY: Use proper copy of DATA command in error message.
226 Could leak key material. Remotely exploitable. CVE-2016-9963.
232 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
233 and 3.4.4 - once the server is enabled to respond to an OCSP request
234 it does even when not requested, resulting in a stapling non-aware
235 client dropping the TLS connection.
237 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
238 support variable-length bit vectors. No functional change.
240 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
241 The I= interface field on outgoing lines is now after the H= remote
242 host field, same as incoming lines. There is a separate
243 outgoing_interface log selector which allows you to disable the
246 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
247 If not running log_selector +smtp_connection the mainlog would be held
248 open indefinitely after a "too many connections" event, including to a
249 deleted file after a log rotate. Leave the per net connection logging
250 leaving it open for efficiency as that will be quickly detected by the
251 check on the next write.
253 HS/01 Bug 1671: Fix post transport crash.
254 Processing the wait-<transport> messages could crash the delivery
255 process if the message IDs didn't exist for some reason. When
256 using 'split_spool_directory=yes' the construction of the spool
257 file name failed already, exposing the same netto behaviour.
259 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
260 mime_regex ACL conditions.
262 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
263 to DSN fail messages (bounces): remote IP, remote greeting, remote response
264 to HELO, local diagnostic string.
266 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
267 log line to debug. Even when configured with a tls authenticator many
268 client connections are expected to not authenticate in this way, so
269 an authenticate fail is not an error.
271 HS/02 Add the Exim version string to the process info. This way exiwhat
272 gives some more detail about the running daemon.
274 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
275 matter for fast-change records such as DNSBLs.
277 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
278 retry record, even if constant. There may be multiple transports with
279 different interface settings and the retry behaviour needs to be kept
282 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
284 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
286 JH/10 Bug 840: fix log_defer_output option of pipe transport
288 JH/11 Bug 830: use same host for all RCPTS of a message, even under
289 hosts_randomize. This matters a lot when combined with mua_wrapper.
291 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
292 ${quote_pgsql:<string>} operator.
294 JH/13 Bug 1708: avoid misaligned access in cached lookup.
296 JH/14 Change header file name for freeradius-client. Relevant if compiling
297 with Radius support; from the Gentoo tree and checked under Fedora.
299 JH/15 Bug 1712: Introduce $prdr_requested flag variable
301 JH/16 Bug 1714: Permit an empty string as expansion result for transport
302 option transport_filter, meaning no filtering.
304 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
306 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
307 defaults to "*" (all hosts). The variable is now available when not built
308 with TLS, default unset, mainly to enable keeping the testsuite sane.
309 If a server certificate is not supplied (via tls_certificate) an error is
310 logged, and clients will find TLS connections fail on startup. Presumably
311 they will retry in-clear.
312 Packagers of Exim are strongly encouraged to create a server certificate
313 at installation time.
315 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
316 with the $config_file variable.
318 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
319 in transport context, after the attempt, and per-recipient. The latter type
320 is per host attempted. The event data is the error message, and the errno
321 information encodes the lookup type (A vs. MX) used for the (first) host,
322 and the trailing two digits of the smtp 4xx response.
324 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
325 to write to mainlog (or rejectlog, paniclog) in the window between file
326 creation and permissions/ownership being changed. Particularly affects
327 installations where exicyclog is run as root, rather than exim user;
328 result is that the running daemon panics and dies.
330 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
332 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
333 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
334 "pri" and "weight". Note that the previous implicit priority given by the
335 list order is no longer honoured.
337 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
340 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
341 by defining SUPPORT_SOCKS.
343 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
344 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
345 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
346 variables are renamed to proxy_{local,external}_{address,port}.
348 JH/27 Move Internationalisation support from Experimental to mainline, enabled
349 for a build by defining SUPPORT_I18N
351 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
352 of the query string, and make ${quote_redis:} do that quoting.
354 JH/29 Move Events support from Experimental to mainline, enabled by default
355 and removable for a build by defining DISABLE_EVENT.
357 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
359 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
360 cached by the daemon.
362 JH/32 Move Redis support from Experimental to mainline, enabled for a build
363 by defining LOOKUP_REDIS. The libhiredis library is required.
365 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
366 keys are given for lookup.
368 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
369 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
370 only supported when built with TLS support. The PolarSSL SHA routines
371 are still used when the TLS library is too old for convenient support.
373 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
374 openssl_options), for security. OpenSSL forces this from version 1.1.0
375 server-side so match that on older versions.
377 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
378 allocation for $value could be released as the expansion processing
379 concluded, but leaving the global pointer active for it.
381 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
382 and to use the domains and local_parts ACL conditions.
384 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
385 incorrectly not doubled on cutthrough transmission, hence seen as a
386 body-termination at the receiving system - resulting in truncated mails.
387 Commonly the sender saw a TCP-level error, and retransmitted the message
388 via the normal store-and-forward channel. This could result in duplicates
389 received - but deduplicating mailstores were liable to retain only the
390 initial truncated version.
392 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
394 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
396 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
397 we're in there, support oversigning also; bug 1309.
399 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
401 HS/04 Add support for keep_environment and add_environment options.
403 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
404 either intentional arithmetic overflow during PRNG, or testing config-
407 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
408 delivery resulted in actual delivery. Cancel cutthrough before DATA
411 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
412 rejects a recipient: pass the reject to the originator.
414 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
415 Many were false-positives and ignorable, but it's worth fixing the
418 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
419 for the new environment-manipulation done at startup. Move the routines
420 from being local to tls.c to being global via the os.c file.
422 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
423 an extract embedded as result-arg for a map, the first arg for extract
424 is unavailable so we cannot tell if this is a numbered or keyed
425 extraction. Accept either.
431 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
434 JH/02 The smtp transport option "multi_domain" is now expanded.
436 JH/03 The smtp transport now requests PRDR by default, if the server offers
439 JH/04 Certificate name checking on server certificates, when exim is a client,
440 is now done by default. The transport option tls_verify_cert_hostnames
441 can be used to disable this per-host. The build option
442 EXPERIMENTAL_CERTNAMES is withdrawn.
444 JH/05 The value of the tls_verify_certificates smtp transport and main options
445 default to the word "system" to access the system default CA bundle.
446 For GnuTLS, only version 3.0.20 or later.
448 JH/06 Verification of the server certificate for a TLS connection is now tried
449 (but not required) by default. The verification status is now logged by
450 default, for both outbound TLS and client-certificate supplying inbound
453 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
456 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
457 Status Notification (bounce) messages are now MIME format per RFC 3464.
458 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
459 under the control of the dsn_advertise_hosts option, and routers may
460 have a dsn_lasthop option.
462 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
463 default, modifiable by a malware= option. The list separator for
464 the options can now be changed in the usual way. Bug 68.
466 JH/10 The smtp_receive_timeout main option is now expanded before use.
468 JH/11 The incoming_interface log option now also enables logging of the
469 local interface on delivery outgoing connections.
471 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
472 if the interface and destination host and port all match.
474 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
477 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
478 Patch from Andrew Lewis.
480 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
481 now supports optional time-restrictions, weighting, and priority
482 modifiers per server. Patch originally by <rommer@active.by>.
484 JH/16 The spamd_address main option now supports a mixed list of local
485 and remote servers. Remote servers can be IPv6 addresses, and
486 specify a port-range.
488 JH/17 Bug 68: The spamd_address main option now supports an optional
489 timeout value per server.
491 JH/18 Bug 1581: Router and transport options headers_add/remove can
492 now have the list separator specified.
494 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
497 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
500 JH/21 Support for the A6 type of dns record is withdrawn.
502 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
503 rather than the verbs used.
505 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
506 from 255 to 1024 chars.
508 JH/24 Verification callouts now attempt to use TLS by default.
510 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
511 are generic router options now. The defaults didn't change.
513 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
514 Original patch from Alexander Shikoff, worked over by JH.
516 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
517 files if ZCAT_COMMAND is not executable.
519 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
521 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
523 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
524 Normally benign, it bites when the pair was led to by a CNAME;
525 modern usage is to not canonicalize the domain to a CNAME target
526 (and we were inconsistent anyway for A-only vs AAAA+A).
528 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
530 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
531 when evaluating $sender_host_dnssec.
533 JH/31 Check the HELO verification lookup for DNSSEC, adding new
534 $sender_helo_dnssec variable.
536 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
538 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
540 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
542 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
543 documented as working, but never had. Support all but $spam_report.
545 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
546 added for tls authenticator.
548 HS/03 Add perl_taintmode main config option
554 TL/01 When running the test suite, the README says that variables such as
555 no_msglog_check are global and can be placed anywhere in a specific
556 test's script, however it was observed that placement needed to be near
557 the beginning for it to behave that way. Changed the runtest perl
558 script to read through the entire script once to detect and set these
559 variables, reset to the beginning of the script, and then run through
560 the script parsing/test process like normal.
562 TL/02 The BSD's have an arc4random API. One of the functions to induce
563 adding randomness was arc4random_stir(), but it has been removed in
564 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
565 function when detected.
567 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
568 cause callback expansion.
570 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
571 syntax errors in an expansion can be treated as a string instead of
572 logging or causing an error, due to the internal use of bool_lax
573 instead of bool when processing it.
575 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
576 server certificates when making smtp deliveries.
578 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
580 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
582 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
584 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
585 Merged patch from Sebastian Wiedenroth.
587 JH/05 Fix results-pipe from transport process. Several recipients, combined
588 with certificate use, exposed issues where response data items split
589 over buffer boundaries were not parsed properly. This eventually
590 resulted in duplicates being sent. This issue only became common enough
591 to notice due to the introduction of connection certificate information,
592 the item size being so much larger. Found and fixed by Wolfgang Breyha.
594 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
595 size buffer was used, resulting in syntax errors when an expansion
598 JH/07 Add support for directories of certificates when compiled with a GnuTLS
599 version 3.3.6 or later.
601 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
602 is EXPERIMENTAL_EVENT, the main-configuration and transport options
603 both become "event_action", the variables become $event_name, $event_data
604 and $event_defer_errno. There is a new variable $verify_mode, usable in
605 routers, transports and related events. The tls:cert event is now also
606 raised for inbound connections, if the main configuration event_action
609 TL/06 In test suite, disable OCSP for old versions of openssl which contained
610 early OCSP support, but no stapling (appears to be less than 1.0.0).
612 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
613 server certificate names available under the smtp transport option
614 "tls_verify_cert_hostname" now do not permit multi-component wildcard
617 JH/10 Time-related extraction expansions from certificates now use the main
618 option "timezone" setting for output formatting, and are consistent
619 between OpenSSL and GnuTLS compilations. Bug 1541.
621 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
622 encoded parameter in the incoming message. Bug 1558.
624 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
625 include certificate info, eximon was claiming there were spoolfile
628 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
630 JH/14 Log delivery-related information more consistently, using the sequence
631 "H=<name> [<ip>]" wherever possible.
633 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
634 are problematic for Debian distribution, omit them from the release
637 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
639 JH/16 Fix string representation of time values on 64bit time_t architectures.
642 JH/17 Fix a null-indirection in certextract expansions when a nondefault
643 output list separator was used.
648 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
649 checkers that were complaining about end of non-void function with no
652 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
653 This was a regression introduced in 4.83 by another bugfix.
655 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
657 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
658 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
664 TF/01 Correctly close the server side of TLS when forking for delivery.
666 When a message was received over SMTP with TLS, Exim failed to clear up
667 the incoming connection properly after forking off the child process to
668 deliver the message. In some situations the subsequent outgoing
669 delivery connection happened to have the same fd number as the incoming
670 connection previously had. Exim would try to use TLS and fail, logging
671 a "Bad file descriptor" error.
673 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
674 utilities have not been installed.
676 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
677 temporary space as the ACL may create new global variables.
679 TL/01 LDAP support uses per connection or global context settings, depending
680 upon the detected version of the libraries at build time.
682 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
683 to extract and use the src ip:port in logging and expansions as if it
684 were a direct connection from the outside internet. PPv2 support was
685 updated based on HAProxy spec change in May 2014.
687 JH/02 Add ${listextract {number}{list}{success}{fail}}.
689 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
690 Properly escape header and check for NULL return.
692 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
695 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
697 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
698 characters in header names, implemented as a verify condition.
699 Contributed by Michael Fischer v. Mollard.
701 TL/05 Rename SPF condition results err_perm and err_temp to standardized
702 results permerror and temperror. Previous values are deprecated but
703 still accepted. In a future release, err_perm and err_temp will be
704 completely removed, which will be a backward incompatibility if the
705 ACL tests for either of these two old results. Patch contributed by
706 user bes-internal on the mailing list.
708 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
710 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
711 selectors, in both main and reject logs.
713 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
716 JH/07 Add malware type "sock" for talking to simple daemon.
718 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
720 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
721 routers/transports under cutthrough routing.
723 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
724 numbers. Touch up "bool" conditional to keep the same definition.
726 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
728 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
730 JH/12 Expand items in router/transport headers_add or headers_remove lists
731 individually rather than the list as a whole. Bug 1452.
733 Required for reasonable handling of multiple headers_ options when
734 they may be empty; requires that headers_remove items with embedded
735 colons must have them doubled (or the list-separator changed).
737 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
738 view the policy declared in the DMARC record. Currently, $dmarc_status
739 is a combined value of both the record presence and the result of the
742 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
744 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
745 dnslookup router and the smtp transport (applying to the forward
748 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
749 of ldap servers used for a specific lookup. Patch provided by Heiko
752 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
753 New variable $lookup_dnssec_authenticated for observability.
755 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
756 Patch submitted by Lars Timman.
758 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
760 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
761 Requires trusted mode and valid format message id, aborts otherwise.
762 Patch contributed by Heiko Schlichting.
764 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
765 certextract with support for various fields. Bug 1358.
767 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
768 is requested by default, modifiable by smtp transport option
771 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
772 operate on certificate variables to give certificate fingerprints
773 Also new ${sha256:cert_variable}.
775 JH/23 The PRDR feature is moved from being Experimental into the mainline.
777 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
778 Christian Aistleitner.
780 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
782 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
783 file. Patch from Wolfgang Breyha.
785 JH/25 Expand the coverage of the delivery $host and $host_address to
786 client authenticators run in verify callout. Bug 1476.
788 JH/26 Port service names are now accepted for tls_on_connect_ports, to
789 align with daemon_smtp_ports. Bug 72.
791 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
792 support and error reporting did not work properly.
794 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
795 and is readable. Patch from Andrew Colin Kissa.
797 TL/14 Enhance documentation of ${run expansion and how it parses the
798 commandline after expansion, particularly in the case when an
799 unquoted variable expansion results in an empty value.
801 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
803 PP/02 Fix internal collision of T_APL on systems which support RFC3123
804 by renaming away from it. Addresses GH issue 15, reported by
807 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
809 TL/15 SECURITY: prevent double expansion in math comparison functions
810 (can expand unsanitized data). Not remotely exploitable.
817 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
819 PP/02 Make -n do something, by making it not do something.
820 When combined with -bP, the name of an option is not output.
822 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
825 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
826 $sender_host_name and config options to manage this, and basic check
829 PP/05 DSCP support for outbound connections and control modifier for inbound.
831 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
832 (Only plugin which currently uses this is kerberos4, which nobody should
833 be using, but we should make it available and other future plugins might
834 conceivably use it, even though it would break NAT; stuff *should* be
835 using channel bindings instead).
837 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
838 name; added for Sendmail compatibility; requires admin caller.
839 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
840 just ignore it); requires trusted caller.
841 Also parse but ignore: -Ac -Am -X<logfile>
844 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
846 TL/02 Add +smtp_confirmation as a default logging option.
848 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
849 Patch by Magnus Holmgren from 2007-02-20.
851 TL/04 Bugzilla 1281 - Spec typo.
852 Bugzilla 1283 - Spec typo.
853 Bugzilla 1290 - Spec grammar fixes.
855 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
857 TL/06 Add Experimental DMARC support using libopendmarc libraries.
859 TL/07 Fix an out of order global option causing a segfault. Reported to dev
860 mailing list by by Dmitry Isaikin.
862 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
864 JH/02 Support "G" suffix to numbers in ${if comparisons.
866 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
868 NM/01 Bugzilla 1197 - Spec typo
869 Bugzilla 1196 - Spec examples corrections
871 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
873 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
874 gnutls_enable_pkcs11, but renamed to more accurately indicate its
877 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
878 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
880 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
881 "acl {{name}{arg}...}", and optional args on acl condition
884 JH/05 Permit multiple router/transport headers_add/remove lines.
886 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
888 JH/07 Avoid using a waiting database for a single-message-only transport.
889 Performance patch from Paul Fisher. Bugzilla 1262.
891 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
894 JH/09 Add $headers_added variable, with content from use of ACL modifier
895 add_header (but not yet added to the message). Bugzilla 199.
897 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
898 Pulled from Bugzilla 817 by Wolfgang Breyha.
900 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
902 (nb: this is the same fix as in Exim 4.80.1)
904 JH/11 Add A= logging on delivery lines, and a client_set_id option on
907 JH/12 Add optional authenticated_sender logging to A= and a log_selector
910 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
912 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
913 advertise SMTP AUTH mechanism to us, instead of a generic
914 protocol violation error. Also, make Exim more robust to bad
915 data from the Dovecot auth socket.
917 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
919 When a queue runner is handling a message, Exim first routes the
920 recipient addresses, during which it prunes them based on the retry
921 hints database. After that it attempts to deliver the message to
922 any remaining recipients. It then updates the hints database using
925 So if a recipient address works intermittently, it can get repeatedly
926 deferred at routing time. The retry hints record remains fresh so the
927 address never reaches the final cutoff time.
929 This is a fairly common occurrence when a user is bumping up against
930 their storage quota. Exim had some logic in its local delivery code
931 to deal with this. However it did not apply to per-recipient defers
932 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
934 This change adds a proper retry rule check during routing so that the
935 final cutoff time is checked against the message's age. We only do
936 this check if there is an address retry record and there is not a
937 domain retry record; this implies that previous attempts to handle
938 the address had the retry_use_local_parts option turned on. We use
939 this as an approximation for the destination being like a local
940 delivery, as in LMTP.
942 I suspect this new check makes the old local delivery cutoff check
943 redundant, but I have not verified this so I left the code in place.
945 TF/02 Correct gecos expansion when From: is a prefix of the username.
947 Test 0254 submits a message to Exim with the header
951 When I ran the test suite under the user fanf2, Exim expanded
952 the header to contain my full name, whereas it should have added
953 a Resent-Sender: header. It erroneously treats any prefix of the
954 username as equal to the username.
956 This change corrects that bug.
958 GF/01 DCC debug and logging tidyup
959 Error conditions log to paniclog rather than rejectlog.
960 Debug lines prefixed by "DCC: " to remove any ambiguity.
962 TF/03 Avoid unnecessary rebuilds of lookup-related code.
964 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
965 Bug spotted by Jeremy Harris; was flawed since initial commit.
966 Would have resulted in OCSP responses post-SNI triggering an Exim
967 NULL dereference and crash.
969 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
971 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
972 Bug detection, analysis and fix by Samuel Thibault.
973 Bugzilla 1331, Debian bug #698092.
975 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
977 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
978 Server implementation by Todd Lyons, client by JH.
979 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
980 config variable "prdr_enable" controls whether the server
981 advertises the facility. If the client requests PRDR a new
982 acl_data_smtp_prdr ACL is called once for each recipient, after
983 the body content is received and before the acl_smtp_data ACL.
984 The client is controlled by both of: a hosts_try_prdr option
985 on the smtp transport, and the server advertisement.
986 Default client logging of deliveries and rejections involving
987 PRDR are flagged with the string "PRDR".
989 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
990 fclose(). Diagnosis by Todd Lyons.
992 PP/17 Update configure.default to handle IPv6 localhost better.
993 Patch by Alain Williams (plus minor tweaks).
996 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
997 This is now consistent with GnuTLS, and is now documented: the
998 previous undocumented portable approach to treating the option as
999 unset was to force an expansion failure. That still works, and
1000 an empty string is now equivalent.
1002 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1003 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1004 not performing validation itself.
1006 PP/20 Added force_command boolean option to pipe transport.
1007 Patch from Nick Koston, of cPanel Inc.
1009 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1012 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1014 PP/21 Fix eximon continuous updating with timestamped log-files.
1015 Broken in a format-string cleanup in 4.80, missed when I repaired the
1016 other false fix of the same issue.
1017 Report and fix from Heiko Schlichting.
1020 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1021 Report from Prashanth Katuri.
1023 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1024 It's SecureTransport, so affects any MacOS clients which use the
1025 system-integrated TLS libraries, including email clients.
1027 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1028 using a MIME ACL for non-SMTP local injection.
1029 Report and assistance in diagnosis by Warren Baker.
1031 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1033 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1035 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1036 last id that failed so it may be referenced in subsequent ACL's.
1038 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1041 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1042 ldap library initialization, allowing self-signed CA's to be
1043 used. Also properly sets require_cert option later in code by
1044 using NULL (global ldap config) instead of ldap handle (per
1045 session). Bug diagnosis and testing by alxgomz.
1047 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1048 the src/util/ subdirectory.
1050 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1051 renamed to Transport Post Delivery Action by Jeremy Harris, as
1054 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1055 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1056 redis_servers = needs to be configured which will be used by the redis
1057 lookup. Patch from Warren Baker, of The Packet Hub.
1059 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1061 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1062 hostname or reverse DNS when processing a host list. Used suggestions
1063 from multiple comments on this bug.
1065 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1067 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1068 Missed a few lines, added it to make the runtest require no keyboard
1071 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1072 contains upper case chars. Make router use caseful_local_part.
1074 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1075 support when GnuTLS has been built with p11-kit.
1081 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1083 This, or similar/improved, will also be change PP/11 of 4.82.
1089 PP/01 Handle short writes when writing local log-files.
1090 In practice, only affects FreeBSD (8 onwards).
1091 Bugzilla 1053, with thanks to Dmitry Isaikin.
1093 NM/01 Bugzilla 949 - Documentation tweak
1095 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1098 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1100 PP/02 Implemented gsasl authenticator.
1102 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1104 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1105 `pkg-config foo` for cflags/libs.
1107 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1108 with rest of GSASL and with heimdal_gssapi.
1110 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1111 `pkg-config foo` for cflags/libs for the TLS implementation.
1113 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1114 properties get this fed in as external SSF. A number of robustness
1115 and debugging improvements to the cyrus_sasl authenticator.
1117 PP/08 cyrus_sasl server now expands the server_realm option.
1119 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1120 Patch by Jeremy Harris.
1122 PP/10 Added dbmjz lookup type.
1124 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1126 PP/12 MAIL args handles TAB as well as SP, for better interop with
1127 non-compliant senders.
1128 Analysis and variant patch by Todd Lyons.
1130 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1131 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1132 Patch from Dirk Mueller <dmueller@suse.com>
1134 PP/13 tls_peerdn now print-escaped for spool files.
1135 Observed some $tls_peerdn in wild which contained \n, which resulted
1136 in spool file corruption.
1138 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1139 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1140 or write after TLS renegotiation, which otherwise led to messages
1143 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1144 as a tracking header (ie: a signed header comes before the signature).
1145 Patch from Wolfgang Breyha.
1147 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1148 comma-sep list; embedded commas doubled.
1150 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1152 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1154 Report and patch from Dmitry Banschikov.
1156 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1157 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1158 failures. We appear to now support TLS1.1+ with Exim.
1160 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1161 lets Exim select keys and certificates based upon TLS SNI from client.
1162 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1163 before an outbound SMTP session. New log_selector, +tls_sni.
1165 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1166 NULL dereference. Report and patch from Alun Jones.
1168 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1169 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1170 Not seeing resolver debug output on NetBSD, but suspect this is a
1171 resolver implementation change.
1173 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1174 Left warnings. Added "eximon gdb" invocation mode.
1176 PP/21 Defaulting "accept_8bitmime" to true, not false.
1178 PP/22 Added -bw for inetd wait mode support.
1180 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1181 locate the relevant includes and libraries. Made this the default.
1183 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1184 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1186 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1187 This may cause build issues on older platforms.
1189 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1190 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1191 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1192 Added SNI support via GnuTLS too.
1193 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1195 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1197 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1198 Applied second patch from Janne, implementing suggestion to default
1199 multiple-strings-in-record handling to match SPF spec.
1201 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1203 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1204 read-only, out of scope).
1205 Patch from Wolfgang Breyha, report from Stuart Northfield.
1207 PP/29 Fix three issues highlighted by clang analyser static analysis.
1208 Only crash-plausible issue would require the Cambridge-specific
1209 iplookup router and a misconfiguration.
1210 Report from Marcin Mirosław.
1212 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1214 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1215 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1216 As part of this, removing so much warning spew let me fix some minor
1217 real issues in debug logging.
1219 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1220 assignment on my part. Fixed.
1222 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1223 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1224 Janne Snabb (who went above and beyond: thank you).
1226 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1227 string otherwise requires a connection and a bunch more work and it's
1228 relatively easy to get wrong. Should also expose TLS library linkage
1231 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1232 64-bit ${eval} (JH/03).
1234 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1235 GNU libc to support some of the 64-bit stuff, should not lead to
1236 conflicts. Defined before os.h is pulled in, so if a given platform
1237 needs to override this, it can.
1239 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1240 protection layer was required, which is not implemented.
1241 Bugzilla 1254, patch from Wolfgang Breyha.
1243 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1244 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1245 tls_dhparam take prime identifiers. Also unbreak combination of
1246 OpenSSL+DH_params+TLSSNI.
1248 PP/39 Disable SSLv2 by default in OpenSSL support.
1254 PP/01 Solaris build fix for Oracle's LDAP libraries.
1255 Bugzilla 1109, patch from Stephen Usher.
1257 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1259 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1262 TF/02 Fix a couple more cases where we did not log the error message
1263 when unlink() failed. See also change 4.74-TF/03.
1265 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1266 lock up or crash if it happened to be inside a call to libc when it
1267 got a SIGUSR1 from exiwhat.
1269 The SIGUSR1 handler appends the current process status to the process
1270 log which is later printed by exiwhat. It used to use the general
1271 purpose logging code to do this, but several functions it calls are
1272 not safe for signals.
1274 The new output code in the SIGUSR1 handler is specific to the process
1275 log, and simple enough that it's easy to inspect for signal safety.
1276 Removing some special cases also simplifies the general logging code.
1277 Removing the spurious timestamps from the process log simplifies
1280 TF/04 Improved ratelimit ACL condition.
1282 The /noupdate option has been deprecated in favour of /readonly which
1283 has clearer semantics. The /leaky, /strict, and /readonly update modes
1284 are mutually exclusive. The update mode is no longer included in the
1285 database key; it just determines when the database is updated. (This
1286 means that when you upgrade Exim will forget old rate measurements.)
1288 Exim now checks that the per_* options are used with an update mode that
1289 makes sense for the current ACL. For example, when Exim is processing a
1290 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1291 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1292 must specify per_mail/readonly. If you omit the update mode it defaults to
1293 /leaky where that makes sense (as before) or /readonly where required.
1295 The /noupdate option is now undocumented but still supported for
1296 backwards compatibility. It is equivalent to /readonly except that in
1297 ACLs where /readonly is required you may specify /leaky/noupdate or
1298 /strict/noupdate which are treated the same as /readonly.
1300 A useful new feature is the /count= option. This is a generalization
1301 of the per_byte option, so that you can measure the throughput of other
1302 aggregate values. For example, the per_byte option is now equivalent
1303 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1305 The per_rcpt option has been generalized using the /count= mechanism
1306 (though it's more complicated than the per_byte equivalence). When it is
1307 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1308 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1309 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1310 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1311 Note that using per_rcpt with a non-readonly update mode in more than
1312 one ACL will cause the recipients to be double-counted. (The per_mail
1313 and per_byte options don't have this problem.)
1315 The handling of very low rates has changed slightly. If the computed rate
1316 is less than the event's count (usually one) then this event is the first
1317 after a long gap. In this case the rate is set to the same as this event's
1318 count, so that the first message of a spam run is counted properly.
1320 The major new feature is a mechanism for counting the rate of unique
1321 events. The new per_addr option counts the number of different
1322 recipients that someone has sent messages to in the last time period. It
1323 behaves like per_rcpt if all the recipient addresses are different, but
1324 duplicate recipient addresses do not increase the measured rate. Like
1325 the /count= option this is a general mechanism, so the per_addr option
1326 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1327 example, measure the rate that a client uses different sender addresses
1328 with the options per_mail/unique=$sender_address. There are further
1329 details in the main documentation.
1331 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1333 TF/06 Removed a few PCRE remnants.
1335 TF/07 Automatically extract Exim's version number from tags in the git
1336 repository when doing development or release builds.
1338 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1339 Bugzilla 879. Patch from Paul Fisher.
1341 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1342 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1345 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1347 PP/05 Variable $av_failed, true if the AV scanner deferred.
1348 Bugzilla 1078. Patch from John Horne.
1350 PP/06 Stop make process more reliably on build failure.
1351 Bugzilla 1087. Patch from Heiko Schlittermann.
1353 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1354 Bugzilla 1089. Patch from Heiko Schlittermann.
1356 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1357 Bugzilla 1131. Patch from Holger Weiß.
1359 PP/09 Handle IPv6 addresses with SPF.
1360 Bugzilla 860. Patch from Wolfgang Breyha.
1362 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1364 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1367 PP/11 match_* no longer expand right-hand-side by default.
1368 New compile-time build option, EXPAND_LISTMATCH_RHS.
1369 New expansion conditions, "inlist", "inlisti".
1371 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1373 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1375 PP/14 fix log_write() format string regression from TF/03.
1376 Bugzilla 1152. Patch from Dmitry Isaikin.
1382 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1384 PP/02 Harmonised TLS library version reporting; only show if debugging.
1385 Layout now matches that introduced for other libraries in 4.74 PP/03.
1387 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1389 PP/04 New "dns_use_edns0" global option.
1391 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1394 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1395 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1397 TK/01 Updated PolarSSL code to 0.14.2.
1398 Bugzilla 1097. Patch from Andreas Metzler.
1400 PP/07 Catch divide-by-zero in ${eval:...}.
1401 Fixes bugzilla 1102.
1403 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1406 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1407 format-string attack -- SECURITY: remote arbitrary code execution.
1409 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1410 time unintentionally subject to list matching rules, letting the header
1411 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1412 arbitrary string expansion). This allowed for information disclosure.
1414 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1415 INT_MIN/-1 -- value coerced to INT_MAX.
1421 NM/01 Workaround for PCRE version dependency in version reporting
1424 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1425 This fixes portability to compilers other than gcc, notably
1426 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1428 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1429 makefiles for portability to HP-UX and POSIX correctness.
1431 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1432 Also via indented variable definition in the Makefile.
1433 (Debugging by Oliver Heesakkers).
1435 PP/02 Restore caching of spamd results with expanded spamd_address.
1436 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1438 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1439 Improves build reliability. Fix from: Frank Elsner
1441 NM/02 Fix wide character breakage in the rfc2047 coding
1442 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1444 NM/03 Allow underscore in dnslist lookups
1445 Fixes bug 1026. Patch from Graeme Fowler
1447 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1448 Code patches from Adam Ciarcinski of NetBSD.
1450 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1453 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1454 is logged, to avoid truncation. Patch from John Horne.
1456 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1457 Patch from Jakob Hirsch.
1459 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1460 SQL string expansion failure details.
1461 Patch from Andrey Oktyabrski.
1463 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1464 Patch from Simon Arlott.
1466 PP/09 New lookups functionality failed to compile on old gcc which rejects
1467 extern declarations in function scope.
1468 Patch from Oliver Fleischmann
1470 PP/10 Use sig_atomic_t for flags set from signal handlers.
1471 Check getgroups() return and improve debugging.
1472 Fixed developed for diagnosis in bug 927 (which turned out to be
1475 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1476 Patch from Mark Zealey.
1478 PP/12 Bugzilla 1056: Improved spamd server selection.
1479 Patch from Mark Zealey.
1481 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1482 Based on patch from Heiko Schlittermann.
1484 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1485 Patch from Uwe Doering, sign-off by Michael Haardt.
1487 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1488 variable declaration deep within a block. Bug and patch from
1491 PP/15 lookups-Makefile IRIX compatibility coercion.
1493 PP/16 Make DISABLE_DKIM build knob functional.
1495 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1496 Patch by Simon Arlott
1498 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1499 variable argument macros. Our copy now differs from upstream.
1505 TF/01 Failure to get a lock on a hints database can have serious
1506 consequences so log it to the panic log.
1508 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1509 controlled using the smtp_confirmation log selector.
1511 TF/03 Include the error message when we fail to unlink a spool file.
1513 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1514 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1515 for maintaining out-of-tree patches for some time.
1517 PP/01 Bugzilla 139: Documentation and portability issues.
1518 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1519 Handle per-OS dynamic-module compilation flags.
1521 PP/02 Let /dev/null have normal permissions.
1522 The 4.73 fixes were a little too stringent and complained about the
1523 permissions on /dev/null. Exempt it from some checks.
1524 Reported by Andreas M. Kirchwitz.
1526 PP/03 Report version information for many libraries, including
1527 Exim version information for dynamically loaded libraries. Created
1528 version.h, now support a version extension string for distributors
1529 who patch heavily. Dynamic module ABI change.
1531 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1532 privilege escalation vulnerability whereby the Exim run-time user
1533 can cause root to append content of the attacker's choosing to
1536 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1539 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1540 If dropping privileges for untrusted macros, we disabled normal logging
1541 on the basis that it would fail; for the Exim run-time user, this is not
1542 the case, and it resulted in successful deliveries going unlogged.
1543 Fixed. Reported by Andreas Metzler.
1549 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1550 only prepend for the Resent-* case. Fixes regression introduced in
1551 Exim 4.70 by NM/22 for Bugzilla 607.
1553 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1554 increasing numbers of administrators be bitten by this.
1556 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1558 PP/03 Bugzilla 994: added openssl_options main configuration option.
1560 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1562 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1564 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1566 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1567 without a peer certificate, leading to a segfault because of an
1568 assumption that peers always have certificates. Be a little more
1569 paranoid. Problem reported by Martin Tscholak.
1571 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1572 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1573 NB: ClamAV planning to remove STREAM in "middle of 2010".
1574 CL also introduces -bmalware, various -d+acl logging additions and
1575 more caution in buffer sizes.
1577 PP/09 Implemented reverse_ip expansion operator.
1579 PP/10 Bugzilla 937: provide a "debug" ACL control.
1581 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1583 PP/12 Bugzilla 973: Implement --version.
1585 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1587 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1589 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1591 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1592 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1593 ignore trailing whitespace.
1595 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1597 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1600 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1601 Notification from Dr Andrew Aitchison.
1603 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1604 ExtendedDetectionInfo response format.
1605 Notification from John Horne.
1607 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1610 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1611 XSL and documented dependency on system catalogs, with examples of how
1614 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1617 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1618 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1619 configuration file which is writeable by the Exim user or group.
1621 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1622 of configuration files to cover files specified with the -C option if
1623 they are going to be used with root privileges, not just the default
1626 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1627 option (effectively making it always true).
1629 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1630 files to be used while preserving root privileges.
1632 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1633 that rogue child processes cannot use them.
1635 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1636 run-time user, instead of root.
1638 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1639 Exim run-time user without dropping privileges.
1641 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1642 result string, instead of calling string_vformat() twice with the same
1645 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1646 for other users. Others should always drop root privileges if they use
1647 -C on the command line, even for a whitelisted configure file.
1649 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1651 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1657 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1658 $data_path, and $header_path variables; fixed documentation bugs and
1661 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1662 exipick to access non-standard spools, including the "frozen" queue
1665 NM/01 Bugzilla 965: Support mysql stored procedures.
1666 Patch from Alain Williams
1668 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1670 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1671 Patch from Andreas Metzler
1673 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1674 Patch from Kirill Miazine
1676 NM/05 Bugzilla 671: Added umask to procmail example.
1678 JJ/03 installed exipick 20100323.0, fixing doc bug
1680 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1681 directory. Notification and patch from Dan Rosenberg.
1683 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1685 TK/02 Improve log output when DKIM signing operation fails.
1687 MH/01 Treat the transport option dkim_domain as a colon separated
1688 list, not as a single string, and sign the message with each element,
1689 omitting multiple occurences of the same signer.
1691 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1692 Bugzilla 985, 986. Patch by Simon Arlott
1694 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1695 Patch by Simon Arlott
1697 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1698 MBX locking. Notification from Dan Rosenberg.
1704 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1706 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1708 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1710 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1712 NM/04 Bugzilla 915: Flag broken perl installation during build.
1718 TK/01 Added patch by Johannes Berg that expands the main option
1719 "spamd_address" if it starts with a dollar sign.
1721 TK/02 Write list of recipients to X-Envelope-Sender header when building
1722 the mbox-format spool file for content scanning (suggested by Jakob
1725 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1726 (http://www.dcc-servers.net/) support via dccifd. Activated by
1727 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1729 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1730 by Mark Daniel Reidel <mr@df.eu>.
1732 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1733 When building exim an external PCRE library is now needed -
1734 PCRE is a system library on the majority of modern systems.
1735 See entry on PCRE_LIBS in EDITME file.
1737 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1738 conversation. Added nologin parameter to request.
1739 Patch contributed by Kirill Miazine.
1741 TF/01 Do not log submission mode rewrites if they do not change the address.
1743 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1745 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1746 log files in place. Contributed by Roberto Lima.
1748 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1750 TF/03 Bugzilla 615: When checking the local_parts router precondition
1751 after a local_part_suffix or local_part_prefix option, Exim now
1752 does not use the address's named list lookup cache, since this
1753 contains cached lookups for the whole local part.
1755 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1756 Robert Millan. Documentation is in experimental-spec.txt.
1758 TF/04 Bugzilla 668: Fix parallel build (make -j).
1760 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1762 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1763 Patch provided by Jan Srzednicki.
1765 TF/05 Leading white space used to be stripped from $spam_report which
1766 wrecked the formatting. Now it is preserved.
1768 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1769 that they are available at delivery time.
1771 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1773 TF/08 TLS error reporting now respects the incoming_interface and
1774 incoming_port log selectors.
1776 TF/09 Produce a more useful error message if an SMTP transport's hosts
1777 setting expands to an empty string.
1779 NM/06 Bugzilla 744: EXPN did not work under TLS.
1780 Patch provided by Phil Pennock.
1782 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1783 Patch provided by Richard Godbee.
1785 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1786 acl_smtp_notquit, added index entry.
1788 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1789 Patch provided by Eugene Bujak.
1791 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1792 accept(). Patch provided by Maxim Dounin.
1794 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1795 Patch provided by Phil Pennock.
1797 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1799 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1800 Patch provided by Brad "anomie" Jorsch.
1802 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1803 Patch provided by Dean Brooks.
1805 TK/05 Add native DKIM support (does not depend on external libraries).
1807 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1808 Patch provided by Graeme Fowler.
1810 NM/16 Bugzilla 851: Documentation example syntax fix.
1812 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1814 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1817 NM/19 Bugzilla 745: TLS version reporting.
1818 Patch provided by Phil Pennock.
1820 NM/20 Bugzilla 167: bool: condition support.
1821 Patch provided by Phil Pennock.
1823 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1824 clients. Patch provided by Phil Pennock.
1826 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1827 Patch provided by Brad "anomie" Jorsch.
1829 NM/23 Bugzilla 687: Fix misparses in eximstats.
1830 Patch provided by Heiko Schlittermann.
1832 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1833 Patch provided by Heiko Schlittermann.
1835 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1836 plus update to original patch.
1838 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1840 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1841 Patch provided by David Brownlee.
1843 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1845 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1847 NM/30 Bugzilla 888: TLS documentation bugfixes.
1849 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1851 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1852 Unlike the original bugzilla I have changed all shell scripts in src tree.
1854 NM/33 Bugzilla 898: Transport filter timeout fix.
1855 Patch by Todd Rinaldo.
1857 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1858 Patch by Serge Demonchaux.
1860 NM/35 Bugzilla 39: Base64 decode bug fixes.
1861 Patch by Jakob Hirsch.
1863 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1865 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1867 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1869 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1875 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1876 ALT-N's libdkim that I have put here:
1877 http://duncanthrax.net/exim-experimental/
1879 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1880 were called 'true' and it seems that C99 defines that as a reserved
1881 keyword to be used with 'bool' variable types. That means you could
1882 not include C99-style headers which use bools without triggering
1883 build errors in sieve.c.
1885 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1886 as mailq or other aliases. Changed the --help handling significantly
1887 to do whats expected. exim_usage() emits usage/help information.
1889 SC/01 Added the -bylocaldomain option to eximstats.
1891 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1893 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1895 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1901 PH/01 Another patch from the Sieve maintainer.
1903 PH/02 When an IPv6 address is converted to a string for single-key lookup
1904 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1905 dots are used instead of colons so that keys in lsearch files need not
1906 contain colons. This was done some time before quoting was made available
1907 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1908 (notated using the quote facility) so as to distinguish them from IPv4
1909 keys. This meant that lookups for IP addresses in host lists did not work
1910 for iplsearch lookups.
1912 This has been fixed by arranging for IPv6 addresses to be expressed with
1913 colons if the lookup type is iplsearch. This is not incompatible, because
1914 previously such lookups could never work.
1916 The situation is now rather anomalous, since one *can* have colons in
1917 ordinary lsearch keys. However, making the change in all cases is
1918 incompatible and would probably break a number of configurations.
1920 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1923 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1924 conversion specification without a maximum field width, thereby enabling
1925 a rogue spamd server to cause a buffer overflow. While nobody in their
1926 right mind would setup Exim to query an untrusted spamd server, an
1927 attacker that gains access to a server running spamd could potentially
1928 exploit this vulnerability to run arbitrary code as the Exim user.
1930 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1931 $primary_hostname instead of what libspf2 thinks the hosts name is.
1933 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1934 a directory entry by the name of the lookup key. Previously, if a
1935 symlink pointed to a non-existing file or a file in a directory that
1936 Exim lacked permissions to read, a lookup for a key matching that
1937 symlink would fail. Now it is enough that a matching directory entry
1938 exists, symlink or not. (Bugzilla 503.)
1940 PH/03 The body_linecount and body_zerocount variables are now exported in the
1943 PH/04 Added the $dnslist_matched variable.
1945 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1946 This means they are set thereafter only if the connection becomes
1949 PH/06 Added the client_condition to authenticators so that some can be skipped
1950 by clients under certain conditions.
1952 PH/07 The error message for a badly-placed control=no_multiline_responses left
1953 "_responses" off the end of the name.
1955 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1957 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1958 (without spaces) instead of just copying the configuration text.
1960 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1962 PH/11 Added $max_received_linelength.
1964 PH/12 Added +ignore_defer and +include_defer to host lists.
1966 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1967 way in which PCRE > 7.0 is built.
1969 PH/14 Implemented queue_only_load_latch.
1971 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1972 MAIL command. The effect was to mangle the value on 64-bit systems.
1974 PH/16 Another patch from the Sieve maintainer.
1976 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1978 PH/18 If a system quota error occurred while trying to create the file for
1979 a maildir delivery, the message "Mailbox is full" was not appended to the
1980 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1981 only to a quota excession during the actual writing of the file.
1983 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1984 characters?) which causes problems in log lines. The DN values are now
1985 passed through string_printing() before being added to log lines.
1987 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1988 and InterBase are left for another time.)
1990 PH/21 Added message_body_newlines option.
1992 PH/22 Guard against possible overflow in moan_check_errorcopy().
1994 PH/23 POSIX allows open() to be a macro; guard against that.
1996 PH/24 If the recipient of an error message contained an @ in the local part
1997 (suitably quoted, of course), incorrect values were put in $domain and
1998 $local_part during the evaluation of errors_copy.
2004 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2005 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2008 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2009 issue a MAIL command.
2011 PH/02 In an ACL statement such as
2013 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2015 if a client was not listed at all, or was listed with a value other than
2016 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2017 the condition was not true (as it should be), so access was not denied.
2018 The bug was that the ! inversion was incorrectly passed on to the second
2019 item. This has been fixed.
2021 PH/03 Added additional dnslists conditions == and =& which are different from
2022 = and & when the dns lookup returns more than one IP address.
2024 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2025 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2027 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2028 FSYNC, which compiles an option called disable_fsync that allows for
2029 bypassing fsync(). The documentation is heavily laced with warnings.
2031 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2033 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2034 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2035 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2036 including adding "make clean"; (3) Added -fPIC when compiling the test
2037 dynamically loaded module, to get rid of a warning.
2039 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2040 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2041 = 0s. The bug is otherwise harmless.
2043 PH/07 There was a bug in the dovecot authenticator such that the value of
2044 $auth1 could be overwritten, and so not correctly preserved, after a
2045 successful authentication. This usually meant that the value preserved by
2046 the server_setid option was incorrect.
2048 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2050 PH/09 Installed PCRE release 7.0.
2052 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2053 run for batched SMTP input. It is now run at the start of every message
2054 in the batch. While fixing this I discovered that the process information
2055 (output by running exiwhat) was not always getting set for -bs and -bS
2056 input. This is fixed, and it now also says "batched" for BSMTP.
2058 PH/11 Added control=no_pipelining.
2060 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2061 patch, slightly modified), and move the expansion of helo_data till after
2062 the connection is made in the smtp transport (so it can use these
2065 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2067 PH/14 Added log_selector = +pid.
2069 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2071 PH/16 Add ${if forany and ${if forall.
2073 PH/17 Added dsn_from option to vary the From: line in DSNs.
2075 PH/18 Flush SMTP output before performing a callout, unless control =
2076 no_callout_flush is set.
2078 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2079 was true (the default) a successful delivery failed to delete the retry
2080 item, thus causing premature timeout of the address. The bug is now
2083 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2085 PH/21 Long custom messages for fakedefer and fakereject are now split up
2086 into multiline responses in the same way that messages for "deny" and
2087 other ACL rejections are.
2089 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2090 with slight modification.
2092 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2093 draft, changing the syntax and factoring some duplicate code".
2095 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2096 for deliveries of the second and subsequent messages over the same SMTP
2099 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2100 ${reduce, with only minor "tidies".
2102 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2104 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2105 expansion side effects.
2107 PH/27 When a message times out after an over-quota error from an Exim-imposed
2108 quota, the bounce message says "mailbox is full". This message was not
2109 being given when it was a system quota that was exceeded. It now should
2112 MH/03 Made $recipients available in local_scan(). local_scan() already has
2113 better access to the recipient list through recipients_list[], but
2114 $recipients can be useful in postmaster-provided expansion strings.
2116 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2117 in the case of a MAIL command with additional options following the
2118 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2119 were accidentally chopped off.
2121 PH/29 SMTP synchronization checks are implemented when a command is read -
2122 there is a check that no more input is waiting when there shouldn't be
2123 any. However, for some commands, a delay in an ACL can mean that it is
2124 some time before the response is written. In this time, more input might
2125 arrive, invalidly. So now there are extra checks after an ACL has run for
2126 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2127 pipelining has not been advertised.
2129 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2131 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2132 returned in the SMTP response when smtp_return_error_details was false.
2133 This has been fixed.
2135 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2136 instead of the C I/O that was originally supplied, because problems were
2137 reported on Solaris.
2139 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2140 Exim which did not show up earlier: it was assuming that a call to
2141 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2142 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2143 was a macro that became an assignment, so it seemed to work. This has
2144 changed to a proper function call with a void return, hence the compile
2145 error. Exim's code has been fixed.
2147 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2150 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2152 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2154 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2155 because it turns out to be too big or there is a timeout) there is no
2156 "Completed" line in the log. When some messages of this type were
2157 selected by exigrep, they were listed as "not completed". Others were
2158 picked up by some special patterns. I have improved the selection
2159 criteria to be more general.
2161 PH/38 The host_find_failed option in the manualroute router can now be set
2162 to "ignore", to completely ignore a host whose IP address cannot be
2163 found. If all hosts are ignored, the behaviour is controlled by the new
2164 host_all_ignored option.
2166 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2167 homing or because of multiple MX records with /mx) generated more than
2168 one IP address, and the following item turned out to be the local host,
2169 all the secondary addresses of the first item were incorrectly removed
2170 from the list, along with the local host and any following hosts (which
2171 is what is supposed to happen).
2173 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2174 whoever called Exim into the -H file. In the case of the daemon it was
2175 behaving confusingly. When first started, it used values for whoever
2176 started the daemon, but after a SIGHUP it used the Exim user (because it
2177 calls itself on a restart). I have changed the code so that it now always
2180 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2181 message are rejected with the same error (e.g. no authentication or bad
2182 sender address), and a DATA command is nevertheless sent (as can happen
2183 with PIPELINING or a stupid MUA), the error message that was given to the
2184 RCPT commands is included in the rejection of the DATA command. This is
2185 intended to be helpful for MUAs that show only the final error to their
2188 PH/42 Another patch from the Sieve maintainer.
2190 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2191 Eximstats - Fixed some broken HTML links and added missing column headers
2193 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2194 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2196 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2198 PH/43 Yet another patch from the Sieve maintainer.
2200 PH/44 I found a way to check for a TCP/IP connection going away before sending
2201 the response to the final '.' that terminates a message, but only in the
2202 case where the client has not sent further data following the '.'
2203 (unfortunately, this is allowed). However, in many cases there won't be
2204 any further data because there won't be any more messages to send. A call
2205 to select() can be used: if it shows that the input is "ready", there is
2206 either input waiting, or the socket has been closed. An attempt to read
2207 the next input character can distinguish the two cases. Previously, Exim
2208 would have sent an OK response which the client would never have see.
2209 This could lead to message repetition. This fix should cure that, at
2210 least in a lot of common cases.
2212 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2213 advertised in response to EHLO.
2219 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2220 fixed by 4.65/MH/01 (is this a record?) are fixed:
2222 (i) An empty string was always treated as zero by the numeric comparison
2223 operators. This behaviour has been restored.
2225 (ii) It is documented that the numeric comparison operators always treat
2226 their arguments as decimal numbers. This was broken in that numbers
2227 starting with 0 were being interpreted as octal.
2229 While fixing these problems I realized that there was another issue that
2230 hadn't been noticed. Values of message_size_limit (both the global option
2231 and the transport option) were treated as octal if they started with 0.
2232 The documentation was vague. These values are now always treated as
2233 decimal, and I will make that clear in the documentation.
2239 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2240 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2243 MH/01 Don't check that the operands of numeric comparison operators are
2244 integers when their expansion is in "skipping" mode (fixes bug
2245 introduced by 4.64-PH/07).
2247 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2248 child addresses, Exim now panics and dies. Previously, because the count
2249 is held in a short int, deliveries were likely to be lost. As such a
2250 large number of recipients for a single message is ridiculous
2251 (performance will be very, very poor), I have chosen to impose a limit
2252 rather than extend the field.
2258 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2259 leftover -K file (the existence of which was triggered by #402).
2260 While we were at it, introduced process PID as part of the -K
2261 filename. This should rule out race conditions when creating
2264 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2265 processing considerably. Previous code took too long for large mails,
2266 triggering a timeout which in turn triggers #401.
2268 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2269 in the DK code in transports.c. sendfile() is not really portable,
2270 hence the _LINUX specificness.
2272 TF/01 In the add_headers option to the mail command in an Exim filter,
2273 there was a bug that Exim would claim a syntax error in any
2274 header after the first one which had an odd number of characters
2277 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2278 callout verification, Exim cached a "reject" for the entire domain. This
2279 is correct for most verifications, but it is not correct for a recipient
2280 verification with use_sender or use_postmaster set, because in that case
2281 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2282 case of MAIL FROM:<> rejection from other early rejections (e.g.
2283 rejection of HELO). When verifying a recipient using a non-null MAIL
2284 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2285 Whatever the result of the callout, the value of the domain cache is
2286 left unchanged (for any other kind of callout, getting as far as trying
2287 RCPT means that the domain itself is ok).
2289 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2292 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2293 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2294 session, the daemon ignores ECONNECT errors and logs others; it now
2295 ignores EPIPE as well.
2297 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2298 (quoted-printable decoding).
2300 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2301 later the small subsequent patch to fix an introduced bug.
2303 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2305 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2307 PH/08 An error is now given if message_size_limit is specified negative.
2309 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2310 to be given (somewhat) arbitrary names.
2312 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2315 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2316 miscellaneous code fixes
2318 PH/10 Added the log_reject_target ACL modifier to specify where to log
2321 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2322 hostname. This is wrong, because it relates to the incoming message (and
2323 probably the interface on which it is arriving) and not to the outgoing
2324 callout (which could be using a different interface). This has been
2325 changed to use the value of the helo_data option from the smtp transport
2326 instead - this is what is used when a message is actually being sent. If
2327 there is no remote transport (possible with a router that sets up host
2328 addresses), $smtp_active_hostname is used.
2330 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2331 tweaks were necessary in order to get it to work (see also 21 below):
2332 (a) The code assumed that strncpy() returns a negative number on buffer
2333 overflow, which isn't the case. Replaced with Exim's string_format()
2335 (b) There were several signed/unsigned issues. I just did the minimum
2336 hacking in of casts. There is scope for a larger refactoring.
2337 (c) The code used strcasecmp() which is not a standard C function.
2338 Replaced with Exim's strcmpic() function.
2339 (d) The code set only $1; it now sets $auth1 as well.
2340 (e) A simple test gave the error "authentication client didn't specify
2341 service in request". It would seem that Dovecot has changed its
2342 interface. Fortunately there's a specification; I followed it and
2343 changed what the client sends and it appears to be working now.
2345 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2348 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2349 address A is aliased to B and C, where B exists and C does not. Without
2350 -v the output is "A verified" because verification stops after a
2351 successful redirection if more than one address is generated. However,
2352 with -v the child addresses are also verified. Exim was outputting "A
2353 failed to verify" and then showing the successful verification for C,
2354 with its parentage. It now outputs "B failed to verify", showing B's
2355 parentage before showing the successful verification of C.
2357 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2358 look up a TXT record in a specific list after matching in a combined
2361 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2362 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2363 they consult the DNS. I had assumed they would set it the way they
2364 wanted; and indeed my experiments on Linux seem to show that in some
2365 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2366 To be on the safe side, however, I have now made the interface to
2367 host_find_byname() similar to host_find_bydns(), with an argument
2368 containing the DNS resolver options. The host_find_byname() function now
2369 sets these options at its start, just as host_find_bydns() does. The smtp
2370 transport options dns_qualify_single and dns_search_parents are passed to
2371 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2372 of host_find_byname() use the default settings of RES_DEFNAMES
2373 (qualify_single) but not RES_DNSRCH (search_parents).
2375 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2376 spool_read_header() do less string testing, by means of a preliminary
2377 switch on the second character of optional "-foo" lines. (This is
2378 overdue, caused by the large number of possibilities that now exist.
2379 Originally there were few.) While I was there, I also converted the
2380 str(n)cmp tests so they don't re-test the leading "-" and the first
2381 character, in the hope this might squeeze out yet more improvement.
2383 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2384 flag allowing group syntax was set by the header_syntax check but not
2385 turned off, possible causing trouble later; (2) The flag was not being
2386 set at all for the header_verify test, causing "group"-style headers to
2387 be rejected. I have now set it in this case, and also caused header_
2388 verify to ignore an empty address taken from a group. While doing this, I
2389 came across some other cases where the code for allowing group syntax
2390 while scanning a header line wasn't quite right (mostly, not resetting
2391 the flag correctly in the right place). These bugs could have caused
2392 trouble for malformed header lines. I hope it is now all correct.
2394 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2395 with the "reply" argument non-NULL. The code, however (which originally
2396 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2397 but it didn't always do it. This confused somebody who was copying the
2398 code for some other use. I have removed all the tests.
2400 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2401 feature that was used to support insecure browsers during the U.S. crypto
2402 embargo. It requires special client support, and Exim is probably the
2403 only MTA that supported it -- and would never use it because real RSA is
2404 always available. This code has been removed, because it had the bad
2405 effect of slowing Exim down by computing (never used) parameters for the
2406 RSA_EXPORT functionality.
2408 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2409 authenticator to fail if there's a tab character in the incoming data
2410 (there should never be unless someone is messing about, as it's supposed
2411 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2412 if the connection is using TLS or if the remote IP is the same as the
2413 local IP, and the "valid-client-cert option" if a client certificate has
2416 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2417 authenticators. This can be used for authorization after authentication
2418 succeeds. (In the case of plaintext, it servers for both authentication
2421 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2422 if any retry times were supplied.
2424 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2425 connection, where there is no client IP address to check. In this
2426 situation, the verify now always succeeds.
2428 PH/25 Applied John Jetmore's -Mset patch.
2430 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2432 PH/27 In a string expansion for a processed (not raw) header when multiple
2433 headers of the same name were present, leading whitespace was being
2434 removed from all of them, but trailing whitespace was being removed only
2435 from the last one. Now trailing whitespace is removed from each header
2436 before concatenation. Completely empty headers in a concatenation (as
2437 before) are ignored.
2439 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2440 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2442 PH/29 [Removed. This was a change that I later backed out, and forgot to
2443 correct the ChangeLog entry (that I had efficiently created) before
2444 committing the later change.]
2446 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2447 address errors (4xx response to RCPT) over the same connection as other
2448 messages routed to the same hosts. Such deliveries are always "forced",
2449 so retry times are not inspected. This resulted in far too many retries
2450 for the affected addresses. The effect occurred only when there were more
2451 hosts than the hosts_max_try setting in the smtp transport when it had
2452 the 4xx errors. Those hosts that it had tried were not added to the list
2453 of hosts for which the message was waiting, so if all were tried, there
2454 was no problem. Two fixes have been applied:
2456 (i) If there are any address or message errors in an SMTP delivery, none
2457 of the hosts (tried or untried) are now added to the list of hosts
2458 for which the message is waiting, so the message should not be a
2459 candidate for sending over the same connection that was used for a
2460 successful delivery of some other message. This seems entirely
2461 reasonable: after all the message is NOT "waiting for some host".
2462 This is so "obvious" that I'm not sure why it wasn't done
2463 previously. Hope I haven't missed anything, but it can't do any
2464 harm, as the worst effect is to miss an optimization.
2466 (ii) If, despite (i), such a delivery is accidentally attempted, the
2467 routing retry time is respected, so at least it doesn't keep
2468 hammering the server.
2470 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2471 in ${readsocket because some servers need this prod.
2473 PH/32 Added some extra debug output when updating a wait-xxx database.
2475 PH/33 The hint "could be header name not terminated by colon", which has been
2476 given for certain expansion errors for a long time, was not being given
2477 for the ${if def:h_colon_omitted{... case.
2479 PH/34 The spec says: "With one important exception, whenever a domain list is
2480 being scanned, $domain contains the subject domain." There was at least
2481 one case where this was not true.
2483 PH/35 The error "getsockname() failed: connection reset by peer" was being
2484 written to the panic log as well as the main log, but it isn't really
2485 panic-worthy as it just means the connection died rather early on. I have
2486 removed the panic log writing for the ECONNRESET error when getsockname()
2489 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2490 runs only) independently of the message's sender address. This meant
2491 that, if the 4xx error was in fact related to the sender, a different
2492 message to the same recipient with a different sender could confuse
2493 things. In particular, this can happen when sending to a greylisting
2494 server, but other circumstances could also provoke similar problems.
2495 I have changed the default so that the retry time for these errors is now
2496 based a combination of the sender and recipient addresses. This change
2497 can be overridden by setting address_retry_include_sender=false in the
2500 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2501 remote server are returned as part of bounce messages. This was not
2502 happening for LMTP over a pipe (the lmtp transport), but now it is the
2503 same for both kinds of LMTP.
2505 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2506 in header lines that were in fact CNAMEs. This is no longer the case.
2508 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2509 and queue runs started by the daemon processed all messages. This has
2510 been fixed so that -R and -S can now usefully be given with -q<time>.
2512 PH/40 Import PCRE release 6.7 (fixes some bugs).
2514 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2516 PH/42 Give an error if -q is specified more than once.
2518 PH/43 Renamed the variables $interface_address and $interface_port as
2519 $received_ip_address and $received_port, to make it clear that these
2520 values apply to message reception, and not to the outgoing interface when
2521 a message is delivered. (The old names remain recognized, of course.)
2523 PH/44 There was no timeout on the connect() call when using a Unix domain
2524 socket in the ${readsocket expansion. There now is.
2526 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2527 be meaningful with "accept".
2529 SC/01 Eximstats V1.43
2530 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2532 SC/02 Eximstats V1.44
2533 Use a glob alias rather than an array ref in the generated
2534 parser. This improves both readability and performance.
2536 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2537 Collect SpamAssassin and rejection statistics.
2538 Don't display local sender or destination tables unless
2539 there is data to show.
2540 Added average volumes into the top table text output.
2542 SC/04 Eximstats V1.46
2543 Collect data on the number of addresses (recipients)
2544 as well as the number of messages.
2546 SC/05 Eximstats V1.47
2547 Added 'Message too big' to the list of mail rejection
2548 reasons (thanks to Marco Gaiarin).
2550 SC/06 Eximstats V1.48
2551 Mainlog lines which have GMT offsets and are too short to
2552 have a flag are now skipped.
2554 SC/07 Eximstats V1.49 (Alain Williams)
2555 Added the -emptyok flag.
2557 SC/08 Eximstats V1.50
2558 Fixes for obtaining the IP address from reject messages.
2560 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2561 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2562 whitespace changes from 4.64-PH/27
2564 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2567 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2568 are found, allow negative numbers in numeric criteria)
2570 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2572 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2573 to match changes made in 4.64-PH/43
2575 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2577 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2578 "domain" is actually an IP address, and force a failure. This locks out
2579 those revolvers/nameservers that support "A-for-A" lookups, in
2580 contravention of the specifications.
2582 PH/48 When a host name was looked up from an IP address, and the subsequent
2583 forward lookup of the name timed out, the host name was left in
2584 $sender_host_name, contrary to the specification.
2586 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2587 restricted to single-key lookups, Exim was not diagnosing an error if
2588 * or *@ was used with a query-style lookup.
2590 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2592 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2593 long ago, but noone interested enough thought of it. Let's just say that
2594 the "1.1" means that there are some new functions that weren't there at
2595 some point in the past.
2597 PH/51 Error processing for expansion failure of helo_data from an smtp
2598 transport during callout processing was broken.
2600 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2601 tested/used via the -bh/-bhc/-bs options.
2603 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2604 bug, fixed in subsequent PCRE releases).
2606 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2607 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2609 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2615 SC/01 Use a glob alias rather than an array ref in eximstats generated
2616 parser. This improves both readability and performance.
2618 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2619 Don't display local sender or destination tables in eximstats unless
2620 there is data to show.
2621 Added average volumes into the eximstats top table text output.
2623 SC/03 Collect data on the number of addresses (recipients) as well
2624 as the number of messages in eximstats.
2626 TF/01 Correct an error in the documentation for the redirect router. Exim
2627 does (usually) call initgroups() when daemonizing.
2629 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2630 with consistent privilege compared to when running as a daemon.
2632 TF/03 Note in the spec that $authenticated_id is not set for local
2633 submissions from trusted users.
2635 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2636 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2638 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2639 by adding some example configuration directives to the default
2640 configuration file. A little bit of work is required to uncomment the
2641 directives and define how usernames and passwords are checked, but
2642 there is now a framework to start from.
2644 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2645 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2646 without this. I don't know how relevant this is to other LDAP libraries.
2648 PH/02 Add the verb name to the "unknown ACL verb" error.
2650 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2652 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2654 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2655 directory not expanded when it should be if an expanded home directory
2656 was set for the address (which is overridden by the transport).
2658 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2661 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2662 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2663 because it is too late at that time, and has no effect.
2665 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2666 security issue with \' (bugzilla #107). I could not use the
2667 PQescapeStringConn() function, because it needs a PGconn value as one of
2670 PH/08 When testing addresses using -bt, indicate those final addresses that
2671 are duplicates that would not cause an additional delivery. At least one
2672 person was confused, thinking that -bt output corresponded to deliveries.
2673 (Suppressing duplicates isn't a good idea as you lose the information
2674 about possibly different redirections that led to the duplicates.)
2676 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2677 systems where poll() doesn't work, in particular OS X.
2679 PH/10 Added more information to debugging output for retry time not reached.
2681 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2682 operations in malware.c.
2684 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2687 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2688 syslog_duplication set false, log lines that would normally be written
2689 both the the main log and to the reject log were not written to syslog at
2692 PH/14 In the default configuration, change the use of "message" in ACL warn
2693 statements to "add_header".
2695 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2696 not followed by a command (e.g. "seen endif").
2698 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2699 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2702 PH/17 Added extra conditions to the default value of delay_warning_condition
2706 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2707 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2708 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2711 The Auto-Submitted: and various List- headers are standardised, whereas I
2712 don't think Precedence: ever was.
2714 PH/18 Refactored debugging code in route_finduser() to show more information,
2715 in particular, the error code if getpwnam() issues one.
2717 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2718 This is apparently needed in addition to the PH/07 change above to avoid
2719 any possible encoding problems.
2721 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2722 but not after initializing Perl.
2724 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2725 output them only if debugging. By default they are written stderr,
2726 apparently, which is not desirable.
2728 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2731 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2734 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2736 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2737 authenticated or an ident call has been made. Suppress the default
2738 values for $authenticated_id and $authenticated_sender (but permit -oMai
2739 and -oMas) when testing with -bh.
2741 PH/24 Re-jigged the order of the tests in the default configuration so that the
2742 tests for valid domains and recipients precede the DNS black list and CSA
2743 tests, on the grounds that those ones are more expensive.
2745 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2746 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2747 This bug exists in every version of Exim that I still have, right back to
2750 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2751 However, an attempt to turn on case-sensitivity in a regex key by
2752 including (?-i) didn't work because the subject string was already
2753 lowercased, and the effects were non-intuitive. It turns out that a
2754 one-line patch can be used to allow (?-i) to work as expected.
2760 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2761 other effects) broke the use of negated acl sub-conditions.
2763 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2766 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2767 "Deny" causes Exim to reject the incoming connection with a 554 error.
2768 Unfortunately, if there is a major crisis, such as a disk failure,
2769 tcp-wrappers gives "deny", whereas what one would like would be some
2770 kind of temporary error. A kludge has been added to help with this.
2771 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2772 554 error is used if errno is still zero or contains ENOENT (which occurs
2773 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2776 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2778 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2779 errors. Otherwise a message that provokes a temporary error (when other
2780 messages do not) can cause a whole host to time out.
2782 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2783 addresses were routed directly to files or pipes from a redirect router.
2784 File deliveries just didn't batch; pipe deliveries might have suffered
2787 PH/06 A failure to get a lock for a hints database would erroneously always say
2788 "Failed to get write lock", even when it was really a read lock.
2790 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2791 of 0600. This has been changed to use the value of the lockfile_mode
2792 option (which defaults to 0600).
2794 PH/08 Applied small patch from the Sieve maintainer.
2796 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2797 folder from quota calculations, a direct delivery into this folder messed
2798 up the contents of the maildirsize file. This was because the regex was
2799 used only to exclude .Trash (or whatever) when the size of the mailbox
2800 was calculated. There was no check that a delivery was happening into an
2801 excluded directory. This bug has been fixed by ignoring all quota
2802 processing for deliveries into excluded directories.
2804 PH/10 Added the maildirfolder_create_regex option to appendfile.
2810 PH/01 The code for finding all the local interface addresses on a FreeBSD
2811 system running IPv6 was broken. This may well have applied to all BSD
2812 systems, as well as to others that have similar system calls. The broken
2813 code found IPv4 interfaces correctly, but gave incorrect values for the
2814 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2815 that it would not match correctly against @[] and not recognize the IPv6
2818 PH/02 The ipliteral router was not recognizing addresses of the form user@
2819 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2821 PH/03 Added disable_ipv6.
2823 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2824 decoded form, because it is most often used to construct To: headers
2825 lines in autoreplies, and the decoded form may well be syntactically
2826 invalid. However, $reply_address has leading white space removed, and all
2827 newlines turned into spaces so that the autoreply transport does not
2830 PH/05 If group was specified without a user on a router, and no group or user
2831 was specified on a transport, the group from the router was ignored.
2833 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2834 for visible compile-time settings that can be used to change these
2835 numbers, for those that want even more. Backwards compatibility with old
2836 spool files has been maintained. However, going back to a previous Exim
2837 release will lost any variables that are in spool files.
2839 PH/07 Two small changes when running in the test harness: increase delay when
2840 passing a TCP/IP connection to a new process, in case the original
2841 process has to generate a bounce, and remove special handling of
2842 127.0.0.2 (sic), which is no longer necessary.
2844 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2845 be the same on different OS.
2847 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2850 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2851 whether --show-vars was specified or not
2853 JJ/02 exipick: Added support for new ACL variable spool format introduced
2856 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2857 syntactically invalid From: or Reply-to: line, and a filter used this to
2858 generate an autoreply, and therefore failed to obtain an address for the
2859 autoreply, Exim could try to deliver to a non-existent relative file
2860 name, causing unrelated and misleading errors. What now happens is that
2861 it logs this as a hard delivery error, but does not attempt to create a
2864 PH/11 The exinext utility has a -C option for testing purposes, but although
2865 the given file was scanned by exinext itself; it wasn't being passed on
2866 when Exim was called.
2868 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2869 an end-of-file indication when reading a command response.
2871 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2872 compiled. In many other places in Exim, IPv6 addresses are always
2873 recognized, so I have changed this. It also means that IPv4 domain
2874 literals of the form [IPV4:n.n.n.n] are now always recognized.
2876 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2877 used if the router is not running as root, for example, when verifying at
2878 ACL time, or when using -bh. The debugging output from this situation was
2879 non-existent - all you got was a failure to exec. I have made two
2882 (a) Failures to set uid/gid, the current directory, or a process leader
2883 in a subprocess such as that created by queryprogram now generate
2884 suitable debugging output when -d is set.
2886 (b) The queryprogram router detects when it is not running as root,
2887 outputs suitable debugging information if -d is set, and then runs
2888 the subprocess without attempting to change uid/gid.
2890 PH/15 Minor change to Makefile for building test_host (undocumented testing
2893 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2894 additional section of a DNS packet that returns MX or SRV records.
2895 Instead, it always explicitly searches for A/AAAA records. This avoids
2896 major problems that occur when a DNS server includes only records of one
2897 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2898 fixed another bug: if SRV records were looked up and the corresponding
2899 address records were *not* found in the additional section, the port
2900 values from the SRV records were lost.
2902 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2903 using the correct key (the original address) when searching the retry
2904 rules in order to find which one to use for generating the retry hint.
2906 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2907 adding the default one. Similarly, if it contains a Reply-To: header, the
2908 errors_reply_to option, if set, is not used.
2910 PH/19 When calculating a retry time, Exim used to measure the "time since
2911 failure" by looking at the "first failed" field in the retry record. Now
2912 it does not use this if it is later than than the arrival time of the
2913 message. Instead it uses the arrival time. This makes for better
2914 behaviour in cases where some deliveries succeed, thus re-setting the
2915 "first failed" field. An example is a quota failure for a huge message
2916 when small messages continue to be delivered. Without this change, the
2917 "time since failure" will always be short, possible causing more frequent
2918 delivery attempts for the huge message than are intended.
2919 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2921 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2922 $1, $2, $3) because the numerical variables can be reset during some
2923 expansion items (e.g. "match"), thereby losing the authentication data.
2925 PH/21 Make -bV show the size of off_t variables so that the test suite can
2926 decide whether to run tests for quotas > 2G.
2928 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2929 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2930 filecount value is greater than 2G or if a quota value is greater than 2G
2931 on a system where the size of off_t is not greater than 4, a panic error
2934 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2935 never match. The debug and -bh output now contains an explicit error
2936 message indicating a malformed IPv4 address or mask.
2938 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2939 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2940 PH/23 above applies.
2942 PH/25 Do not write to syslog when running in the test harness. The only
2943 occasion when this arises is a failure to open the main or panic logs
2944 (for which there is an explicit test).
2946 PH/26 Added the /no_tell option to "control=freeze".
2948 PH/27 If a host name lookup failed very early in a connection, for example, if
2949 the IP address matched host_lookup and the reverse lookup yielded a name
2950 that did not have a forward lookup, an error message of the form "no IP
2951 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2952 could be logged. Now it outputs the IP address instead of "NULL".
2954 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2955 allows the sender and the authenticated sender to be set when
2956 submitting a message from within Exim. Since child_open_exim() is
2957 documented for local_scan(), the new function should be too.
2959 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2960 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2961 results in an empty string is now treated as unset.
2963 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2965 PH/31 Added sender_verify_fail logging option.
2967 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2968 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2969 tidied the source and removed it altogether.
2971 PH/33 When a queue run was abandoned because the load average was too high, a
2972 log line was always written; now it is written only if the queue_run log
2973 selector is set. In addition, the log line for abandonment now contains
2974 information about the queue run such as the pid. This is always present
2975 in "start" and "stop" lines but was omitted from the "abandon" line.
2977 PH/34 Omit spaces between a header name and the colon in the error message that
2978 is given when verify = headers_syntax fails (if there are lots of them,
2979 the message gets confusing).
2981 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2982 names, as there are now some PTR records that contain slashes. This check
2983 is only to protect against broken name servers that fall over on strange
2984 characters, so the fact that it applies to all lookups doesn't matter.
2986 PH/36 Now that the new test suite is complete, we can remove some of the
2987 special code in Exim that was needed for the old test suite. For example,
2988 sorting DNS records because real resolvers return them in an arbitrary
2989 order. The new test suite's fake resolver always returns records in the
2992 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2993 bounces) except when queue_only is set, to avoid logging races between
2994 the different processes.
2996 PH/38 Panic-die if .include specifies a non-absolute path.
2998 PH/39 A tweak to the "H" retry rule from its user.
3000 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3001 a label. They prevented compilation on older perls.
3003 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3004 a warning to be raised on newish perls.
3006 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3007 on queue. Changes to match documented behaviour of showing count of
3008 messages matching specified criteria.
3010 PH/40 Changed the default ident timeout from 30s to 5s.
3012 PH/41 Added support for the use of login_cap features, on those BSD systems
3013 that have them, for controlling the resources used by pipe deliveries.
3015 PH/42 The content-scanning code uses fopen() to create files in which to put
3016 message data. Previously it was not paying any attention to the mode of
3017 the files. Exim runs with umask(0) because the rest of the code creates
3018 files with open(), and sets the required mode explicitly. Thus, these
3019 files were ending up world-writeable. This was not a big issue, because,
3020 being within the spool directory, they were not world-accessible. I have
3021 created a function called modefopen, which takes an additional mode
3022 argument. It sets umask(777), creates the file, chmods it to the required
3023 mode, then resets the umask. All the relevant calls to fopen() in the
3024 content scanning code have been changed to use this function.
3026 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3027 to 24 hours. This avoids potential overflow problems when processing G
3028 and H retry rules. I suspect nobody ever tinkers with this value.
3030 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3032 PH/45 When the plaintext authenticator is running as a client, the server's
3033 challenges are checked to ensure they are valid base64 strings. By
3034 default, the authentication attempt is cancelled if an invalid string is
3035 received. Setting client_ignore_invalid_base64 true ignores these errors.
3036 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3037 they are received. Thus, the responses can be made to depend on the
3038 challenges. If an invalid string is ignored, an empty string is placed in
3041 PH/46 Messages that are created by the autoreply transport now contains a
3042 References: header, in accordance with RFCs 2822 and 3834.
3044 PH/47 Added authenticated_sender_force to the smtp transport.
3046 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3048 PH/49 Installed latest patch from the Sieve maintainer.
3050 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3051 was also set, the appendfile transport was unnecessarily scanning a
3052 directory of message files (e.g. for maildir delivery) to find the count
3053 of files (along with the size), even though it did not need this
3054 information. It now does the scan only if it needs to find either the
3055 size of the count of files.
3057 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3059 PH/52 Two bugs concerned with error handling when the smtp transport is
3062 (i) Exim was not creating retry information for temporary errors given
3063 for individual recipients after the DATA command when the smtp transport
3064 was used in LMTP mode. This meant that they could be retried too
3065 frequently, and not timed out correctly.
3067 (ii) Exim was setting the flag that allows error details to be returned
3068 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3069 recipients that were returned after the DATA command.
3071 PH/53 This is related to PH/52, but is more general: for any failing address,
3072 when detailed error information was permitted to be returned to the
3073 sender, but the error was temporary, then after the final timeout, only
3074 "retry timeout exceeded" was returned. Now it returns the full error as
3075 well as "retry timeout exceeded".
3077 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3078 do this, and (what is worse) MTAs that accept it.
3080 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3081 will now be deprecated.
3083 PH/56 New os.c-cygwin from the Cygwin maintainer.
3085 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3086 formats (previously only available in exim formats via -bpr, -bpru,
3087 and -bpra. Now also available in native and exiqgrep formats)
3089 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3090 with very large, slow to parse queues
3092 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3094 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3096 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3097 responses to authentication challenges, though it was showing the
3098 challenges; (ii) I've removed the CR characters from the debug output for
3101 PH/58 Allow for the insertion of a newline as well as a space when a string
3102 is turned into more than one encoded-word during RFC 2047 encoding. The
3103 Sieve code now uses this.
3105 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3106 data_4xx, lost_connection, tls_required.
3108 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3109 message was being sent as an SMTP response.
3111 PH/61 Add -l and -k options to exicyclog.
3113 PH/62 When verifying, if an address was redirected to one new address, so that
3114 verification continued, and the new address failed or deferred after
3115 having set something in $address_data, the value of $address_data was not
3116 passed back to the ACL. This was different to the case when no
3117 redirection occurred. The value is now passed back in both cases.
3119 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3120 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3121 use login_cap.h, so on its own it isn't the distinguishing feature. The
3122 new name refers directly to the setclassresources() function.
3124 PH/65 Added configuration files for NetBSD3.
3126 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3128 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3129 is preferred over IPv4.
3131 PH/68 The bounce_return_message and bounce_return_body options were not being
3132 honoured for bounces generated during the reception of non-SMTP messages.
3133 In particular, this applied to messages rejected by the ACL. This bug has
3134 been fixed. However, if bounce_return_message is true and bounce_return_
3135 body is false, the headers that are returned for a non-SMTP message
3136 include only those that have been read before the error was detected.
3137 (In the case of an ACL rejection, they have all been read.)
3139 PH/69 The HTML version of the specification is now built in a directory called
3140 spec_html instead of spec.html, because the latter looks like a path with
3141 a MIME-type, and this confuses some software.
3143 PH/70 Catch two compiler warnings in sieve.c.
3145 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3146 function verify_get_ident() calls ip_connect() to connect a socket, but
3147 if the "connect()" function timed out, ip_connect() used to close the
3148 socket. However, verify_get_ident() also closes the socket later, and in
3149 between Exim writes to the log, which may get opened at this point. When
3150 the socket was closed in ip_connect(), the log could get the same file
3151 descriptor number as the socket. This naturally causes chaos. The fix is
3152 not to close the socket in ip_connect(); the socket should be closed by
3153 the function that creates it. There was only one place in the code where
3154 this was missing, in the iplookup router, which I don't think anybody now
3155 uses, but I've fixed it anyway.
3157 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3158 well as to direct DNS lookups. Otherwise the handling of names in host
3159 lists is inconsistent and therefore confusing.
3165 PH/01 Two changes to the default runtime configuration:
3167 (1) Move the checks for relay_from_hosts and authenticated clients from
3168 after to before the (commented out) DNS black list checks.
3170 (2) Add control=submission to the relay_from_hosts and authenticated
3171 clients checks, on the grounds that messages accepted by these
3172 statements are most likely to be submissions.
3174 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3176 (1) Generate an error if the third argument for the ${prvs expansion is
3179 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3182 (3) Reset the variables that are obtained from the first argument of
3183 ${prvscheck and used in the second argument before leaving the code,
3184 because their memory is reclaimed, so using them afterwards may do
3187 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3188 one (it's much easier than Tom thought :-).
3190 (5) Because of (4), we can now allow for the use of $prvscheck_result
3191 inside the third argument.
3193 PH/03 For some reason, the default setting of PATH when running a command from
3194 a pipe transport was just "/usr/bin". I have changed it to
3197 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3198 anything to be listed in the output from -bV.
3200 PH/05 When a filter generated an autoreply, the entire To: header line was
3201 quoted in the delivery log line, like this:
3203 => >A.N.Other <ano@some.domain> <original@ddress> ...
3205 This has been changed so that it extracts the operative address. There
3206 may be more than one such address. If so, they are comma-separated, like
3209 => >ano@some.domain,ona@other.domain <original@ddress> ...
3211 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3212 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3213 not being looked up in the rDNS to get a host name, Exim was showing the
3214 IP address twice in Received: lines, even though the IP addresses were
3215 identical. For example:
3217 Received: from [1.2.3.4] (helo=[1.2.3.4])
3219 However, if the real host name was known, it was omitting the HELO data
3220 if it matched the actual IP address. This has been tidied up so that it
3221 doesn't show the same IP address twice.
3223 PH/07 When both +timestamp and +memory debugging was on, the value given by
3224 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3225 called by the debug printing, thereby overwriting the timestamp buffer.
3226 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3228 PH/08 When the original message was included in an autoreply transport, it
3229 always said "this is a copy of the message, including all the headers",
3230 even if body_only or headers_only was set. It now gives an appropriate
3233 PH/09 Applied a patch from the Sieve maintainer which:
3235 o fixes some comments
3236 o adds the (disabled) notify extension core
3237 o adds some debug output for the result of if/elsif tests
3238 o points to the current vacation draft in the documentation
3239 and documents the missing references header update
3243 o fixes a bug in processing the envelope test (when testing
3244 multiple envelope elements, the last element determined the
3247 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3248 Electronic Mail") by including:
3250 Auto-submitted: auto-generated
3252 in the messages that it generates (bounce messages and others, such as
3253 warnings). In the case of bounce messages for non-SMTP messages, there was
3254 also a typo: it was using "Auto_submitted" (underscore instead of
3255 hyphen). Since every message generated by Exim is necessarily in response
3256 to another message, thes have all been changed to:
3258 Auto-Submitted: auto-replied
3260 in accordance with these statements in the RFC:
3262 The auto-replied keyword:
3264 - SHOULD be used on messages sent in direct response to another
3265 message by an automatic process,
3267 - MUST NOT be used on manually-generated messages,
3269 - MAY be used on Delivery Status Notifications (DSNs) and Message
3270 Disposition Notifications (MDNs),
3272 - MUST NOT be used on messages generated by automatic or periodic
3273 processes, except for messages which are automatic responses to
3276 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3277 to the default Received: header definition.
3279 PH/12 Added log selector acl_warn_skipped (default on).
3281 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3282 variables because (a) they are in the wrong storage pool and (b) even if
3283 they were copied, it wouldn't work properly because of the caching.
3285 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3286 checking when decoding. Apparently there are clients that generate
3287 overlong encoded strings. Why am I not surprised?
3289 PH/15 If the first argument of "${if match_address" was not empty, but did not
3290 contain an "@" character, Exim crashed. Now it writes a panic log message
3291 and treats the condition as false.
3293 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3295 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3296 extension "envelope-auth". The code is finished and in agreement with
3297 other implementations, but there is no documentation so far and in fact,
3298 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3299 not changing the active code.
3301 Print executed "if" and "elsif" statements when debugging is used. This
3302 helps a great deal to understand what a filter does.
3304 Document more things not specified clearly in RFC3028. I had all this
3305 sorted out, when out of a sudden new issues came to my mind. Oops."
3307 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3310 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3311 canonical form (as documented). However, after a host name lookup from
3312 the IP address, check_host() was doing a simple string comparison with
3313 addresses acquired from the DNS when checking that the found name did
3314 have the original IP as one of its addresses. Since any found IPv6
3315 addresses are likely to be in abbreviated form, the comparison could
3316 fail. Luckily, there already exists a function for doing the comparison
3317 by converting both addresses to binary, so now that is used instead of
3318 the text comparison.
3320 PH/20 There was another similar case to PH/19, when a complete host name was
3321 given in a host list; looking up its IP address could give an abbreviated
3322 form, whereas the current host's name might or might not be abbreviated.
3323 The same fix has been applied.
3329 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3330 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3333 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3334 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3336 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3338 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3339 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3340 Also, for "Auto-Submitted", treat anything other than "no" as
3341 non-personal, in accordance with RFC 3834. (Previously it treated
3342 anything starting "auto-" as non-personal.)
3344 TF/01 The control=submission/name=... option had a problem with syntax
3345 errors if the name included a slash character. The /name= option
3346 now slurps the rest of the string, so it can include any characters
3347 but it must come last in the list of options (after /sender_retain
3350 PH/05 Some modifications to the interface to the fake nameserver for the new
3358 TK/01 Added the "success_on_redirect" address verification option. See
3359 NewStuff for rationale and an example.
3361 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3363 PH/02 Patch to exigrep to allow it to work on syslog lines.
3365 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3366 fread() to skip over the body file's header line, because in Cygwin the
3367 header line is locked and is inaccessible.
3369 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3370 co-exist for some time) to make it clear that it is the Exim ID that is
3371 referenced, not the Message-ID: header line.
3373 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3374 string_format() function, because snprintf() does not exist on all
3377 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3378 ${stat: expansion item.
3380 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3381 protocol synchronization error", to keep the pedants happy.
3383 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3384 well as for IRIX systems, when gcc is being used. See the host.c source
3387 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3389 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3392 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3393 transport and to the smtp transport in LMTP mode.
3395 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3397 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3398 run a filter in a subprocess. This could lead to confusion in subsequent
3399 lookups in the parent process. There should also be a search_tidyup() at
3400 the end of the subprocess.
3402 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3403 only if the host matched helo_try_verify_hosts, which caused the
3404 verification to occur when the EHLO/HELO command was issued. The ACL just
3405 tested the remembered result. Now, if a previous verification attempt has
3406 not happened, "verify = helo" does it there and then.
3408 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3410 TK/03 Fix log output including CR from clamd.
3412 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3413 exist provoked a memory error which could cause a segfault.
3415 PH/15 Installed PCRE 6.2
3417 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3419 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3420 of the problem. Specifically, suggested +O2 rather than +O1 for the
3423 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3425 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3426 record, the port was not correctly logged when the outgoing_port log
3427 selector was set (it logged the transort's default port).
3429 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3430 fallback_hosts, and "hosts" in the smtp transport.
3432 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3433 host errors such as "Connection refused".
3435 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3436 authentication with radiusclient 0.4.9:
3438 - Error returned from rc_read_config was caught wrongly
3439 - Username/password not passed on to radius server due to wrong length.
3441 The presumption is that some radiusclient API changes for 4.51/PH/17
3442 were not taken care of correctly. The code is still untested by me (my
3443 Linux distribution still has 0.3.2 of radiusclient), but it was
3444 contributed by a Radius user.
3446 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3447 expanding the "port" option of the smtp transport.
3449 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3450 while reading a MIME header. Thanks to Tom Hughes for a patch.
3452 PH/24 Include config.h inside local_scan.h so that configuration settings are
3455 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3456 that in an ACL for RCPT (for example), you can examine exactly what was
3459 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3460 commands, but it was not correctly comparing the address with the actual
3461 client host address. Thus, it would show the EHLO address in Received:
3462 header lines when this was not necessary.
3464 PH/27 Added the % operator to ${eval:}.
3466 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3467 it should be ignoring failures (because with -C, for example, it has lost
3468 privilege). It wasn't ignoring creation failures other than "already
3471 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3474 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3475 because an input line was too long, either on its own, or by virtue of
3476 too many continuations, the temporary file was not being removed, and the
3477 return code was incorrect.
3479 PH/31 Missing "BOOL" in function definition in filtertest.c.
3481 PH/32 Applied Sieve patches from the maintainer.
3483 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3485 PH/33 Added "verify = not_blind".
3487 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3488 Local/Makefile (with some defaults set). These are used in built scripts
3489 such as exicyclog, but they have never been used in the exim_install
3490 script (though there are many overriding facilities there). I have
3491 arranged that the exim_install script now takes note of these two
3494 PH/35 Installed configuration files for Dragonfly.
3496 PH/36 When a locally submitted message by a trusted user did not contain a
3497 From: header, and the sender address was obtained from -f or from an SMTP
3498 MAIL command, and the trusted user did not use -F to supply a sender
3499 name, $originator_name was incorrectly used when constructing a From:
3500 header. Furthermore, $originator_name was used for submission mode
3501 messages from external hosts without From: headers in a similar way,
3502 which is clearly wrong.
3504 PH/37 Added control=suppress_local_fixups.
3506 PH/38 When log_selector = +received_sender was set, and the addition of the
3507 sender made the log line's construction buffer exactly full, or one byte
3508 less than full, an overflow happened when the terminating "\n" was
3511 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3512 when the result of a list match is failure because a DNS lookup failed.
3514 PH/40 RM_COMMAND is now used in the building process.
3516 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3517 the "build-* directories that it finds.
3519 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3520 address in a domain literal was a prefix of an interface address.
3522 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3523 when verifying a sender address, unless rewrite_headers is false.
3525 PH/44 Wrote a long comment about why errors_to addresses are verified as
3526 recipients, not senders.
3528 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3529 the ratelimit ACL was added.
3531 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3533 PH/46 Added extra information about PostgreSQL errors to the error string.
3535 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3536 avoiding the need to install special zones in a real server. This is
3537 backwards compatible; if it can't find the fake resolver, it drops back.
3538 Thus, both old and new test suites can be run.
3540 TF/02 Added util/ratelimit.pl
3542 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3543 clock is set back in time.
3545 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3546 Candler <B.Candler@pobox.com>.
3548 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3549 OK for addresses that are the result of redirections.
3551 PH/48 A number of further additions for the benefit of the new test suite,
3552 including a fake gethostbyname() that interfaces to the fake DNS resolver
3555 TF/06 The fix for widen_domains has also been applied to qualify_single and
3556 search_parents which are the other dnslookup options that can cause
3559 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3562 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3564 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3565 service" if it encountered an MX record with an empty target hostname.
3566 The message is now "an MX or SRV record indicated no SMTP service".
3568 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3569 if the DNS of the sending site is misconfigured. This is quite a
3570 common situation. This change restores the behaviour of treating a
3571 helo verification defer as a failure.
3573 PH/51 If self=fail was set on a router, the bounce message did not include the
3574 actual error message.
3580 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3582 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3583 command itself ended in error, the underlying message about the transport
3584 filter timeout was being overwritten with the pipe command error. Now the
3585 underlying error message should be appended to the second error message.
3587 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3589 PH/02 Reduce the amount of output that "make" produces by default. Full output
3590 can still be requested.
3592 PH/03 The warning log line about a condition test deferring for a "warn" verb
3593 was being output only once per connection, rather than after each
3594 occurrence (because it was using the same function as for successful
3595 "warn" verbs). This seems wrong, so I have changed it.
3597 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3598 it should not have, which might have caused a crash in the right
3599 circumstances, but probably never did.
3601 PH/04 Installed a modified version of Tony Finch's patch to make submission
3602 mode fix the return path as well as the Sender: header line, and to
3603 add a /name= option so that you can make the user's friendly name appear
3606 TF/03 Added the control = fakedefer ACL modifier.
3608 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3609 Mark Lowes for thorough testing.
3611 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3613 TK/03 Merged latest SRS patch from Miles Wilton.
3615 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3616 with the definition in sysexits.h (which is #included earlier).
3617 Fortunately, Exim does not actually use EX_OK. The code used to try to
3618 preserve the sysexits.h value, by assuming that macro definitions were
3619 scanned for macro replacements. I have been disabused of this notion,
3620 so now the code just undefines EX_OK before #including unistd.h.
3622 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3623 in the smtp transport. When a block could not be written in a single
3624 write() function, the timeout was being re-applied to each part-write.
3625 This seems wrong - if the receiver was accepting one byte at a time it
3626 would take for ever. The timeout is now adjusted when this happens. It
3627 doesn't have to be particularly precise.
3629 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3630 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3632 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3633 without a domain if the check to <postmaster@domain> fails.
3635 SC/01 Eximstats: added -xls and the ability to specify output files
3636 (patch written by Frank Heydlauf).
3638 SC/02 Eximstats: use FileHandles for outputting results.
3640 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3642 SC/04 Eximstats: fixed display of large numbers with -nvr option
3644 SC/05 Eximstats: fixed merging of reports with empty tables.
3646 SC/06 Eximstats: added the -include_original_destination flag
3648 SC/07 Eximstats: removed tabs and trailing whitespace.
3650 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3652 TK/06 MBOX spool code: Add real "From " MBOX separator line
3653 so the .eml file is really in mbox format (even though
3654 most programs do not really care). Patch from Alex Miller.
3656 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3657 The latter is generated from $received_to and is only set if the
3658 message has one envelope recipient. SA can use these headers,
3659 obviously out-of-the-box. Patch from Alex Miller.
3661 PH/08 The ${def test on a variable was returning false if the variable's
3662 value was "0", contrary to what the specification has always said!
3663 The result should be true unless the variable is empty.
3665 PH/09 The syntax error of a character other than { following "${if
3666 def:variable_name" (after optional whitespace) was not being diagnosed.
3667 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3668 accidental colon was present, for example, could give incorrect results.
3670 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3671 result is used (not including appendfile, where other changes are about
3674 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3675 This involved changing a lot of size variables from int to off_t. It
3676 should work with maildirs and everything.
3678 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3679 spamd dying while we are connected to it.
3681 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3684 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3685 function for BDB 4.3.
3687 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3689 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3690 the sources for PCRE have been re-arranged and more files are now
3693 PH/15 The code I had for printing potentially long long variables in PH/11
3694 above was not the best (it lost precision). The length of off_t variables
3695 is now inspected at build time, and an appropriate printing format (%ld
3696 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3697 to be "long long int" or "long int". This is needed for the internal
3698 formatting function string_vformat().
3700 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3701 the configuration file to be ":syslog", then the script "guesses" where
3702 the logs files are, rather than using the compiled in default. In our
3703 case the guess is not the same as the compiled default, so the script
3704 suddenly stopped working when I started to use syslog. The patch checks
3705 to see if log_file_path is "". If so, it attempts to read it from exim
3706 with no configuration file to get the compiled in version, before it
3707 falls back to the previous guessing code."
3709 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3710 implementing BATV in an Exim configuration. See NewStuff for the gory
3713 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3714 Makefile that are specific to HP-UX.
3716 PH/18 If the "use_postmaster" option was set for a recipient callout together
3717 with the "random" option, the postmaster address was used as the MAIL
3718 FROM address for the random test, but not for the subsequent recipient
3719 test. It is now used for both.
3721 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3722 patch removes a few documentation additions to RFC 3028, because the
3723 latest draft now contains them. It adds the new en;ascii-case comparator
3724 and a new error check for 8bit text in MIME parts. Comparator and
3725 require names are now matched exactly. I enabled the subaddress
3726 extension, but it is not well tested yet (read: it works for me)."
3728 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3729 rework some of the code of TK/09 above to avoid the hardwired use of
3730 "%lld" and "long long". Replaced the call to snprintf() with a call to
3733 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3734 records point to non-existent hosts", "retry timeout exceeded", and
3735 "retry time not reached for any host after a long failure period".
3737 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3738 experimental DomainKeys support:
3740 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3741 (2) On an error such as an illegally used "control", the wrong name for
3742 the control was given.
3744 These problems did NOT occur unless DomainKeys support was compiled.
3746 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3748 PH/24 Added ${if match_ip condition.
3750 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3751 databases so that it will be absolutely obvious if a crash occurs in the
3752 DB library. This is a regular occurrence (often caused by mis-matched
3755 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3756 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3757 picked up on a user's system that detects such things. There doesn't seem
3758 to be a gcc warning option for this - only an attribute that has to be
3759 put on the function's prototype. It seems that in Fedora Core 4 they have
3760 set this on a number of new functions. No doubt there will be more in due
3763 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3764 specify a transport. However, if an address that was verified by such a
3765 router was the subject of a callout, Exim crashed because it tried to
3766 read the rcpt_include_affixes from the non-existent transport. Now it
3767 just assumes that the setting of that option is false. This bug was
3768 introduced by 4.51/PH/31.
3770 PH/28 Changed -d+all to exclude +memory, because that information is very
3771 rarely of interest, but it makes the output a lot bigger. People tend to
3772 do -d+all out of habit.
3774 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3775 code in os-type was giving problems when libc.so lives in lib64, like on
3778 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3779 aren't the modern standard, and it seems that some systems' include files
3780 don't always have them. Exim was already checking for some of the newer
3781 ones like T_AAAA, and defining it itself. I've added checks for all the
3782 record types that Exim uses.
3784 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3785 not automatically generating a new one, as it is supposed to. This
3786 prevented TLS from working. If the file did exist, but contained invalid
3787 data, a new version was generated, as expected. It was only the case of a
3788 non-existent file that was broken.
3790 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3791 with a change in libdomainkeys > 0.64.
3793 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3794 from DNS. If the selector record carries the flag, it now has
3795 precedence over the domain-wide flag.
3797 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3799 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3800 the use of a port name that isn't defined in /etc/services) occurred, the
3801 message was deferred as in a normal delivery, and thus remained on the
3802 spool, instead of being failed because of the mua_wrapper setting. This
3803 is now fixed, and I tidied up some of the mua_wrapper messages at the
3806 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3807 the messages in a hash of arrays rather than using individual hashes.
3808 This is a bit cleaner and results in dramatic memory savings, albeit
3809 at a slight CPU cost.
3811 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3812 as requested by Marc Sherman.
3814 SC/10 Eximstats: added histograms for user specified patterns as requested
3817 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3819 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3820 fopen() in the content-scanning modules that did not already have it.
3826 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3827 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3829 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3831 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3833 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3834 to test Sieve filters that use "vacation".
3836 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3837 that changes the way the GnuTLS parameters are stored in the cache file.
3838 The new format can be generated externally. For backward compatibility,
3839 if the data in the cache doesn't make sense, Exim assumes it has read an
3840 old-format file, and it generates new data and writes a new file. This
3841 means that you can't go back to an older release without removing the
3844 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3845 work if there are any delivery delays because "one_time" forces the
3846 parent to be marked "delivered", so its unseen clone is never tried
3847 again. For this reason, Exim now forbids the simultaneous setting of
3850 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3851 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3852 entry if you want to know the details. The fix, however, neglected to
3853 consider the case when local delivery batching is involved. The test for
3854 "previously delivered" was not happening when checking to see if an
3855 address could be batched with a previous (undelivered) one; under
3856 certain circumstances this could lead to multiple deliveries to the same
3859 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3860 in its include files, and this causes problems building Exim.
3862 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3863 header_syntax) but Exim was just ignoring anything given after a slash.
3864 In particular, this caused confusion with an attempt to use "verify =
3865 reverse_host_lookup/defer_ok". An error is now given when options are
3866 supplied for verify items that do not have them. (Maybe reverse_host_
3867 lookup should have a defer_ok option, but that's a different point.)
3869 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3870 defined by RFC 821) to 2048, because there were problems with some AUTH
3871 commands, and RFC 1869 says the size should be increased for extended
3872 SMTP commands that take arguments.
3874 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3877 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3878 "unknown" error; now it says that the functionality isn't in the binary.
3880 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3881 an address' error message when a string expansion fails (syntax or
3882 whatever). Otherwise the password may appear in the log. Following change
3883 PH/42 below, there is no longer a chance of it appearing in a bounce
3886 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3888 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3889 of its addresses was ever tried. (Bugzilla bug #2.)
3891 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3892 the result incorrectly in the debug output. (It correctly added a newline
3893 to what was transported.)
3895 TF/01 Added $received_time.
3897 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3898 commented out examples of how to interface to a virus scanner and to
3899 SpamAssassin. Also added commented examples of av_scanner and
3900 spamd_address settings.
3902 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3903 and controls are allowed in which ACLs. There were a couple of minor
3904 errors. Some of the entries in the conditions table (which is a table of
3905 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3906 negation of where the condition IS allowed.
3908 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3910 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3911 header file does not have a version number, so I've had to invent a new
3912 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3913 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3914 radiusclient), but it was contributed by a Radius user.
3916 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3917 files or directories, for OpenSSL.
3919 PH/19 When an Exim process that is running as root has to create an Exim log
3920 file, it does so in a subprocess that runs as exim:exim so as to get the
3921 ownership right at creation (otherwise, other Exim processes might see
3922 the file with the wrong ownership). There was no test for failure of this
3923 fork() call, which would lead to the process getting stuck as it waited
3924 for a non-existent subprocess. Forks do occasionally fail when resources
3925 run out. I reviewed all the other calls to fork(); they all seem to check
3928 PH/20 When checking for unexpected SMTP input at connect time (before writing
3929 the banner), Exim was not dealing correctly with a non-positive return
3930 from the read() function. If the client had disconnected by this time,
3931 the result was a log entry for a synchronization error with an empty
3932 string after "input=" when read() returned zero. If read() returned -1
3933 (an event I could not check), uninitialized data bytes were printed.
3934 There were reports of junk text (parts of files, etc) appearing after
3937 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3939 PH/22 Added support for macro redefinition, and (re)definition in between
3940 driver and ACL definitions.
3942 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3943 forgetting to use the resulting value; it was using the unexpanded value.
3945 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3946 hadn't been configured. The fix is from Juergen Kreileder, who
3947 understands it better than I do:
3949 "Here's what I see happening with three configured cyrus_sasl
3950 authenticators configured (plain, login, cram-md5):
3952 On startup auth_cyrus_sasl_init() gets called for each of these.
3953 This means three calls to sasl_listmech() without a specified mech_list.
3954 => SASL tests which mechs of all available mechs actually work
3955 => three warnings about OTP not working
3956 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3958 With the patch, sasl_listmech() also gets called three times. But now
3959 SASL's mech_list option is set to the server_mech specified in the the
3960 authenticator. Or in other words, the answer from sasl_listmech()
3961 gets limited to just the mech you're testing for (which is different
3963 => the return list contains just 'plain' or 'login', 'cram-md5' or
3964 nothing depending on the value of ob->server_mech.
3966 I've just tested the patch: Authentication still works fine,
3967 unavailable mechs specified in the exim configuration are still
3968 caught, and the auth.log warnings about OTP are gone."
3970 PH/25 When debugging is enabled, the contents of the command line are added
3971 to the debugging output, even when log_selector=+arguments is not
3974 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3975 answer is "GNU", and only if the return is "GNU/something" is the answer
3978 PH/27 $acl_verify_message is now set immediately after the failure of a
3979 verification in an ACL, and so is available in subsequent modifiers. In
3980 particular, the message can be preserved by coding like this:
3982 warn !verify = sender
3983 set acl_m0 = $acl_verify_message
3985 Previously, $acl_verify_message was set only while expanding "message"
3986 and "log_message" when a very denied access.
3988 PH/28 Modified OS/os.c-Linux with
3990 -#ifndef OS_LOAD_AVERAGE
3991 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3993 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3994 nomenclature these days.)
3996 PH/29 Installed patch from the Sieve maintainer that adds the options
3997 sieve_useraddress and sieve_subaddress to the redirect router.
3999 PH/30 In these circumstances:
4000 . Two addresses routed to the same list of hosts;
4001 . First host does not offer TLS;
4002 . First host accepts first address;
4003 . First host gives temporary error to second address;
4004 . Second host offers TLS and a TLS session is established;
4005 . Second host accepts second address.
4006 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4007 and peerdn, if requested) that were in fact used only for the second
4010 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4011 attention to any local part prefix or suffix that was matched by the
4012 router that accepted the address. It now behaves in the same way as it
4013 does for delivery: the affixes are removed from the local part unless
4014 rcpt_include_affixes is set on the transport.
4016 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4017 timeout during the DATA phase of an incoming message.
4019 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4020 applied a patch sanctioned by the Sieve maintainer.
4022 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4023 the uid or gid is negative. A case of a negative gid caused this to be
4024 noticed. The fix allows for either to be negative.
4026 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4027 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4030 PH/36 The change PH/12 above was broken. Fixed it.
4032 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4033 the grounds that routing the same address twice would always produce the
4034 same answer. This might have been true once, but it is certainly no
4035 longer true now. Routing a child address may depend on the previous
4036 routing that produced that child. Some complicated redirection strategies
4037 went wrong when messages had multiple recipients, and made Exim's
4038 behaviour dependent on the order in which the addresses were given.
4040 I have moved the duplicate checking until after the routing is complete.
4041 Exim scans the addresses that are assigned to local and remote
4042 transports, and removes any duplicates. This means that more work will be
4043 done, as duplicates will always all be routed, but duplicates are
4044 presumably rare, so I don't expect this is of any significance.
4046 For deliveries to pipes, files, and autoreplies, the duplicate checking
4047 still happens during the routing process, since they are not going to be
4050 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4051 It corrects a timeout issue with spamd. This is Ian's comment: "The
4052 background is that sometimes spamd either never reads data from a
4053 connection it has accepted, or it never writes response data. The exiscan
4054 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4055 blindly assumes that writes won't block so it may never time out."
4057 PH/39 Allow G after quota size as well as K and M.
4059 PH/40 The value set for $authenticated_id in an authenticator may not contain
4060 binary zeroes or newlines because the value is written to log lines and
4061 to spool files. There was no check on this. Now the value is run through
4062 the string_printing() function so that such characters are converted to
4063 printable escape sequences.
4065 PH/41 $message_linecount is a new variable that contains the total number of
4066 lines in the message. Compare $body_linecount, which is the count for the
4069 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4070 bounce and delay warning messages, except in certain special cases, which
4073 (a) An SMTP error message from a remote host;
4074 (b) A message specified in a :fail: redirection;
4075 (c) A message specified in a "fail" command in a system filter;
4076 (d) A message specified in a FAIL return from the queryprogram router;
4077 (e) A message specified by the cannot_route_message router option.
4079 In these cases only, Exim does include the error details in bounce and
4080 warning messages. There are also a few cases where bland messages such
4081 as "unrouteable address" or "local delivery error" are given.
4083 PH/43 $value is now also set for the "else" part of a ${run expansion.
4085 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4086 being worked on, but at least Exim now implements the latest version to
4089 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4090 process to complete was treated as a delivery failure, a timeout while
4091 writing the message to the pipe was logged, but erroneously treated as a
4092 successful delivery. Such timeouts include transport filter timeouts. For
4093 consistency with the overall process timeout, these timeouts are now
4094 treated as errors, giving rise to delivery failures by default. However,
4095 there is now a new Boolean option for the pipe transport called
4096 timeout_defer, which, if set TRUE, converts the failures into defers for
4097 both kinds of timeout. A transport filter timeout is now identified in
4100 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4101 systems where "make" and "gmake" are different, calling "gmake" at top
4102 level broke things. I've arranged for the value of $(MAKE) to be passed
4103 from the Makefile to this script so that it can call the same version of
4107 A note about Exim versions 4.44 and 4.50
4108 ----------------------------------------
4110 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4111 changes of various kinds. As a consequence, a big documentation update was
4112 needed. This delayed the release for rather longer than seemed good, especially
4113 in the light of a couple of (minor) security issues. Therefore, the changes
4114 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4115 release. So 4.44 and 4.50 are in effect two different branches that both start
4118 I have left the 4.50 change log unchanged; it contains all the changes since
4119 4.43. The change log for 4.44 is below; many of its items are identical to
4120 those for 4.50. This seems to be the most sensible way to preserve the
4121 historical information.
4127 1. Minor wording change to the doc/README.SIEVE file.
4129 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4130 computation of the current number of files was incorrect.
4132 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4133 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4136 4. Give more explanation in the error message when the command for a transport
4137 filter fails to execute.
4139 5. There are several places where Exim runs a non-Exim command in a
4140 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4141 was being done only for the command run by the queryprogram router. It is
4142 now done for all such subprocesses. The other cases are: ${run, transport
4143 filters, and the commands run by the lmtp and pipe transports.
4145 6. Added CONFIGURE_GROUP build-time option.
4147 7. Some older OS have a limit of 256 on the maximum number of file
4148 descriptors. Exim was using setrlimit() to set 1000 as a large value
4149 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4150 systems. I've change it so that if it can't get 1000, it tries for 256.
4152 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4153 was an oversight, and furthermore, ever since the addition of extra
4154 controls (e.g. 4.43/32), the checks on when to allow different forms of
4155 "control" were broken. There should now be diagnostics for all cases when a
4156 control that does not make sense is encountered.
4158 9. Added the /retain_sender option to "control=submission".
4160 10. $recipients is now available in the predata ACL (oversight).
4162 11. Tidy the search cache before the fork to do a delivery from a message
4163 received from the command line. Otherwise the child will trigger a lookup
4164 failure and thereby defer the delivery if it tries to use (for example) a
4165 cached ldap connection that the parent has called unbind on.
4167 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4168 of $address_data from the recipient verification was clobbered by the
4169 sender verification.
4171 13. The value of address_data from a sender verification is now available in
4172 $sender_address_data in subsequent conditions in the ACL statement.
4174 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4176 15. Added a new option "connect=<time>" to callout options, to set a different
4179 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4180 was its contents. (It was OK if the option was not defined at all.)
4182 17. A "Completed" log line is now written for messages that are removed from
4183 the spool by the -Mrm option.
4185 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4186 information about exactly what failed.
4188 19. Added -dd to debug only the daemon process.
4190 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4191 handles timeouts, both on the server side and network timeouts. Renamed the
4192 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4194 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4195 It is now set to "smtps".
4197 22. $host_address is now set to the target address during the checking of
4198 ignore_target_hosts.
4200 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4201 being passed; this would have caused $sender_host_name to have been used if
4202 matching the list had actually called for a host name (not very likely,
4203 since this list is usually IP addresses). A host name is now passed as
4206 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4207 code that specifies a non-restarting handler (typically sigaction() in
4208 modern systems) in an attempt to fix a rare and obscure crash bug.
4210 25. Narrowed the window for a race in the daemon that could cause it to ignore
4211 SIGCHLD signals. This is not a major problem, because they are used only to
4212 wake it up if nothing else does.
4214 26. A malformed maildirsize file could cause Exim to calculate negative values
4215 for the mailbox size or file count. Odd effects could occur as a result.
4216 The maildirsize information is now recalculated if the size or filecount
4219 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4220 support for a long time. Removed HAVE_SYS_VFS_H.
4222 28. Installed the latest version of exipick from John Jetmore.
4224 29. In an address list, if the pattern was not a regular expression, an empty
4225 subject address (from a bounce message) matched only if the pattern was an
4226 empty string. Non-empty patterns were not even tested. This was the wrong
4227 because it is perfectly reasonable to use an empty address as part of a
4228 database query. An empty address is now tested by patterns that are
4229 lookups. However, all the other forms of pattern expect the subject to
4230 contain a local part and a domain, and therefore, for them, an empty
4231 address still always fails if the pattern is not itself empty.
4233 30. Exim went into a mad DNS loop when attempting to do a callout where the
4234 host was specified on an smtp transport, and looking it up yielded more
4235 than one IP address.
4237 31. Re-factored the code for checking spool and log partition space into a
4238 function that finds that data and another that does the check. The former
4239 is then used to implement four new variables: $spool_space, $log_space,
4240 $spool_inodes, and $log_inodes.
4242 32. The RFC2047 encoding function was originally intended for short strings
4243 such as real names; it was not keeping to the 75-character limit for
4244 encoded words that the RFC imposes. It now respects the limit, and
4245 generates multiple encoded words if necessary. To be on the safe side, I
4246 have increased the buffer size for the ${rfc2047: expansion operator from
4249 33. It is now permitted to omit both strings after an "if" condition; if the
4250 condition is true, the result is "true". As before, when the second string
4251 is omitted, a false condition yields an empty string. This makes it less
4252 cumbersome to write custom ACL and router conditions.
4254 34. Failure to deliver a bounce message always caused it to be frozen, even if
4255 there was an errors_to setting on the router. The errors_to setting is now
4258 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4259 canonical form (fully expanded) before being placed in
4260 $sender_host_address.
4262 36. The table in the code that translates DNS record types into text (T_A to
4263 "A" for instance) was missing entries for NS and CNAME. It is just possible
4264 that this could have caused confusion if both these types were looked up
4265 for the same domain, because the text type is used as part of Exim's
4266 per-process caching. But the chance of anyone hitting this buglet seems
4269 37. The dnsdb lookup has been extended in a number of ways.
4271 (1) There is a new type, "zns", which walks up the domain tree until it
4272 finds some nameserver records. It should be used with care.
4274 (2) There is a new type, "mxh", which is like "mx" except that it returns
4275 just the host names, not the priorities.
4277 (3) It is now possible to give a list of domains (or IP addresses) to be
4278 looked up. The behaviour when one of the lookups defers can be
4279 controlled by a keyword.
4281 (4) It is now possible to specify the separator character for use when
4282 multiple records are returned.
4284 38. The dnslists ACL condition has been extended: it is now possible to supply
4285 a list of IP addresses and/or domains to be looked up in a particular DNS
4288 39. Added log_selector=+queue_time_overall.
4290 40. When running the queue in the test harness, wait just a tad after forking a
4291 delivery process, to get repeatability of debugging output.
4293 41. Include certificate and key file names in error message when GnuTLS fails
4294 to set them up, because the GnuTLS error message doesn't include the name
4295 of the failing file when there is a problem reading it.
4297 42. Allow both -bf and -bF in the same test run.
4299 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4301 44. The "Exiscan patch" is now merged into the mainline Exim source.
4303 45. Sometimes the final signoff response after QUIT could fail to get
4304 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4305 < 0 before doing a fflush(). This bug looks as though it goes back to the
4306 introduction of TLS in release 3.20, but "sometimes" must have been rare
4307 because the tests only now provoked it.
4309 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4310 (this can affect the format of dates).
4312 47. exim_tidydb, when checking for the continued existence of a message for
4313 which it has found a message-specific retry record, was not finding
4314 messages that were in split spool directories. Consequently, it was
4315 deleting retry records that should have stayed in existence.
4317 48. Steve fixed some bugs in eximstats.
4319 49. The SPA authentication driver was not abandoning authentication and moving
4320 on to the next authenticator when an expansion was forced to fail,
4321 contradicting the general specification for all authenticators. Instead it
4322 was generating a temporary error. It now behaves as specified.
4324 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4325 (the order specifies the preference for clients). The order is now AES256,
4326 AES128, 3DES, ARCFOUR128.
4328 51. Small patch to Sieve code - explicitly set From: when generating an
4331 52. Exim crashed if a remote delivery caused a very long error message to be
4332 recorded - for instance if somebody sent an entire SpamAssassin report back
4333 as a large number of 550 error lines. This bug was coincidentally fixed by
4334 increasing the size of one of Exim's internal buffers (big_buffer) that
4335 happened as part of the Exiscan merge. However, to be on the safe side, I
4336 have made the code more robust (and fixed the comments that describe what
4339 53. Now that there can be additional text after "Completed" in log lines (if
4340 the queue_time_overall log selector is set), a one-byte patch to exigrep
4341 was needed to allow it to recognize "Completed" as not the last thing in
4344 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4345 patch that reportedly fixes this has been added. I am not expert enough to
4346 create a test for it. This is what the patch creator wrote:
4348 "I found a little strange behaviour of ldap code when working with
4349 Windows 2003 AD Domain, where users was placed in more than one
4350 Organization Units. When I tried to give exim partial DN, the exit code
4351 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4352 But simultaneously result of request was absolutely normal ldap result,
4353 so I produce this patch..."
4355 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4356 REFERENCE, so I have modified the code to exclude the patch when that macro
4359 55. Some experimental protocols are using DNS PTR records for new purposes. The
4360 keys for these records are domain names, not reversed IP addresses. The
4361 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4362 leaves it alone. Component reversal etc. now happens only for IP addresses.
4365 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4367 57. Double the size of the debug message buffer (to 2048) so that more of very
4368 long debug lines gets shown.
4370 58. The exicyclog utility now does better if the number of log files to keep
4371 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4373 59. Two changes related to the smtp_active_hostname option:
4375 (1) $smtp_active_hostname is now available as a variable.
4376 (2) The default for smtp_banner uses $smtp_active_hostname instead
4377 of $primary_hostname.
4379 60. The host_aton() function is supposed to be passed a string that is known
4380 to be a valid IP address. However, in the case of IPv6 addresses, it was
4381 not checking this. This is a hostage to fortune. Exim now panics and dies
4382 if the condition is not met. A case was found where this could be provoked
4383 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4384 components; fortuitously, this particular loophole had already been fixed
4385 by change 4.50/55 above.
4387 If there are any other similar loopholes, the new check in host_aton()
4388 itself should stop them being exploited. The report I received stated that
4389 data on the command line could provoke the exploit when Exim was running as
4390 exim, but did not say which command line option was involved. All I could
4391 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4392 running as the user.
4395 61. There was a buffer overflow vulnerability in the SPA authentication code
4396 (which came originally from the Samba project). I have added a test to the
4397 spa_base64_to_bits() function which I hope fixes it.
4400 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4401 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4403 63. The daemon start-up calls getloadavg() while still root for those OS that
4404 need the first call to be done as root, but it missed one case: when
4405 deliver_queue_load_max is set with deliver_drop_privilege. This is
4406 necessary for the benefit of the queue runner, because there is no re-exec
4407 when deliver_drop_privilege is set.
4409 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4410 This has been fixed.
4412 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4413 was in use, was not putting the data itself into the right store pool;
4414 consequently, it could be overwritten for a subsequent message in the same
4415 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4418 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4420 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4421 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4422 the function were treating the return as a boolean value, which happened to
4423 work because 0=false and not-0=true, but is not correct code.
4425 68. The host_aton() function was not handling scoped IPv6 addresses (those
4426 with, for example, "%eth0" on the end) correctly.
4428 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4429 negated items (that is, ~something) in unsigned ints. Some compilers
4430 apparently mutter when there is no cast.
4432 70. If an address verification called from an ACL failed, and did not produce a
4433 user-specific message (i.e. there was only a "system" message), nothing was
4434 put in $acl_verify_message. In this situation, it now puts the system
4437 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4438 session; change 4.31/43 added the unwanted input to the log line - except
4439 that it did not do this in the start of session case. It now does.
4441 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4442 This is wrong and can cause the other end to generate a synchronization
4443 error if it is another Exim or anything else that does the synchronization
4444 check. A QUIT command is no longer sent after a timeout.
4446 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4447 during host lookups.
4449 74. The defer_ok option of callout verification was not working if it was used
4450 when verifying addresses in header lines, that is, for this case:
4452 verify = header_sender/callout=defer_ok
4454 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4455 those file descriptors could be used for SMTP connections. If anything
4456 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4457 could be sent to the SMTP client, causing chaos. The daemon now opens
4458 stdin, stdout, and stderr to /dev/null when it puts itself into the
4461 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4462 log by default. The user can override this with suitable Perl magic.
4464 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4465 the log message when discard triggers, was not working for the DATA ACL or
4466 for the non-SMTP ACL.
4468 78. Error message wording change in sieve.c.
4470 79. If smtp_accept_max_per_host was set, the number of connections could be
4471 restricted to fewer than expected, because the daemon was trying to set up
4472 a new connection before checking whether the processes handling previous
4473 connections had finished. The check for completed processes is now done
4474 earlier. On busy systems, this bug wouldn't be noticed because something
4475 else would have woken the daemon, and it would have reaped the completed
4478 80. If a message was submitted locally by a user whose login name contained one
4479 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4480 It caused a spool format error. I have fixed the spool reading code. A
4481 related problem was that the "from" clause in the Received: line became
4482 illegal because of the space(s). It is now covered by ${quote_local_part.
4484 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4487 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4489 83. Patch from Sieve maintainer to fix unterminated string problem in
4490 "vacation" handling.
4492 84. Some minor changes to the Linux configuration files to help with other
4493 OS variants using glibc.
4495 85. One more patch for Sieve to update vacation handling to latest spec.
4498 ----------------------------------------------------
4499 See the note above about the 4.44 and 4.50 releases.
4500 ----------------------------------------------------
4506 1. Change 4.43/35 introduced a bug that caused file counts to be
4507 incorrectly computed when quota_filecount was set in an appendfile
4510 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4511 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4514 3. Give more explanation in the error message when the command for a transport
4515 filter fails to execute.
4517 4. There are several places where Exim runs a non-Exim command in a
4518 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4519 was being done only for the command run by the queryprogram router. It is
4520 now done for all such subprocesses. The other cases are: ${run, transport
4521 filters, and the commands run by the lmtp and pipe transports.
4523 5. Some older OS have a limit of 256 on the maximum number of file
4524 descriptors. Exim was using setrlimit() to set 1000 as a large value
4525 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4526 systems. I've change it so that if it can't get 1000, it tries for 256.
4528 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4529 was an oversight, and furthermore, ever since the addition of extra
4530 controls (e.g. 4.43/32), the checks on when to allow different forms of
4531 "control" were broken. There should now be diagnostics for all cases when a
4532 control that does not make sense is encountered.
4534 7. $recipients is now available in the predata ACL (oversight).
4536 8. Tidy the search cache before the fork to do a delivery from a message
4537 received from the command line. Otherwise the child will trigger a lookup
4538 failure and thereby defer the delivery if it tries to use (for example) a
4539 cached ldap connection that the parent has called unbind on.
4541 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4542 of $address_data from the recipient verification was clobbered by the
4543 sender verification.
4545 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4546 was its contents. (It was OK if the option was not defined at all.)
4548 11. A "Completed" log line is now written for messages that are removed from
4549 the spool by the -Mrm option.
4551 12. $host_address is now set to the target address during the checking of
4552 ignore_target_hosts.
4554 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4555 being passed; this would have caused $sender_host_name to have been used if
4556 matching the list had actually called for a host name (not very likely,
4557 since this list is usually IP addresses). A host name is now passed as
4560 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4561 code that specifies a non-restarting handler (typically sigaction() in
4562 modern systems) in an attempt to fix a rare and obscure crash bug.
4564 15. Narrowed the window for a race in the daemon that could cause it to ignore
4565 SIGCHLD signals. This is not a major problem, because they are used only to
4566 wake it up if nothing else does.
4568 16. A malformed maildirsize file could cause Exim to calculate negative values
4569 for the mailbox size or file count. Odd effects could occur as a result.
4570 The maildirsize information is now recalculated if the size or filecount
4573 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4574 support for a long time. Removed HAVE_SYS_VFS_H.
4576 18. Updated exipick to current release from John Jetmore.
4578 19. Allow an empty sender to be matched against a lookup in an address list.
4579 Previously the only cases considered were a regular expression, or an
4582 20. Exim went into a mad DNS lookup loop when doing a callout where the
4583 host was specified on the transport, if the DNS lookup yielded more than
4586 21. The RFC2047 encoding function was originally intended for short strings
4587 such as real names; it was not keeping to the 75-character limit for
4588 encoded words that the RFC imposes. It now respects the limit, and
4589 generates multiple encoded words if necessary. To be on the safe side, I
4590 have increased the buffer size for the ${rfc2047: expansion operator from
4593 22. Failure to deliver a bounce message always caused it to be frozen, even if
4594 there was an errors_to setting on the router. The errors_to setting is now
4597 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4598 canonical form (fully expanded) before being placed in
4599 $sender_host_address.
4601 24. Updated eximstats to version 1.33
4603 25. Include certificate and key file names in error message when GnuTLS fails
4604 to set them up, because the GnuTLS error message doesn't include the name
4605 of the failing file when there is a problem reading it.
4607 26. Expand error message when OpenSSL has problems setting up cert/key files.
4610 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4611 (this can affect the format of dates).
4613 28. exim_tidydb, when checking for the continued existence of a message for
4614 which it has found a message-specific retry record, was not finding
4615 messages that were in split spool directories. Consequently, it was
4616 deleting retry records that should have stayed in existence.
4618 29. eximstats updated to version 1.35
4619 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4620 1.35 - bugfix such that pie charts by volume are generated correctly
4622 30. The SPA authentication driver was not abandoning authentication and moving
4623 on to the next authenticator when an expansion was forced to fail,
4624 contradicting the general specification for all authenticators. Instead it
4625 was generating a temporary error. It now behaves as specified.
4627 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4628 (the order specifies the preference for clients). The order is now AES256,
4629 AES128, 3DES, ARCFOUR128.
4631 31. Small patch to Sieve code - explicitly set From: when generating an
4634 32. Exim crashed if a remote delivery caused a very long error message to be
4635 recorded - for instance if somebody sent an entire SpamAssassin report back
4636 as a large number of 550 error lines. This bug was coincidentally fixed by
4637 increasing the size of one of Exim's internal buffers (big_buffer) that
4638 happened as part of the Exiscan merge. However, to be on the safe side, I
4639 have made the code more robust (and fixed the comments that describe what
4642 33. Some experimental protocols are using DNS PTR records for new purposes. The
4643 keys for these records are domain names, not reversed IP addresses. The
4644 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4645 leaves it alone. Component reversal etc. now happens only for IP addresses.
4648 34. The host_aton() function is supposed to be passed a string that is known
4649 to be a valid IP address. However, in the case of IPv6 addresses, it was
4650 not checking this. This is a hostage to fortune. Exim now panics and dies
4651 if the condition is not met. A case was found where this could be provoked
4652 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4653 components; fortuitously, this particular loophole had already been fixed
4654 by change 4.50/55 or 4.44/33 above.
4656 If there are any other similar loopholes, the new check in host_aton()
4657 itself should stop them being exploited. The report I received stated that
4658 data on the command line could provoke the exploit when Exim was running as
4659 exim, but did not say which command line option was involved. All I could
4660 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4661 running as the user.
4664 35. There was a buffer overflow vulnerability in the SPA authentication code
4665 (which came originally from the Samba project). I have added a test to the
4666 spa_base64_to_bits() function which I hope fixes it.
4669 36. The daemon start-up calls getloadavg() while still root for those OS that
4670 need the first call to be done as root, but it missed one case: when
4671 deliver_queue_load_max is set with deliver_drop_privilege. This is
4672 necessary for the benefit of the queue runner, because there is no re-exec
4673 when deliver_drop_privilege is set.
4675 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4676 was in use, was not putting the data itself into the right store pool;
4677 consequently, it could be overwritten for a subsequent message in the same
4678 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4681 38. Sometimes the final signoff response after QUIT could fail to get
4682 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4683 < 0 before doing a fflush(). This bug looks as though it goes back to the
4684 introduction of TLS in release 3.20, but "sometimes" must have been rare
4685 because the tests only now provoked it.
4691 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4692 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4693 Now it returns an integer. A number of calls were still expecting a T/F
4694 return. Fortuitously, in all cases, the tests worked in OK situations,
4695 which is the norm. However, things would have gone wrong on any write
4696 failures on the smtp file descriptor. This function is used when sending
4697 messages over SMTP and also when doing verify callouts.
4699 2. When Exim is called to do synchronous delivery of a locally submitted
4700 message (the -odf or -odi options), it no longer closes stderr before doing
4703 3. Implemented the mua_wrapper option.
4705 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4707 5. Implemented the functions header_remove(), header_testname(),
4708 header_add_at_position(), and receive_remove_recipient(), and exported them
4711 6. If an ACL "warn" statement specified the addition of headers, Exim already
4712 inserted X-ACL-Warn: at the start if there was no header name. However, it
4713 was not making this test for the second and subsequent header lines if
4714 there were newlines in the string. This meant that an invalid header could
4715 be inserted if Exim was badly configured.
4717 7. Allow an ACL "warn" statement to add header lines at the start or after all
4718 the Received: headers, as well as at the end.
4720 8. Added the rcpt_4xx retry error code.
4722 9. Added postmaster_mailfrom=xxx to callout verification option.
4724 10. Added mailfrom=xxxx to the callout verification option, for verify=
4727 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4728 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4730 12. Inserted some casts to stop certain compilers warning when using pointer
4731 differences as field lengths or precisions in printf-type calls (mostly
4732 affecting debugging statements).
4734 13. Added optional readline() support for -be (dynamically loaded).
4736 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4737 same clock tick as a message's arrival, so that its received time was the
4738 same as the "first fail" time on the retry record, and that message
4739 remained on the queue past the ultimate address timeout, every queue runner
4740 would try a delivery (because it was past the ultimate address timeout) but
4741 after another failure, the ultimate address timeout, which should have then
4742 bounced the address, did not kick in. This was a "< instead of <=" error;
4743 in most cases the first failure would have been in the next clock tick
4744 after the received time, and all would be well.
4746 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4747 being recognized when the domain list was tested by the match_domain
4748 condition in an expansion string.
4750 16. Added the ${str2b64: operator.
4752 17. Exim was always calling setrlimit() to set a large limit for the number of
4753 processes, without checking whether the existing limit was already
4754 adequate. (It did check for the limit on file descriptors.) Furthermore,
4755 errors from getrlimit() and setrlimit() were being ignored. Now they are
4756 logged to the main and panic logs, but Exim does carry on, to try to do its
4757 job under whatever limits there are.
4759 18. Imported PCRE 5.0.
4761 19. Trivial typo in log message " temporarily refused connection" (the leading
4764 20. If the log selector return_path_on_delivery was set and an address was
4765 redirected to /dev/null, the delivery process crashed because it assumed
4766 that a return path would always be set for a "successful" delivery. In this
4767 case, the whole delivery is bypassed as an optimization, and therefore no
4770 21. Internal re-arrangement: the function for sending a challenge and reading
4771 a response while authentication was assuming a zero-terminated challenge
4772 string. It's now changed to take a pointer and a length, to allow for
4773 binary data in such strings.
4775 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4777 23. Exim was not respecting finduser_retries when seeking the login of the
4778 uid under which it was called; it was always trying 10 times. (The default
4779 setting of finduser_retries is zero.) Also, it was sleeping after the final
4780 failure, which is pointless.
4782 24. Implemented tls_on_connect_ports.
4784 25. Implemented acl_smtp_predata.
4786 26. If the domain in control=submission is set empty, Exim assumes that the
4787 authenticated id is a complete email address when it generates From: or
4788 Sender: header lines.
4790 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4791 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4792 chown and chgrp in /bin and hostname in /usr/bin.
4794 28. Exim was keeping the "process log" file open after each use, just as it
4795 does for the main log. This opens the possibility of it remaining open for
4796 long periods when the USR1 signal hits a daemon. Occasional processlog
4797 errors were reported, that could have been caused by this. Anyway, it seems
4798 much more sensible not to leave this file open at all, so that is what now
4801 29. The long-running daemon process does not normally write to the log once it
4802 has entered its main loop, and it closes the log before doing so. This is
4803 so that log files can straightforwardly be renamed and moved. However,
4804 there are a couple of unusual error situations where the daemon does write
4805 log entries, and I had neglected to close the log afterwards.
4807 30. The text of an SMTP error response that was received during a remote
4808 delivery was being truncated at 512 bytes. This is too short for some of
4809 the long messages that one sometimes sees. I've increased the limit to
4812 31. It is now possible to make retry rules that apply only when a message has a
4813 specific sender, in particular, an empty sender.
4815 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4816 it possible to be selective about when SMTP synchronization is enforced.
4818 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4820 32. Implemented hosts_connection_nolog.
4822 33. Added an ACL for QUIT.
4824 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4827 35. Added mailbox_size and mailbox_filecount to appendfile.
4829 36. Added control = no_multiline_responses to ACLs.
4831 37. There was a bug in the logic of the code that waits for the clock to tick
4832 in the case where the clock went backwards by a substantial amount such
4833 that the microsecond fraction of "now" was more than the microsecond
4834 fraction of "then" (but the whole seconds number was less).
4836 38. Added support for the libradius Radius client library this is found on
4837 FreeBSD (previously only the radiusclient library was supported).
4843 1. When certain lookups returned multiple values in the form name=value, the
4844 quoting of the values was not always being done properly. Specifically:
4845 (a) If the value started with a double quote, but contained no whitespace,
4847 (b) If the value contained whitespace other than a space character (i.e.
4848 tabs or newlines or carriage returns) it was not quoted.
4849 This fix has been applied to the mysql and pgsql lookups by writing a
4850 separate quoting function and calling it from the lookup code. The fix
4851 should probably also be applied to nisplus, ibase and oracle lookups, but
4852 since I cannot test any of those, I have not disturbed their existing code.
4854 2. A hit in the callout cache for a specific address caused a log line with no
4855 reason for rejecting RCPT. Now it says "Previous (cached) callout
4856 verification failure".
4858 3. There was an off-by-one bug in the queryprogram router. An over-long
4859 return line was truncated at 256 instead of 255 characters, thereby
4860 overflowing its buffer with the terminating zero. As well as fixing this, I
4861 have increased the buffer size to 1024 (and made a note to document this).
4863 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4864 when Exim is waiting for an SMTP response from a remote server, Exim
4865 restarts its select() call on the socket, thereby resetting its timeout.
4866 This is not a problem when such interrupts are rare. Somebody set up a cron
4867 job to run exiwhat every 2 minutes, which is less than the normal select()
4868 timeout (5 or 10 minutes). This meant that the select() timeout never
4869 kicked in because it was always reset. I have fixed this by comparing the
4870 time when an interrupt arrives with the time at the start of the first call
4871 to select(). If more time than the timeout has elapsed, the interrupt is
4872 treated as a timeout.
4874 5. Some internal re-factoring in preparation for the addition of Sieve
4875 extensions (by MH). In particular, the "personal" test is moved to a
4876 separate function, and given an option for scanning Cc: and Bcc: (which is
4877 not set for Exim filters).
4879 6. When Exim created an email address using the login of the caller as the
4880 local part (e.g. when creating a From: or Sender: header line), it was not
4881 quoting the local part when it contained special characters such as @.
4883 7. Installed new OpenBSD configuration files.
4885 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4886 try to make them clearer.
4888 9. Callout options, other than the timeout value, were being ignored when
4889 verifying sender addresses in header lines. For example, when using
4891 verify = header_sender/callout=no_cache
4893 the cache was (incorrectly) being used.
4895 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4896 only the Cygwin environment.
4898 11. When return_path_on_delivery was set as a log selector, if different remote
4899 addresses in the same message used different return paths and parallel
4900 remote delivery occurred, the wrong values would sometimes be logged.
4901 (Whenever a remote delivery process finished, the return path value from
4902 the most recently started remote delivery process was logged.)
4904 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4905 lines when AUTH and/or TLS are in use. This is the "received protocol"
4906 field. Exim used to use "asmtp" for authenticated SMTP, without any
4907 indication (in the protocol name) for TLS use. Now it follows the RFC and
4908 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4909 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4910 names appear in log lines as well as in Received: header lines.
4912 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4913 extensions, and comparison tests, and to fix some bugs.
4915 14. Changes to the "personal" filter test:
4917 (1) The test was buggy in that it was just doing the equivalent of
4918 "contains" tests on header lines. For example, if a user's address was
4919 anne@some.where, the "personal" test would incorrectly be true for
4921 To: susanne@some.where
4923 This test is now done by extracting each address from the header in turn,
4924 and checking the entire address. Other tests that are part of "personal"
4925 are now done using regular expressions (for example, to check local parts
4926 of addresses in From: header lines).
4928 (2) The list of non-personal local parts in From: addresses has been
4929 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4930 taken from the Sieve specification recommendations.
4932 (3) If the message contains any header line starting with "List-" it is
4933 treated as non-personal.
4935 (4) The test for "circular" in the Subject: header line has been removed
4936 because it now seems ill-conceived.
4938 15. Minor typos in src/EDITME comments corrected.
4940 16. Installed latest exipick from John Jetmore.
4942 17. If headers_add on a router specified a text string that was too long for
4943 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4944 of string_sprintf() is now avoided.
4946 18. $message_body_size was not set (it was always zero) when running the DATA
4947 ACL and the local_scan() function.
4949 19. For the "mail" command in an Exim filter, no default was being set for
4950 the once_repeat time, causing a random time value to be used if "once" was
4951 specified. (If the value happened to be <= 0, no repeat happened.) The
4952 default is now 0s, meaning "never repeat". The "vacation" command was OK
4953 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4954 (I found it when inspecting the code).
4956 20. There is now an overall timeout for performing a callout verification. It
4957 defaults to 4 times the callout timeout, which applies to individual SMTP
4958 commands during the callout. The overall timeout applies when there is more
4959 than one host that can be tried. The timeout is checked before trying the
4960 next host. This prevents very long delays if there are a large number of
4961 hosts and all are timing out (e.g. when the network connections are timing
4962 out). The value of the overall timeout can be changed by specifying an
4963 additional sub-option for "callout", called "maxwait". For example:
4965 verify = sender/callout=5s,maxwait=20s
4967 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4968 to the end before writing, but this should make it even safer).
4970 22. Exim was forgetting that it had advertised PIPELINING for the second and
4971 subsequent messages on an SMTP connection. It was also not resetting its
4972 memory on STARTTLS and an internal HELO.
4974 23. When Exim logs an SMTP synchronization error within a session, it now
4975 records whether PIPELINING has been advertised or not.
4977 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4978 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4979 rather than long int.
4981 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4983 26. Added the never_mail option to autoreply.
4989 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4990 crash if the getsockname() call failed; this can happen if a connection is
4991 closed very soon after it is established. The problem was simply in the
4992 order in which certain operations were done, causing Exim to try to write
4993 to the SMTP stream before it had set up the file descriptor. The bug has
4994 been fixed by making things happen in the correct order.
5000 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5001 before the connection was closed, thus losing the rejection response.
5003 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5004 some early Solaris releases, but causes trouble in current releases where
5005 socklen_t is defined.
5007 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5010 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5013 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5014 that releasing the top of it at the end releases what was used for sub-
5015 expansions (unless the block got too big). However, discard this block if
5016 the first thing is a variable or header, so that we can use its block when
5017 it is dynamic (useful for very large $message_headers, for example).
5019 6. Lookups now cache *every* query, not just the most recent. A new, separate
5020 store pool is used for this. It can be recovered when all lookup caches are
5021 flushed. Lookups now release memory at the end of their result strings.
5022 This has involved some general refactoring of the lookup sources.
5024 7. Some code has been added to the store_xxx() functions to reduce the amount
5025 of flapping under certain conditions.
5027 8. log_incoming_interface used to affect only the <= reception log lines. Now
5028 it causes the local interface and port to be added to several more SMTP log
5029 lines, for example "SMTP connection from", and rejection lines.
5031 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5033 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5035 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5036 could be overwritten at the end of the current message (or the start of a
5037 new message if it was set in a HELO ACL). The value is now preserved for
5038 the duration of the SMTP connection.
5040 12. If a transport had a headers_rewrite setting, and a matching header line
5041 contained an unqualified address, that address was qualified, even if it
5042 did not match any rewriting rules. The underlying bug was that the values
5043 of the flags that permit the existence of unqualified sender and recipient
5044 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5045 non-local messages, and by -bnq for local messages) were not being
5046 preserved with the message after it was received.
5048 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5049 "next input=" as part of the text comprising the host identity instead of
5050 the correct text. The code was using the same buffer for two different
5051 strings. However, depending on which order the printing function evaluated
5052 its arguments, the bug did not always show up. Under Linux, for example, my
5053 test suite worked just fine.
5055 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5056 doesn't work with some older versions of Perl. It has been changed to "my",
5057 which in any case is probably the better facility to use.
5059 15. A really picky compiler found some instances of statements for creating
5060 error messages that either had too many or two few arguments for the format
5063 16. The size of the buffer for calls to the DNS resolver has been increased
5064 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5065 for addresses that have a lot of PTR records. This alleviates a problem; it
5066 does not fully solve it.
5068 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5069 buffer now truncates the list and logs the incident, which is the same
5070 action as happens when Exim is looking up a host name and its aliases.
5071 Previously in this situation something unpredictable would happen;
5072 sometimes it was "internal error: store_reset failed".
5074 18. If a server dropped the connection unexpectedly when an Exim client was
5075 using GnuTLS and trying to read a response, the client delivery process
5076 crashed while trying to generate an error log message.
5078 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5079 string, for example:
5081 warn message = H1: something\nH2: something
5083 the text was added as a single header line from Exim's point of view
5084 though it ended up OK in the delivered message. However, searching for the
5085 second and subsequent header lines using $h_h2: did not work. This has been
5086 fixed. Similarly, if a system filter added multiple headers in this way,
5087 the routers could not see them.
5089 20. Expanded the error message when iplsearch is called with an invalid key to
5090 suggest using net-iplsearch in a host list.
5092 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5093 ACLs are no longer actually imposed (and a message to that effect is
5096 22. If a "gecos" field in a passwd entry contained escaped characters, in
5097 particular, if it contained a \" sequence, Exim got it wrong when building
5098 a From: or a Sender: header from that name. A second bug also caused
5099 incorrect handling when an unquoted " was present following a character
5100 that needed quoting.
5102 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5103 was not being matched caselessly.
5105 24. Arranged for all hyphens in the exim.8 source to be escaped with
5108 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5109 properly. Recipient callout cache records were still being keyed to include
5110 the sender, even when use_sender was set false. This led to far more
5111 callouts that were necessary. The sender is no longer included in the key
5112 when use_sender is false.
5114 26. Added "control = submission" modifier to ACLs.
5116 27. Added the ${base62d: operator to decode base 62 numbers.
5118 28. dnsdb lookups can now access SRV records.
5120 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5121 the configuration file.
5123 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5124 (-v) mode. This makes the output for a verbose queue run more intelligible.
5126 31. Added a use_postmaster feature to recipient callouts.
5128 32. Added the $body_zerocount variable, containing the number of binary zero
5129 bytes in the message body.
5131 33. The time of last modification of the "new" subdirectory is now used as the
5132 "mailbox time last read" when there is a quota error for a maildir
5135 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5137 35. Added +ignore_unknown as a special item in host lists.
5139 36. Code for decoding IPv6 addresses in host lists is now included, even if
5140 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5141 address was recognized as an IP address, but was then not correctly decoded
5142 into binary, causing unexpected and incorrect effects when compared with
5149 1. Very minor rewording of debugging text in manualroute to say "list of
5150 hosts" instead of "hostlist".
5152 2. If verify=header_syntax was set, and a header line with an unqualified
5153 address (no domain) and a large number of spaces between the end of the
5154 name and the colon was received, the reception process suffered a buffer
5155 overflow, and (when I tested it) crashed. This was caused by some obsolete
5156 code that should have been removed. The fix is to remove it!
5158 3. When running in the test harness, delay a bit after writing a bounce
5159 message to get a bit more predictability in the log output.
5161 4. Added a call to search_tidyup() just before forking a reception process. In
5162 theory, someone could use a lookup in the expansion of smtp_accept_max_
5163 per_host which, without the tidyup, could leave open a database connection.
5165 5. Added the variables $recipient_data and $sender_data which get set from a
5166 lookup success in an ACL "recipients" or "senders" condition, or a router
5167 "senders" option, similar to $domain_data and $local_part_data.
5169 6. Moved the writing of debug_print from before to after the "senders" test
5172 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5173 problems for message scanning, either using a data ACL, or using
5174 local_scan() because the Received: header was not generated till after they
5175 were called (in order to set the time as the time of reception completion).
5176 I have revised the way this works. The header is now generated after the
5177 body is received, but before the ACL or local_scan() are called. After they
5178 are run, the timestamp in the header is updated.
5184 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5185 before starting a queue runner without re-exec. This happened only when
5186 deliver_drop_privilege was set or when the Exim user was set to root. The
5187 effect of the bug was that timeouts during subsequent deliveries caused
5188 crashes instead of being properly handled. The handler is now left at its
5189 default (and expected) setting.
5191 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5192 message, again when deliver_drop_privilege is set or Exim is run as root.
5193 The bug described in (1) was not present in this case, but the tidying up
5194 of the other signals was missing. I have made the two cases consistent.
5196 3. The ignore_target_hosts setting on a manualroute router was being ignored
5197 for hosts that were looked up using the /MX notation.
5199 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5202 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5203 operated on the sender address. After changing the $sender_address to <>
5204 for the sender address verify, Exim was re-instated it as the original
5205 (before rewriting) address, but remembering that it had rewritten it, so it
5206 wasn't rewriting it again. This bug also had the effect of breaking the
5207 sender address verification caching when the sender address was rewritten.
5209 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5210 This has been changed so that if the ip literal address matches
5211 ignore_target_hosts, the router declines.
5213 7. Added expansion conditions match_domain, match_address, and match_local_
5214 part (NOT match_host).
5216 8. The placeholder for the Received: header didn't have a length field set.
5218 9. Added code to Exim itself and to exim_lock to test for a specific race
5219 condition that could lead to file corruption when using MBX delivery. The
5220 issue is with the lockfile that is created in /tmp. If this file is removed
5221 after a process has opened it but before that process has acquired a lock,
5222 there is the potential for a second process to recreate the file and also
5223 acquire a lock. This could lead to two Exim processes writing to the file
5224 at the same time. The added code performs the same test as UW imapd; it
5225 checks after acquiring the lock that its file descriptor still refers to
5226 the same named file.
5228 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5229 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5232 11. Added the smtp_active_hostname option. If used, this will typically be made
5233 to depend on the incoming interface address. Because $interface_address is
5234 not set up until the daemon has forked a reception process, error responses
5235 that can happen earlier (such as "too many connections") no longer contain
5238 12. If an expansion in a condition on a "warn" statement fails because a lookup
5239 defers, the "warn" statement is abandoned, and the next ACL statement is
5240 processed. Previously this caused the whole ACL to be aborted.
5242 13. Added the iplsearch lookup type.
5244 14. Added ident_timeout as a log selector.
5246 15. Added tls_certificate_verified as a log selector.
5248 16. Added a global option tls_require_ciphers (compare the smtp transport
5249 option of the same name). This controls incoming TLS connections.
5251 17. I finally figured out how to make tls_require_ciphers do a similar thing
5252 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5253 before starting the TLS session.
5255 18. Tabs are now shown as \t in -bP output.
5257 19. If the log selector return_path_on_delivery was set, Exim crashed when
5258 bouncing a message because it had too many Received: header lines.
5260 20. If two routers both had headers_remove settings, and the first one included
5261 a superfluous trailing colon, the final name in the first list and the
5262 first name in the second list were incorrectly joined into one item (with a
5263 colon in the middle).
5269 1. Added -C and -D options to the exinext utility, mainly to make it easier
5270 to include in the automated testing, but these could be helpful when
5271 multiple configurations are in use.
5273 2. The exinext utility was not formatting the output nicely when there was
5274 an alternate port involved in the retry record key, nor when there was a
5275 message id as well (for retries that were specific to a specific message
5276 and a specific host). It was also confused by IPv6 addresses, because of
5277 the additional colons they contain. I have fixed the IPv4 problem, and
5278 patched it up to do a reasonable job for IPv6.
5280 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5281 delivery, the log line now contains "pipelined" if PIPELINING was used.
5283 4. An SMTP transport process used to panic and die if the bind() call to set
5284 an explicit outgoing interface failed. This has been changed; it is now
5285 treated in the same way as a connect() failure.
5287 5. A reference to $sender_host_name in the part of a conditional expansion
5288 that was being skipped was still causing a DNS lookup. This no longer
5291 6. The def: expansion condition was not recognizing references to header lines
5292 that used bh_ and bheader_.
5294 7. Added the _cache feature to named lists.
5296 8. The code for checking quota_filecount in the appendfile transport was
5297 allowing one more file than it should have been.
5299 9. For compatibility with Sendmail, the command line option
5307 and sets the incoming protocol and host name (for trusted callers). The
5308 host name and its colon can be omitted when only the protocol is to be set.
5309 Note the Exim already has two private options, -pd and -ps, that refer to
5310 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5311 "s", but I don't think that's a major issue.
5313 10. A number of refactoring changes to the code, none of which should affect
5316 (a) The number of logging options was getting close to filling up the
5317 32-bit word that was used as a bit map. I have split them into two classes:
5318 those that are passed in the argument to log_write(), and those that are
5319 only ever tested independently outside of that function. These are now in
5320 separate 32-bit words, so there is plenty of room for expansion again.
5321 There is no change in the user interface or the logging behaviour.
5323 (b) When building, for example, log lines, the code previously used a
5324 macro that called string_cat() twice, in order to add two strings. This is
5325 not really sufficiently general. Furthermore, there was one instance where
5326 it was actually wrong because one of the argument was used twice, and in
5327 one call a function was used. (As it happened, calling the function twice
5328 did not affect the overall behaviour.) The macro has been replaced by a
5329 function that can join an arbitrary number of extra strings onto a growing
5332 (c) The code for expansion conditions now uses a table and a binary chop
5333 instead of a serial search (which was left over from when there were very
5334 few conditions). Also, it now recognizes conditions like "pam" even when
5335 the relevant support is not compiled in: a suitably worded error message is
5336 given if an attempt is made to use such a condition.
5338 11. Added ${time_interval:xxxxx}.
5340 12. A bug was causing one of the ddress fields not to be passed back correctly
5341 from remote delivery subprocesses. The field in question was not being
5342 subsequently used, so this caused to problems in practice.
5344 13. Added new log selectors queue_time and deliver_time.
5346 14. Might have fixed a bug in maildirsizefile handling that threw up
5347 "unexpected character" debug warnings, and recalculated the data
5348 unnecessarily. In any case, I expanded the warning message to give more
5351 15. Added the message "Restricted characters in address" to the statements in
5352 the default ACL that block characters like @ and % in local parts.
5354 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5355 Three changes have been made:
5357 (a) There was a serious bug; a negative response to MAIL caused the whole
5358 recipient domain to be cached as invalid, thereby blocking all messages
5359 to all local parts at the same domain, from all senders. This bug has
5360 been fixed. The domain is no longer cached after a negative response to
5361 MAIL if the sender used is not empty.
5363 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5366 (c) A new callout option, "use_sender" has been added for people who want
5367 the modified behaviour.
5373 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5376 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5377 indeed breaks things for older releases.
5379 3. Added additional logging to the case where there is a problem reading data
5380 from a filter that is running in a subprocess using a pipe, in order to
5381 try to track down a specific problem.
5383 4. Testing facility fudge: when running in the test harness and attempting
5384 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5385 getting "No route to host". Convert this to a timeout.
5387 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5390 6. Some OS don't have socklen_t but use size_t instead. This affects the
5391 fifth argument of getsockopt() amongst other things. This is now
5392 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5393 can be set for individual OS. I have set it for SunOS5, OSF1, and
5394 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5395 some earlier ones do not.
5397 7. Change 4.30/15 was not doing the test caselessly.
5399 8. The standard form for an IPv6 address literal was being rejected by address
5400 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5401 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5402 this, as well as the form without the "IPv6" on the front (but only when
5403 address literals are enabled, of course).
5405 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5407 10. Exim crashed if a message with an empty sender address specified by -f
5408 encountered a router with an errors_to setting. This could be provoked only
5409 by a command such as
5413 where an empty string was supplied; "<>" did not hit this bug.
5415 11. Installed PCRE release 4.5.
5417 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5418 remained set. It is now erased.
5420 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5421 times from message ids (which are base 36 rather than the normal 62).
5423 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5424 were being counted as actual protocol errors, and logged if the log
5425 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5426 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5427 and DATA following a set of rejected RCPTs do not count as protocol errors.
5428 In other words, Exim assumes they were pipelined, though this may not
5429 actually be the case. Of course, in all cases the client gets an
5430 appropriate error code.
5432 15. If a lookup fails in an ACL condition, a message about the failure may
5433 be available; it is used if testing the ACL cannot continue, because most
5434 such messages specify what the cause of the deferral is. However, some
5435 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5436 that caused an old message to be retained and used if a later statement
5437 caused a defer, replacing the real cause of the deferral.
5439 16. If an IP address had so many PTR records that the DNS lookup buffer
5440 was not large enough to hold them, Exim could crash while trying to process
5441 the truncated data. It now detects and logs this case.
5443 17. Further to 4.21/58, another change has been made: if (and only if) the
5444 first line of a message (the first header line) ends with CRLF, a bare LF
5445 in a subsequent header line has a space inserted after it, so as not to
5446 terminate the header.
5448 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5449 unnecessarily, used atoi() instead of strtol(), and didn't check the
5450 termination when getting file sizes from file names by regex.
5452 19. Completely re-implemented the support for maildirsize files, in the light
5453 of a number of problems with the previous contributed implementation
5454 (4.30/29). In particular:
5456 . If the quota is zero, the maildirsize file is maintained, but no quota is
5459 . If the maildir directory does not exist, it is created before any attempt
5460 to write a maildirsize file.
5462 . The quota value in the file is just a cache; if the quota is changed in
5463 the transport, the new value overrides.
5465 . A regular expression is available for excluding directories from the
5468 20. The autoreply transport checks the characters in options that define the
5469 message's headers; it allows continued headers, but it was checking with
5470 isspace() after an embedded newline instead of explicitly looking for a
5473 21. If all the "regular" hosts to which an address was routed had passed their
5474 expiry times, and had not reached their retry times, the address was
5475 bounced, even if fallback hosts were defined. Now Exim should go on to try
5478 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5479 equivalent code in the SMTP transport. Some hosts send humungous responses
5480 to HELO/EHLO, more than 1024 it seems.
5482 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5483 gives compiler warnings in some environments. I've now done it "properly",
5486 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5487 (because of problems with the built-in one) was declared to return uschar *
5488 instead of char *, causing compiler failure.
5490 25. Fixed a file descriptor leak when processing alias/forward files.
5492 26. Fixed a minor format string issue in dbfn.c.
5494 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5496 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5497 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5500 29. When a sender address is verified, it is cached, to save repeating the test
5501 when there is more than one recipient in a message. However, when the
5502 verification involves a callout, it is possible for different callout
5503 options to be set for different recipients. It is too complicated to keep
5504 track of this in the cache, so now Exim always runs a verification when a
5505 callout is required, relying on the callout cache for the optimization.
5506 The overhead is duplication of the address routing, but this should not be
5509 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5510 to be verified with callout=postmaster, and the main callout worked but the
5511 postmaster check failed, the verification correctly failed. However, if a
5512 subsequent RCPT command asked for sender verification *without* the
5513 postmaster check, incorrect caching caused this verification also to fail,
5516 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5517 it was not caching the DNS options (qualify_single, search_parents) that
5518 were used when the lookup failed. A subsequent lookup with different
5519 options therefore always gave the same answer, though there were cases
5520 where it should not have. (Example: a "domains = !$mx_any" option on a
5521 dnslookup router: the "domains" option is always processed without any
5522 widening, but the router might have qualify_single set.) Now Exim uses the
5523 cached value only when the same options are set.
5525 32. Added John Jetmore's "exipick" utility to the distribution.
5527 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5528 than a timeout (e.g. a certificate is required, and is not provided), an
5529 Exim server now closes the connection immediately. Previously it waited for
5530 the client to close - but if the client is SSL, it seems that they each
5531 wait for each other, leading to a delay before one of them times out.
5533 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5534 maintained 0.8.x compatibility because I don't think many are using it, and
5535 it is clearly obsolete.
5537 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5540 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5541 client certificate was expired. A simple patch fixes this, though I don't
5542 understand the full logic of why the verify callback is called multiple
5545 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5546 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5547 which causes problems with some clients (such as the Certicom SSL Plus
5548 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5549 disables the coutermeasure allowing Eudora to connect."
5551 38. Exim was not checking that a write() to a log file succeeded. This could
5552 lead to Bad Things if a log got too big, in particular if it hit a file
5553 size limit. Exim now panics and dies if it cannot write to a log file, just
5554 as it does if it cannot open a log file.
5556 39. Modified OS/Makefile-Linux so that it now contains
5558 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5560 The two -D definitions ensure that Exim is compiled with large file
5561 support, which makes it possible to handle log files that are bigger than
5564 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5565 instance) a domain was checked against a named list that involved a lookup,
5566 causing $domain_data to be set, then another domain was checked against the
5567 same list, then the first domain was re-checked, the value of $domain_data
5568 after the final check could be wrong. In particular, if the second check
5569 failed, it could be set empty. This bug probably also applied to
5572 41. The strip_trailing_dot option was not being applied to the address given
5573 with the -f command-line option.
5575 42. The code for reading a message's header from the spool was incrementing
5576 $received_count, but never initializing it. This meant that the value was
5577 incorrect (doubled) while delivering a message in the same process in which
5578 it was received. In the most common configuration of Exim, this never
5579 happens - a fresh exec is done - but it can happen when
5580 deliver_drop_privilege is set.
5582 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5583 it now includes up to 150 characters of the unexpected data in the log
5586 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5587 and building data strings. The size of both of these buffers was 10 000
5588 bytes - far larger than anybody would *ever* want, thought I. Needless to
5589 say, somebody hit the limit. I have increased the maximum line length to
5590 20 000 and the maximum data length of concatenated lines to 100 000. I have
5591 also fixed two bugs, because there was no checking on these buffers. Tsk,
5592 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5593 buffer is too small.
5595 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5596 lsearch lookups. Now it does.
5598 46. When parsing a route_list item in a manualroute router, a fixed-length
5599 buffer was used for the list of hosts. I made this 1024 bytes long,
5600 thinking that nobody would ever have a list of hosts that long. Wrong.
5601 Somebody had a whole pile of complicated expansion conditions, and the
5602 string was silently truncated, leading to an expansion error. It turns out
5603 that it is easier to change to an unlimited length (owing to other changes
5604 that have happened since this code was originally written) than to build
5605 structure for giving a limitation error. The length of the item that
5606 expands into the list of hosts is now unlimited.
5608 47. The lsearch lookup could not handle data where the length of text line was
5609 more than 4095 characters. Such lines were truncated, leading to shortened
5610 data being returned. It should now handle lines of any length.
5612 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5613 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5616 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5617 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5618 no longer happens in comment lines. A list of replacements is now placed
5619 at the head of all of the source files, except those whose only change is
5620 to replace PERL_COMMAND in the very first #! line.
5622 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5623 messages on the queue, with a bottom-up merge sort, using code contributed
5624 by Michael Haardt. This should make operations like -bp somewhat faster on
5625 large queues. It won't affect queue runners, except when queue_run_in_order
5628 51. Installed eximstats 1.31 in the distribution.
5630 52. Added support for SRV lookups to the dnslookup router.
5632 53. If an ACL referred to $message_body or $message_body_end, the value was not
5633 reset for any messages that followed in the same SMTP session.
5635 54. The store-handling optimization for building very long strings was not
5636 differentiating between the different store pools. I don't think this
5637 actually made any difference in practice, but I've tidied it.
5639 55. While running the routers to verify a sender address, $sender_address
5640 was still set to the sender address. This is wrong, because when routing to
5641 send a bounce to the sender, it would be empty. Therefore, I have changed
5642 it so that, while verifying a sender address, $sender_address is set to <>.
5643 (There is no change to what happens when verifying a recipient address.)
5645 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5646 target A or AAAA records (if not already returned) without resetting the
5647 qualify_single or search_parents options of the DNS resolver. These are
5648 inappropriate in this case because the targets of MX and SRV records must
5649 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5650 target that, when qualified, matched something in the local domain. These
5651 two options are now turned off when doing these lookups.
5653 57. It seems that at least some releases of Reiserfs (which does not have the
5654 concept of a fixed number of inodes) returns zero and not -1 for the
5655 number of available inodes. This interacted badly with check_spool_inodes,
5656 which assumed that -1 was the "no such thing" setting. What I have done is
5657 to check that the total number of inodes is greater than zero before doing
5658 the test of how many are available.
5660 58. When a "warn" ACL statement has a log_message modifier, the message is
5661 remembered, and not repeated. This is to avoid a lot of repetition when a
5662 message has many recipients that cause the same warning to be written.
5663 However, Exim was preserving the list of already written lines for an
5664 entire SMTP session, which doesn't seem right. The memory is now reset if a
5665 new message is started.
5667 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5668 debugging output unless log_rewrite was also set.
5670 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5671 of (int)(handle) when we know that handle contains (void *)(-1).
5673 61. The Exim daemon panic-logs an error return when it closes the incoming
5674 connection. However "connection reset by peer" seems to be common, and
5675 isn't really an error worthy of noting specially, so that particular error
5678 62. When Exim is trying to find all the local interfaces, it used to panic and
5679 die if the ioctl to get the interface flags failed. However, it seems that
5680 on at least one OS (Solaris 9) it is possible to have an interface that is
5681 included in the list of interfaces, but for which you get a failure error
5682 for this call. This happens when the interface is not "plumbed" into a
5683 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5684 failure of the "get flags" call assumes that the interface is down.
5686 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5687 makes life easier for people who are doing arithmetic on fields extracted
5688 from dates, where you often get leading zeros that should not be
5689 interpreted as octal.
5691 64. Added qualify_domain to the redirect router, to override the global
5694 65. If a pathologically long header line contained very many addresses (the
5695 report of this problem mentioned 10 000) and each of them was rewritten,
5696 Exim could use up a very large amount of memory. (It kept on making new
5697 copies of the header line as it rewrote, and never released the old ones.)
5698 At the expense of a bit more processing, the header rewriting function has
5699 been changed so that it no longer eats memory in this way.
5701 66. The generation of the Received: header has been moved from the time that a
5702 message starts to be received, to the time that it finishes. The timestamp
5703 in the Received: header should now be very close to that of the <= log
5704 line. There are two side-effects of this change:
5706 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5707 logged header lines no longer include the local Received: line, because
5708 it has not yet been created. The same applies to a copy of the message
5709 that is returned to a non-SMTP sender when a message is rejected.
5711 (b) When a filter file is tested using -bf, no additional Received: header
5712 is added to the test message. After some thought, I decided that this
5715 This change does not affect the value of $received_for. It is still set
5716 after address rewriting, but before local_scan() is called.
5718 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5720 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5721 gave an unhelpful panic error message, and a defer error. I have managed to
5722 change this behaviour so that it now rejects any supplied certificate,
5723 which seems right, as the list of acceptable certificates is empty.
5725 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5726 gave an unhelpful defer error. I have not managed to make this reject any
5727 supplied certificates, but the error message it gives is "no certificate
5728 supplied", which is not helpful.
5730 70. exigrep's output now also includes lines that are not associated with any
5731 message, but which match the given pattern. Implemented by a patch from
5732 Martin Sluka, which also tidied up the Perl a bit.
5734 71. Recipient callout verification, like sender verification, was using <> in
5735 the MAIL FROM command. This isn't really the right thing, since the actual
5736 sender may affect whether the remote host accepts the recipient or not. I
5737 have changed it to use the actual sender in the callout; this means that
5738 the cache record is now keyed on a recipient/sender pair, not just the
5739 recipient address. There doesn't seem to be a real danger of callout loops,
5740 since a callout by the remote host to check the sender would use <>.
5741 [SEE ABOVE: changed after hitting problems.]
5743 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5744 temporary errors. However, in the case of such a code being given after
5745 the end of a data transmission (i.e. after ".") Exim was failing to write
5746 a retry record for the message. (Yes, there was some broken host that was
5747 actually sending 8xx at this point.)
5749 73. An unknown lookup type in a host list could cause Exim to panic-die when
5750 the list was checked. (An example that provoked this was putting <; in the
5751 middle of a list instead of at the start.) If this happened during a DATA
5752 ACL check, a -D file could be left lying around. This kind of configuration
5753 error no longer causes Exim to die; instead it causes a defer error. The
5754 incident is still logged to the main and panic logs.
5756 74. Buglet left over from Exim 3 conversion. The message "too many messages
5757 in one connection" was written to the rejectlog but not the mainlog, except
5758 when address rewriting (yes!) was being logged.
5760 75. Added write_rejectlog option.
5762 76. When a system filter was run not as root (that is, when system_filter_user
5763 was set), the values of the $n variables were not being returned to the
5764 main process; thus, they were not subsequently available in the $sn
5767 77. Added +return_path_on_delivery log selector.
5769 78. A connection timeout was being treated differently from recipients deferred
5770 when testing hosts_max_try with a message that was older than the host's
5771 retry timeout. (The host should not be counted, thus allowing all hosts to
5772 be tried at least once before bouncing.) This may have been the cause of an
5773 occasionally reported bug whereby a message would remain on the queue
5774 longer than the retry timeout, but would be bounced if a delivery was
5775 forced. I say "may" because I never totally pinned down the problem;
5776 setting up timeout/retry tests is difficult. See also the next item.
5778 79. The ultimate address timeout was not being applied to errors that involved
5779 a combination of host plus message (for example, a timeout on a MAIL
5780 command). When an address resolved to a number of possible hosts, and they
5781 were not all tried for each delivery (e.g. because of hosts_max_try), a
5782 message could remain on the queue longer than the retry timeout.
5784 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5787 81. Fixed an obscure SMTP outgoing bug which required at least the following
5788 conditions: (a) there was another message waiting for the same server;
5789 (b) the server returned 5xx to all RCPT commands in the first message so
5790 that the message was not completed; (c) the server dropped the connection
5791 or gave a negative response to the RSET that Exim sends to abort the
5792 transaction. The observed case was a dropped connection after DATA that had
5793 been sent in pipelining mode. That is, the server had advertised PIPELINING
5794 but was not implementing it correctly. The effect of the bug was incorrect
5795 behaviour, such as trying another host, and this could lead to a crash.
5801 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5802 and daemon.c were passed as pointers to ints; they should have been
5803 pointers to socklen_t variables (which are typically unsigned ints).
5805 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5808 3. Fixed a really odd bug that affected only the testing scheme; patching a
5809 certain fixed string in the binary changed the value of another string that
5810 happened to be identical to the end of the original first string.
5812 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5813 name", it returns that address as the IP address. On some operating
5814 systems (e.g. Solaris), it also passes back the IP address string as the
5815 "host name". However, on others (e.g. Linux), it passes back an empty
5816 string. Exim wasn't checking for this, and was changing the host name to an
5817 empty string, assuming it had been canonicalized.
5819 5. Although rare, it is permitted to have more than one PTR record for a given
5820 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5821 all the names associated with an address, because they do in Solaris.
5822 However, it seems that they do not in Linux for data that comes from the
5823 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5824 I found this out when I moved to a new Linux workstation and tried to run
5825 the Exim test suite.
5827 To get round this problem I have changed the code so that it now does its
5828 own call to the DNS to look up PTR records when searching for a host name.
5829 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5830 addresses that are only in /etc/hosts are still found.
5832 This behaviour is, however, controlled by an option called host_lookup_
5833 order, which defaults to "bydns:byaddr". If people want to use the other
5834 order, or indeed, just use one or the other means of lookup, they can
5835 specify it in this variable.
5837 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5838 some operating systems, this comes back from gethostbyaddr() as an empty
5839 string, and this is what Exim used to test for. However, it seems that in
5840 other systems, "." is yielded. Exim now tests for this case too.
5842 7. The values of check_spool_space and check_log_space are now held internally
5843 as a number of kilobytes instead of an absolute number of bytes. If a
5844 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5845 kilobyte. This means that much larger values can be stored.
5847 8. Exim monitor: an attempt to get the action menu when not actually pointing
5848 at a message produces an empty menu entitled "No message selected". This
5849 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5850 no entries in it ("Shell widget menu has zero width and/or height"). So I
5851 have added a single, blank menu entry in this case.
5853 9. Added ${quote_local_part.
5855 10. MIME decoding is now applied to the contents of Subject: header lines when
5858 11. Now that a reference to $sender_host_address automatically causes a reverse
5859 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5860 host lookup before query-style lookups in lists that might use this
5861 variable. This has therefore been abolished, and the "net-" prefix is no
5862 longer necessary for query-style lookups.
5864 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5865 have been a typo for LFLAGS, so it has been changed.
5867 13. The install script calls Exim with "-C /dev/null" in order to find the
5868 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5869 to be output. However, since Exim outputs its version number before the
5870 error, it didn't break the script. It just looked ugly. I fixed this by
5871 always allowing "-C /dev/null" if the caller is root.
5873 14. Ignore overlarge ACL variable number when reading spool file - insurance
5874 against a later release with more variables having written the file.
5876 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5877 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5878 this, as well as the form without the "IPv6" on the front.
5880 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5881 OS/Makefile-Darwin file.
5883 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5884 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5886 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5887 variable. However, some OS use ints for inodes. I've added cast to long int
5888 to get rid of the compiler warning.
5890 19. I had forgotten to lock out "/../" in configuration file names when
5891 ALT_CONFIG_PREFIX was set.
5893 20. Routers used for verification do not need to specify transports. However,
5894 if such a router generated a host list, and callout was configured, Exim
5895 crashed, because it could not find a port number from the (non-existent)
5896 transport. It now assumes port 25 in this circumstance.
5898 21. Added the -t option to exigrep.
5900 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5901 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5902 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5903 them is set, LOOKUP_LSEARCH is forced.
5905 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5907 24. Added sender and host information to the "rejected by local_scan()" log
5908 line; previously there was no indication of these.
5910 25. Added .include_if_exists.
5912 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5913 Linux. It turns out that not all file systems support this. Apparently some
5914 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5915 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5916 not supported on the file descriptor, is now ignored when Exim is trying to
5917 sync a directory. This applies only to Linux.
5919 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5921 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5922 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5925 29. Added support for maildirsize files from supplied patch (modified a bit).
5927 30. The use of :fail: followed by an empty string could lead Exim to respond to
5928 sender verification failures with (e.g.):
5930 550 Verification failed for <xxx>
5931 550 Sender verify failed
5933 where the first response line was missing the '-' that indicates it is not
5934 the final line of the response.
5936 31. The loop for finding the name of the user that called Exim had a hardwired
5937 limit of 10; it now uses the value of finduser_retries, which is used for
5938 all other user lookups.
5940 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5943 33. Exim was neglecting to zero errno before one call of strtol() when
5944 expanding a string and expecting an integer value. On some systems this
5945 resulted in spurious "integer overflow" errors. Also, it was casting the
5946 result into an int without checking.
5948 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5949 did not work. The code looks as if it has *never* worked, though it appears
5950 to have been documented since at least release 1.62. I have made it work.
5952 35. The "timeout_DNS" error in retry rules, also documented since at least
5953 1.62, also never worked. As it isn't clear exactly what this means, and
5954 clearly it isn't a major issue, I have abolished the feature by treating it
5955 as "timeout", and writing a warning to the main and panic logs.
5957 36. The display of retry rules for -brt wasn't always showing the error code
5960 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5961 timeout_connect_A, timeout_connect_MX.
5963 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5964 to the empty sender.
5966 39. The daemon was not analysing the content of -oX till after it had closed
5967 stderr and disconnected from the controlling terminal. This meant that any
5968 syntax errors were only noted on the panic log, and the return code from
5969 the command was 0. By re-arranging the code a little, I've made the
5970 decoding happen first, so such errors now appear on stderr, and the return
5971 code is 1. However, the actual setting up of the sockets still happens in
5972 the disconnected process, so errors there are still only recorded on the
5975 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5976 connections (as happens on some IP stacks) was logged at start up time as
5977 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5978 it from "IPv6 and IPv4", which means that two separate sockets are being
5981 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5982 says whether AF_INET or AF_INET6 was passed as an argument.
5984 42. Exiwhat output was messed up when time zones were included in log
5987 43. Exiwhat now gives more information about the daemon's listening ports,
5988 and whether -tls-on-connect was used.
5990 44. The "port" option of the smtp transport is now expanded.
5992 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5993 silently ignored. Now an error message is written to the main and panic
5996 46. There's a new ACL modifier called "logwrite" which writes to a log file
5997 as soon as it is encountered.
5999 47. Added $local_user_uid and $local_user_gid at routing time.
6001 48. Exim crashed when trying to verify a sender address that was being
6004 49. Exim was recognizing only a space character after ".include". It now also
6005 recognizes a tab character.
6007 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6008 extracting the relevant information from the specification. The man page no
6009 longer contains scrambled data for the -d option, and I've added a section
6010 at the front about calling Exim under different names.
6012 51. Added "extra_headers" argument to the "mail" command in filter files.
6014 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6017 53. Installed eximstats 1.29.
6019 54. Added transport_filter_timeout as a generic transport option.
6021 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6022 Cc: header lines. This was required by RFC 822, but it not required by RFC
6025 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6026 incoming messages that did not have them. Now it does so only if the
6027 message originates locally, that is, if there is no associated remote host
6028 address. When Resent- header lines are present, this applies to the Resent-
6029 lines rather than the non-Resent- lines.
6031 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6032 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6034 58. Messages for configuration errors now include the name of the main
6035 configuration file - useful now that there may be more than one file in a
6036 list (.included file names were always shown).
6038 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6039 for those rare installations that do not start the daemon as root or run it
6040 setuid root. I've cut out the call to initgroups() if the daemon is not
6043 60. The Exim user and group can now be bound into the binary as text strings
6044 that are looked up at the start of Exim's processing.
6046 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6048 62. Added $mailstore_basename variable.
6050 63. Installed patch to sieve.c from Michael Haardt.
6052 64. When Exim failed to open the panic log after failing to open the main log,
6053 the original message it was trying to log was written to stderr and debug
6054 output, but if they were not available (the usual case in production), it
6055 was lost. Now it is written to syslog before the two lines that record the
6056 failures to open the logs.
6058 65. Users' Exim filters run in subprocesses under the user's uid. It is
6059 possible for a "deliver" command or an alias in a "personal" command to
6060 provoke an address rewrite. If logging of address rewriting is configured,
6061 this fails because the process is not running as root or exim. There may be
6062 a better way of dealing with this, but for the moment (because 4.30 needs
6063 to be released), I have disabled address rewrite logging when running a
6064 filter in a non-root, non-exim process.
6070 1. The buildconfig auxiliary program wasn't quoting the value set for
6071 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6072 not defined. This bug was masked in 4.22 by the effect that was fixed in
6075 2. Some messages that were rejected after a message id was allocated were
6076 shown as "incomplete" by exigrep. It no longer does this for messages that
6077 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6079 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6080 have allow_domain_literals set, the ID did not get logged in the <= line.
6081 Domain literals are now always recognized in Message-ID: header lines.
6083 4. The first argument for a ${extract expansion item is the key name or field
6084 number. Leading and trailing spaces in this item were not being ignored,
6085 causing some misleading effects.
6087 5. When deliver_drop_privilege was set, single queue runner processes started
6088 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6089 same command in the process it spins off) were not dropping privilege.
6091 6. When the daemon running as "exim" started a queue runner, it always
6092 re-executed Exim in the spun-off process. This is a waste of effort when
6093 deliver_drop_privilege is set. The new process now just calls the
6094 queue-runner function directly.
6100 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6103 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6104 ignored. Though the use of -oP was forcing the writing of a pid file, it
6105 was always written to the default place.
6107 3. If the message "no IP address found for host xxxx" is generated during
6108 incoming verification, it is now followed by identification of the incoming
6109 connection (so you can more easily find what provoked it).
6111 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6113 5. Added some features to "harden" Exim a bit more against certain attacks:
6115 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6116 be put in Local/Makefile. This is like the never_users runtime option,
6117 but it cannot be overridden. The default setting is "root".
6119 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6120 prefix string with which any file named in a -C command line option
6123 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6124 is retained for -C and -D only if the caller of Exim is root. Without
6125 it, the exim user may also use -C and -D and retain privilege.
6127 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6128 command line option is disabled.
6130 6. Macro names set by the -D option must start with an upper case letter, just
6131 like macro names defined in the configuration file.
6133 7. Added "dereference=" facility to LDAP.
6135 8. Two instances of the typo "uknown" in the source files are fixed.
6137 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6138 the Configure-Makefile script screwed up while processing it.
6140 10. Incorporated PCRE 4.4.
6142 11. The SMTP synchronization check was not operating right at the start of an
6143 SMTP session. For example, it could not catch a HELO sent before the client
6144 waited for the greeting. There is now a check for outstanding input at the
6145 point when the greeting is written. Because of the duplex, asynchronous
6146 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6147 way, but not yet received, when the check is performed.
6149 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6150 on TCP/IP sockets, because this apparently causes some broken clients to
6153 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6154 unchanged) from the Cygwin maintainer.
6156 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6157 was supported instead of testing for mbx. Effectively a typo.
6159 15. The spa authenticator server code was not checking that the input it
6160 received was valid base64.
6162 16. The debug output line for the "set" modifier in ACLs was not showing the
6163 name of the variable that was being set.
6165 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6167 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6168 lookup on its own. Something else was needed to trigger the lookup. For
6169 example, a match in host_lookup or the need for a host name in a host list.
6170 Now, if $sender_host_name is referenced and the host name has not yet been
6171 looked up, a lookup is performed. If the lookup fails, the variable remains
6172 empty, and $host_lookup_failed is set to "1".
6174 19. Added "eqi" as a case-independent comparison operator.
6176 20. The saslauthd authentication condition could segfault if neither service
6177 nor realm was specified.
6179 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6180 error message that was logged was misleading, and incoming SMTP
6181 connections were dropped. The message is now more accurate, and temporary
6182 errors are given to SMTP connections.
6184 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6185 (except RSET) with a 421 error, until QUIT is received. However, it was
6186 failing to send a response to QUIT.
6188 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6189 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6190 ACL is now run after the helo_try_verify_hosts code.
6192 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6193 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6194 case-independent, so other case variants are also recognized). Apparently
6195 some systems use these upper case variants.
6197 25. If more than two messages were waiting for the same host, and a transport
6198 filter was specified for the transport, Exim sent two messages over the
6199 same TCP/IP connection, and then failed with "socket operation on non-
6200 socket" when it tried to send the third.
6202 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6204 27. The extern definition of crypt16() in expand.c was not being excluded when
6205 the OS had its own crypt16() function.
6207 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6208 as a preferred synonym for return_size_limit, both as an option and as an
6211 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6213 30. Changed the default configuration ACL to relax the local part checking rule
6214 for addresses that are not in any local domains. For these addresses,
6215 slashes and pipe symbols are allowed within local parts, but the sequence
6216 /../ is explicitly forbidden.
6218 31. SPA server authentication was not clearing the challenge buffer before
6221 32. log_message in a "warn" ACL statement was writing to the reject log as
6222 well as to the main log, which contradicts the documentation and doesn't
6223 seem right (because no rejection is happening). So I have stopped it.
6225 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6226 However, I am unable to do any testing of this.
6228 34. Fixed an infelicity in the appendfile transport. When checking directories
6229 for a mailbox, to see if any needed to be created, it was accidentally
6230 using path names with one or more superfluous leading slashes; tracing
6231 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6233 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6234 contents are added to the log line that is written for every discarded
6235 recipient. (Previously a log_message setting was ignored.)
6237 36. The ${quote: operator now quotes the string if it is empty.
6239 37. The install script runs exim in order to find its version number. If for
6240 some reason other than non-existence or emptiness, which it checks, it
6241 could not run './exim', it was installing it with an empty version number,
6242 i.e. as "exim-". This error state is now caught, and the installation is
6245 38. An argument was missing from the function that creates an error message
6246 when Exim fails to connect to the socket for saslauthd authentication.
6247 This could cause Exim to crash, or give a corrupted message.
6249 39. Added isip, isip4, and isip6 to ${if conditions.
6251 40. The ACL variables $acl_xx are now saved with the message, and can be
6252 accessed later in routers, transports, and filters.
6254 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6255 strings in the file are not string-expanded.
6257 42. If a MAIL command specified a SIZE value that was too large to fit into an
6258 int variable, the check against message_size_limit failed. Such values are
6259 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6260 day this will have to be increased, but I don't think I want to be around
6261 when emails are that large.
6268 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6269 iconv() is not standard in FreeBSD.
6271 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6272 IPv6 enabled. The observed symptom was a segmentation fault on return from
6273 the function os_common_find_running_interfaces() in src/os.c.
6275 3. In the check_special_case() function in daemon.c I had used "errno" as an
6276 argument name, which causes warnings on some systems. This was basically a
6277 typo, since it was named "eno" in the comments!
6279 4. The code that waits for the clock to tick (at a resolution of some fraction
6280 of a second) so as to ensure message-id uniqueness was always waiting for
6281 at least one whole tick, when it could have waited for less. [This is
6282 almost certainly not relevant at current processor speeds, where it is
6283 unlikely to ever wait at all. But we try to future-proof.]
6285 5. The function that sleeps for a time interval that includes fractions of a
6286 second contained a race. It did not block SIGALRM between setting the
6287 timer, and suspending (a couple of lines later). If the interval was short
6288 and the sigsuspend() was delayed until after it had expired, the suspension
6289 never ended. On busy systems this could lead to processes getting stuck for
6292 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6293 process, before it forks any delivery processes. The open lookup caching
6294 mechanism meant that the open file or database connection was passed into
6295 the delivery process. The problem was that delivery processes always tidy
6296 up cached lookup data. This could cause a problem for the next delivery
6297 process started by the queue runner, because the external queue runner
6298 process does not know about the closure. So the next delivery process
6299 still has data in the lookup cache. In the case of a file lookup, there was
6300 no problem because closing a file descriptor in a subprocess doesn't affect
6301 the parent. However, if the lookup was caching a connection to a database,
6302 the connection was closed, and the second delivery process was likely to
6303 see errors such as "PGSQL: query failed: server closed the connection
6304 unexpectedly". The problem has been fixed by closing all cached lookups
6305 in a queue runner before running a delivery process.
6307 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6308 seem to have the "const" qualifier which it has on other OS. I've
6311 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6312 *of the same type* (client or server) with the same public name that an
6313 error should be diagnosed.
6315 9. When Exim looked up a host name for an IP address, but failed to find the
6316 original IP address when looking up the host name (a safety check), it
6317 output the message "<ip address> does not match any IP for NULL", which was
6318 confusing, to say the least. The bug was that the host name should have
6319 appeared instead of "NULL".
6321 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6322 user that is built into the binary, and the -C or -D options is used, root
6323 privilege is dropped before the configuration file is read. In addition,
6324 logging is switched to stderr instead of the normal log files. If the
6325 configuration then re-defines the Exim user, the unprivileged environment
6326 is probably not what is expected, so Exim logs a panic warning message (but
6329 However, if deliver_drop_privilege is set, the unprivileged state may well
6330 be exactly what is intended, so the warning has been cut out in that case,
6331 and Exim is allowed to try to write to its normal log files.
6337 1. smtp_return_error_details was not giving details for temporary sender
6338 or receiver verification errors.
6340 2. Diagnose a configuration error if two authenticators have the same public
6343 3. Exim used not to create the message log file for a message until the first
6344 delivery attempt. This could be confusing when incoming messages were held
6345 for policy or load reasons. The message log file is now created at the time
6346 the message is received, and an initial "Received" line is written to it.
6348 4. The automatically generated man page for command line options had a minor
6349 bug that caused no ill effects; however, a more serious problem was that
6350 the procedure for building the man page automatically didn't always
6351 operate. Consequently, release 4.20 contains an out-of-date version. This
6352 shouldn't happen again.
6354 5. When building Exim with embedded Perl support, the script that builds the
6355 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6356 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6358 6. The freeze_tell option was not being used for messages that were frozen on
6359 arrival, either by an ACL or by local_scan().
6361 7. Added the smtp_incomplete_transaction log selector.
6363 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6364 was accepting AUTH without a new EHLO.
6366 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6367 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6370 10. Logging of TCP/IP connections (when configured) now happens in the main
6371 daemon process instead of the child process, so that the TCP/IP connection
6372 count is more accurate (but it can never be perfect).
6374 11. The use of "drop" in a nested ACL was not being handled correctly in the
6375 outer ACL. Now, if condition failure induced by the nested "drop" causes
6376 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6377 or "require"), the connection is dropped.
6379 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6380 that yield "discard" can now be used with an "accept" or a "discard" verb,
6381 but an error is generated for any others (because I can't see a useful way
6382 to define what should happen).
6384 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6385 are now processed in the same way as lines in the Exim configuration file.
6386 In particular, continuation lines are supported.
6388 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6390 15. Added -ti meaning -t -i.
6392 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6393 domains, and warn by logging if others are found.
6395 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6396 returned from GIFCONF when Exim is trying to find the list of interfaces on
6397 a host. The code in os.c has been modified to copy each ifreq to an aligned
6398 structure in all cases.
6400 Also, in some cases, the returned ifreq's were being copied to a 'struct
6401 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6402 means the last couple of bytes of an IPv6 address could be chopped if the
6403 ifreq contained only a normal sockaddr (14 bytes storage).
6405 18. Named domain lists were not supported in the hosts_treat_as_local option.
6406 An entry such as +xxxx was not recognized, and was treated as a literal
6409 19. Ensure that header lines added by a DATA ACL are included in the reject log
6410 if the ACL subsequently rejects the message.
6412 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6413 MD5 (which is deprecated).
6415 21. When testing a filter file using -bf, Exim was writing a message when it
6416 took the sender from a "From " line in the message, but it was not doing so
6417 when it took $return_path from a Return-Path: header line. It now does.
6419 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6420 with a valid header line field name (a series of printing characters
6421 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6423 23. Changed "disc" in the source to "disk" to conform to the documentation and
6424 the book and for uniformity.
6426 24. Ignore Sendmail's -Ooption=value command line item.
6428 25. When execve() failed while trying to run a command in a pipe transport,
6429 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6430 could be confused with a return value of 69 from the command itself. This
6431 has been changed to 127, the value the shell returns if it is asked to run
6432 a non-existent command. The wording for the related log line suggests a
6433 non-existent command as the problem.
6435 26. If received_header_text expands to an empty string, do not add a Received:
6436 header line to the message. (Well, it adds a token one on the spool, but
6437 marks it "old" so that it doesn't get used or transmitted.)
6439 27. Installed eximstats 1.28 (addition of -nt option).
6441 28. There was no check for failure on the call to getsockname() in the daemon
6442 code. This can fail if there is a shortage of resources on the system, with
6443 ENOMEM, for example. A temporary error is now given on failure.
6445 29. Contrary to the C standard, it seems that in some environments, the
6446 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6447 program. Exim now does this explicitly; it affects the formatting of
6448 timestamps using strftime().
6450 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6451 complaints. I've now initialized all the variables, to avoid this.
6453 32. Header lines added by a system filter were not being "seen" during
6454 transport-time rewrites.
6456 33. The info_callback() function passed to OpenSSL is set up with type void
6457 (*)(SSL *, int, int), as described somewhere. However, when calling the
6458 function (actually a macro) that sets it up, the type void(*)() is
6459 expected. I've put in a cast to prevent warnings from picky compilers.
6461 34. If a DNS black list lookup found a CNAME record, but there were no A
6462 records associated with the domain it pointed at, Exim crashed.
6464 35. If a DNS black list lookup returned more than one A record, Exim ignored
6465 all but the first. It now scans all returned addresses if a particular IP
6466 value is being sought. In this situation, the contents of the
6467 $dnslist_value variable are a list of all the addresses, separated by a
6470 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6471 to accept a host name and all its aliases if the forward lookup for any of
6472 them yielded the IP address of the incoming connection. Now it accepts only
6473 those names whose forward lookup yields the correct IP address. Any other
6474 names are discarded. This closes a loophole whereby a rogue DNS
6475 administrator could create reverse DNS records to break through a
6476 wildcarded host restriction in an ACL.
6478 37. If a user filter or a system filter that ran in a subprocess used any of
6479 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6480 the wrong values were passed to the pipe command ($thisaddress had the
6481 value of $0, $0 had the value of $1, etc). This bug was introduced by
6482 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6484 38. Improved the line breaking for long SMTP error messages from ACLs.
6485 Previously, if there was no break point between 40 and 75 characters, Exim
6486 left the rest of the message alone. Two changes have been made: (a) I've
6487 reduced the minimum length to 35 characters; (b) if it can't find a break
6488 point between 35 and 75 characters, it looks ahead and uses the first one
6489 that it finds. This may give the occasional overlong line, but at least the
6490 remaining text gets split now.
6492 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6493 file descriptors might be low, and that setting 1000 would always raise it.
6494 It turns out that in some environments, the limit is already over 1000 and
6495 that lowering it causes trouble. So now Exim takes care not to decrease it.
6497 40. When delivering a message, the value of $return_path is set to $sender_
6498 address at the start of routing (routers may change the value). By an
6499 oversight, this default was not being set up when an address was tested by
6500 -bt or -bv, which affected the outcome if any router or filter referred to
6503 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6504 log only when "message" or "log_message" are set. However, if one of the
6505 conditions was an address verification, or a call to a nested ACL, the
6506 messages generated by the underlying test were being passed through. This
6507 no longer happens. The underlying message is available in $acl_verify_
6508 message for both "message" and "log_message" expansions, so it can be
6509 passed through if needed.
6511 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6512 new expansion $bh_ to give the encoded byte string without charset
6513 translation. Translation happens only if iconv() is available; HAVE_ICONV
6514 indicates this at build time. HEADERS_CHARSET gives the charset to
6515 translate to; headers_charset can change it in the configuration, and
6516 "headers charset" can change it in an individual filter file.
6518 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6519 that creates RFC 2047 encoded "words" labels them as that charset instead
6520 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6521 expansion operator; (ii) when Exim creates a From: line for a local
6522 message; (iii) when a header line is rewritten to include a "phrase" part.
6524 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6525 buggy, causing it to skip the first lines of messages whose message ID
6526 ended in 'D'. This would not have bitten before Exim release 4.14, because
6527 message IDs were unlikely to end in 'D' before then. The effect was to have
6528 incorrect size information for certain domains.
6530 45. #include "config.h" was missing at the start of the crypt16.c module. This
6531 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6534 46. If there was a timeout during a "random" callout check, Exim treated it as
6535 a failure of the random address, and carried on sending RSET and the real
6536 address. If the delay was just some slowness somewhere, the response to the
6537 original RCPT would be taken as a response to RSET and so on, causing
6538 mayhem of various kinds.
6540 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6541 when I implemented it. It didn't allow for the fact that some option values
6542 may legitimately be negative (e.g. size_addition), and it didn't even do
6543 the right test for positive values.
6545 48. Domain names in DNS records are case-independent. Exim always looks them up
6546 in lower case. Some resolvers return domain names in exactly the case they
6547 appear in the zone file, that is, they may contain uppercase letters. Not
6548 all resolvers do this - some return always lower case. Exim was treating a
6549 change of case by a resolver as a change of domain, similar to a widening
6550 of a domain abbreviation. This triggered its re-routing code and so it was
6551 trying to route what was effectively the same domain again. This normally
6552 caused routing to fail (because the router wouldn't handle the domain
6553 twice). Now Exim checks for this case specially, and just changes the
6554 casing of the domain that it ultimately uses when it transmits the message
6557 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6560 50. If a filter generated a file delivery with a non-absolute name (possible if
6561 no home directory exists for the router), the forbid_file option was not
6564 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6565 the existing equality matching.
6567 52. Exim was using ints instead of ino_t variables in some places where it was
6568 dealing with inode numbers.
6570 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6571 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6572 called TMPDIR, and if it finds it is different, it changes its value.
6574 54. The smtp_printf() function is now made available to local_scan() so
6575 additional output lines can be written before returning. There is also an
6576 smtp_fflush() function to enable the detection of a dropped connection.
6577 The variables smtp_input and smtp_batched_input are exported to
6580 55. Changed the default runtime configuration: the message "Unknown user"
6581 has been removed from the ACL, and instead placed on the localuser router,
6582 using the cannot_route_message feature. This means that any verification
6583 failures that generate their own messages won't get overridden. Similarly,
6584 the "Unrouteable address" message that was in the ACL for unverifiable
6585 relay addresses has also been removed.
6587 56. Added hosts_avoid_esmtp to the smtp transport.
6589 57. The exicyclog script was not checking for the esoteric option
6590 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6591 will work only if exicyclog is run under the appropriate euid.
6593 58. Following a discussion on the list, the rules by which Exim recognises line
6594 endings on incoming messages have been changed. The -dropcr and drop_cr
6595 options are now no-ops, retained only for backwards compatibility. The
6596 following line terminators are recognized: LF CRLF CR. However, special
6597 processing applies to CR:
6599 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6600 nor a local message in the state where . is a terminator.
6602 (ii) If a bare CR is encountered in a header line, an extra space is added
6603 after the line terminator so as not to end the header. The reasoning
6604 behind this is that bare CRs in header lines are most likely either
6605 to be mistakes, or people trying to play silly games.
6607 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6608 was being incorrectly given as 18 bytes larger than it should have been.
6609 This is a VOB (very old bug).
6611 60. This may never have affected anything current, but just in case it has:
6612 When the local host is found other than at the start of a list of hosts,
6613 the local host, those with the same MX, and any that follow, are discarded.
6614 When the list in question was part of a longer list of hosts, the following
6615 hosts (not currently being processed) were also being discarded. This no
6616 longer happens. I'm not sure if this situation could ever has previously
6619 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6622 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6623 header. This is something that is recommended in a new Internet Draft, and
6624 is something that is documented as being done by Sendmail. There are two
6625 possible values. For messages generated by the autoreply transport, Exim
6628 Auto-Submitted: auto-replied
6630 whereas for all other generated messages (e.g. bounces) it adds
6632 Auto-Submitted: auto-generated
6634 63. The "personal" condition in filters now includes a test for the
6635 Auto-Submitted: header. If it contains the string "auto-" the message it
6636 not considered personal.
6638 64. Added rcpt_include_affixes as a generic transport option.
6640 65. Added queue_only_override (default true).
6642 66. Added the syslog_duplication option.
6644 67. If what should have been the first header line of a message consisted of
6645 a space followed by a colon, Exim was mis-interpreting it as a header line.
6646 It isn't of course - it is syntactically invalid and should therefore be
6647 treated as the start of the message body. The misbehaviour could have
6648 caused a number of strange effects, including loss of data in subsequent
6649 header lines, and spool format errors.
6651 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6652 client host had authenticated. This control can now be exercised by an ACL
6653 for more flexibility.
6655 69. By default, callouts do not happen when testing with -bh. There is now a
6656 variant, -bhc, which does actually run the callout code, including
6657 consulting and updating the callout cache.
6659 70. Added support for saslauthd authentication, courtesy of Alexander
6662 71. If statvfs() failed on the spool or log directories while checking their
6663 size for availability, Exim confusingly gave the error "space shortage".
6664 Furthermore, in debugging mode it crashed with a floating point exception.
6665 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6666 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6667 serious problem, Exim now writes to the main and panic logs when this
6668 happens, with details of the failure. It then refuses to accept the
6669 incoming message, giving the message "spool directory problem" or "log
6670 directory problem" with a 421 code for SMTP messages.
6672 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6673 0, 1, and 2 exist, because some OS complain for execs without them (see
6674 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6675 descriptors. However, the code omitted to check that the open succeeded,
6676 causing mysterious errors if for some reason the permissions on /dev/null
6677 got screwed. Now Exim writes a message to the main and panic logs, and
6678 bombs out if it can't open /dev/null.
6680 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6681 interact so that it is all more flexible. It is supposed to remain
6682 backwards compatible. Also added extra_local_interfaces.
6684 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6685 to bomb out with an assertion failure - to the client this appears as a
6686 connection drop. This problem occurs in the part of the code that was taken
6687 from the Samba project. Fortunately, the assertion is in a very simple
6688 function, so I have fixed this by reproducing the function inline in the
6689 one place where it is called, and arranging for authentication to fail
6690 instead of killing the process with assert().
6692 75. The SPA client code was not working when the server requested OEM rather
6693 than Unicode encoding.
6695 76. Added code to make require_files with a specific uid setting more usable in
6696 the case where statting the file as root fails - usually a non-root-mounted
6697 NFS file system. When this happens and the failure is EACCES, Exim now
6698 forks a subprocess and does the per-uid checking as the relevant uid.
6700 77. Added process_log_path.
6702 78. If log_file_path was not explicitly set, a setting of check_log_space or
6703 check_log_inodes was ignored.
6705 79. If a space check for the spool or log partitions fails, the incident is now
6706 logged. Of course, in the latter case the data may get lost...
6708 80. Added the %p formatting code to string_format() so that it can be used to
6709 print addresses in debug_print(). Adjusted all the address printing in the
6710 debugging in store.c to use %p rather than %d.
6712 81. There was a concern that a line of code in smtp_in.c could overflow a
6713 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6714 initially expressed, the concern was not well-founded, because trailing
6715 spaces are removed early. However, if the trailing spaces were followed by
6716 a NULL, they did not get removed, so the overflow was possible. Two fixes
6719 (a) I re-wrote the offending code in a cleaner fashion.
6720 (b) If an incoming SMTP command contains a NULL character, it is rejected
6723 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6724 runs initgroups(), so that if the Exim user is in any additional groups,
6725 they will be used during message reception.
6731 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob