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.
56 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
57 supports it and a size is available (ie. the sending peer gave us one).
59 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
60 years of being deprecated). The replacements are the ACLs
61 acl_smtp_mime and acl_not_smtp_mime.
63 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
64 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
65 or even an in-clear connection were permitted. Now, if the host lookup
66 was dnssec and dane was requested then the host is only used if the
67 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
68 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
69 if one fails this test.
70 This means that a poorly-configured remote DNS will make it incommunicado;
71 but it protects against a DNS-interception attack on it.
73 JH/04 Bug 1810: make continued-use of an open smtp transport connection
74 non-noisy when a race steals the message being considered.
76 JH/05 If main configuration option tls_certificate is unset, generate a
77 self-signed certificate for inbound TLS connections.
79 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
80 in rewrites and routers.
82 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
83 and logged a warning sing 4.83; now they are a configuration file error.
85 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
86 (lacking @domain). Apply the same qualification processing as RCPT.
88 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
90 JH/10 Support ${sha256:} applied to a string (as well as the previous
93 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
94 a cutthrough deliver is pending, as we always want to make a connection.
95 This also avoids re-routing the message when later placing the cutthrough
96 connection after a verify cache hit.
97 Do not update it with the verify result either.
99 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
100 when routing results in more than one destination address.
102 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
103 signing (which inhibits the cutthrough capability). Previously only
104 the presence of an option was tested; now an expansion evaluating as
105 empty is permissible (obviously it should depend only on data available
106 when the cutthrough connection is made).
108 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
109 the relevant preceding SMTP command did not note the pipelining mode.
111 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
112 Previously they were not counted.
114 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
115 as one having no matching records. Previously we deferred the message
116 that needed the lookup.
118 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
119 distinguished as "(=".
121 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
122 for missing MX records. Previously it only worked for missing A records.
124 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
126 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
127 after the data-go-ahead and data-ack. Patch from Jason Betts.
129 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
130 even for a "none" policy. Patch from Tony Meyer.
132 JH/22 Fix continued use of a connection for further deliveries. If a port was
133 specified by a router, it must also match for the delivery to be
136 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
137 When one of the recipients of a message was unsuitable for the connection
138 (has no matching addresses), we lost track of needing to mark it
139 deferred. As a result mail would be lost.
141 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
143 JH/25 Decoding ACL controls is now done using a binary search; the source code
144 takes up less space and should be simpler to maintain. Merge the ACL
145 condition decode tables also, with similar effect.
147 JH/26 Fix problem with one_time used on a redirect router which returned the
148 parent address unchanged. A retry would see the parent address marked as
149 delivered, so not attempt the (identical) child. As a result mail would
152 JH/27 Fix a possible security hole, wherein a process operating with the Exim
153 UID can gain a root shell. Credit to http://www.halfdog.net/ for
154 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
157 JH/28 Enable {spool,log} filesystem space and inode checks as default.
158 Main config options check_{log,spool}_{inodes,space} are now
159 100 inodes, 10MB unless set otherwise in the configuration.
161 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
162 Previously it only applied to the main-section connection policy
165 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
167 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
168 by me. Added RFC7919 DH primes as an alternative.
170 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
171 are not in the system include path.
173 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
174 GnuTLS, when a session startup failed (eg because the client disconnected)
175 Exim did stdio operations after fclose. This was exposed by a recent
176 change which nulled out the file handle after the fclose.
178 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
179 signed directly by the cert-signing cert, rather than an intermediate
180 OCSP-signing cert. This is the model used by LetsEncrypt.
182 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
184 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
185 an incoming connection.
187 HS/02 Bug 1802: Do not half-close the connection after sending a request
190 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
191 fallback to "prime256v1".
193 JH/34 SECURITY: Use proper copy of DATA command in error message.
194 Could leak key material. Remotely exploitable. CVE-2016-9963.
200 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
201 and 3.4.4 - once the server is enabled to respond to an OCSP request
202 it does even when not requested, resulting in a stapling non-aware
203 client dropping the TLS connection.
205 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
206 support variable-length bit vectors. No functional change.
208 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
209 The I= interface field on outgoing lines is now after the H= remote
210 host field, same as incoming lines. There is a separate
211 outgoing_interface log selector which allows you to disable the
214 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
215 If not running log_selector +smtp_connection the mainlog would be held
216 open indefinitely after a "too many connections" event, including to a
217 deleted file after a log rotate. Leave the per net connection logging
218 leaving it open for efficiency as that will be quickly detected by the
219 check on the next write.
221 HS/01 Bug 1671: Fix post transport crash.
222 Processing the wait-<transport> messages could crash the delivery
223 process if the message IDs didn't exist for some reason. When
224 using 'split_spool_directory=yes' the construction of the spool
225 file name failed already, exposing the same netto behaviour.
227 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
228 mime_regex ACL conditions.
230 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
231 to DSN fail messages (bounces): remote IP, remote greeting, remote response
232 to HELO, local diagnostic string.
234 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
235 log line to debug. Even when configured with a tls authenticator many
236 client connections are expected to not authenticate in this way, so
237 an authenticate fail is not an error.
239 HS/02 Add the Exim version string to the process info. This way exiwhat
240 gives some more detail about the running daemon.
242 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
243 matter for fast-change records such as DNSBLs.
245 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
246 retry record, even if constant. There may be multiple transports with
247 different interface settings and the retry behaviour needs to be kept
250 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
252 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
254 JH/10 Bug 840: fix log_defer_output option of pipe transport
256 JH/11 Bug 830: use same host for all RCPTS of a message, even under
257 hosts_randomize. This matters a lot when combined with mua_wrapper.
259 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
260 ${quote_pgsql:<string>} operator.
262 JH/13 Bug 1708: avoid misaligned access in cached lookup.
264 JH/14 Change header file name for freeradius-client. Relevant if compiling
265 with Radius support; from the Gentoo tree and checked under Fedora.
267 JH/15 Bug 1712: Introduce $prdr_requested flag variable
269 JH/16 Bug 1714: Permit an empty string as expansion result for transport
270 option transport_filter, meaning no filtering.
272 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
274 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
275 defaults to "*" (all hosts). The variable is now available when not built
276 with TLS, default unset, mainly to enable keeping the testsuite sane.
277 If a server certificate is not supplied (via tls_certificate) an error is
278 logged, and clients will find TLS connections fail on startup. Presumably
279 they will retry in-clear.
280 Packagers of Exim are strongly encouraged to create a server certificate
281 at installation time.
283 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
284 with the $config_file variable.
286 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
287 in transport context, after the attempt, and per-recipient. The latter type
288 is per host attempted. The event data is the error message, and the errno
289 information encodes the lookup type (A vs. MX) used for the (first) host,
290 and the trailing two digits of the smtp 4xx response.
292 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
293 to write to mainlog (or rejectlog, paniclog) in the window between file
294 creation and permissions/ownership being changed. Particularly affects
295 installations where exicyclog is run as root, rather than exim user;
296 result is that the running daemon panics and dies.
298 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
300 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
301 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
302 "pri" and "weight". Note that the previous implicit priority given by the
303 list order is no longer honoured.
305 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
308 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
309 by defining SUPPORT_SOCKS.
311 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
312 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
313 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
314 variables are renamed to proxy_{local,external}_{address,port}.
316 JH/27 Move Internationalisation support from Experimental to mainline, enabled
317 for a build by defining SUPPORT_I18N
319 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
320 of the query string, and make ${quote_redis:} do that quoting.
322 JH/29 Move Events support from Experimental to mainline, enabled by default
323 and removable for a build by defining DISABLE_EVENT.
325 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
327 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
328 cached by the daemon.
330 JH/32 Move Redis support from Experimental to mainline, enabled for a build
331 by defining LOOKUP_REDIS. The libhiredis library is required.
333 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
334 keys are given for lookup.
336 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
337 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
338 only supported when built with TLS support. The PolarSSL SHA routines
339 are still used when the TLS library is too old for convenient support.
341 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
342 openssl_options), for security. OpenSSL forces this from version 1.1.0
343 server-side so match that on older versions.
345 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
346 allocation for $value could be released as the expansion processing
347 concluded, but leaving the global pointer active for it.
349 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
350 and to use the domains and local_parts ACL conditions.
352 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
353 incorrectly not doubled on cutthrough transmission, hence seen as a
354 body-termination at the receiving system - resulting in truncated mails.
355 Commonly the sender saw a TCP-level error, and retransmitted the message
356 via the normal store-and-forward channel. This could result in duplicates
357 received - but deduplicating mailstores were liable to retain only the
358 initial truncated version.
360 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
362 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
364 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
365 we're in there, support oversigning also; bug 1309.
367 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
369 HS/04 Add support for keep_environment and add_environment options.
371 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
372 either intentional arithmetic overflow during PRNG, or testing config-
375 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
376 delivery resulted in actual delivery. Cancel cutthrough before DATA
379 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
380 rejects a recipient: pass the reject to the originator.
382 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
383 Many were false-positives and ignorable, but it's worth fixing the
386 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
387 for the new environment-manipulation done at startup. Move the routines
388 from being local to tls.c to being global via the os.c file.
390 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
391 an extract embedded as result-arg for a map, the first arg for extract
392 is unavailable so we cannot tell if this is a numbered or keyed
393 extraction. Accept either.
399 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
402 JH/02 The smtp transport option "multi_domain" is now expanded.
404 JH/03 The smtp transport now requests PRDR by default, if the server offers
407 JH/04 Certificate name checking on server certificates, when exim is a client,
408 is now done by default. The transport option tls_verify_cert_hostnames
409 can be used to disable this per-host. The build option
410 EXPERIMENTAL_CERTNAMES is withdrawn.
412 JH/05 The value of the tls_verify_certificates smtp transport and main options
413 default to the word "system" to access the system default CA bundle.
414 For GnuTLS, only version 3.0.20 or later.
416 JH/06 Verification of the server certificate for a TLS connection is now tried
417 (but not required) by default. The verification status is now logged by
418 default, for both outbound TLS and client-certificate supplying inbound
421 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
424 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
425 Status Notification (bounce) messages are now MIME format per RFC 3464.
426 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
427 under the control of the dsn_advertise_hosts option, and routers may
428 have a dsn_lasthop option.
430 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
431 default, modifiable by a malware= option. The list separator for
432 the options can now be changed in the usual way. Bug 68.
434 JH/10 The smtp_receive_timeout main option is now expanded before use.
436 JH/11 The incoming_interface log option now also enables logging of the
437 local interface on delivery outgoing connections.
439 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
440 if the interface and destination host and port all match.
442 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
445 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
446 Patch from Andrew Lewis.
448 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
449 now supports optional time-restrictions, weighting, and priority
450 modifiers per server. Patch originally by <rommer@active.by>.
452 JH/16 The spamd_address main option now supports a mixed list of local
453 and remote servers. Remote servers can be IPv6 addresses, and
454 specify a port-range.
456 JH/17 Bug 68: The spamd_address main option now supports an optional
457 timeout value per server.
459 JH/18 Bug 1581: Router and transport options headers_add/remove can
460 now have the list separator specified.
462 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
465 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
468 JH/21 Support for the A6 type of dns record is withdrawn.
470 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
471 rather than the verbs used.
473 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
474 from 255 to 1024 chars.
476 JH/24 Verification callouts now attempt to use TLS by default.
478 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
479 are generic router options now. The defaults didn't change.
481 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
482 Original patch from Alexander Shikoff, worked over by JH.
484 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
485 files if ZCAT_COMMAND is not executable.
487 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
489 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
491 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
492 Normally benign, it bites when the pair was led to by a CNAME;
493 modern usage is to not canonicalize the domain to a CNAME target
494 (and we were inconsistent anyway for A-only vs AAAA+A).
496 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
498 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
499 when evaluating $sender_host_dnssec.
501 JH/31 Check the HELO verification lookup for DNSSEC, adding new
502 $sender_helo_dnssec variable.
504 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
506 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
508 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
510 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
511 documented as working, but never had. Support all but $spam_report.
513 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
514 added for tls authenticator.
516 HS/03 Add perl_taintmode main config option
522 TL/01 When running the test suite, the README says that variables such as
523 no_msglog_check are global and can be placed anywhere in a specific
524 test's script, however it was observed that placement needed to be near
525 the beginning for it to behave that way. Changed the runtest perl
526 script to read through the entire script once to detect and set these
527 variables, reset to the beginning of the script, and then run through
528 the script parsing/test process like normal.
530 TL/02 The BSD's have an arc4random API. One of the functions to induce
531 adding randomness was arc4random_stir(), but it has been removed in
532 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
533 function when detected.
535 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
536 cause callback expansion.
538 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
539 syntax errors in an expansion can be treated as a string instead of
540 logging or causing an error, due to the internal use of bool_lax
541 instead of bool when processing it.
543 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
544 server certificates when making smtp deliveries.
546 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
548 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
550 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
552 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
553 Merged patch from Sebastian Wiedenroth.
555 JH/05 Fix results-pipe from transport process. Several recipients, combined
556 with certificate use, exposed issues where response data items split
557 over buffer boundaries were not parsed properly. This eventually
558 resulted in duplicates being sent. This issue only became common enough
559 to notice due to the introduction of connection certificate information,
560 the item size being so much larger. Found and fixed by Wolfgang Breyha.
562 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
563 size buffer was used, resulting in syntax errors when an expansion
566 JH/07 Add support for directories of certificates when compiled with a GnuTLS
567 version 3.3.6 or later.
569 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
570 is EXPERIMENTAL_EVENT, the main-configuration and transport options
571 both become "event_action", the variables become $event_name, $event_data
572 and $event_defer_errno. There is a new variable $verify_mode, usable in
573 routers, transports and related events. The tls:cert event is now also
574 raised for inbound connections, if the main configuration event_action
577 TL/06 In test suite, disable OCSP for old versions of openssl which contained
578 early OCSP support, but no stapling (appears to be less than 1.0.0).
580 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
581 server certificate names available under the smtp transport option
582 "tls_verify_cert_hostname" now do not permit multi-component wildcard
585 JH/10 Time-related extraction expansions from certificates now use the main
586 option "timezone" setting for output formatting, and are consistent
587 between OpenSSL and GnuTLS compilations. Bug 1541.
589 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
590 encoded parameter in the incoming message. Bug 1558.
592 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
593 include certificate info, eximon was claiming there were spoolfile
596 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
598 JH/14 Log delivery-related information more consistently, using the sequence
599 "H=<name> [<ip>]" wherever possible.
601 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
602 are problematic for Debian distribution, omit them from the release
605 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
607 JH/16 Fix string representation of time values on 64bit time_t architectures.
610 JH/17 Fix a null-indirection in certextract expansions when a nondefault
611 output list separator was used.
616 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
617 checkers that were complaining about end of non-void function with no
620 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
621 This was a regression introduced in 4.83 by another bugfix.
623 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
625 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
626 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
632 TF/01 Correctly close the server side of TLS when forking for delivery.
634 When a message was received over SMTP with TLS, Exim failed to clear up
635 the incoming connection properly after forking off the child process to
636 deliver the message. In some situations the subsequent outgoing
637 delivery connection happened to have the same fd number as the incoming
638 connection previously had. Exim would try to use TLS and fail, logging
639 a "Bad file descriptor" error.
641 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
642 utilities have not been installed.
644 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
645 temporary space as the ACL may create new global variables.
647 TL/01 LDAP support uses per connection or global context settings, depending
648 upon the detected version of the libraries at build time.
650 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
651 to extract and use the src ip:port in logging and expansions as if it
652 were a direct connection from the outside internet. PPv2 support was
653 updated based on HAProxy spec change in May 2014.
655 JH/02 Add ${listextract {number}{list}{success}{fail}}.
657 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
658 Properly escape header and check for NULL return.
660 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
663 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
665 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
666 characters in header names, implemented as a verify condition.
667 Contributed by Michael Fischer v. Mollard.
669 TL/05 Rename SPF condition results err_perm and err_temp to standardized
670 results permerror and temperror. Previous values are deprecated but
671 still accepted. In a future release, err_perm and err_temp will be
672 completely removed, which will be a backward incompatibility if the
673 ACL tests for either of these two old results. Patch contributed by
674 user bes-internal on the mailing list.
676 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
678 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
679 selectors, in both main and reject logs.
681 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
684 JH/07 Add malware type "sock" for talking to simple daemon.
686 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
688 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
689 routers/transports under cutthrough routing.
691 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
692 numbers. Touch up "bool" conditional to keep the same definition.
694 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
696 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
698 JH/12 Expand items in router/transport headers_add or headers_remove lists
699 individually rather than the list as a whole. Bug 1452.
701 Required for reasonable handling of multiple headers_ options when
702 they may be empty; requires that headers_remove items with embedded
703 colons must have them doubled (or the list-separator changed).
705 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
706 view the policy declared in the DMARC record. Currently, $dmarc_status
707 is a combined value of both the record presence and the result of the
710 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
712 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
713 dnslookup router and the smtp transport (applying to the forward
716 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
717 of ldap servers used for a specific lookup. Patch provided by Heiko
720 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
721 New variable $lookup_dnssec_authenticated for observability.
723 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
724 Patch submitted by Lars Timman.
726 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
728 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
729 Requires trusted mode and valid format message id, aborts otherwise.
730 Patch contributed by Heiko Schlichting.
732 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
733 certextract with support for various fields. Bug 1358.
735 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
736 is requested by default, modifiable by smtp transport option
739 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
740 operate on certificate variables to give certificate fingerprints
741 Also new ${sha256:cert_variable}.
743 JH/23 The PRDR feature is moved from being Experimental into the mainline.
745 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
746 Christian Aistleitner.
748 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
750 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
751 file. Patch from Wolfgang Breyha.
753 JH/25 Expand the coverage of the delivery $host and $host_address to
754 client authenticators run in verify callout. Bug 1476.
756 JH/26 Port service names are now accepted for tls_on_connect_ports, to
757 align with daemon_smtp_ports. Bug 72.
759 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
760 support and error reporting did not work properly.
762 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
763 and is readable. Patch from Andrew Colin Kissa.
765 TL/14 Enhance documentation of ${run expansion and how it parses the
766 commandline after expansion, particularly in the case when an
767 unquoted variable expansion results in an empty value.
769 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
771 PP/02 Fix internal collision of T_APL on systems which support RFC3123
772 by renaming away from it. Addresses GH issue 15, reported by
775 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
777 TL/15 SECURITY: prevent double expansion in math comparison functions
778 (can expand unsanitized data). Not remotely exploitable.
785 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
787 PP/02 Make -n do something, by making it not do something.
788 When combined with -bP, the name of an option is not output.
790 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
793 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
794 $sender_host_name and config options to manage this, and basic check
797 PP/05 DSCP support for outbound connections and control modifier for inbound.
799 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
800 (Only plugin which currently uses this is kerberos4, which nobody should
801 be using, but we should make it available and other future plugins might
802 conceivably use it, even though it would break NAT; stuff *should* be
803 using channel bindings instead).
805 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
806 name; added for Sendmail compatibility; requires admin caller.
807 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
808 just ignore it); requires trusted caller.
809 Also parse but ignore: -Ac -Am -X<logfile>
812 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
814 TL/02 Add +smtp_confirmation as a default logging option.
816 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
817 Patch by Magnus Holmgren from 2007-02-20.
819 TL/04 Bugzilla 1281 - Spec typo.
820 Bugzilla 1283 - Spec typo.
821 Bugzilla 1290 - Spec grammar fixes.
823 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
825 TL/06 Add Experimental DMARC support using libopendmarc libraries.
827 TL/07 Fix an out of order global option causing a segfault. Reported to dev
828 mailing list by by Dmitry Isaikin.
830 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
832 JH/02 Support "G" suffix to numbers in ${if comparisons.
834 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
836 NM/01 Bugzilla 1197 - Spec typo
837 Bugzilla 1196 - Spec examples corrections
839 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
841 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
842 gnutls_enable_pkcs11, but renamed to more accurately indicate its
845 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
846 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
848 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
849 "acl {{name}{arg}...}", and optional args on acl condition
852 JH/05 Permit multiple router/transport headers_add/remove lines.
854 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
856 JH/07 Avoid using a waiting database for a single-message-only transport.
857 Performance patch from Paul Fisher. Bugzilla 1262.
859 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
862 JH/09 Add $headers_added variable, with content from use of ACL modifier
863 add_header (but not yet added to the message). Bugzilla 199.
865 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
866 Pulled from Bugzilla 817 by Wolfgang Breyha.
868 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
870 (nb: this is the same fix as in Exim 4.80.1)
872 JH/11 Add A= logging on delivery lines, and a client_set_id option on
875 JH/12 Add optional authenticated_sender logging to A= and a log_selector
878 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
880 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
881 advertise SMTP AUTH mechanism to us, instead of a generic
882 protocol violation error. Also, make Exim more robust to bad
883 data from the Dovecot auth socket.
885 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
887 When a queue runner is handling a message, Exim first routes the
888 recipient addresses, during which it prunes them based on the retry
889 hints database. After that it attempts to deliver the message to
890 any remaining recipients. It then updates the hints database using
893 So if a recipient address works intermittently, it can get repeatedly
894 deferred at routing time. The retry hints record remains fresh so the
895 address never reaches the final cutoff time.
897 This is a fairly common occurrence when a user is bumping up against
898 their storage quota. Exim had some logic in its local delivery code
899 to deal with this. However it did not apply to per-recipient defers
900 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
902 This change adds a proper retry rule check during routing so that the
903 final cutoff time is checked against the message's age. We only do
904 this check if there is an address retry record and there is not a
905 domain retry record; this implies that previous attempts to handle
906 the address had the retry_use_local_parts option turned on. We use
907 this as an approximation for the destination being like a local
908 delivery, as in LMTP.
910 I suspect this new check makes the old local delivery cutoff check
911 redundant, but I have not verified this so I left the code in place.
913 TF/02 Correct gecos expansion when From: is a prefix of the username.
915 Test 0254 submits a message to Exim with the header
919 When I ran the test suite under the user fanf2, Exim expanded
920 the header to contain my full name, whereas it should have added
921 a Resent-Sender: header. It erroneously treats any prefix of the
922 username as equal to the username.
924 This change corrects that bug.
926 GF/01 DCC debug and logging tidyup
927 Error conditions log to paniclog rather than rejectlog.
928 Debug lines prefixed by "DCC: " to remove any ambiguity.
930 TF/03 Avoid unnecessary rebuilds of lookup-related code.
932 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
933 Bug spotted by Jeremy Harris; was flawed since initial commit.
934 Would have resulted in OCSP responses post-SNI triggering an Exim
935 NULL dereference and crash.
937 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
939 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
940 Bug detection, analysis and fix by Samuel Thibault.
941 Bugzilla 1331, Debian bug #698092.
943 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
945 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
946 Server implementation by Todd Lyons, client by JH.
947 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
948 config variable "prdr_enable" controls whether the server
949 advertises the facility. If the client requests PRDR a new
950 acl_data_smtp_prdr ACL is called once for each recipient, after
951 the body content is received and before the acl_smtp_data ACL.
952 The client is controlled by both of: a hosts_try_prdr option
953 on the smtp transport, and the server advertisement.
954 Default client logging of deliveries and rejections involving
955 PRDR are flagged with the string "PRDR".
957 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
958 fclose(). Diagnosis by Todd Lyons.
960 PP/17 Update configure.default to handle IPv6 localhost better.
961 Patch by Alain Williams (plus minor tweaks).
964 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
965 This is now consistent with GnuTLS, and is now documented: the
966 previous undocumented portable approach to treating the option as
967 unset was to force an expansion failure. That still works, and
968 an empty string is now equivalent.
970 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
971 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
972 not performing validation itself.
974 PP/20 Added force_command boolean option to pipe transport.
975 Patch from Nick Koston, of cPanel Inc.
977 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
980 TF/04 Added udpsend ACL modifier and hexquote expansion operator
982 PP/21 Fix eximon continuous updating with timestamped log-files.
983 Broken in a format-string cleanup in 4.80, missed when I repaired the
984 other false fix of the same issue.
985 Report and fix from Heiko Schlichting.
988 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
989 Report from Prashanth Katuri.
991 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
992 It's SecureTransport, so affects any MacOS clients which use the
993 system-integrated TLS libraries, including email clients.
995 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
996 using a MIME ACL for non-SMTP local injection.
997 Report and assistance in diagnosis by Warren Baker.
999 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1001 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1003 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1004 last id that failed so it may be referenced in subsequent ACL's.
1006 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1009 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1010 ldap library initialization, allowing self-signed CA's to be
1011 used. Also properly sets require_cert option later in code by
1012 using NULL (global ldap config) instead of ldap handle (per
1013 session). Bug diagnosis and testing by alxgomz.
1015 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1016 the src/util/ subdirectory.
1018 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1019 renamed to Transport Post Delivery Action by Jeremy Harris, as
1022 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1023 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1024 redis_servers = needs to be configured which will be used by the redis
1025 lookup. Patch from Warren Baker, of The Packet Hub.
1027 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1029 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1030 hostname or reverse DNS when processing a host list. Used suggestions
1031 from multiple comments on this bug.
1033 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1035 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1036 Missed a few lines, added it to make the runtest require no keyboard
1039 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1040 contains upper case chars. Make router use caseful_local_part.
1042 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1043 support when GnuTLS has been built with p11-kit.
1049 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1051 This, or similar/improved, will also be change PP/11 of 4.82.
1057 PP/01 Handle short writes when writing local log-files.
1058 In practice, only affects FreeBSD (8 onwards).
1059 Bugzilla 1053, with thanks to Dmitry Isaikin.
1061 NM/01 Bugzilla 949 - Documentation tweak
1063 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1066 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1068 PP/02 Implemented gsasl authenticator.
1070 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1072 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1073 `pkg-config foo` for cflags/libs.
1075 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1076 with rest of GSASL and with heimdal_gssapi.
1078 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1079 `pkg-config foo` for cflags/libs for the TLS implementation.
1081 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1082 properties get this fed in as external SSF. A number of robustness
1083 and debugging improvements to the cyrus_sasl authenticator.
1085 PP/08 cyrus_sasl server now expands the server_realm option.
1087 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1088 Patch by Jeremy Harris.
1090 PP/10 Added dbmjz lookup type.
1092 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1094 PP/12 MAIL args handles TAB as well as SP, for better interop with
1095 non-compliant senders.
1096 Analysis and variant patch by Todd Lyons.
1098 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1099 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1100 Patch from Dirk Mueller <dmueller@suse.com>
1102 PP/13 tls_peerdn now print-escaped for spool files.
1103 Observed some $tls_peerdn in wild which contained \n, which resulted
1104 in spool file corruption.
1106 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1107 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1108 or write after TLS renegotiation, which otherwise led to messages
1111 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1112 as a tracking header (ie: a signed header comes before the signature).
1113 Patch from Wolfgang Breyha.
1115 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1116 comma-sep list; embedded commas doubled.
1118 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1120 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1122 Report and patch from Dmitry Banschikov.
1124 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1125 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1126 failures. We appear to now support TLS1.1+ with Exim.
1128 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1129 lets Exim select keys and certificates based upon TLS SNI from client.
1130 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1131 before an outbound SMTP session. New log_selector, +tls_sni.
1133 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1134 NULL dereference. Report and patch from Alun Jones.
1136 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1137 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1138 Not seeing resolver debug output on NetBSD, but suspect this is a
1139 resolver implementation change.
1141 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1142 Left warnings. Added "eximon gdb" invocation mode.
1144 PP/21 Defaulting "accept_8bitmime" to true, not false.
1146 PP/22 Added -bw for inetd wait mode support.
1148 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1149 locate the relevant includes and libraries. Made this the default.
1151 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1152 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1154 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1155 This may cause build issues on older platforms.
1157 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1158 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1159 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1160 Added SNI support via GnuTLS too.
1161 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1163 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1165 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1166 Applied second patch from Janne, implementing suggestion to default
1167 multiple-strings-in-record handling to match SPF spec.
1169 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1171 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1172 read-only, out of scope).
1173 Patch from Wolfgang Breyha, report from Stuart Northfield.
1175 PP/29 Fix three issues highlighted by clang analyser static analysis.
1176 Only crash-plausible issue would require the Cambridge-specific
1177 iplookup router and a misconfiguration.
1178 Report from Marcin Mirosław.
1180 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1182 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1183 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1184 As part of this, removing so much warning spew let me fix some minor
1185 real issues in debug logging.
1187 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1188 assignment on my part. Fixed.
1190 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1191 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1192 Janne Snabb (who went above and beyond: thank you).
1194 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1195 string otherwise requires a connection and a bunch more work and it's
1196 relatively easy to get wrong. Should also expose TLS library linkage
1199 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1200 64-bit ${eval} (JH/03).
1202 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1203 GNU libc to support some of the 64-bit stuff, should not lead to
1204 conflicts. Defined before os.h is pulled in, so if a given platform
1205 needs to override this, it can.
1207 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1208 protection layer was required, which is not implemented.
1209 Bugzilla 1254, patch from Wolfgang Breyha.
1211 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1212 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1213 tls_dhparam take prime identifiers. Also unbreak combination of
1214 OpenSSL+DH_params+TLSSNI.
1216 PP/39 Disable SSLv2 by default in OpenSSL support.
1222 PP/01 Solaris build fix for Oracle's LDAP libraries.
1223 Bugzilla 1109, patch from Stephen Usher.
1225 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1227 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1230 TF/02 Fix a couple more cases where we did not log the error message
1231 when unlink() failed. See also change 4.74-TF/03.
1233 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1234 lock up or crash if it happened to be inside a call to libc when it
1235 got a SIGUSR1 from exiwhat.
1237 The SIGUSR1 handler appends the current process status to the process
1238 log which is later printed by exiwhat. It used to use the general
1239 purpose logging code to do this, but several functions it calls are
1240 not safe for signals.
1242 The new output code in the SIGUSR1 handler is specific to the process
1243 log, and simple enough that it's easy to inspect for signal safety.
1244 Removing some special cases also simplifies the general logging code.
1245 Removing the spurious timestamps from the process log simplifies
1248 TF/04 Improved ratelimit ACL condition.
1250 The /noupdate option has been deprecated in favour of /readonly which
1251 has clearer semantics. The /leaky, /strict, and /readonly update modes
1252 are mutually exclusive. The update mode is no longer included in the
1253 database key; it just determines when the database is updated. (This
1254 means that when you upgrade Exim will forget old rate measurements.)
1256 Exim now checks that the per_* options are used with an update mode that
1257 makes sense for the current ACL. For example, when Exim is processing a
1258 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1259 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1260 must specify per_mail/readonly. If you omit the update mode it defaults to
1261 /leaky where that makes sense (as before) or /readonly where required.
1263 The /noupdate option is now undocumented but still supported for
1264 backwards compatibility. It is equivalent to /readonly except that in
1265 ACLs where /readonly is required you may specify /leaky/noupdate or
1266 /strict/noupdate which are treated the same as /readonly.
1268 A useful new feature is the /count= option. This is a generalization
1269 of the per_byte option, so that you can measure the throughput of other
1270 aggregate values. For example, the per_byte option is now equivalent
1271 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1273 The per_rcpt option has been generalized using the /count= mechanism
1274 (though it's more complicated than the per_byte equivalence). When it is
1275 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1276 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1277 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1278 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1279 Note that using per_rcpt with a non-readonly update mode in more than
1280 one ACL will cause the recipients to be double-counted. (The per_mail
1281 and per_byte options don't have this problem.)
1283 The handling of very low rates has changed slightly. If the computed rate
1284 is less than the event's count (usually one) then this event is the first
1285 after a long gap. In this case the rate is set to the same as this event's
1286 count, so that the first message of a spam run is counted properly.
1288 The major new feature is a mechanism for counting the rate of unique
1289 events. The new per_addr option counts the number of different
1290 recipients that someone has sent messages to in the last time period. It
1291 behaves like per_rcpt if all the recipient addresses are different, but
1292 duplicate recipient addresses do not increase the measured rate. Like
1293 the /count= option this is a general mechanism, so the per_addr option
1294 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1295 example, measure the rate that a client uses different sender addresses
1296 with the options per_mail/unique=$sender_address. There are further
1297 details in the main documentation.
1299 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1301 TF/06 Removed a few PCRE remnants.
1303 TF/07 Automatically extract Exim's version number from tags in the git
1304 repository when doing development or release builds.
1306 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1307 Bugzilla 879. Patch from Paul Fisher.
1309 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1310 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1313 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1315 PP/05 Variable $av_failed, true if the AV scanner deferred.
1316 Bugzilla 1078. Patch from John Horne.
1318 PP/06 Stop make process more reliably on build failure.
1319 Bugzilla 1087. Patch from Heiko Schlittermann.
1321 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1322 Bugzilla 1089. Patch from Heiko Schlittermann.
1324 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1325 Bugzilla 1131. Patch from Holger Weiß.
1327 PP/09 Handle IPv6 addresses with SPF.
1328 Bugzilla 860. Patch from Wolfgang Breyha.
1330 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1332 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1335 PP/11 match_* no longer expand right-hand-side by default.
1336 New compile-time build option, EXPAND_LISTMATCH_RHS.
1337 New expansion conditions, "inlist", "inlisti".
1339 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1341 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1343 PP/14 fix log_write() format string regression from TF/03.
1344 Bugzilla 1152. Patch from Dmitry Isaikin.
1350 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1352 PP/02 Harmonised TLS library version reporting; only show if debugging.
1353 Layout now matches that introduced for other libraries in 4.74 PP/03.
1355 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1357 PP/04 New "dns_use_edns0" global option.
1359 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1362 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1363 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1365 TK/01 Updated PolarSSL code to 0.14.2.
1366 Bugzilla 1097. Patch from Andreas Metzler.
1368 PP/07 Catch divide-by-zero in ${eval:...}.
1369 Fixes bugzilla 1102.
1371 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1374 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1375 format-string attack -- SECURITY: remote arbitrary code execution.
1377 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1378 time unintentionally subject to list matching rules, letting the header
1379 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1380 arbitrary string expansion). This allowed for information disclosure.
1382 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1383 INT_MIN/-1 -- value coerced to INT_MAX.
1389 NM/01 Workaround for PCRE version dependency in version reporting
1392 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1393 This fixes portability to compilers other than gcc, notably
1394 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1396 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1397 makefiles for portability to HP-UX and POSIX correctness.
1399 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1400 Also via indented variable definition in the Makefile.
1401 (Debugging by Oliver Heesakkers).
1403 PP/02 Restore caching of spamd results with expanded spamd_address.
1404 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1406 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1407 Improves build reliability. Fix from: Frank Elsner
1409 NM/02 Fix wide character breakage in the rfc2047 coding
1410 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1412 NM/03 Allow underscore in dnslist lookups
1413 Fixes bug 1026. Patch from Graeme Fowler
1415 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1416 Code patches from Adam Ciarcinski of NetBSD.
1418 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1421 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1422 is logged, to avoid truncation. Patch from John Horne.
1424 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1425 Patch from Jakob Hirsch.
1427 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1428 SQL string expansion failure details.
1429 Patch from Andrey Oktyabrski.
1431 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1432 Patch from Simon Arlott.
1434 PP/09 New lookups functionality failed to compile on old gcc which rejects
1435 extern declarations in function scope.
1436 Patch from Oliver Fleischmann
1438 PP/10 Use sig_atomic_t for flags set from signal handlers.
1439 Check getgroups() return and improve debugging.
1440 Fixed developed for diagnosis in bug 927 (which turned out to be
1443 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1444 Patch from Mark Zealey.
1446 PP/12 Bugzilla 1056: Improved spamd server selection.
1447 Patch from Mark Zealey.
1449 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1450 Based on patch from Heiko Schlittermann.
1452 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1453 Patch from Uwe Doering, sign-off by Michael Haardt.
1455 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1456 variable declaration deep within a block. Bug and patch from
1459 PP/15 lookups-Makefile IRIX compatibility coercion.
1461 PP/16 Make DISABLE_DKIM build knob functional.
1463 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1464 Patch by Simon Arlott
1466 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1467 variable argument macros. Our copy now differs from upstream.
1473 TF/01 Failure to get a lock on a hints database can have serious
1474 consequences so log it to the panic log.
1476 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1477 controlled using the smtp_confirmation log selector.
1479 TF/03 Include the error message when we fail to unlink a spool file.
1481 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1482 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1483 for maintaining out-of-tree patches for some time.
1485 PP/01 Bugzilla 139: Documentation and portability issues.
1486 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1487 Handle per-OS dynamic-module compilation flags.
1489 PP/02 Let /dev/null have normal permissions.
1490 The 4.73 fixes were a little too stringent and complained about the
1491 permissions on /dev/null. Exempt it from some checks.
1492 Reported by Andreas M. Kirchwitz.
1494 PP/03 Report version information for many libraries, including
1495 Exim version information for dynamically loaded libraries. Created
1496 version.h, now support a version extension string for distributors
1497 who patch heavily. Dynamic module ABI change.
1499 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1500 privilege escalation vulnerability whereby the Exim run-time user
1501 can cause root to append content of the attacker's choosing to
1504 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1507 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1508 If dropping privileges for untrusted macros, we disabled normal logging
1509 on the basis that it would fail; for the Exim run-time user, this is not
1510 the case, and it resulted in successful deliveries going unlogged.
1511 Fixed. Reported by Andreas Metzler.
1517 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1518 only prepend for the Resent-* case. Fixes regression introduced in
1519 Exim 4.70 by NM/22 for Bugzilla 607.
1521 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1522 increasing numbers of administrators be bitten by this.
1524 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1526 PP/03 Bugzilla 994: added openssl_options main configuration option.
1528 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1530 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1532 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1534 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1535 without a peer certificate, leading to a segfault because of an
1536 assumption that peers always have certificates. Be a little more
1537 paranoid. Problem reported by Martin Tscholak.
1539 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1540 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1541 NB: ClamAV planning to remove STREAM in "middle of 2010".
1542 CL also introduces -bmalware, various -d+acl logging additions and
1543 more caution in buffer sizes.
1545 PP/09 Implemented reverse_ip expansion operator.
1547 PP/10 Bugzilla 937: provide a "debug" ACL control.
1549 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1551 PP/12 Bugzilla 973: Implement --version.
1553 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1555 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1557 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1559 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1560 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1561 ignore trailing whitespace.
1563 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1565 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1568 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1569 Notification from Dr Andrew Aitchison.
1571 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1572 ExtendedDetectionInfo response format.
1573 Notification from John Horne.
1575 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1578 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1579 XSL and documented dependency on system catalogs, with examples of how
1582 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1585 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1586 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1587 configuration file which is writeable by the Exim user or group.
1589 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1590 of configuration files to cover files specified with the -C option if
1591 they are going to be used with root privileges, not just the default
1594 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1595 option (effectively making it always true).
1597 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1598 files to be used while preserving root privileges.
1600 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1601 that rogue child processes cannot use them.
1603 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1604 run-time user, instead of root.
1606 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1607 Exim run-time user without dropping privileges.
1609 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1610 result string, instead of calling string_vformat() twice with the same
1613 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1614 for other users. Others should always drop root privileges if they use
1615 -C on the command line, even for a whitelisted configure file.
1617 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1619 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1625 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1626 $data_path, and $header_path variables; fixed documentation bugs and
1629 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1630 exipick to access non-standard spools, including the "frozen" queue
1633 NM/01 Bugzilla 965: Support mysql stored procedures.
1634 Patch from Alain Williams
1636 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1638 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1639 Patch from Andreas Metzler
1641 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1642 Patch from Kirill Miazine
1644 NM/05 Bugzilla 671: Added umask to procmail example.
1646 JJ/03 installed exipick 20100323.0, fixing doc bug
1648 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1649 directory. Notification and patch from Dan Rosenberg.
1651 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1653 TK/02 Improve log output when DKIM signing operation fails.
1655 MH/01 Treat the transport option dkim_domain as a colon separated
1656 list, not as a single string, and sign the message with each element,
1657 omitting multiple occurences of the same signer.
1659 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1660 Bugzilla 985, 986. Patch by Simon Arlott
1662 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1663 Patch by Simon Arlott
1665 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1666 MBX locking. Notification from Dan Rosenberg.
1672 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1674 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1676 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1678 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1680 NM/04 Bugzilla 915: Flag broken perl installation during build.
1686 TK/01 Added patch by Johannes Berg that expands the main option
1687 "spamd_address" if it starts with a dollar sign.
1689 TK/02 Write list of recipients to X-Envelope-Sender header when building
1690 the mbox-format spool file for content scanning (suggested by Jakob
1693 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1694 (http://www.dcc-servers.net/) support via dccifd. Activated by
1695 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1697 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1698 by Mark Daniel Reidel <mr@df.eu>.
1700 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1701 When building exim an external PCRE library is now needed -
1702 PCRE is a system library on the majority of modern systems.
1703 See entry on PCRE_LIBS in EDITME file.
1705 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1706 conversation. Added nologin parameter to request.
1707 Patch contributed by Kirill Miazine.
1709 TF/01 Do not log submission mode rewrites if they do not change the address.
1711 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1713 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1714 log files in place. Contributed by Roberto Lima.
1716 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1718 TF/03 Bugzilla 615: When checking the local_parts router precondition
1719 after a local_part_suffix or local_part_prefix option, Exim now
1720 does not use the address's named list lookup cache, since this
1721 contains cached lookups for the whole local part.
1723 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1724 Robert Millan. Documentation is in experimental-spec.txt.
1726 TF/04 Bugzilla 668: Fix parallel build (make -j).
1728 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1730 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1731 Patch provided by Jan Srzednicki.
1733 TF/05 Leading white space used to be stripped from $spam_report which
1734 wrecked the formatting. Now it is preserved.
1736 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1737 that they are available at delivery time.
1739 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1741 TF/08 TLS error reporting now respects the incoming_interface and
1742 incoming_port log selectors.
1744 TF/09 Produce a more useful error message if an SMTP transport's hosts
1745 setting expands to an empty string.
1747 NM/06 Bugzilla 744: EXPN did not work under TLS.
1748 Patch provided by Phil Pennock.
1750 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1751 Patch provided by Richard Godbee.
1753 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1754 acl_smtp_notquit, added index entry.
1756 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1757 Patch provided by Eugene Bujak.
1759 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1760 accept(). Patch provided by Maxim Dounin.
1762 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1763 Patch provided by Phil Pennock.
1765 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1767 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1768 Patch provided by Brad "anomie" Jorsch.
1770 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1771 Patch provided by Dean Brooks.
1773 TK/05 Add native DKIM support (does not depend on external libraries).
1775 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1776 Patch provided by Graeme Fowler.
1778 NM/16 Bugzilla 851: Documentation example syntax fix.
1780 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1782 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1785 NM/19 Bugzilla 745: TLS version reporting.
1786 Patch provided by Phil Pennock.
1788 NM/20 Bugzilla 167: bool: condition support.
1789 Patch provided by Phil Pennock.
1791 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1792 clients. Patch provided by Phil Pennock.
1794 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1795 Patch provided by Brad "anomie" Jorsch.
1797 NM/23 Bugzilla 687: Fix misparses in eximstats.
1798 Patch provided by Heiko Schlittermann.
1800 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1801 Patch provided by Heiko Schlittermann.
1803 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1804 plus update to original patch.
1806 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1808 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1809 Patch provided by David Brownlee.
1811 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1813 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1815 NM/30 Bugzilla 888: TLS documentation bugfixes.
1817 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1819 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1820 Unlike the original bugzilla I have changed all shell scripts in src tree.
1822 NM/33 Bugzilla 898: Transport filter timeout fix.
1823 Patch by Todd Rinaldo.
1825 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1826 Patch by Serge Demonchaux.
1828 NM/35 Bugzilla 39: Base64 decode bug fixes.
1829 Patch by Jakob Hirsch.
1831 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1833 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1835 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1837 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1843 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1844 ALT-N's libdkim that I have put here:
1845 http://duncanthrax.net/exim-experimental/
1847 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1848 were called 'true' and it seems that C99 defines that as a reserved
1849 keyword to be used with 'bool' variable types. That means you could
1850 not include C99-style headers which use bools without triggering
1851 build errors in sieve.c.
1853 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1854 as mailq or other aliases. Changed the --help handling significantly
1855 to do whats expected. exim_usage() emits usage/help information.
1857 SC/01 Added the -bylocaldomain option to eximstats.
1859 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1861 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1863 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1869 PH/01 Another patch from the Sieve maintainer.
1871 PH/02 When an IPv6 address is converted to a string for single-key lookup
1872 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1873 dots are used instead of colons so that keys in lsearch files need not
1874 contain colons. This was done some time before quoting was made available
1875 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1876 (notated using the quote facility) so as to distinguish them from IPv4
1877 keys. This meant that lookups for IP addresses in host lists did not work
1878 for iplsearch lookups.
1880 This has been fixed by arranging for IPv6 addresses to be expressed with
1881 colons if the lookup type is iplsearch. This is not incompatible, because
1882 previously such lookups could never work.
1884 The situation is now rather anomalous, since one *can* have colons in
1885 ordinary lsearch keys. However, making the change in all cases is
1886 incompatible and would probably break a number of configurations.
1888 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1891 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1892 conversion specification without a maximum field width, thereby enabling
1893 a rogue spamd server to cause a buffer overflow. While nobody in their
1894 right mind would setup Exim to query an untrusted spamd server, an
1895 attacker that gains access to a server running spamd could potentially
1896 exploit this vulnerability to run arbitrary code as the Exim user.
1898 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1899 $primary_hostname instead of what libspf2 thinks the hosts name is.
1901 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1902 a directory entry by the name of the lookup key. Previously, if a
1903 symlink pointed to a non-existing file or a file in a directory that
1904 Exim lacked permissions to read, a lookup for a key matching that
1905 symlink would fail. Now it is enough that a matching directory entry
1906 exists, symlink or not. (Bugzilla 503.)
1908 PH/03 The body_linecount and body_zerocount variables are now exported in the
1911 PH/04 Added the $dnslist_matched variable.
1913 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1914 This means they are set thereafter only if the connection becomes
1917 PH/06 Added the client_condition to authenticators so that some can be skipped
1918 by clients under certain conditions.
1920 PH/07 The error message for a badly-placed control=no_multiline_responses left
1921 "_responses" off the end of the name.
1923 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1925 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1926 (without spaces) instead of just copying the configuration text.
1928 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1930 PH/11 Added $max_received_linelength.
1932 PH/12 Added +ignore_defer and +include_defer to host lists.
1934 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1935 way in which PCRE > 7.0 is built.
1937 PH/14 Implemented queue_only_load_latch.
1939 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1940 MAIL command. The effect was to mangle the value on 64-bit systems.
1942 PH/16 Another patch from the Sieve maintainer.
1944 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1946 PH/18 If a system quota error occurred while trying to create the file for
1947 a maildir delivery, the message "Mailbox is full" was not appended to the
1948 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1949 only to a quota excession during the actual writing of the file.
1951 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1952 characters?) which causes problems in log lines. The DN values are now
1953 passed through string_printing() before being added to log lines.
1955 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1956 and InterBase are left for another time.)
1958 PH/21 Added message_body_newlines option.
1960 PH/22 Guard against possible overflow in moan_check_errorcopy().
1962 PH/23 POSIX allows open() to be a macro; guard against that.
1964 PH/24 If the recipient of an error message contained an @ in the local part
1965 (suitably quoted, of course), incorrect values were put in $domain and
1966 $local_part during the evaluation of errors_copy.
1972 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1973 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1976 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1977 issue a MAIL command.
1979 PH/02 In an ACL statement such as
1981 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1983 if a client was not listed at all, or was listed with a value other than
1984 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1985 the condition was not true (as it should be), so access was not denied.
1986 The bug was that the ! inversion was incorrectly passed on to the second
1987 item. This has been fixed.
1989 PH/03 Added additional dnslists conditions == and =& which are different from
1990 = and & when the dns lookup returns more than one IP address.
1992 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1993 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1995 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1996 FSYNC, which compiles an option called disable_fsync that allows for
1997 bypassing fsync(). The documentation is heavily laced with warnings.
1999 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2001 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2002 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2003 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2004 including adding "make clean"; (3) Added -fPIC when compiling the test
2005 dynamically loaded module, to get rid of a warning.
2007 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2008 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2009 = 0s. The bug is otherwise harmless.
2011 PH/07 There was a bug in the dovecot authenticator such that the value of
2012 $auth1 could be overwritten, and so not correctly preserved, after a
2013 successful authentication. This usually meant that the value preserved by
2014 the server_setid option was incorrect.
2016 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2018 PH/09 Installed PCRE release 7.0.
2020 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2021 run for batched SMTP input. It is now run at the start of every message
2022 in the batch. While fixing this I discovered that the process information
2023 (output by running exiwhat) was not always getting set for -bs and -bS
2024 input. This is fixed, and it now also says "batched" for BSMTP.
2026 PH/11 Added control=no_pipelining.
2028 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2029 patch, slightly modified), and move the expansion of helo_data till after
2030 the connection is made in the smtp transport (so it can use these
2033 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2035 PH/14 Added log_selector = +pid.
2037 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2039 PH/16 Add ${if forany and ${if forall.
2041 PH/17 Added dsn_from option to vary the From: line in DSNs.
2043 PH/18 Flush SMTP output before performing a callout, unless control =
2044 no_callout_flush is set.
2046 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2047 was true (the default) a successful delivery failed to delete the retry
2048 item, thus causing premature timeout of the address. The bug is now
2051 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2053 PH/21 Long custom messages for fakedefer and fakereject are now split up
2054 into multiline responses in the same way that messages for "deny" and
2055 other ACL rejections are.
2057 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2058 with slight modification.
2060 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2061 draft, changing the syntax and factoring some duplicate code".
2063 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2064 for deliveries of the second and subsequent messages over the same SMTP
2067 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2068 ${reduce, with only minor "tidies".
2070 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2072 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2073 expansion side effects.
2075 PH/27 When a message times out after an over-quota error from an Exim-imposed
2076 quota, the bounce message says "mailbox is full". This message was not
2077 being given when it was a system quota that was exceeded. It now should
2080 MH/03 Made $recipients available in local_scan(). local_scan() already has
2081 better access to the recipient list through recipients_list[], but
2082 $recipients can be useful in postmaster-provided expansion strings.
2084 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2085 in the case of a MAIL command with additional options following the
2086 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2087 were accidentally chopped off.
2089 PH/29 SMTP synchronization checks are implemented when a command is read -
2090 there is a check that no more input is waiting when there shouldn't be
2091 any. However, for some commands, a delay in an ACL can mean that it is
2092 some time before the response is written. In this time, more input might
2093 arrive, invalidly. So now there are extra checks after an ACL has run for
2094 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2095 pipelining has not been advertised.
2097 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2099 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2100 returned in the SMTP response when smtp_return_error_details was false.
2101 This has been fixed.
2103 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2104 instead of the C I/O that was originally supplied, because problems were
2105 reported on Solaris.
2107 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2108 Exim which did not show up earlier: it was assuming that a call to
2109 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2110 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2111 was a macro that became an assignment, so it seemed to work. This has
2112 changed to a proper function call with a void return, hence the compile
2113 error. Exim's code has been fixed.
2115 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2118 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2120 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2122 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2123 because it turns out to be too big or there is a timeout) there is no
2124 "Completed" line in the log. When some messages of this type were
2125 selected by exigrep, they were listed as "not completed". Others were
2126 picked up by some special patterns. I have improved the selection
2127 criteria to be more general.
2129 PH/38 The host_find_failed option in the manualroute router can now be set
2130 to "ignore", to completely ignore a host whose IP address cannot be
2131 found. If all hosts are ignored, the behaviour is controlled by the new
2132 host_all_ignored option.
2134 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2135 homing or because of multiple MX records with /mx) generated more than
2136 one IP address, and the following item turned out to be the local host,
2137 all the secondary addresses of the first item were incorrectly removed
2138 from the list, along with the local host and any following hosts (which
2139 is what is supposed to happen).
2141 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2142 whoever called Exim into the -H file. In the case of the daemon it was
2143 behaving confusingly. When first started, it used values for whoever
2144 started the daemon, but after a SIGHUP it used the Exim user (because it
2145 calls itself on a restart). I have changed the code so that it now always
2148 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2149 message are rejected with the same error (e.g. no authentication or bad
2150 sender address), and a DATA command is nevertheless sent (as can happen
2151 with PIPELINING or a stupid MUA), the error message that was given to the
2152 RCPT commands is included in the rejection of the DATA command. This is
2153 intended to be helpful for MUAs that show only the final error to their
2156 PH/42 Another patch from the Sieve maintainer.
2158 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2159 Eximstats - Fixed some broken HTML links and added missing column headers
2161 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2162 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2164 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2166 PH/43 Yet another patch from the Sieve maintainer.
2168 PH/44 I found a way to check for a TCP/IP connection going away before sending
2169 the response to the final '.' that terminates a message, but only in the
2170 case where the client has not sent further data following the '.'
2171 (unfortunately, this is allowed). However, in many cases there won't be
2172 any further data because there won't be any more messages to send. A call
2173 to select() can be used: if it shows that the input is "ready", there is
2174 either input waiting, or the socket has been closed. An attempt to read
2175 the next input character can distinguish the two cases. Previously, Exim
2176 would have sent an OK response which the client would never have see.
2177 This could lead to message repetition. This fix should cure that, at
2178 least in a lot of common cases.
2180 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2181 advertised in response to EHLO.
2187 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2188 fixed by 4.65/MH/01 (is this a record?) are fixed:
2190 (i) An empty string was always treated as zero by the numeric comparison
2191 operators. This behaviour has been restored.
2193 (ii) It is documented that the numeric comparison operators always treat
2194 their arguments as decimal numbers. This was broken in that numbers
2195 starting with 0 were being interpreted as octal.
2197 While fixing these problems I realized that there was another issue that
2198 hadn't been noticed. Values of message_size_limit (both the global option
2199 and the transport option) were treated as octal if they started with 0.
2200 The documentation was vague. These values are now always treated as
2201 decimal, and I will make that clear in the documentation.
2207 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2208 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2211 MH/01 Don't check that the operands of numeric comparison operators are
2212 integers when their expansion is in "skipping" mode (fixes bug
2213 introduced by 4.64-PH/07).
2215 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2216 child addresses, Exim now panics and dies. Previously, because the count
2217 is held in a short int, deliveries were likely to be lost. As such a
2218 large number of recipients for a single message is ridiculous
2219 (performance will be very, very poor), I have chosen to impose a limit
2220 rather than extend the field.
2226 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2227 leftover -K file (the existence of which was triggered by #402).
2228 While we were at it, introduced process PID as part of the -K
2229 filename. This should rule out race conditions when creating
2232 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2233 processing considerably. Previous code took too long for large mails,
2234 triggering a timeout which in turn triggers #401.
2236 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2237 in the DK code in transports.c. sendfile() is not really portable,
2238 hence the _LINUX specificness.
2240 TF/01 In the add_headers option to the mail command in an Exim filter,
2241 there was a bug that Exim would claim a syntax error in any
2242 header after the first one which had an odd number of characters
2245 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2246 callout verification, Exim cached a "reject" for the entire domain. This
2247 is correct for most verifications, but it is not correct for a recipient
2248 verification with use_sender or use_postmaster set, because in that case
2249 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2250 case of MAIL FROM:<> rejection from other early rejections (e.g.
2251 rejection of HELO). When verifying a recipient using a non-null MAIL
2252 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2253 Whatever the result of the callout, the value of the domain cache is
2254 left unchanged (for any other kind of callout, getting as far as trying
2255 RCPT means that the domain itself is ok).
2257 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2260 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2261 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2262 session, the daemon ignores ECONNECT errors and logs others; it now
2263 ignores EPIPE as well.
2265 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2266 (quoted-printable decoding).
2268 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2269 later the small subsequent patch to fix an introduced bug.
2271 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2273 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2275 PH/08 An error is now given if message_size_limit is specified negative.
2277 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2278 to be given (somewhat) arbitrary names.
2280 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2283 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2284 miscellaneous code fixes
2286 PH/10 Added the log_reject_target ACL modifier to specify where to log
2289 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2290 hostname. This is wrong, because it relates to the incoming message (and
2291 probably the interface on which it is arriving) and not to the outgoing
2292 callout (which could be using a different interface). This has been
2293 changed to use the value of the helo_data option from the smtp transport
2294 instead - this is what is used when a message is actually being sent. If
2295 there is no remote transport (possible with a router that sets up host
2296 addresses), $smtp_active_hostname is used.
2298 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2299 tweaks were necessary in order to get it to work (see also 21 below):
2300 (a) The code assumed that strncpy() returns a negative number on buffer
2301 overflow, which isn't the case. Replaced with Exim's string_format()
2303 (b) There were several signed/unsigned issues. I just did the minimum
2304 hacking in of casts. There is scope for a larger refactoring.
2305 (c) The code used strcasecmp() which is not a standard C function.
2306 Replaced with Exim's strcmpic() function.
2307 (d) The code set only $1; it now sets $auth1 as well.
2308 (e) A simple test gave the error "authentication client didn't specify
2309 service in request". It would seem that Dovecot has changed its
2310 interface. Fortunately there's a specification; I followed it and
2311 changed what the client sends and it appears to be working now.
2313 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2316 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2317 address A is aliased to B and C, where B exists and C does not. Without
2318 -v the output is "A verified" because verification stops after a
2319 successful redirection if more than one address is generated. However,
2320 with -v the child addresses are also verified. Exim was outputting "A
2321 failed to verify" and then showing the successful verification for C,
2322 with its parentage. It now outputs "B failed to verify", showing B's
2323 parentage before showing the successful verification of C.
2325 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2326 look up a TXT record in a specific list after matching in a combined
2329 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2330 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2331 they consult the DNS. I had assumed they would set it the way they
2332 wanted; and indeed my experiments on Linux seem to show that in some
2333 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2334 To be on the safe side, however, I have now made the interface to
2335 host_find_byname() similar to host_find_bydns(), with an argument
2336 containing the DNS resolver options. The host_find_byname() function now
2337 sets these options at its start, just as host_find_bydns() does. The smtp
2338 transport options dns_qualify_single and dns_search_parents are passed to
2339 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2340 of host_find_byname() use the default settings of RES_DEFNAMES
2341 (qualify_single) but not RES_DNSRCH (search_parents).
2343 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2344 spool_read_header() do less string testing, by means of a preliminary
2345 switch on the second character of optional "-foo" lines. (This is
2346 overdue, caused by the large number of possibilities that now exist.
2347 Originally there were few.) While I was there, I also converted the
2348 str(n)cmp tests so they don't re-test the leading "-" and the first
2349 character, in the hope this might squeeze out yet more improvement.
2351 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2352 flag allowing group syntax was set by the header_syntax check but not
2353 turned off, possible causing trouble later; (2) The flag was not being
2354 set at all for the header_verify test, causing "group"-style headers to
2355 be rejected. I have now set it in this case, and also caused header_
2356 verify to ignore an empty address taken from a group. While doing this, I
2357 came across some other cases where the code for allowing group syntax
2358 while scanning a header line wasn't quite right (mostly, not resetting
2359 the flag correctly in the right place). These bugs could have caused
2360 trouble for malformed header lines. I hope it is now all correct.
2362 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2363 with the "reply" argument non-NULL. The code, however (which originally
2364 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2365 but it didn't always do it. This confused somebody who was copying the
2366 code for some other use. I have removed all the tests.
2368 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2369 feature that was used to support insecure browsers during the U.S. crypto
2370 embargo. It requires special client support, and Exim is probably the
2371 only MTA that supported it -- and would never use it because real RSA is
2372 always available. This code has been removed, because it had the bad
2373 effect of slowing Exim down by computing (never used) parameters for the
2374 RSA_EXPORT functionality.
2376 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2377 authenticator to fail if there's a tab character in the incoming data
2378 (there should never be unless someone is messing about, as it's supposed
2379 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2380 if the connection is using TLS or if the remote IP is the same as the
2381 local IP, and the "valid-client-cert option" if a client certificate has
2384 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2385 authenticators. This can be used for authorization after authentication
2386 succeeds. (In the case of plaintext, it servers for both authentication
2389 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2390 if any retry times were supplied.
2392 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2393 connection, where there is no client IP address to check. In this
2394 situation, the verify now always succeeds.
2396 PH/25 Applied John Jetmore's -Mset patch.
2398 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2400 PH/27 In a string expansion for a processed (not raw) header when multiple
2401 headers of the same name were present, leading whitespace was being
2402 removed from all of them, but trailing whitespace was being removed only
2403 from the last one. Now trailing whitespace is removed from each header
2404 before concatenation. Completely empty headers in a concatenation (as
2405 before) are ignored.
2407 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2408 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2410 PH/29 [Removed. This was a change that I later backed out, and forgot to
2411 correct the ChangeLog entry (that I had efficiently created) before
2412 committing the later change.]
2414 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2415 address errors (4xx response to RCPT) over the same connection as other
2416 messages routed to the same hosts. Such deliveries are always "forced",
2417 so retry times are not inspected. This resulted in far too many retries
2418 for the affected addresses. The effect occurred only when there were more
2419 hosts than the hosts_max_try setting in the smtp transport when it had
2420 the 4xx errors. Those hosts that it had tried were not added to the list
2421 of hosts for which the message was waiting, so if all were tried, there
2422 was no problem. Two fixes have been applied:
2424 (i) If there are any address or message errors in an SMTP delivery, none
2425 of the hosts (tried or untried) are now added to the list of hosts
2426 for which the message is waiting, so the message should not be a
2427 candidate for sending over the same connection that was used for a
2428 successful delivery of some other message. This seems entirely
2429 reasonable: after all the message is NOT "waiting for some host".
2430 This is so "obvious" that I'm not sure why it wasn't done
2431 previously. Hope I haven't missed anything, but it can't do any
2432 harm, as the worst effect is to miss an optimization.
2434 (ii) If, despite (i), such a delivery is accidentally attempted, the
2435 routing retry time is respected, so at least it doesn't keep
2436 hammering the server.
2438 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2439 in ${readsocket because some servers need this prod.
2441 PH/32 Added some extra debug output when updating a wait-xxx database.
2443 PH/33 The hint "could be header name not terminated by colon", which has been
2444 given for certain expansion errors for a long time, was not being given
2445 for the ${if def:h_colon_omitted{... case.
2447 PH/34 The spec says: "With one important exception, whenever a domain list is
2448 being scanned, $domain contains the subject domain." There was at least
2449 one case where this was not true.
2451 PH/35 The error "getsockname() failed: connection reset by peer" was being
2452 written to the panic log as well as the main log, but it isn't really
2453 panic-worthy as it just means the connection died rather early on. I have
2454 removed the panic log writing for the ECONNRESET error when getsockname()
2457 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2458 runs only) independently of the message's sender address. This meant
2459 that, if the 4xx error was in fact related to the sender, a different
2460 message to the same recipient with a different sender could confuse
2461 things. In particular, this can happen when sending to a greylisting
2462 server, but other circumstances could also provoke similar problems.
2463 I have changed the default so that the retry time for these errors is now
2464 based a combination of the sender and recipient addresses. This change
2465 can be overridden by setting address_retry_include_sender=false in the
2468 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2469 remote server are returned as part of bounce messages. This was not
2470 happening for LMTP over a pipe (the lmtp transport), but now it is the
2471 same for both kinds of LMTP.
2473 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2474 in header lines that were in fact CNAMEs. This is no longer the case.
2476 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2477 and queue runs started by the daemon processed all messages. This has
2478 been fixed so that -R and -S can now usefully be given with -q<time>.
2480 PH/40 Import PCRE release 6.7 (fixes some bugs).
2482 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2484 PH/42 Give an error if -q is specified more than once.
2486 PH/43 Renamed the variables $interface_address and $interface_port as
2487 $received_ip_address and $received_port, to make it clear that these
2488 values apply to message reception, and not to the outgoing interface when
2489 a message is delivered. (The old names remain recognized, of course.)
2491 PH/44 There was no timeout on the connect() call when using a Unix domain
2492 socket in the ${readsocket expansion. There now is.
2494 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2495 be meaningful with "accept".
2497 SC/01 Eximstats V1.43
2498 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2500 SC/02 Eximstats V1.44
2501 Use a glob alias rather than an array ref in the generated
2502 parser. This improves both readability and performance.
2504 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2505 Collect SpamAssassin and rejection statistics.
2506 Don't display local sender or destination tables unless
2507 there is data to show.
2508 Added average volumes into the top table text output.
2510 SC/04 Eximstats V1.46
2511 Collect data on the number of addresses (recipients)
2512 as well as the number of messages.
2514 SC/05 Eximstats V1.47
2515 Added 'Message too big' to the list of mail rejection
2516 reasons (thanks to Marco Gaiarin).
2518 SC/06 Eximstats V1.48
2519 Mainlog lines which have GMT offsets and are too short to
2520 have a flag are now skipped.
2522 SC/07 Eximstats V1.49 (Alain Williams)
2523 Added the -emptyok flag.
2525 SC/08 Eximstats V1.50
2526 Fixes for obtaining the IP address from reject messages.
2528 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2529 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2530 whitespace changes from 4.64-PH/27
2532 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2535 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2536 are found, allow negative numbers in numeric criteria)
2538 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2540 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2541 to match changes made in 4.64-PH/43
2543 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2545 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2546 "domain" is actually an IP address, and force a failure. This locks out
2547 those revolvers/nameservers that support "A-for-A" lookups, in
2548 contravention of the specifications.
2550 PH/48 When a host name was looked up from an IP address, and the subsequent
2551 forward lookup of the name timed out, the host name was left in
2552 $sender_host_name, contrary to the specification.
2554 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2555 restricted to single-key lookups, Exim was not diagnosing an error if
2556 * or *@ was used with a query-style lookup.
2558 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2560 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2561 long ago, but noone interested enough thought of it. Let's just say that
2562 the "1.1" means that there are some new functions that weren't there at
2563 some point in the past.
2565 PH/51 Error processing for expansion failure of helo_data from an smtp
2566 transport during callout processing was broken.
2568 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2569 tested/used via the -bh/-bhc/-bs options.
2571 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2572 bug, fixed in subsequent PCRE releases).
2574 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2575 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2577 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2583 SC/01 Use a glob alias rather than an array ref in eximstats generated
2584 parser. This improves both readability and performance.
2586 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2587 Don't display local sender or destination tables in eximstats unless
2588 there is data to show.
2589 Added average volumes into the eximstats top table text output.
2591 SC/03 Collect data on the number of addresses (recipients) as well
2592 as the number of messages in eximstats.
2594 TF/01 Correct an error in the documentation for the redirect router. Exim
2595 does (usually) call initgroups() when daemonizing.
2597 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2598 with consistent privilege compared to when running as a daemon.
2600 TF/03 Note in the spec that $authenticated_id is not set for local
2601 submissions from trusted users.
2603 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2604 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2606 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2607 by adding some example configuration directives to the default
2608 configuration file. A little bit of work is required to uncomment the
2609 directives and define how usernames and passwords are checked, but
2610 there is now a framework to start from.
2612 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2613 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2614 without this. I don't know how relevant this is to other LDAP libraries.
2616 PH/02 Add the verb name to the "unknown ACL verb" error.
2618 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2620 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2622 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2623 directory not expanded when it should be if an expanded home directory
2624 was set for the address (which is overridden by the transport).
2626 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2629 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2630 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2631 because it is too late at that time, and has no effect.
2633 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2634 security issue with \' (bugzilla #107). I could not use the
2635 PQescapeStringConn() function, because it needs a PGconn value as one of
2638 PH/08 When testing addresses using -bt, indicate those final addresses that
2639 are duplicates that would not cause an additional delivery. At least one
2640 person was confused, thinking that -bt output corresponded to deliveries.
2641 (Suppressing duplicates isn't a good idea as you lose the information
2642 about possibly different redirections that led to the duplicates.)
2644 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2645 systems where poll() doesn't work, in particular OS X.
2647 PH/10 Added more information to debugging output for retry time not reached.
2649 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2650 operations in malware.c.
2652 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2655 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2656 syslog_duplication set false, log lines that would normally be written
2657 both the the main log and to the reject log were not written to syslog at
2660 PH/14 In the default configuration, change the use of "message" in ACL warn
2661 statements to "add_header".
2663 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2664 not followed by a command (e.g. "seen endif").
2666 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2667 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2670 PH/17 Added extra conditions to the default value of delay_warning_condition
2674 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2675 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2676 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2679 The Auto-Submitted: and various List- headers are standardised, whereas I
2680 don't think Precedence: ever was.
2682 PH/18 Refactored debugging code in route_finduser() to show more information,
2683 in particular, the error code if getpwnam() issues one.
2685 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2686 This is apparently needed in addition to the PH/07 change above to avoid
2687 any possible encoding problems.
2689 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2690 but not after initializing Perl.
2692 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2693 output them only if debugging. By default they are written stderr,
2694 apparently, which is not desirable.
2696 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2699 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2702 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2704 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2705 authenticated or an ident call has been made. Suppress the default
2706 values for $authenticated_id and $authenticated_sender (but permit -oMai
2707 and -oMas) when testing with -bh.
2709 PH/24 Re-jigged the order of the tests in the default configuration so that the
2710 tests for valid domains and recipients precede the DNS black list and CSA
2711 tests, on the grounds that those ones are more expensive.
2713 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2714 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2715 This bug exists in every version of Exim that I still have, right back to
2718 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2719 However, an attempt to turn on case-sensitivity in a regex key by
2720 including (?-i) didn't work because the subject string was already
2721 lowercased, and the effects were non-intuitive. It turns out that a
2722 one-line patch can be used to allow (?-i) to work as expected.
2728 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2729 other effects) broke the use of negated acl sub-conditions.
2731 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2734 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2735 "Deny" causes Exim to reject the incoming connection with a 554 error.
2736 Unfortunately, if there is a major crisis, such as a disk failure,
2737 tcp-wrappers gives "deny", whereas what one would like would be some
2738 kind of temporary error. A kludge has been added to help with this.
2739 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2740 554 error is used if errno is still zero or contains ENOENT (which occurs
2741 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2744 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2746 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2747 errors. Otherwise a message that provokes a temporary error (when other
2748 messages do not) can cause a whole host to time out.
2750 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2751 addresses were routed directly to files or pipes from a redirect router.
2752 File deliveries just didn't batch; pipe deliveries might have suffered
2755 PH/06 A failure to get a lock for a hints database would erroneously always say
2756 "Failed to get write lock", even when it was really a read lock.
2758 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2759 of 0600. This has been changed to use the value of the lockfile_mode
2760 option (which defaults to 0600).
2762 PH/08 Applied small patch from the Sieve maintainer.
2764 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2765 folder from quota calculations, a direct delivery into this folder messed
2766 up the contents of the maildirsize file. This was because the regex was
2767 used only to exclude .Trash (or whatever) when the size of the mailbox
2768 was calculated. There was no check that a delivery was happening into an
2769 excluded directory. This bug has been fixed by ignoring all quota
2770 processing for deliveries into excluded directories.
2772 PH/10 Added the maildirfolder_create_regex option to appendfile.
2778 PH/01 The code for finding all the local interface addresses on a FreeBSD
2779 system running IPv6 was broken. This may well have applied to all BSD
2780 systems, as well as to others that have similar system calls. The broken
2781 code found IPv4 interfaces correctly, but gave incorrect values for the
2782 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2783 that it would not match correctly against @[] and not recognize the IPv6
2786 PH/02 The ipliteral router was not recognizing addresses of the form user@
2787 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2789 PH/03 Added disable_ipv6.
2791 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2792 decoded form, because it is most often used to construct To: headers
2793 lines in autoreplies, and the decoded form may well be syntactically
2794 invalid. However, $reply_address has leading white space removed, and all
2795 newlines turned into spaces so that the autoreply transport does not
2798 PH/05 If group was specified without a user on a router, and no group or user
2799 was specified on a transport, the group from the router was ignored.
2801 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2802 for visible compile-time settings that can be used to change these
2803 numbers, for those that want even more. Backwards compatibility with old
2804 spool files has been maintained. However, going back to a previous Exim
2805 release will lost any variables that are in spool files.
2807 PH/07 Two small changes when running in the test harness: increase delay when
2808 passing a TCP/IP connection to a new process, in case the original
2809 process has to generate a bounce, and remove special handling of
2810 127.0.0.2 (sic), which is no longer necessary.
2812 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2813 be the same on different OS.
2815 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2818 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2819 whether --show-vars was specified or not
2821 JJ/02 exipick: Added support for new ACL variable spool format introduced
2824 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2825 syntactically invalid From: or Reply-to: line, and a filter used this to
2826 generate an autoreply, and therefore failed to obtain an address for the
2827 autoreply, Exim could try to deliver to a non-existent relative file
2828 name, causing unrelated and misleading errors. What now happens is that
2829 it logs this as a hard delivery error, but does not attempt to create a
2832 PH/11 The exinext utility has a -C option for testing purposes, but although
2833 the given file was scanned by exinext itself; it wasn't being passed on
2834 when Exim was called.
2836 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2837 an end-of-file indication when reading a command response.
2839 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2840 compiled. In many other places in Exim, IPv6 addresses are always
2841 recognized, so I have changed this. It also means that IPv4 domain
2842 literals of the form [IPV4:n.n.n.n] are now always recognized.
2844 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2845 used if the router is not running as root, for example, when verifying at
2846 ACL time, or when using -bh. The debugging output from this situation was
2847 non-existent - all you got was a failure to exec. I have made two
2850 (a) Failures to set uid/gid, the current directory, or a process leader
2851 in a subprocess such as that created by queryprogram now generate
2852 suitable debugging output when -d is set.
2854 (b) The queryprogram router detects when it is not running as root,
2855 outputs suitable debugging information if -d is set, and then runs
2856 the subprocess without attempting to change uid/gid.
2858 PH/15 Minor change to Makefile for building test_host (undocumented testing
2861 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2862 additional section of a DNS packet that returns MX or SRV records.
2863 Instead, it always explicitly searches for A/AAAA records. This avoids
2864 major problems that occur when a DNS server includes only records of one
2865 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2866 fixed another bug: if SRV records were looked up and the corresponding
2867 address records were *not* found in the additional section, the port
2868 values from the SRV records were lost.
2870 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2871 using the correct key (the original address) when searching the retry
2872 rules in order to find which one to use for generating the retry hint.
2874 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2875 adding the default one. Similarly, if it contains a Reply-To: header, the
2876 errors_reply_to option, if set, is not used.
2878 PH/19 When calculating a retry time, Exim used to measure the "time since
2879 failure" by looking at the "first failed" field in the retry record. Now
2880 it does not use this if it is later than than the arrival time of the
2881 message. Instead it uses the arrival time. This makes for better
2882 behaviour in cases where some deliveries succeed, thus re-setting the
2883 "first failed" field. An example is a quota failure for a huge message
2884 when small messages continue to be delivered. Without this change, the
2885 "time since failure" will always be short, possible causing more frequent
2886 delivery attempts for the huge message than are intended.
2887 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2889 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2890 $1, $2, $3) because the numerical variables can be reset during some
2891 expansion items (e.g. "match"), thereby losing the authentication data.
2893 PH/21 Make -bV show the size of off_t variables so that the test suite can
2894 decide whether to run tests for quotas > 2G.
2896 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2897 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2898 filecount value is greater than 2G or if a quota value is greater than 2G
2899 on a system where the size of off_t is not greater than 4, a panic error
2902 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2903 never match. The debug and -bh output now contains an explicit error
2904 message indicating a malformed IPv4 address or mask.
2906 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2907 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2908 PH/23 above applies.
2910 PH/25 Do not write to syslog when running in the test harness. The only
2911 occasion when this arises is a failure to open the main or panic logs
2912 (for which there is an explicit test).
2914 PH/26 Added the /no_tell option to "control=freeze".
2916 PH/27 If a host name lookup failed very early in a connection, for example, if
2917 the IP address matched host_lookup and the reverse lookup yielded a name
2918 that did not have a forward lookup, an error message of the form "no IP
2919 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2920 could be logged. Now it outputs the IP address instead of "NULL".
2922 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2923 allows the sender and the authenticated sender to be set when
2924 submitting a message from within Exim. Since child_open_exim() is
2925 documented for local_scan(), the new function should be too.
2927 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2928 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2929 results in an empty string is now treated as unset.
2931 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2933 PH/31 Added sender_verify_fail logging option.
2935 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2936 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2937 tidied the source and removed it altogether.
2939 PH/33 When a queue run was abandoned because the load average was too high, a
2940 log line was always written; now it is written only if the queue_run log
2941 selector is set. In addition, the log line for abandonment now contains
2942 information about the queue run such as the pid. This is always present
2943 in "start" and "stop" lines but was omitted from the "abandon" line.
2945 PH/34 Omit spaces between a header name and the colon in the error message that
2946 is given when verify = headers_syntax fails (if there are lots of them,
2947 the message gets confusing).
2949 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2950 names, as there are now some PTR records that contain slashes. This check
2951 is only to protect against broken name servers that fall over on strange
2952 characters, so the fact that it applies to all lookups doesn't matter.
2954 PH/36 Now that the new test suite is complete, we can remove some of the
2955 special code in Exim that was needed for the old test suite. For example,
2956 sorting DNS records because real resolvers return them in an arbitrary
2957 order. The new test suite's fake resolver always returns records in the
2960 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2961 bounces) except when queue_only is set, to avoid logging races between
2962 the different processes.
2964 PH/38 Panic-die if .include specifies a non-absolute path.
2966 PH/39 A tweak to the "H" retry rule from its user.
2968 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2969 a label. They prevented compilation on older perls.
2971 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2972 a warning to be raised on newish perls.
2974 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2975 on queue. Changes to match documented behaviour of showing count of
2976 messages matching specified criteria.
2978 PH/40 Changed the default ident timeout from 30s to 5s.
2980 PH/41 Added support for the use of login_cap features, on those BSD systems
2981 that have them, for controlling the resources used by pipe deliveries.
2983 PH/42 The content-scanning code uses fopen() to create files in which to put
2984 message data. Previously it was not paying any attention to the mode of
2985 the files. Exim runs with umask(0) because the rest of the code creates
2986 files with open(), and sets the required mode explicitly. Thus, these
2987 files were ending up world-writeable. This was not a big issue, because,
2988 being within the spool directory, they were not world-accessible. I have
2989 created a function called modefopen, which takes an additional mode
2990 argument. It sets umask(777), creates the file, chmods it to the required
2991 mode, then resets the umask. All the relevant calls to fopen() in the
2992 content scanning code have been changed to use this function.
2994 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2995 to 24 hours. This avoids potential overflow problems when processing G
2996 and H retry rules. I suspect nobody ever tinkers with this value.
2998 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3000 PH/45 When the plaintext authenticator is running as a client, the server's
3001 challenges are checked to ensure they are valid base64 strings. By
3002 default, the authentication attempt is cancelled if an invalid string is
3003 received. Setting client_ignore_invalid_base64 true ignores these errors.
3004 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3005 they are received. Thus, the responses can be made to depend on the
3006 challenges. If an invalid string is ignored, an empty string is placed in
3009 PH/46 Messages that are created by the autoreply transport now contains a
3010 References: header, in accordance with RFCs 2822 and 3834.
3012 PH/47 Added authenticated_sender_force to the smtp transport.
3014 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3016 PH/49 Installed latest patch from the Sieve maintainer.
3018 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3019 was also set, the appendfile transport was unnecessarily scanning a
3020 directory of message files (e.g. for maildir delivery) to find the count
3021 of files (along with the size), even though it did not need this
3022 information. It now does the scan only if it needs to find either the
3023 size of the count of files.
3025 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3027 PH/52 Two bugs concerned with error handling when the smtp transport is
3030 (i) Exim was not creating retry information for temporary errors given
3031 for individual recipients after the DATA command when the smtp transport
3032 was used in LMTP mode. This meant that they could be retried too
3033 frequently, and not timed out correctly.
3035 (ii) Exim was setting the flag that allows error details to be returned
3036 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3037 recipients that were returned after the DATA command.
3039 PH/53 This is related to PH/52, but is more general: for any failing address,
3040 when detailed error information was permitted to be returned to the
3041 sender, but the error was temporary, then after the final timeout, only
3042 "retry timeout exceeded" was returned. Now it returns the full error as
3043 well as "retry timeout exceeded".
3045 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3046 do this, and (what is worse) MTAs that accept it.
3048 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3049 will now be deprecated.
3051 PH/56 New os.c-cygwin from the Cygwin maintainer.
3053 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3054 formats (previously only available in exim formats via -bpr, -bpru,
3055 and -bpra. Now also available in native and exiqgrep formats)
3057 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3058 with very large, slow to parse queues
3060 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3062 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3064 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3065 responses to authentication challenges, though it was showing the
3066 challenges; (ii) I've removed the CR characters from the debug output for
3069 PH/58 Allow for the insertion of a newline as well as a space when a string
3070 is turned into more than one encoded-word during RFC 2047 encoding. The
3071 Sieve code now uses this.
3073 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3074 data_4xx, lost_connection, tls_required.
3076 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3077 message was being sent as an SMTP response.
3079 PH/61 Add -l and -k options to exicyclog.
3081 PH/62 When verifying, if an address was redirected to one new address, so that
3082 verification continued, and the new address failed or deferred after
3083 having set something in $address_data, the value of $address_data was not
3084 passed back to the ACL. This was different to the case when no
3085 redirection occurred. The value is now passed back in both cases.
3087 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3088 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3089 use login_cap.h, so on its own it isn't the distinguishing feature. The
3090 new name refers directly to the setclassresources() function.
3092 PH/65 Added configuration files for NetBSD3.
3094 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3096 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3097 is preferred over IPv4.
3099 PH/68 The bounce_return_message and bounce_return_body options were not being
3100 honoured for bounces generated during the reception of non-SMTP messages.
3101 In particular, this applied to messages rejected by the ACL. This bug has
3102 been fixed. However, if bounce_return_message is true and bounce_return_
3103 body is false, the headers that are returned for a non-SMTP message
3104 include only those that have been read before the error was detected.
3105 (In the case of an ACL rejection, they have all been read.)
3107 PH/69 The HTML version of the specification is now built in a directory called
3108 spec_html instead of spec.html, because the latter looks like a path with
3109 a MIME-type, and this confuses some software.
3111 PH/70 Catch two compiler warnings in sieve.c.
3113 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3114 function verify_get_ident() calls ip_connect() to connect a socket, but
3115 if the "connect()" function timed out, ip_connect() used to close the
3116 socket. However, verify_get_ident() also closes the socket later, and in
3117 between Exim writes to the log, which may get opened at this point. When
3118 the socket was closed in ip_connect(), the log could get the same file
3119 descriptor number as the socket. This naturally causes chaos. The fix is
3120 not to close the socket in ip_connect(); the socket should be closed by
3121 the function that creates it. There was only one place in the code where
3122 this was missing, in the iplookup router, which I don't think anybody now
3123 uses, but I've fixed it anyway.
3125 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3126 well as to direct DNS lookups. Otherwise the handling of names in host
3127 lists is inconsistent and therefore confusing.
3133 PH/01 Two changes to the default runtime configuration:
3135 (1) Move the checks for relay_from_hosts and authenticated clients from
3136 after to before the (commented out) DNS black list checks.
3138 (2) Add control=submission to the relay_from_hosts and authenticated
3139 clients checks, on the grounds that messages accepted by these
3140 statements are most likely to be submissions.
3142 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3144 (1) Generate an error if the third argument for the ${prvs expansion is
3147 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3150 (3) Reset the variables that are obtained from the first argument of
3151 ${prvscheck and used in the second argument before leaving the code,
3152 because their memory is reclaimed, so using them afterwards may do
3155 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3156 one (it's much easier than Tom thought :-).
3158 (5) Because of (4), we can now allow for the use of $prvscheck_result
3159 inside the third argument.
3161 PH/03 For some reason, the default setting of PATH when running a command from
3162 a pipe transport was just "/usr/bin". I have changed it to
3165 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3166 anything to be listed in the output from -bV.
3168 PH/05 When a filter generated an autoreply, the entire To: header line was
3169 quoted in the delivery log line, like this:
3171 => >A.N.Other <ano@some.domain> <original@ddress> ...
3173 This has been changed so that it extracts the operative address. There
3174 may be more than one such address. If so, they are comma-separated, like
3177 => >ano@some.domain,ona@other.domain <original@ddress> ...
3179 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3180 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3181 not being looked up in the rDNS to get a host name, Exim was showing the
3182 IP address twice in Received: lines, even though the IP addresses were
3183 identical. For example:
3185 Received: from [1.2.3.4] (helo=[1.2.3.4])
3187 However, if the real host name was known, it was omitting the HELO data
3188 if it matched the actual IP address. This has been tidied up so that it
3189 doesn't show the same IP address twice.
3191 PH/07 When both +timestamp and +memory debugging was on, the value given by
3192 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3193 called by the debug printing, thereby overwriting the timestamp buffer.
3194 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3196 PH/08 When the original message was included in an autoreply transport, it
3197 always said "this is a copy of the message, including all the headers",
3198 even if body_only or headers_only was set. It now gives an appropriate
3201 PH/09 Applied a patch from the Sieve maintainer which:
3203 o fixes some comments
3204 o adds the (disabled) notify extension core
3205 o adds some debug output for the result of if/elsif tests
3206 o points to the current vacation draft in the documentation
3207 and documents the missing references header update
3211 o fixes a bug in processing the envelope test (when testing
3212 multiple envelope elements, the last element determined the
3215 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3216 Electronic Mail") by including:
3218 Auto-submitted: auto-generated
3220 in the messages that it generates (bounce messages and others, such as
3221 warnings). In the case of bounce messages for non-SMTP messages, there was
3222 also a typo: it was using "Auto_submitted" (underscore instead of
3223 hyphen). Since every message generated by Exim is necessarily in response
3224 to another message, thes have all been changed to:
3226 Auto-Submitted: auto-replied
3228 in accordance with these statements in the RFC:
3230 The auto-replied keyword:
3232 - SHOULD be used on messages sent in direct response to another
3233 message by an automatic process,
3235 - MUST NOT be used on manually-generated messages,
3237 - MAY be used on Delivery Status Notifications (DSNs) and Message
3238 Disposition Notifications (MDNs),
3240 - MUST NOT be used on messages generated by automatic or periodic
3241 processes, except for messages which are automatic responses to
3244 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3245 to the default Received: header definition.
3247 PH/12 Added log selector acl_warn_skipped (default on).
3249 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3250 variables because (a) they are in the wrong storage pool and (b) even if
3251 they were copied, it wouldn't work properly because of the caching.
3253 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3254 checking when decoding. Apparently there are clients that generate
3255 overlong encoded strings. Why am I not surprised?
3257 PH/15 If the first argument of "${if match_address" was not empty, but did not
3258 contain an "@" character, Exim crashed. Now it writes a panic log message
3259 and treats the condition as false.
3261 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3263 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3264 extension "envelope-auth". The code is finished and in agreement with
3265 other implementations, but there is no documentation so far and in fact,
3266 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3267 not changing the active code.
3269 Print executed "if" and "elsif" statements when debugging is used. This
3270 helps a great deal to understand what a filter does.
3272 Document more things not specified clearly in RFC3028. I had all this
3273 sorted out, when out of a sudden new issues came to my mind. Oops."
3275 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3278 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3279 canonical form (as documented). However, after a host name lookup from
3280 the IP address, check_host() was doing a simple string comparison with
3281 addresses acquired from the DNS when checking that the found name did
3282 have the original IP as one of its addresses. Since any found IPv6
3283 addresses are likely to be in abbreviated form, the comparison could
3284 fail. Luckily, there already exists a function for doing the comparison
3285 by converting both addresses to binary, so now that is used instead of
3286 the text comparison.
3288 PH/20 There was another similar case to PH/19, when a complete host name was
3289 given in a host list; looking up its IP address could give an abbreviated
3290 form, whereas the current host's name might or might not be abbreviated.
3291 The same fix has been applied.
3297 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3298 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3301 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3302 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3304 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3306 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3307 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3308 Also, for "Auto-Submitted", treat anything other than "no" as
3309 non-personal, in accordance with RFC 3834. (Previously it treated
3310 anything starting "auto-" as non-personal.)
3312 TF/01 The control=submission/name=... option had a problem with syntax
3313 errors if the name included a slash character. The /name= option
3314 now slurps the rest of the string, so it can include any characters
3315 but it must come last in the list of options (after /sender_retain
3318 PH/05 Some modifications to the interface to the fake nameserver for the new
3326 TK/01 Added the "success_on_redirect" address verification option. See
3327 NewStuff for rationale and an example.
3329 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3331 PH/02 Patch to exigrep to allow it to work on syslog lines.
3333 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3334 fread() to skip over the body file's header line, because in Cygwin the
3335 header line is locked and is inaccessible.
3337 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3338 co-exist for some time) to make it clear that it is the Exim ID that is
3339 referenced, not the Message-ID: header line.
3341 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3342 string_format() function, because snprintf() does not exist on all
3345 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3346 ${stat: expansion item.
3348 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3349 protocol synchronization error", to keep the pedants happy.
3351 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3352 well as for IRIX systems, when gcc is being used. See the host.c source
3355 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3357 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3360 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3361 transport and to the smtp transport in LMTP mode.
3363 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3365 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3366 run a filter in a subprocess. This could lead to confusion in subsequent
3367 lookups in the parent process. There should also be a search_tidyup() at
3368 the end of the subprocess.
3370 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3371 only if the host matched helo_try_verify_hosts, which caused the
3372 verification to occur when the EHLO/HELO command was issued. The ACL just
3373 tested the remembered result. Now, if a previous verification attempt has
3374 not happened, "verify = helo" does it there and then.
3376 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3378 TK/03 Fix log output including CR from clamd.
3380 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3381 exist provoked a memory error which could cause a segfault.
3383 PH/15 Installed PCRE 6.2
3385 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3387 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3388 of the problem. Specifically, suggested +O2 rather than +O1 for the
3391 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3393 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3394 record, the port was not correctly logged when the outgoing_port log
3395 selector was set (it logged the transort's default port).
3397 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3398 fallback_hosts, and "hosts" in the smtp transport.
3400 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3401 host errors such as "Connection refused".
3403 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3404 authentication with radiusclient 0.4.9:
3406 - Error returned from rc_read_config was caught wrongly
3407 - Username/password not passed on to radius server due to wrong length.
3409 The presumption is that some radiusclient API changes for 4.51/PH/17
3410 were not taken care of correctly. The code is still untested by me (my
3411 Linux distribution still has 0.3.2 of radiusclient), but it was
3412 contributed by a Radius user.
3414 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3415 expanding the "port" option of the smtp transport.
3417 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3418 while reading a MIME header. Thanks to Tom Hughes for a patch.
3420 PH/24 Include config.h inside local_scan.h so that configuration settings are
3423 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3424 that in an ACL for RCPT (for example), you can examine exactly what was
3427 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3428 commands, but it was not correctly comparing the address with the actual
3429 client host address. Thus, it would show the EHLO address in Received:
3430 header lines when this was not necessary.
3432 PH/27 Added the % operator to ${eval:}.
3434 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3435 it should be ignoring failures (because with -C, for example, it has lost
3436 privilege). It wasn't ignoring creation failures other than "already
3439 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3442 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3443 because an input line was too long, either on its own, or by virtue of
3444 too many continuations, the temporary file was not being removed, and the
3445 return code was incorrect.
3447 PH/31 Missing "BOOL" in function definition in filtertest.c.
3449 PH/32 Applied Sieve patches from the maintainer.
3451 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3453 PH/33 Added "verify = not_blind".
3455 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3456 Local/Makefile (with some defaults set). These are used in built scripts
3457 such as exicyclog, but they have never been used in the exim_install
3458 script (though there are many overriding facilities there). I have
3459 arranged that the exim_install script now takes note of these two
3462 PH/35 Installed configuration files for Dragonfly.
3464 PH/36 When a locally submitted message by a trusted user did not contain a
3465 From: header, and the sender address was obtained from -f or from an SMTP
3466 MAIL command, and the trusted user did not use -F to supply a sender
3467 name, $originator_name was incorrectly used when constructing a From:
3468 header. Furthermore, $originator_name was used for submission mode
3469 messages from external hosts without From: headers in a similar way,
3470 which is clearly wrong.
3472 PH/37 Added control=suppress_local_fixups.
3474 PH/38 When log_selector = +received_sender was set, and the addition of the
3475 sender made the log line's construction buffer exactly full, or one byte
3476 less than full, an overflow happened when the terminating "\n" was
3479 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3480 when the result of a list match is failure because a DNS lookup failed.
3482 PH/40 RM_COMMAND is now used in the building process.
3484 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3485 the "build-* directories that it finds.
3487 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3488 address in a domain literal was a prefix of an interface address.
3490 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3491 when verifying a sender address, unless rewrite_headers is false.
3493 PH/44 Wrote a long comment about why errors_to addresses are verified as
3494 recipients, not senders.
3496 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3497 the ratelimit ACL was added.
3499 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3501 PH/46 Added extra information about PostgreSQL errors to the error string.
3503 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3504 avoiding the need to install special zones in a real server. This is
3505 backwards compatible; if it can't find the fake resolver, it drops back.
3506 Thus, both old and new test suites can be run.
3508 TF/02 Added util/ratelimit.pl
3510 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3511 clock is set back in time.
3513 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3514 Candler <B.Candler@pobox.com>.
3516 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3517 OK for addresses that are the result of redirections.
3519 PH/48 A number of further additions for the benefit of the new test suite,
3520 including a fake gethostbyname() that interfaces to the fake DNS resolver
3523 TF/06 The fix for widen_domains has also been applied to qualify_single and
3524 search_parents which are the other dnslookup options that can cause
3527 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3530 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3532 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3533 service" if it encountered an MX record with an empty target hostname.
3534 The message is now "an MX or SRV record indicated no SMTP service".
3536 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3537 if the DNS of the sending site is misconfigured. This is quite a
3538 common situation. This change restores the behaviour of treating a
3539 helo verification defer as a failure.
3541 PH/51 If self=fail was set on a router, the bounce message did not include the
3542 actual error message.
3548 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3550 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3551 command itself ended in error, the underlying message about the transport
3552 filter timeout was being overwritten with the pipe command error. Now the
3553 underlying error message should be appended to the second error message.
3555 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3557 PH/02 Reduce the amount of output that "make" produces by default. Full output
3558 can still be requested.
3560 PH/03 The warning log line about a condition test deferring for a "warn" verb
3561 was being output only once per connection, rather than after each
3562 occurrence (because it was using the same function as for successful
3563 "warn" verbs). This seems wrong, so I have changed it.
3565 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3566 it should not have, which might have caused a crash in the right
3567 circumstances, but probably never did.
3569 PH/04 Installed a modified version of Tony Finch's patch to make submission
3570 mode fix the return path as well as the Sender: header line, and to
3571 add a /name= option so that you can make the user's friendly name appear
3574 TF/03 Added the control = fakedefer ACL modifier.
3576 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3577 Mark Lowes for thorough testing.
3579 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3581 TK/03 Merged latest SRS patch from Miles Wilton.
3583 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3584 with the definition in sysexits.h (which is #included earlier).
3585 Fortunately, Exim does not actually use EX_OK. The code used to try to
3586 preserve the sysexits.h value, by assuming that macro definitions were
3587 scanned for macro replacements. I have been disabused of this notion,
3588 so now the code just undefines EX_OK before #including unistd.h.
3590 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3591 in the smtp transport. When a block could not be written in a single
3592 write() function, the timeout was being re-applied to each part-write.
3593 This seems wrong - if the receiver was accepting one byte at a time it
3594 would take for ever. The timeout is now adjusted when this happens. It
3595 doesn't have to be particularly precise.
3597 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3598 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3600 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3601 without a domain if the check to <postmaster@domain> fails.
3603 SC/01 Eximstats: added -xls and the ability to specify output files
3604 (patch written by Frank Heydlauf).
3606 SC/02 Eximstats: use FileHandles for outputting results.
3608 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3610 SC/04 Eximstats: fixed display of large numbers with -nvr option
3612 SC/05 Eximstats: fixed merging of reports with empty tables.
3614 SC/06 Eximstats: added the -include_original_destination flag
3616 SC/07 Eximstats: removed tabs and trailing whitespace.
3618 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3620 TK/06 MBOX spool code: Add real "From " MBOX separator line
3621 so the .eml file is really in mbox format (even though
3622 most programs do not really care). Patch from Alex Miller.
3624 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3625 The latter is generated from $received_to and is only set if the
3626 message has one envelope recipient. SA can use these headers,
3627 obviously out-of-the-box. Patch from Alex Miller.
3629 PH/08 The ${def test on a variable was returning false if the variable's
3630 value was "0", contrary to what the specification has always said!
3631 The result should be true unless the variable is empty.
3633 PH/09 The syntax error of a character other than { following "${if
3634 def:variable_name" (after optional whitespace) was not being diagnosed.
3635 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3636 accidental colon was present, for example, could give incorrect results.
3638 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3639 result is used (not including appendfile, where other changes are about
3642 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3643 This involved changing a lot of size variables from int to off_t. It
3644 should work with maildirs and everything.
3646 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3647 spamd dying while we are connected to it.
3649 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3652 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3653 function for BDB 4.3.
3655 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3657 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3658 the sources for PCRE have been re-arranged and more files are now
3661 PH/15 The code I had for printing potentially long long variables in PH/11
3662 above was not the best (it lost precision). The length of off_t variables
3663 is now inspected at build time, and an appropriate printing format (%ld
3664 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3665 to be "long long int" or "long int". This is needed for the internal
3666 formatting function string_vformat().
3668 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3669 the configuration file to be ":syslog", then the script "guesses" where
3670 the logs files are, rather than using the compiled in default. In our
3671 case the guess is not the same as the compiled default, so the script
3672 suddenly stopped working when I started to use syslog. The patch checks
3673 to see if log_file_path is "". If so, it attempts to read it from exim
3674 with no configuration file to get the compiled in version, before it
3675 falls back to the previous guessing code."
3677 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3678 implementing BATV in an Exim configuration. See NewStuff for the gory
3681 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3682 Makefile that are specific to HP-UX.
3684 PH/18 If the "use_postmaster" option was set for a recipient callout together
3685 with the "random" option, the postmaster address was used as the MAIL
3686 FROM address for the random test, but not for the subsequent recipient
3687 test. It is now used for both.
3689 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3690 patch removes a few documentation additions to RFC 3028, because the
3691 latest draft now contains them. It adds the new en;ascii-case comparator
3692 and a new error check for 8bit text in MIME parts. Comparator and
3693 require names are now matched exactly. I enabled the subaddress
3694 extension, but it is not well tested yet (read: it works for me)."
3696 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3697 rework some of the code of TK/09 above to avoid the hardwired use of
3698 "%lld" and "long long". Replaced the call to snprintf() with a call to
3701 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3702 records point to non-existent hosts", "retry timeout exceeded", and
3703 "retry time not reached for any host after a long failure period".
3705 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3706 experimental DomainKeys support:
3708 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3709 (2) On an error such as an illegally used "control", the wrong name for
3710 the control was given.
3712 These problems did NOT occur unless DomainKeys support was compiled.
3714 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3716 PH/24 Added ${if match_ip condition.
3718 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3719 databases so that it will be absolutely obvious if a crash occurs in the
3720 DB library. This is a regular occurrence (often caused by mis-matched
3723 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3724 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3725 picked up on a user's system that detects such things. There doesn't seem
3726 to be a gcc warning option for this - only an attribute that has to be
3727 put on the function's prototype. It seems that in Fedora Core 4 they have
3728 set this on a number of new functions. No doubt there will be more in due
3731 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3732 specify a transport. However, if an address that was verified by such a
3733 router was the subject of a callout, Exim crashed because it tried to
3734 read the rcpt_include_affixes from the non-existent transport. Now it
3735 just assumes that the setting of that option is false. This bug was
3736 introduced by 4.51/PH/31.
3738 PH/28 Changed -d+all to exclude +memory, because that information is very
3739 rarely of interest, but it makes the output a lot bigger. People tend to
3740 do -d+all out of habit.
3742 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3743 code in os-type was giving problems when libc.so lives in lib64, like on
3746 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3747 aren't the modern standard, and it seems that some systems' include files
3748 don't always have them. Exim was already checking for some of the newer
3749 ones like T_AAAA, and defining it itself. I've added checks for all the
3750 record types that Exim uses.
3752 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3753 not automatically generating a new one, as it is supposed to. This
3754 prevented TLS from working. If the file did exist, but contained invalid
3755 data, a new version was generated, as expected. It was only the case of a
3756 non-existent file that was broken.
3758 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3759 with a change in libdomainkeys > 0.64.
3761 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3762 from DNS. If the selector record carries the flag, it now has
3763 precedence over the domain-wide flag.
3765 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3767 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3768 the use of a port name that isn't defined in /etc/services) occurred, the
3769 message was deferred as in a normal delivery, and thus remained on the
3770 spool, instead of being failed because of the mua_wrapper setting. This
3771 is now fixed, and I tidied up some of the mua_wrapper messages at the
3774 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3775 the messages in a hash of arrays rather than using individual hashes.
3776 This is a bit cleaner and results in dramatic memory savings, albeit
3777 at a slight CPU cost.
3779 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3780 as requested by Marc Sherman.
3782 SC/10 Eximstats: added histograms for user specified patterns as requested
3785 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3787 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3788 fopen() in the content-scanning modules that did not already have it.
3794 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3795 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3797 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3799 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3801 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3802 to test Sieve filters that use "vacation".
3804 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3805 that changes the way the GnuTLS parameters are stored in the cache file.
3806 The new format can be generated externally. For backward compatibility,
3807 if the data in the cache doesn't make sense, Exim assumes it has read an
3808 old-format file, and it generates new data and writes a new file. This
3809 means that you can't go back to an older release without removing the
3812 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3813 work if there are any delivery delays because "one_time" forces the
3814 parent to be marked "delivered", so its unseen clone is never tried
3815 again. For this reason, Exim now forbids the simultaneous setting of
3818 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3819 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3820 entry if you want to know the details. The fix, however, neglected to
3821 consider the case when local delivery batching is involved. The test for
3822 "previously delivered" was not happening when checking to see if an
3823 address could be batched with a previous (undelivered) one; under
3824 certain circumstances this could lead to multiple deliveries to the same
3827 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3828 in its include files, and this causes problems building Exim.
3830 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3831 header_syntax) but Exim was just ignoring anything given after a slash.
3832 In particular, this caused confusion with an attempt to use "verify =
3833 reverse_host_lookup/defer_ok". An error is now given when options are
3834 supplied for verify items that do not have them. (Maybe reverse_host_
3835 lookup should have a defer_ok option, but that's a different point.)
3837 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3838 defined by RFC 821) to 2048, because there were problems with some AUTH
3839 commands, and RFC 1869 says the size should be increased for extended
3840 SMTP commands that take arguments.
3842 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3845 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3846 "unknown" error; now it says that the functionality isn't in the binary.
3848 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3849 an address' error message when a string expansion fails (syntax or
3850 whatever). Otherwise the password may appear in the log. Following change
3851 PH/42 below, there is no longer a chance of it appearing in a bounce
3854 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3856 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3857 of its addresses was ever tried. (Bugzilla bug #2.)
3859 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3860 the result incorrectly in the debug output. (It correctly added a newline
3861 to what was transported.)
3863 TF/01 Added $received_time.
3865 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3866 commented out examples of how to interface to a virus scanner and to
3867 SpamAssassin. Also added commented examples of av_scanner and
3868 spamd_address settings.
3870 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3871 and controls are allowed in which ACLs. There were a couple of minor
3872 errors. Some of the entries in the conditions table (which is a table of
3873 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3874 negation of where the condition IS allowed.
3876 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3878 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3879 header file does not have a version number, so I've had to invent a new
3880 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3881 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3882 radiusclient), but it was contributed by a Radius user.
3884 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3885 files or directories, for OpenSSL.
3887 PH/19 When an Exim process that is running as root has to create an Exim log
3888 file, it does so in a subprocess that runs as exim:exim so as to get the
3889 ownership right at creation (otherwise, other Exim processes might see
3890 the file with the wrong ownership). There was no test for failure of this
3891 fork() call, which would lead to the process getting stuck as it waited
3892 for a non-existent subprocess. Forks do occasionally fail when resources
3893 run out. I reviewed all the other calls to fork(); they all seem to check
3896 PH/20 When checking for unexpected SMTP input at connect time (before writing
3897 the banner), Exim was not dealing correctly with a non-positive return
3898 from the read() function. If the client had disconnected by this time,
3899 the result was a log entry for a synchronization error with an empty
3900 string after "input=" when read() returned zero. If read() returned -1
3901 (an event I could not check), uninitialized data bytes were printed.
3902 There were reports of junk text (parts of files, etc) appearing after
3905 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3907 PH/22 Added support for macro redefinition, and (re)definition in between
3908 driver and ACL definitions.
3910 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3911 forgetting to use the resulting value; it was using the unexpanded value.
3913 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3914 hadn't been configured. The fix is from Juergen Kreileder, who
3915 understands it better than I do:
3917 "Here's what I see happening with three configured cyrus_sasl
3918 authenticators configured (plain, login, cram-md5):
3920 On startup auth_cyrus_sasl_init() gets called for each of these.
3921 This means three calls to sasl_listmech() without a specified mech_list.
3922 => SASL tests which mechs of all available mechs actually work
3923 => three warnings about OTP not working
3924 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3926 With the patch, sasl_listmech() also gets called three times. But now
3927 SASL's mech_list option is set to the server_mech specified in the the
3928 authenticator. Or in other words, the answer from sasl_listmech()
3929 gets limited to just the mech you're testing for (which is different
3931 => the return list contains just 'plain' or 'login', 'cram-md5' or
3932 nothing depending on the value of ob->server_mech.
3934 I've just tested the patch: Authentication still works fine,
3935 unavailable mechs specified in the exim configuration are still
3936 caught, and the auth.log warnings about OTP are gone."
3938 PH/25 When debugging is enabled, the contents of the command line are added
3939 to the debugging output, even when log_selector=+arguments is not
3942 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3943 answer is "GNU", and only if the return is "GNU/something" is the answer
3946 PH/27 $acl_verify_message is now set immediately after the failure of a
3947 verification in an ACL, and so is available in subsequent modifiers. In
3948 particular, the message can be preserved by coding like this:
3950 warn !verify = sender
3951 set acl_m0 = $acl_verify_message
3953 Previously, $acl_verify_message was set only while expanding "message"
3954 and "log_message" when a very denied access.
3956 PH/28 Modified OS/os.c-Linux with
3958 -#ifndef OS_LOAD_AVERAGE
3959 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3961 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3962 nomenclature these days.)
3964 PH/29 Installed patch from the Sieve maintainer that adds the options
3965 sieve_useraddress and sieve_subaddress to the redirect router.
3967 PH/30 In these circumstances:
3968 . Two addresses routed to the same list of hosts;
3969 . First host does not offer TLS;
3970 . First host accepts first address;
3971 . First host gives temporary error to second address;
3972 . Second host offers TLS and a TLS session is established;
3973 . Second host accepts second address.
3974 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3975 and peerdn, if requested) that were in fact used only for the second
3978 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3979 attention to any local part prefix or suffix that was matched by the
3980 router that accepted the address. It now behaves in the same way as it
3981 does for delivery: the affixes are removed from the local part unless
3982 rcpt_include_affixes is set on the transport.
3984 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3985 timeout during the DATA phase of an incoming message.
3987 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3988 applied a patch sanctioned by the Sieve maintainer.
3990 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3991 the uid or gid is negative. A case of a negative gid caused this to be
3992 noticed. The fix allows for either to be negative.
3994 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3995 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3998 PH/36 The change PH/12 above was broken. Fixed it.
4000 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4001 the grounds that routing the same address twice would always produce the
4002 same answer. This might have been true once, but it is certainly no
4003 longer true now. Routing a child address may depend on the previous
4004 routing that produced that child. Some complicated redirection strategies
4005 went wrong when messages had multiple recipients, and made Exim's
4006 behaviour dependent on the order in which the addresses were given.
4008 I have moved the duplicate checking until after the routing is complete.
4009 Exim scans the addresses that are assigned to local and remote
4010 transports, and removes any duplicates. This means that more work will be
4011 done, as duplicates will always all be routed, but duplicates are
4012 presumably rare, so I don't expect this is of any significance.
4014 For deliveries to pipes, files, and autoreplies, the duplicate checking
4015 still happens during the routing process, since they are not going to be
4018 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4019 It corrects a timeout issue with spamd. This is Ian's comment: "The
4020 background is that sometimes spamd either never reads data from a
4021 connection it has accepted, or it never writes response data. The exiscan
4022 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4023 blindly assumes that writes won't block so it may never time out."
4025 PH/39 Allow G after quota size as well as K and M.
4027 PH/40 The value set for $authenticated_id in an authenticator may not contain
4028 binary zeroes or newlines because the value is written to log lines and
4029 to spool files. There was no check on this. Now the value is run through
4030 the string_printing() function so that such characters are converted to
4031 printable escape sequences.
4033 PH/41 $message_linecount is a new variable that contains the total number of
4034 lines in the message. Compare $body_linecount, which is the count for the
4037 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4038 bounce and delay warning messages, except in certain special cases, which
4041 (a) An SMTP error message from a remote host;
4042 (b) A message specified in a :fail: redirection;
4043 (c) A message specified in a "fail" command in a system filter;
4044 (d) A message specified in a FAIL return from the queryprogram router;
4045 (e) A message specified by the cannot_route_message router option.
4047 In these cases only, Exim does include the error details in bounce and
4048 warning messages. There are also a few cases where bland messages such
4049 as "unrouteable address" or "local delivery error" are given.
4051 PH/43 $value is now also set for the "else" part of a ${run expansion.
4053 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4054 being worked on, but at least Exim now implements the latest version to
4057 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4058 process to complete was treated as a delivery failure, a timeout while
4059 writing the message to the pipe was logged, but erroneously treated as a
4060 successful delivery. Such timeouts include transport filter timeouts. For
4061 consistency with the overall process timeout, these timeouts are now
4062 treated as errors, giving rise to delivery failures by default. However,
4063 there is now a new Boolean option for the pipe transport called
4064 timeout_defer, which, if set TRUE, converts the failures into defers for
4065 both kinds of timeout. A transport filter timeout is now identified in
4068 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4069 systems where "make" and "gmake" are different, calling "gmake" at top
4070 level broke things. I've arranged for the value of $(MAKE) to be passed
4071 from the Makefile to this script so that it can call the same version of
4075 A note about Exim versions 4.44 and 4.50
4076 ----------------------------------------
4078 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4079 changes of various kinds. As a consequence, a big documentation update was
4080 needed. This delayed the release for rather longer than seemed good, especially
4081 in the light of a couple of (minor) security issues. Therefore, the changes
4082 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4083 release. So 4.44 and 4.50 are in effect two different branches that both start
4086 I have left the 4.50 change log unchanged; it contains all the changes since
4087 4.43. The change log for 4.44 is below; many of its items are identical to
4088 those for 4.50. This seems to be the most sensible way to preserve the
4089 historical information.
4095 1. Minor wording change to the doc/README.SIEVE file.
4097 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4098 computation of the current number of files was incorrect.
4100 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4101 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4104 4. Give more explanation in the error message when the command for a transport
4105 filter fails to execute.
4107 5. There are several places where Exim runs a non-Exim command in a
4108 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4109 was being done only for the command run by the queryprogram router. It is
4110 now done for all such subprocesses. The other cases are: ${run, transport
4111 filters, and the commands run by the lmtp and pipe transports.
4113 6. Added CONFIGURE_GROUP build-time option.
4115 7. Some older OS have a limit of 256 on the maximum number of file
4116 descriptors. Exim was using setrlimit() to set 1000 as a large value
4117 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4118 systems. I've change it so that if it can't get 1000, it tries for 256.
4120 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4121 was an oversight, and furthermore, ever since the addition of extra
4122 controls (e.g. 4.43/32), the checks on when to allow different forms of
4123 "control" were broken. There should now be diagnostics for all cases when a
4124 control that does not make sense is encountered.
4126 9. Added the /retain_sender option to "control=submission".
4128 10. $recipients is now available in the predata ACL (oversight).
4130 11. Tidy the search cache before the fork to do a delivery from a message
4131 received from the command line. Otherwise the child will trigger a lookup
4132 failure and thereby defer the delivery if it tries to use (for example) a
4133 cached ldap connection that the parent has called unbind on.
4135 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4136 of $address_data from the recipient verification was clobbered by the
4137 sender verification.
4139 13. The value of address_data from a sender verification is now available in
4140 $sender_address_data in subsequent conditions in the ACL statement.
4142 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4144 15. Added a new option "connect=<time>" to callout options, to set a different
4147 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4148 was its contents. (It was OK if the option was not defined at all.)
4150 17. A "Completed" log line is now written for messages that are removed from
4151 the spool by the -Mrm option.
4153 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4154 information about exactly what failed.
4156 19. Added -dd to debug only the daemon process.
4158 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4159 handles timeouts, both on the server side and network timeouts. Renamed the
4160 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4162 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4163 It is now set to "smtps".
4165 22. $host_address is now set to the target address during the checking of
4166 ignore_target_hosts.
4168 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4169 being passed; this would have caused $sender_host_name to have been used if
4170 matching the list had actually called for a host name (not very likely,
4171 since this list is usually IP addresses). A host name is now passed as
4174 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4175 code that specifies a non-restarting handler (typically sigaction() in
4176 modern systems) in an attempt to fix a rare and obscure crash bug.
4178 25. Narrowed the window for a race in the daemon that could cause it to ignore
4179 SIGCHLD signals. This is not a major problem, because they are used only to
4180 wake it up if nothing else does.
4182 26. A malformed maildirsize file could cause Exim to calculate negative values
4183 for the mailbox size or file count. Odd effects could occur as a result.
4184 The maildirsize information is now recalculated if the size or filecount
4187 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4188 support for a long time. Removed HAVE_SYS_VFS_H.
4190 28. Installed the latest version of exipick from John Jetmore.
4192 29. In an address list, if the pattern was not a regular expression, an empty
4193 subject address (from a bounce message) matched only if the pattern was an
4194 empty string. Non-empty patterns were not even tested. This was the wrong
4195 because it is perfectly reasonable to use an empty address as part of a
4196 database query. An empty address is now tested by patterns that are
4197 lookups. However, all the other forms of pattern expect the subject to
4198 contain a local part and a domain, and therefore, for them, an empty
4199 address still always fails if the pattern is not itself empty.
4201 30. Exim went into a mad DNS loop when attempting to do a callout where the
4202 host was specified on an smtp transport, and looking it up yielded more
4203 than one IP address.
4205 31. Re-factored the code for checking spool and log partition space into a
4206 function that finds that data and another that does the check. The former
4207 is then used to implement four new variables: $spool_space, $log_space,
4208 $spool_inodes, and $log_inodes.
4210 32. The RFC2047 encoding function was originally intended for short strings
4211 such as real names; it was not keeping to the 75-character limit for
4212 encoded words that the RFC imposes. It now respects the limit, and
4213 generates multiple encoded words if necessary. To be on the safe side, I
4214 have increased the buffer size for the ${rfc2047: expansion operator from
4217 33. It is now permitted to omit both strings after an "if" condition; if the
4218 condition is true, the result is "true". As before, when the second string
4219 is omitted, a false condition yields an empty string. This makes it less
4220 cumbersome to write custom ACL and router conditions.
4222 34. Failure to deliver a bounce message always caused it to be frozen, even if
4223 there was an errors_to setting on the router. The errors_to setting is now
4226 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4227 canonical form (fully expanded) before being placed in
4228 $sender_host_address.
4230 36. The table in the code that translates DNS record types into text (T_A to
4231 "A" for instance) was missing entries for NS and CNAME. It is just possible
4232 that this could have caused confusion if both these types were looked up
4233 for the same domain, because the text type is used as part of Exim's
4234 per-process caching. But the chance of anyone hitting this buglet seems
4237 37. The dnsdb lookup has been extended in a number of ways.
4239 (1) There is a new type, "zns", which walks up the domain tree until it
4240 finds some nameserver records. It should be used with care.
4242 (2) There is a new type, "mxh", which is like "mx" except that it returns
4243 just the host names, not the priorities.
4245 (3) It is now possible to give a list of domains (or IP addresses) to be
4246 looked up. The behaviour when one of the lookups defers can be
4247 controlled by a keyword.
4249 (4) It is now possible to specify the separator character for use when
4250 multiple records are returned.
4252 38. The dnslists ACL condition has been extended: it is now possible to supply
4253 a list of IP addresses and/or domains to be looked up in a particular DNS
4256 39. Added log_selector=+queue_time_overall.
4258 40. When running the queue in the test harness, wait just a tad after forking a
4259 delivery process, to get repeatability of debugging output.
4261 41. Include certificate and key file names in error message when GnuTLS fails
4262 to set them up, because the GnuTLS error message doesn't include the name
4263 of the failing file when there is a problem reading it.
4265 42. Allow both -bf and -bF in the same test run.
4267 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4269 44. The "Exiscan patch" is now merged into the mainline Exim source.
4271 45. Sometimes the final signoff response after QUIT could fail to get
4272 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4273 < 0 before doing a fflush(). This bug looks as though it goes back to the
4274 introduction of TLS in release 3.20, but "sometimes" must have been rare
4275 because the tests only now provoked it.
4277 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4278 (this can affect the format of dates).
4280 47. exim_tidydb, when checking for the continued existence of a message for
4281 which it has found a message-specific retry record, was not finding
4282 messages that were in split spool directories. Consequently, it was
4283 deleting retry records that should have stayed in existence.
4285 48. Steve fixed some bugs in eximstats.
4287 49. The SPA authentication driver was not abandoning authentication and moving
4288 on to the next authenticator when an expansion was forced to fail,
4289 contradicting the general specification for all authenticators. Instead it
4290 was generating a temporary error. It now behaves as specified.
4292 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4293 (the order specifies the preference for clients). The order is now AES256,
4294 AES128, 3DES, ARCFOUR128.
4296 51. Small patch to Sieve code - explicitly set From: when generating an
4299 52. Exim crashed if a remote delivery caused a very long error message to be
4300 recorded - for instance if somebody sent an entire SpamAssassin report back
4301 as a large number of 550 error lines. This bug was coincidentally fixed by
4302 increasing the size of one of Exim's internal buffers (big_buffer) that
4303 happened as part of the Exiscan merge. However, to be on the safe side, I
4304 have made the code more robust (and fixed the comments that describe what
4307 53. Now that there can be additional text after "Completed" in log lines (if
4308 the queue_time_overall log selector is set), a one-byte patch to exigrep
4309 was needed to allow it to recognize "Completed" as not the last thing in
4312 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4313 patch that reportedly fixes this has been added. I am not expert enough to
4314 create a test for it. This is what the patch creator wrote:
4316 "I found a little strange behaviour of ldap code when working with
4317 Windows 2003 AD Domain, where users was placed in more than one
4318 Organization Units. When I tried to give exim partial DN, the exit code
4319 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4320 But simultaneously result of request was absolutely normal ldap result,
4321 so I produce this patch..."
4323 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4324 REFERENCE, so I have modified the code to exclude the patch when that macro
4327 55. Some experimental protocols are using DNS PTR records for new purposes. The
4328 keys for these records are domain names, not reversed IP addresses. The
4329 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4330 leaves it alone. Component reversal etc. now happens only for IP addresses.
4333 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4335 57. Double the size of the debug message buffer (to 2048) so that more of very
4336 long debug lines gets shown.
4338 58. The exicyclog utility now does better if the number of log files to keep
4339 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4341 59. Two changes related to the smtp_active_hostname option:
4343 (1) $smtp_active_hostname is now available as a variable.
4344 (2) The default for smtp_banner uses $smtp_active_hostname instead
4345 of $primary_hostname.
4347 60. The host_aton() function is supposed to be passed a string that is known
4348 to be a valid IP address. However, in the case of IPv6 addresses, it was
4349 not checking this. This is a hostage to fortune. Exim now panics and dies
4350 if the condition is not met. A case was found where this could be provoked
4351 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4352 components; fortuitously, this particular loophole had already been fixed
4353 by change 4.50/55 above.
4355 If there are any other similar loopholes, the new check in host_aton()
4356 itself should stop them being exploited. The report I received stated that
4357 data on the command line could provoke the exploit when Exim was running as
4358 exim, but did not say which command line option was involved. All I could
4359 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4360 running as the user.
4363 61. There was a buffer overflow vulnerability in the SPA authentication code
4364 (which came originally from the Samba project). I have added a test to the
4365 spa_base64_to_bits() function which I hope fixes it.
4368 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4369 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4371 63. The daemon start-up calls getloadavg() while still root for those OS that
4372 need the first call to be done as root, but it missed one case: when
4373 deliver_queue_load_max is set with deliver_drop_privilege. This is
4374 necessary for the benefit of the queue runner, because there is no re-exec
4375 when deliver_drop_privilege is set.
4377 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4378 This has been fixed.
4380 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4381 was in use, was not putting the data itself into the right store pool;
4382 consequently, it could be overwritten for a subsequent message in the same
4383 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4386 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4388 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4389 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4390 the function were treating the return as a boolean value, which happened to
4391 work because 0=false and not-0=true, but is not correct code.
4393 68. The host_aton() function was not handling scoped IPv6 addresses (those
4394 with, for example, "%eth0" on the end) correctly.
4396 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4397 negated items (that is, ~something) in unsigned ints. Some compilers
4398 apparently mutter when there is no cast.
4400 70. If an address verification called from an ACL failed, and did not produce a
4401 user-specific message (i.e. there was only a "system" message), nothing was
4402 put in $acl_verify_message. In this situation, it now puts the system
4405 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4406 session; change 4.31/43 added the unwanted input to the log line - except
4407 that it did not do this in the start of session case. It now does.
4409 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4410 This is wrong and can cause the other end to generate a synchronization
4411 error if it is another Exim or anything else that does the synchronization
4412 check. A QUIT command is no longer sent after a timeout.
4414 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4415 during host lookups.
4417 74. The defer_ok option of callout verification was not working if it was used
4418 when verifying addresses in header lines, that is, for this case:
4420 verify = header_sender/callout=defer_ok
4422 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4423 those file descriptors could be used for SMTP connections. If anything
4424 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4425 could be sent to the SMTP client, causing chaos. The daemon now opens
4426 stdin, stdout, and stderr to /dev/null when it puts itself into the
4429 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4430 log by default. The user can override this with suitable Perl magic.
4432 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4433 the log message when discard triggers, was not working for the DATA ACL or
4434 for the non-SMTP ACL.
4436 78. Error message wording change in sieve.c.
4438 79. If smtp_accept_max_per_host was set, the number of connections could be
4439 restricted to fewer than expected, because the daemon was trying to set up
4440 a new connection before checking whether the processes handling previous
4441 connections had finished. The check for completed processes is now done
4442 earlier. On busy systems, this bug wouldn't be noticed because something
4443 else would have woken the daemon, and it would have reaped the completed
4446 80. If a message was submitted locally by a user whose login name contained one
4447 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4448 It caused a spool format error. I have fixed the spool reading code. A
4449 related problem was that the "from" clause in the Received: line became
4450 illegal because of the space(s). It is now covered by ${quote_local_part.
4452 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4455 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4457 83. Patch from Sieve maintainer to fix unterminated string problem in
4458 "vacation" handling.
4460 84. Some minor changes to the Linux configuration files to help with other
4461 OS variants using glibc.
4463 85. One more patch for Sieve to update vacation handling to latest spec.
4466 ----------------------------------------------------
4467 See the note above about the 4.44 and 4.50 releases.
4468 ----------------------------------------------------
4474 1. Change 4.43/35 introduced a bug that caused file counts to be
4475 incorrectly computed when quota_filecount was set in an appendfile
4478 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4479 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4482 3. Give more explanation in the error message when the command for a transport
4483 filter fails to execute.
4485 4. There are several places where Exim runs a non-Exim command in a
4486 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4487 was being done only for the command run by the queryprogram router. It is
4488 now done for all such subprocesses. The other cases are: ${run, transport
4489 filters, and the commands run by the lmtp and pipe transports.
4491 5. Some older OS have a limit of 256 on the maximum number of file
4492 descriptors. Exim was using setrlimit() to set 1000 as a large value
4493 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4494 systems. I've change it so that if it can't get 1000, it tries for 256.
4496 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4497 was an oversight, and furthermore, ever since the addition of extra
4498 controls (e.g. 4.43/32), the checks on when to allow different forms of
4499 "control" were broken. There should now be diagnostics for all cases when a
4500 control that does not make sense is encountered.
4502 7. $recipients is now available in the predata ACL (oversight).
4504 8. Tidy the search cache before the fork to do a delivery from a message
4505 received from the command line. Otherwise the child will trigger a lookup
4506 failure and thereby defer the delivery if it tries to use (for example) a
4507 cached ldap connection that the parent has called unbind on.
4509 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4510 of $address_data from the recipient verification was clobbered by the
4511 sender verification.
4513 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4514 was its contents. (It was OK if the option was not defined at all.)
4516 11. A "Completed" log line is now written for messages that are removed from
4517 the spool by the -Mrm option.
4519 12. $host_address is now set to the target address during the checking of
4520 ignore_target_hosts.
4522 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4523 being passed; this would have caused $sender_host_name to have been used if
4524 matching the list had actually called for a host name (not very likely,
4525 since this list is usually IP addresses). A host name is now passed as
4528 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4529 code that specifies a non-restarting handler (typically sigaction() in
4530 modern systems) in an attempt to fix a rare and obscure crash bug.
4532 15. Narrowed the window for a race in the daemon that could cause it to ignore
4533 SIGCHLD signals. This is not a major problem, because they are used only to
4534 wake it up if nothing else does.
4536 16. A malformed maildirsize file could cause Exim to calculate negative values
4537 for the mailbox size or file count. Odd effects could occur as a result.
4538 The maildirsize information is now recalculated if the size or filecount
4541 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4542 support for a long time. Removed HAVE_SYS_VFS_H.
4544 18. Updated exipick to current release from John Jetmore.
4546 19. Allow an empty sender to be matched against a lookup in an address list.
4547 Previously the only cases considered were a regular expression, or an
4550 20. Exim went into a mad DNS lookup loop when doing a callout where the
4551 host was specified on the transport, if the DNS lookup yielded more than
4554 21. The RFC2047 encoding function was originally intended for short strings
4555 such as real names; it was not keeping to the 75-character limit for
4556 encoded words that the RFC imposes. It now respects the limit, and
4557 generates multiple encoded words if necessary. To be on the safe side, I
4558 have increased the buffer size for the ${rfc2047: expansion operator from
4561 22. Failure to deliver a bounce message always caused it to be frozen, even if
4562 there was an errors_to setting on the router. The errors_to setting is now
4565 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4566 canonical form (fully expanded) before being placed in
4567 $sender_host_address.
4569 24. Updated eximstats to version 1.33
4571 25. Include certificate and key file names in error message when GnuTLS fails
4572 to set them up, because the GnuTLS error message doesn't include the name
4573 of the failing file when there is a problem reading it.
4575 26. Expand error message when OpenSSL has problems setting up cert/key files.
4578 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4579 (this can affect the format of dates).
4581 28. exim_tidydb, when checking for the continued existence of a message for
4582 which it has found a message-specific retry record, was not finding
4583 messages that were in split spool directories. Consequently, it was
4584 deleting retry records that should have stayed in existence.
4586 29. eximstats updated to version 1.35
4587 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4588 1.35 - bugfix such that pie charts by volume are generated correctly
4590 30. The SPA authentication driver was not abandoning authentication and moving
4591 on to the next authenticator when an expansion was forced to fail,
4592 contradicting the general specification for all authenticators. Instead it
4593 was generating a temporary error. It now behaves as specified.
4595 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4596 (the order specifies the preference for clients). The order is now AES256,
4597 AES128, 3DES, ARCFOUR128.
4599 31. Small patch to Sieve code - explicitly set From: when generating an
4602 32. Exim crashed if a remote delivery caused a very long error message to be
4603 recorded - for instance if somebody sent an entire SpamAssassin report back
4604 as a large number of 550 error lines. This bug was coincidentally fixed by
4605 increasing the size of one of Exim's internal buffers (big_buffer) that
4606 happened as part of the Exiscan merge. However, to be on the safe side, I
4607 have made the code more robust (and fixed the comments that describe what
4610 33. Some experimental protocols are using DNS PTR records for new purposes. The
4611 keys for these records are domain names, not reversed IP addresses. The
4612 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4613 leaves it alone. Component reversal etc. now happens only for IP addresses.
4616 34. The host_aton() function is supposed to be passed a string that is known
4617 to be a valid IP address. However, in the case of IPv6 addresses, it was
4618 not checking this. This is a hostage to fortune. Exim now panics and dies
4619 if the condition is not met. A case was found where this could be provoked
4620 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4621 components; fortuitously, this particular loophole had already been fixed
4622 by change 4.50/55 or 4.44/33 above.
4624 If there are any other similar loopholes, the new check in host_aton()
4625 itself should stop them being exploited. The report I received stated that
4626 data on the command line could provoke the exploit when Exim was running as
4627 exim, but did not say which command line option was involved. All I could
4628 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4629 running as the user.
4632 35. There was a buffer overflow vulnerability in the SPA authentication code
4633 (which came originally from the Samba project). I have added a test to the
4634 spa_base64_to_bits() function which I hope fixes it.
4637 36. The daemon start-up calls getloadavg() while still root for those OS that
4638 need the first call to be done as root, but it missed one case: when
4639 deliver_queue_load_max is set with deliver_drop_privilege. This is
4640 necessary for the benefit of the queue runner, because there is no re-exec
4641 when deliver_drop_privilege is set.
4643 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4644 was in use, was not putting the data itself into the right store pool;
4645 consequently, it could be overwritten for a subsequent message in the same
4646 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4649 38. Sometimes the final signoff response after QUIT could fail to get
4650 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4651 < 0 before doing a fflush(). This bug looks as though it goes back to the
4652 introduction of TLS in release 3.20, but "sometimes" must have been rare
4653 because the tests only now provoked it.
4659 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4660 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4661 Now it returns an integer. A number of calls were still expecting a T/F
4662 return. Fortuitously, in all cases, the tests worked in OK situations,
4663 which is the norm. However, things would have gone wrong on any write
4664 failures on the smtp file descriptor. This function is used when sending
4665 messages over SMTP and also when doing verify callouts.
4667 2. When Exim is called to do synchronous delivery of a locally submitted
4668 message (the -odf or -odi options), it no longer closes stderr before doing
4671 3. Implemented the mua_wrapper option.
4673 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4675 5. Implemented the functions header_remove(), header_testname(),
4676 header_add_at_position(), and receive_remove_recipient(), and exported them
4679 6. If an ACL "warn" statement specified the addition of headers, Exim already
4680 inserted X-ACL-Warn: at the start if there was no header name. However, it
4681 was not making this test for the second and subsequent header lines if
4682 there were newlines in the string. This meant that an invalid header could
4683 be inserted if Exim was badly configured.
4685 7. Allow an ACL "warn" statement to add header lines at the start or after all
4686 the Received: headers, as well as at the end.
4688 8. Added the rcpt_4xx retry error code.
4690 9. Added postmaster_mailfrom=xxx to callout verification option.
4692 10. Added mailfrom=xxxx to the callout verification option, for verify=
4695 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4696 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4698 12. Inserted some casts to stop certain compilers warning when using pointer
4699 differences as field lengths or precisions in printf-type calls (mostly
4700 affecting debugging statements).
4702 13. Added optional readline() support for -be (dynamically loaded).
4704 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4705 same clock tick as a message's arrival, so that its received time was the
4706 same as the "first fail" time on the retry record, and that message
4707 remained on the queue past the ultimate address timeout, every queue runner
4708 would try a delivery (because it was past the ultimate address timeout) but
4709 after another failure, the ultimate address timeout, which should have then
4710 bounced the address, did not kick in. This was a "< instead of <=" error;
4711 in most cases the first failure would have been in the next clock tick
4712 after the received time, and all would be well.
4714 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4715 being recognized when the domain list was tested by the match_domain
4716 condition in an expansion string.
4718 16. Added the ${str2b64: operator.
4720 17. Exim was always calling setrlimit() to set a large limit for the number of
4721 processes, without checking whether the existing limit was already
4722 adequate. (It did check for the limit on file descriptors.) Furthermore,
4723 errors from getrlimit() and setrlimit() were being ignored. Now they are
4724 logged to the main and panic logs, but Exim does carry on, to try to do its
4725 job under whatever limits there are.
4727 18. Imported PCRE 5.0.
4729 19. Trivial typo in log message " temporarily refused connection" (the leading
4732 20. If the log selector return_path_on_delivery was set and an address was
4733 redirected to /dev/null, the delivery process crashed because it assumed
4734 that a return path would always be set for a "successful" delivery. In this
4735 case, the whole delivery is bypassed as an optimization, and therefore no
4738 21. Internal re-arrangement: the function for sending a challenge and reading
4739 a response while authentication was assuming a zero-terminated challenge
4740 string. It's now changed to take a pointer and a length, to allow for
4741 binary data in such strings.
4743 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4745 23. Exim was not respecting finduser_retries when seeking the login of the
4746 uid under which it was called; it was always trying 10 times. (The default
4747 setting of finduser_retries is zero.) Also, it was sleeping after the final
4748 failure, which is pointless.
4750 24. Implemented tls_on_connect_ports.
4752 25. Implemented acl_smtp_predata.
4754 26. If the domain in control=submission is set empty, Exim assumes that the
4755 authenticated id is a complete email address when it generates From: or
4756 Sender: header lines.
4758 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4759 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4760 chown and chgrp in /bin and hostname in /usr/bin.
4762 28. Exim was keeping the "process log" file open after each use, just as it
4763 does for the main log. This opens the possibility of it remaining open for
4764 long periods when the USR1 signal hits a daemon. Occasional processlog
4765 errors were reported, that could have been caused by this. Anyway, it seems
4766 much more sensible not to leave this file open at all, so that is what now
4769 29. The long-running daemon process does not normally write to the log once it
4770 has entered its main loop, and it closes the log before doing so. This is
4771 so that log files can straightforwardly be renamed and moved. However,
4772 there are a couple of unusual error situations where the daemon does write
4773 log entries, and I had neglected to close the log afterwards.
4775 30. The text of an SMTP error response that was received during a remote
4776 delivery was being truncated at 512 bytes. This is too short for some of
4777 the long messages that one sometimes sees. I've increased the limit to
4780 31. It is now possible to make retry rules that apply only when a message has a
4781 specific sender, in particular, an empty sender.
4783 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4784 it possible to be selective about when SMTP synchronization is enforced.
4786 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4788 32. Implemented hosts_connection_nolog.
4790 33. Added an ACL for QUIT.
4792 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4795 35. Added mailbox_size and mailbox_filecount to appendfile.
4797 36. Added control = no_multiline_responses to ACLs.
4799 37. There was a bug in the logic of the code that waits for the clock to tick
4800 in the case where the clock went backwards by a substantial amount such
4801 that the microsecond fraction of "now" was more than the microsecond
4802 fraction of "then" (but the whole seconds number was less).
4804 38. Added support for the libradius Radius client library this is found on
4805 FreeBSD (previously only the radiusclient library was supported).
4811 1. When certain lookups returned multiple values in the form name=value, the
4812 quoting of the values was not always being done properly. Specifically:
4813 (a) If the value started with a double quote, but contained no whitespace,
4815 (b) If the value contained whitespace other than a space character (i.e.
4816 tabs or newlines or carriage returns) it was not quoted.
4817 This fix has been applied to the mysql and pgsql lookups by writing a
4818 separate quoting function and calling it from the lookup code. The fix
4819 should probably also be applied to nisplus, ibase and oracle lookups, but
4820 since I cannot test any of those, I have not disturbed their existing code.
4822 2. A hit in the callout cache for a specific address caused a log line with no
4823 reason for rejecting RCPT. Now it says "Previous (cached) callout
4824 verification failure".
4826 3. There was an off-by-one bug in the queryprogram router. An over-long
4827 return line was truncated at 256 instead of 255 characters, thereby
4828 overflowing its buffer with the terminating zero. As well as fixing this, I
4829 have increased the buffer size to 1024 (and made a note to document this).
4831 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4832 when Exim is waiting for an SMTP response from a remote server, Exim
4833 restarts its select() call on the socket, thereby resetting its timeout.
4834 This is not a problem when such interrupts are rare. Somebody set up a cron
4835 job to run exiwhat every 2 minutes, which is less than the normal select()
4836 timeout (5 or 10 minutes). This meant that the select() timeout never
4837 kicked in because it was always reset. I have fixed this by comparing the
4838 time when an interrupt arrives with the time at the start of the first call
4839 to select(). If more time than the timeout has elapsed, the interrupt is
4840 treated as a timeout.
4842 5. Some internal re-factoring in preparation for the addition of Sieve
4843 extensions (by MH). In particular, the "personal" test is moved to a
4844 separate function, and given an option for scanning Cc: and Bcc: (which is
4845 not set for Exim filters).
4847 6. When Exim created an email address using the login of the caller as the
4848 local part (e.g. when creating a From: or Sender: header line), it was not
4849 quoting the local part when it contained special characters such as @.
4851 7. Installed new OpenBSD configuration files.
4853 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4854 try to make them clearer.
4856 9. Callout options, other than the timeout value, were being ignored when
4857 verifying sender addresses in header lines. For example, when using
4859 verify = header_sender/callout=no_cache
4861 the cache was (incorrectly) being used.
4863 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4864 only the Cygwin environment.
4866 11. When return_path_on_delivery was set as a log selector, if different remote
4867 addresses in the same message used different return paths and parallel
4868 remote delivery occurred, the wrong values would sometimes be logged.
4869 (Whenever a remote delivery process finished, the return path value from
4870 the most recently started remote delivery process was logged.)
4872 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4873 lines when AUTH and/or TLS are in use. This is the "received protocol"
4874 field. Exim used to use "asmtp" for authenticated SMTP, without any
4875 indication (in the protocol name) for TLS use. Now it follows the RFC and
4876 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4877 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4878 names appear in log lines as well as in Received: header lines.
4880 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4881 extensions, and comparison tests, and to fix some bugs.
4883 14. Changes to the "personal" filter test:
4885 (1) The test was buggy in that it was just doing the equivalent of
4886 "contains" tests on header lines. For example, if a user's address was
4887 anne@some.where, the "personal" test would incorrectly be true for
4889 To: susanne@some.where
4891 This test is now done by extracting each address from the header in turn,
4892 and checking the entire address. Other tests that are part of "personal"
4893 are now done using regular expressions (for example, to check local parts
4894 of addresses in From: header lines).
4896 (2) The list of non-personal local parts in From: addresses has been
4897 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4898 taken from the Sieve specification recommendations.
4900 (3) If the message contains any header line starting with "List-" it is
4901 treated as non-personal.
4903 (4) The test for "circular" in the Subject: header line has been removed
4904 because it now seems ill-conceived.
4906 15. Minor typos in src/EDITME comments corrected.
4908 16. Installed latest exipick from John Jetmore.
4910 17. If headers_add on a router specified a text string that was too long for
4911 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4912 of string_sprintf() is now avoided.
4914 18. $message_body_size was not set (it was always zero) when running the DATA
4915 ACL and the local_scan() function.
4917 19. For the "mail" command in an Exim filter, no default was being set for
4918 the once_repeat time, causing a random time value to be used if "once" was
4919 specified. (If the value happened to be <= 0, no repeat happened.) The
4920 default is now 0s, meaning "never repeat". The "vacation" command was OK
4921 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4922 (I found it when inspecting the code).
4924 20. There is now an overall timeout for performing a callout verification. It
4925 defaults to 4 times the callout timeout, which applies to individual SMTP
4926 commands during the callout. The overall timeout applies when there is more
4927 than one host that can be tried. The timeout is checked before trying the
4928 next host. This prevents very long delays if there are a large number of
4929 hosts and all are timing out (e.g. when the network connections are timing
4930 out). The value of the overall timeout can be changed by specifying an
4931 additional sub-option for "callout", called "maxwait". For example:
4933 verify = sender/callout=5s,maxwait=20s
4935 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4936 to the end before writing, but this should make it even safer).
4938 22. Exim was forgetting that it had advertised PIPELINING for the second and
4939 subsequent messages on an SMTP connection. It was also not resetting its
4940 memory on STARTTLS and an internal HELO.
4942 23. When Exim logs an SMTP synchronization error within a session, it now
4943 records whether PIPELINING has been advertised or not.
4945 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4946 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4947 rather than long int.
4949 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4951 26. Added the never_mail option to autoreply.
4957 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4958 crash if the getsockname() call failed; this can happen if a connection is
4959 closed very soon after it is established. The problem was simply in the
4960 order in which certain operations were done, causing Exim to try to write
4961 to the SMTP stream before it had set up the file descriptor. The bug has
4962 been fixed by making things happen in the correct order.
4968 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4969 before the connection was closed, thus losing the rejection response.
4971 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4972 some early Solaris releases, but causes trouble in current releases where
4973 socklen_t is defined.
4975 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4978 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4981 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4982 that releasing the top of it at the end releases what was used for sub-
4983 expansions (unless the block got too big). However, discard this block if
4984 the first thing is a variable or header, so that we can use its block when
4985 it is dynamic (useful for very large $message_headers, for example).
4987 6. Lookups now cache *every* query, not just the most recent. A new, separate
4988 store pool is used for this. It can be recovered when all lookup caches are
4989 flushed. Lookups now release memory at the end of their result strings.
4990 This has involved some general refactoring of the lookup sources.
4992 7. Some code has been added to the store_xxx() functions to reduce the amount
4993 of flapping under certain conditions.
4995 8. log_incoming_interface used to affect only the <= reception log lines. Now
4996 it causes the local interface and port to be added to several more SMTP log
4997 lines, for example "SMTP connection from", and rejection lines.
4999 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5001 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5003 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5004 could be overwritten at the end of the current message (or the start of a
5005 new message if it was set in a HELO ACL). The value is now preserved for
5006 the duration of the SMTP connection.
5008 12. If a transport had a headers_rewrite setting, and a matching header line
5009 contained an unqualified address, that address was qualified, even if it
5010 did not match any rewriting rules. The underlying bug was that the values
5011 of the flags that permit the existence of unqualified sender and recipient
5012 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5013 non-local messages, and by -bnq for local messages) were not being
5014 preserved with the message after it was received.
5016 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5017 "next input=" as part of the text comprising the host identity instead of
5018 the correct text. The code was using the same buffer for two different
5019 strings. However, depending on which order the printing function evaluated
5020 its arguments, the bug did not always show up. Under Linux, for example, my
5021 test suite worked just fine.
5023 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5024 doesn't work with some older versions of Perl. It has been changed to "my",
5025 which in any case is probably the better facility to use.
5027 15. A really picky compiler found some instances of statements for creating
5028 error messages that either had too many or two few arguments for the format
5031 16. The size of the buffer for calls to the DNS resolver has been increased
5032 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5033 for addresses that have a lot of PTR records. This alleviates a problem; it
5034 does not fully solve it.
5036 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5037 buffer now truncates the list and logs the incident, which is the same
5038 action as happens when Exim is looking up a host name and its aliases.
5039 Previously in this situation something unpredictable would happen;
5040 sometimes it was "internal error: store_reset failed".
5042 18. If a server dropped the connection unexpectedly when an Exim client was
5043 using GnuTLS and trying to read a response, the client delivery process
5044 crashed while trying to generate an error log message.
5046 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5047 string, for example:
5049 warn message = H1: something\nH2: something
5051 the text was added as a single header line from Exim's point of view
5052 though it ended up OK in the delivered message. However, searching for the
5053 second and subsequent header lines using $h_h2: did not work. This has been
5054 fixed. Similarly, if a system filter added multiple headers in this way,
5055 the routers could not see them.
5057 20. Expanded the error message when iplsearch is called with an invalid key to
5058 suggest using net-iplsearch in a host list.
5060 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5061 ACLs are no longer actually imposed (and a message to that effect is
5064 22. If a "gecos" field in a passwd entry contained escaped characters, in
5065 particular, if it contained a \" sequence, Exim got it wrong when building
5066 a From: or a Sender: header from that name. A second bug also caused
5067 incorrect handling when an unquoted " was present following a character
5068 that needed quoting.
5070 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5071 was not being matched caselessly.
5073 24. Arranged for all hyphens in the exim.8 source to be escaped with
5076 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5077 properly. Recipient callout cache records were still being keyed to include
5078 the sender, even when use_sender was set false. This led to far more
5079 callouts that were necessary. The sender is no longer included in the key
5080 when use_sender is false.
5082 26. Added "control = submission" modifier to ACLs.
5084 27. Added the ${base62d: operator to decode base 62 numbers.
5086 28. dnsdb lookups can now access SRV records.
5088 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5089 the configuration file.
5091 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5092 (-v) mode. This makes the output for a verbose queue run more intelligible.
5094 31. Added a use_postmaster feature to recipient callouts.
5096 32. Added the $body_zerocount variable, containing the number of binary zero
5097 bytes in the message body.
5099 33. The time of last modification of the "new" subdirectory is now used as the
5100 "mailbox time last read" when there is a quota error for a maildir
5103 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5105 35. Added +ignore_unknown as a special item in host lists.
5107 36. Code for decoding IPv6 addresses in host lists is now included, even if
5108 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5109 address was recognized as an IP address, but was then not correctly decoded
5110 into binary, causing unexpected and incorrect effects when compared with
5117 1. Very minor rewording of debugging text in manualroute to say "list of
5118 hosts" instead of "hostlist".
5120 2. If verify=header_syntax was set, and a header line with an unqualified
5121 address (no domain) and a large number of spaces between the end of the
5122 name and the colon was received, the reception process suffered a buffer
5123 overflow, and (when I tested it) crashed. This was caused by some obsolete
5124 code that should have been removed. The fix is to remove it!
5126 3. When running in the test harness, delay a bit after writing a bounce
5127 message to get a bit more predictability in the log output.
5129 4. Added a call to search_tidyup() just before forking a reception process. In
5130 theory, someone could use a lookup in the expansion of smtp_accept_max_
5131 per_host which, without the tidyup, could leave open a database connection.
5133 5. Added the variables $recipient_data and $sender_data which get set from a
5134 lookup success in an ACL "recipients" or "senders" condition, or a router
5135 "senders" option, similar to $domain_data and $local_part_data.
5137 6. Moved the writing of debug_print from before to after the "senders" test
5140 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5141 problems for message scanning, either using a data ACL, or using
5142 local_scan() because the Received: header was not generated till after they
5143 were called (in order to set the time as the time of reception completion).
5144 I have revised the way this works. The header is now generated after the
5145 body is received, but before the ACL or local_scan() are called. After they
5146 are run, the timestamp in the header is updated.
5152 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5153 before starting a queue runner without re-exec. This happened only when
5154 deliver_drop_privilege was set or when the Exim user was set to root. The
5155 effect of the bug was that timeouts during subsequent deliveries caused
5156 crashes instead of being properly handled. The handler is now left at its
5157 default (and expected) setting.
5159 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5160 message, again when deliver_drop_privilege is set or Exim is run as root.
5161 The bug described in (1) was not present in this case, but the tidying up
5162 of the other signals was missing. I have made the two cases consistent.
5164 3. The ignore_target_hosts setting on a manualroute router was being ignored
5165 for hosts that were looked up using the /MX notation.
5167 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5170 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5171 operated on the sender address. After changing the $sender_address to <>
5172 for the sender address verify, Exim was re-instated it as the original
5173 (before rewriting) address, but remembering that it had rewritten it, so it
5174 wasn't rewriting it again. This bug also had the effect of breaking the
5175 sender address verification caching when the sender address was rewritten.
5177 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5178 This has been changed so that if the ip literal address matches
5179 ignore_target_hosts, the router declines.
5181 7. Added expansion conditions match_domain, match_address, and match_local_
5182 part (NOT match_host).
5184 8. The placeholder for the Received: header didn't have a length field set.
5186 9. Added code to Exim itself and to exim_lock to test for a specific race
5187 condition that could lead to file corruption when using MBX delivery. The
5188 issue is with the lockfile that is created in /tmp. If this file is removed
5189 after a process has opened it but before that process has acquired a lock,
5190 there is the potential for a second process to recreate the file and also
5191 acquire a lock. This could lead to two Exim processes writing to the file
5192 at the same time. The added code performs the same test as UW imapd; it
5193 checks after acquiring the lock that its file descriptor still refers to
5194 the same named file.
5196 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5197 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5200 11. Added the smtp_active_hostname option. If used, this will typically be made
5201 to depend on the incoming interface address. Because $interface_address is
5202 not set up until the daemon has forked a reception process, error responses
5203 that can happen earlier (such as "too many connections") no longer contain
5206 12. If an expansion in a condition on a "warn" statement fails because a lookup
5207 defers, the "warn" statement is abandoned, and the next ACL statement is
5208 processed. Previously this caused the whole ACL to be aborted.
5210 13. Added the iplsearch lookup type.
5212 14. Added ident_timeout as a log selector.
5214 15. Added tls_certificate_verified as a log selector.
5216 16. Added a global option tls_require_ciphers (compare the smtp transport
5217 option of the same name). This controls incoming TLS connections.
5219 17. I finally figured out how to make tls_require_ciphers do a similar thing
5220 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5221 before starting the TLS session.
5223 18. Tabs are now shown as \t in -bP output.
5225 19. If the log selector return_path_on_delivery was set, Exim crashed when
5226 bouncing a message because it had too many Received: header lines.
5228 20. If two routers both had headers_remove settings, and the first one included
5229 a superfluous trailing colon, the final name in the first list and the
5230 first name in the second list were incorrectly joined into one item (with a
5231 colon in the middle).
5237 1. Added -C and -D options to the exinext utility, mainly to make it easier
5238 to include in the automated testing, but these could be helpful when
5239 multiple configurations are in use.
5241 2. The exinext utility was not formatting the output nicely when there was
5242 an alternate port involved in the retry record key, nor when there was a
5243 message id as well (for retries that were specific to a specific message
5244 and a specific host). It was also confused by IPv6 addresses, because of
5245 the additional colons they contain. I have fixed the IPv4 problem, and
5246 patched it up to do a reasonable job for IPv6.
5248 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5249 delivery, the log line now contains "pipelined" if PIPELINING was used.
5251 4. An SMTP transport process used to panic and die if the bind() call to set
5252 an explicit outgoing interface failed. This has been changed; it is now
5253 treated in the same way as a connect() failure.
5255 5. A reference to $sender_host_name in the part of a conditional expansion
5256 that was being skipped was still causing a DNS lookup. This no longer
5259 6. The def: expansion condition was not recognizing references to header lines
5260 that used bh_ and bheader_.
5262 7. Added the _cache feature to named lists.
5264 8. The code for checking quota_filecount in the appendfile transport was
5265 allowing one more file than it should have been.
5267 9. For compatibility with Sendmail, the command line option
5275 and sets the incoming protocol and host name (for trusted callers). The
5276 host name and its colon can be omitted when only the protocol is to be set.
5277 Note the Exim already has two private options, -pd and -ps, that refer to
5278 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5279 "s", but I don't think that's a major issue.
5281 10. A number of refactoring changes to the code, none of which should affect
5284 (a) The number of logging options was getting close to filling up the
5285 32-bit word that was used as a bit map. I have split them into two classes:
5286 those that are passed in the argument to log_write(), and those that are
5287 only ever tested independently outside of that function. These are now in
5288 separate 32-bit words, so there is plenty of room for expansion again.
5289 There is no change in the user interface or the logging behaviour.
5291 (b) When building, for example, log lines, the code previously used a
5292 macro that called string_cat() twice, in order to add two strings. This is
5293 not really sufficiently general. Furthermore, there was one instance where
5294 it was actually wrong because one of the argument was used twice, and in
5295 one call a function was used. (As it happened, calling the function twice
5296 did not affect the overall behaviour.) The macro has been replaced by a
5297 function that can join an arbitrary number of extra strings onto a growing
5300 (c) The code for expansion conditions now uses a table and a binary chop
5301 instead of a serial search (which was left over from when there were very
5302 few conditions). Also, it now recognizes conditions like "pam" even when
5303 the relevant support is not compiled in: a suitably worded error message is
5304 given if an attempt is made to use such a condition.
5306 11. Added ${time_interval:xxxxx}.
5308 12. A bug was causing one of the ddress fields not to be passed back correctly
5309 from remote delivery subprocesses. The field in question was not being
5310 subsequently used, so this caused to problems in practice.
5312 13. Added new log selectors queue_time and deliver_time.
5314 14. Might have fixed a bug in maildirsizefile handling that threw up
5315 "unexpected character" debug warnings, and recalculated the data
5316 unnecessarily. In any case, I expanded the warning message to give more
5319 15. Added the message "Restricted characters in address" to the statements in
5320 the default ACL that block characters like @ and % in local parts.
5322 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5323 Three changes have been made:
5325 (a) There was a serious bug; a negative response to MAIL caused the whole
5326 recipient domain to be cached as invalid, thereby blocking all messages
5327 to all local parts at the same domain, from all senders. This bug has
5328 been fixed. The domain is no longer cached after a negative response to
5329 MAIL if the sender used is not empty.
5331 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5334 (c) A new callout option, "use_sender" has been added for people who want
5335 the modified behaviour.
5341 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5344 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5345 indeed breaks things for older releases.
5347 3. Added additional logging to the case where there is a problem reading data
5348 from a filter that is running in a subprocess using a pipe, in order to
5349 try to track down a specific problem.
5351 4. Testing facility fudge: when running in the test harness and attempting
5352 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5353 getting "No route to host". Convert this to a timeout.
5355 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5358 6. Some OS don't have socklen_t but use size_t instead. This affects the
5359 fifth argument of getsockopt() amongst other things. This is now
5360 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5361 can be set for individual OS. I have set it for SunOS5, OSF1, and
5362 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5363 some earlier ones do not.
5365 7. Change 4.30/15 was not doing the test caselessly.
5367 8. The standard form for an IPv6 address literal was being rejected by address
5368 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5369 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5370 this, as well as the form without the "IPv6" on the front (but only when
5371 address literals are enabled, of course).
5373 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5375 10. Exim crashed if a message with an empty sender address specified by -f
5376 encountered a router with an errors_to setting. This could be provoked only
5377 by a command such as
5381 where an empty string was supplied; "<>" did not hit this bug.
5383 11. Installed PCRE release 4.5.
5385 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5386 remained set. It is now erased.
5388 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5389 times from message ids (which are base 36 rather than the normal 62).
5391 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5392 were being counted as actual protocol errors, and logged if the log
5393 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5394 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5395 and DATA following a set of rejected RCPTs do not count as protocol errors.
5396 In other words, Exim assumes they were pipelined, though this may not
5397 actually be the case. Of course, in all cases the client gets an
5398 appropriate error code.
5400 15. If a lookup fails in an ACL condition, a message about the failure may
5401 be available; it is used if testing the ACL cannot continue, because most
5402 such messages specify what the cause of the deferral is. However, some
5403 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5404 that caused an old message to be retained and used if a later statement
5405 caused a defer, replacing the real cause of the deferral.
5407 16. If an IP address had so many PTR records that the DNS lookup buffer
5408 was not large enough to hold them, Exim could crash while trying to process
5409 the truncated data. It now detects and logs this case.
5411 17. Further to 4.21/58, another change has been made: if (and only if) the
5412 first line of a message (the first header line) ends with CRLF, a bare LF
5413 in a subsequent header line has a space inserted after it, so as not to
5414 terminate the header.
5416 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5417 unnecessarily, used atoi() instead of strtol(), and didn't check the
5418 termination when getting file sizes from file names by regex.
5420 19. Completely re-implemented the support for maildirsize files, in the light
5421 of a number of problems with the previous contributed implementation
5422 (4.30/29). In particular:
5424 . If the quota is zero, the maildirsize file is maintained, but no quota is
5427 . If the maildir directory does not exist, it is created before any attempt
5428 to write a maildirsize file.
5430 . The quota value in the file is just a cache; if the quota is changed in
5431 the transport, the new value overrides.
5433 . A regular expression is available for excluding directories from the
5436 20. The autoreply transport checks the characters in options that define the
5437 message's headers; it allows continued headers, but it was checking with
5438 isspace() after an embedded newline instead of explicitly looking for a
5441 21. If all the "regular" hosts to which an address was routed had passed their
5442 expiry times, and had not reached their retry times, the address was
5443 bounced, even if fallback hosts were defined. Now Exim should go on to try
5446 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5447 equivalent code in the SMTP transport. Some hosts send humungous responses
5448 to HELO/EHLO, more than 1024 it seems.
5450 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5451 gives compiler warnings in some environments. I've now done it "properly",
5454 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5455 (because of problems with the built-in one) was declared to return uschar *
5456 instead of char *, causing compiler failure.
5458 25. Fixed a file descriptor leak when processing alias/forward files.
5460 26. Fixed a minor format string issue in dbfn.c.
5462 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5464 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5465 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5468 29. When a sender address is verified, it is cached, to save repeating the test
5469 when there is more than one recipient in a message. However, when the
5470 verification involves a callout, it is possible for different callout
5471 options to be set for different recipients. It is too complicated to keep
5472 track of this in the cache, so now Exim always runs a verification when a
5473 callout is required, relying on the callout cache for the optimization.
5474 The overhead is duplication of the address routing, but this should not be
5477 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5478 to be verified with callout=postmaster, and the main callout worked but the
5479 postmaster check failed, the verification correctly failed. However, if a
5480 subsequent RCPT command asked for sender verification *without* the
5481 postmaster check, incorrect caching caused this verification also to fail,
5484 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5485 it was not caching the DNS options (qualify_single, search_parents) that
5486 were used when the lookup failed. A subsequent lookup with different
5487 options therefore always gave the same answer, though there were cases
5488 where it should not have. (Example: a "domains = !$mx_any" option on a
5489 dnslookup router: the "domains" option is always processed without any
5490 widening, but the router might have qualify_single set.) Now Exim uses the
5491 cached value only when the same options are set.
5493 32. Added John Jetmore's "exipick" utility to the distribution.
5495 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5496 than a timeout (e.g. a certificate is required, and is not provided), an
5497 Exim server now closes the connection immediately. Previously it waited for
5498 the client to close - but if the client is SSL, it seems that they each
5499 wait for each other, leading to a delay before one of them times out.
5501 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5502 maintained 0.8.x compatibility because I don't think many are using it, and
5503 it is clearly obsolete.
5505 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5508 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5509 client certificate was expired. A simple patch fixes this, though I don't
5510 understand the full logic of why the verify callback is called multiple
5513 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5514 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5515 which causes problems with some clients (such as the Certicom SSL Plus
5516 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5517 disables the coutermeasure allowing Eudora to connect."
5519 38. Exim was not checking that a write() to a log file succeeded. This could
5520 lead to Bad Things if a log got too big, in particular if it hit a file
5521 size limit. Exim now panics and dies if it cannot write to a log file, just
5522 as it does if it cannot open a log file.
5524 39. Modified OS/Makefile-Linux so that it now contains
5526 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5528 The two -D definitions ensure that Exim is compiled with large file
5529 support, which makes it possible to handle log files that are bigger than
5532 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5533 instance) a domain was checked against a named list that involved a lookup,
5534 causing $domain_data to be set, then another domain was checked against the
5535 same list, then the first domain was re-checked, the value of $domain_data
5536 after the final check could be wrong. In particular, if the second check
5537 failed, it could be set empty. This bug probably also applied to
5540 41. The strip_trailing_dot option was not being applied to the address given
5541 with the -f command-line option.
5543 42. The code for reading a message's header from the spool was incrementing
5544 $received_count, but never initializing it. This meant that the value was
5545 incorrect (doubled) while delivering a message in the same process in which
5546 it was received. In the most common configuration of Exim, this never
5547 happens - a fresh exec is done - but it can happen when
5548 deliver_drop_privilege is set.
5550 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5551 it now includes up to 150 characters of the unexpected data in the log
5554 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5555 and building data strings. The size of both of these buffers was 10 000
5556 bytes - far larger than anybody would *ever* want, thought I. Needless to
5557 say, somebody hit the limit. I have increased the maximum line length to
5558 20 000 and the maximum data length of concatenated lines to 100 000. I have
5559 also fixed two bugs, because there was no checking on these buffers. Tsk,
5560 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5561 buffer is too small.
5563 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5564 lsearch lookups. Now it does.
5566 46. When parsing a route_list item in a manualroute router, a fixed-length
5567 buffer was used for the list of hosts. I made this 1024 bytes long,
5568 thinking that nobody would ever have a list of hosts that long. Wrong.
5569 Somebody had a whole pile of complicated expansion conditions, and the
5570 string was silently truncated, leading to an expansion error. It turns out
5571 that it is easier to change to an unlimited length (owing to other changes
5572 that have happened since this code was originally written) than to build
5573 structure for giving a limitation error. The length of the item that
5574 expands into the list of hosts is now unlimited.
5576 47. The lsearch lookup could not handle data where the length of text line was
5577 more than 4095 characters. Such lines were truncated, leading to shortened
5578 data being returned. It should now handle lines of any length.
5580 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5581 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5584 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5585 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5586 no longer happens in comment lines. A list of replacements is now placed
5587 at the head of all of the source files, except those whose only change is
5588 to replace PERL_COMMAND in the very first #! line.
5590 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5591 messages on the queue, with a bottom-up merge sort, using code contributed
5592 by Michael Haardt. This should make operations like -bp somewhat faster on
5593 large queues. It won't affect queue runners, except when queue_run_in_order
5596 51. Installed eximstats 1.31 in the distribution.
5598 52. Added support for SRV lookups to the dnslookup router.
5600 53. If an ACL referred to $message_body or $message_body_end, the value was not
5601 reset for any messages that followed in the same SMTP session.
5603 54. The store-handling optimization for building very long strings was not
5604 differentiating between the different store pools. I don't think this
5605 actually made any difference in practice, but I've tidied it.
5607 55. While running the routers to verify a sender address, $sender_address
5608 was still set to the sender address. This is wrong, because when routing to
5609 send a bounce to the sender, it would be empty. Therefore, I have changed
5610 it so that, while verifying a sender address, $sender_address is set to <>.
5611 (There is no change to what happens when verifying a recipient address.)
5613 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5614 target A or AAAA records (if not already returned) without resetting the
5615 qualify_single or search_parents options of the DNS resolver. These are
5616 inappropriate in this case because the targets of MX and SRV records must
5617 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5618 target that, when qualified, matched something in the local domain. These
5619 two options are now turned off when doing these lookups.
5621 57. It seems that at least some releases of Reiserfs (which does not have the
5622 concept of a fixed number of inodes) returns zero and not -1 for the
5623 number of available inodes. This interacted badly with check_spool_inodes,
5624 which assumed that -1 was the "no such thing" setting. What I have done is
5625 to check that the total number of inodes is greater than zero before doing
5626 the test of how many are available.
5628 58. When a "warn" ACL statement has a log_message modifier, the message is
5629 remembered, and not repeated. This is to avoid a lot of repetition when a
5630 message has many recipients that cause the same warning to be written.
5631 However, Exim was preserving the list of already written lines for an
5632 entire SMTP session, which doesn't seem right. The memory is now reset if a
5633 new message is started.
5635 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5636 debugging output unless log_rewrite was also set.
5638 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5639 of (int)(handle) when we know that handle contains (void *)(-1).
5641 61. The Exim daemon panic-logs an error return when it closes the incoming
5642 connection. However "connection reset by peer" seems to be common, and
5643 isn't really an error worthy of noting specially, so that particular error
5646 62. When Exim is trying to find all the local interfaces, it used to panic and
5647 die if the ioctl to get the interface flags failed. However, it seems that
5648 on at least one OS (Solaris 9) it is possible to have an interface that is
5649 included in the list of interfaces, but for which you get a failure error
5650 for this call. This happens when the interface is not "plumbed" into a
5651 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5652 failure of the "get flags" call assumes that the interface is down.
5654 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5655 makes life easier for people who are doing arithmetic on fields extracted
5656 from dates, where you often get leading zeros that should not be
5657 interpreted as octal.
5659 64. Added qualify_domain to the redirect router, to override the global
5662 65. If a pathologically long header line contained very many addresses (the
5663 report of this problem mentioned 10 000) and each of them was rewritten,
5664 Exim could use up a very large amount of memory. (It kept on making new
5665 copies of the header line as it rewrote, and never released the old ones.)
5666 At the expense of a bit more processing, the header rewriting function has
5667 been changed so that it no longer eats memory in this way.
5669 66. The generation of the Received: header has been moved from the time that a
5670 message starts to be received, to the time that it finishes. The timestamp
5671 in the Received: header should now be very close to that of the <= log
5672 line. There are two side-effects of this change:
5674 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5675 logged header lines no longer include the local Received: line, because
5676 it has not yet been created. The same applies to a copy of the message
5677 that is returned to a non-SMTP sender when a message is rejected.
5679 (b) When a filter file is tested using -bf, no additional Received: header
5680 is added to the test message. After some thought, I decided that this
5683 This change does not affect the value of $received_for. It is still set
5684 after address rewriting, but before local_scan() is called.
5686 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5688 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5689 gave an unhelpful panic error message, and a defer error. I have managed to
5690 change this behaviour so that it now rejects any supplied certificate,
5691 which seems right, as the list of acceptable certificates is empty.
5693 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5694 gave an unhelpful defer error. I have not managed to make this reject any
5695 supplied certificates, but the error message it gives is "no certificate
5696 supplied", which is not helpful.
5698 70. exigrep's output now also includes lines that are not associated with any
5699 message, but which match the given pattern. Implemented by a patch from
5700 Martin Sluka, which also tidied up the Perl a bit.
5702 71. Recipient callout verification, like sender verification, was using <> in
5703 the MAIL FROM command. This isn't really the right thing, since the actual
5704 sender may affect whether the remote host accepts the recipient or not. I
5705 have changed it to use the actual sender in the callout; this means that
5706 the cache record is now keyed on a recipient/sender pair, not just the
5707 recipient address. There doesn't seem to be a real danger of callout loops,
5708 since a callout by the remote host to check the sender would use <>.
5709 [SEE ABOVE: changed after hitting problems.]
5711 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5712 temporary errors. However, in the case of such a code being given after
5713 the end of a data transmission (i.e. after ".") Exim was failing to write
5714 a retry record for the message. (Yes, there was some broken host that was
5715 actually sending 8xx at this point.)
5717 73. An unknown lookup type in a host list could cause Exim to panic-die when
5718 the list was checked. (An example that provoked this was putting <; in the
5719 middle of a list instead of at the start.) If this happened during a DATA
5720 ACL check, a -D file could be left lying around. This kind of configuration
5721 error no longer causes Exim to die; instead it causes a defer error. The
5722 incident is still logged to the main and panic logs.
5724 74. Buglet left over from Exim 3 conversion. The message "too many messages
5725 in one connection" was written to the rejectlog but not the mainlog, except
5726 when address rewriting (yes!) was being logged.
5728 75. Added write_rejectlog option.
5730 76. When a system filter was run not as root (that is, when system_filter_user
5731 was set), the values of the $n variables were not being returned to the
5732 main process; thus, they were not subsequently available in the $sn
5735 77. Added +return_path_on_delivery log selector.
5737 78. A connection timeout was being treated differently from recipients deferred
5738 when testing hosts_max_try with a message that was older than the host's
5739 retry timeout. (The host should not be counted, thus allowing all hosts to
5740 be tried at least once before bouncing.) This may have been the cause of an
5741 occasionally reported bug whereby a message would remain on the queue
5742 longer than the retry timeout, but would be bounced if a delivery was
5743 forced. I say "may" because I never totally pinned down the problem;
5744 setting up timeout/retry tests is difficult. See also the next item.
5746 79. The ultimate address timeout was not being applied to errors that involved
5747 a combination of host plus message (for example, a timeout on a MAIL
5748 command). When an address resolved to a number of possible hosts, and they
5749 were not all tried for each delivery (e.g. because of hosts_max_try), a
5750 message could remain on the queue longer than the retry timeout.
5752 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5755 81. Fixed an obscure SMTP outgoing bug which required at least the following
5756 conditions: (a) there was another message waiting for the same server;
5757 (b) the server returned 5xx to all RCPT commands in the first message so
5758 that the message was not completed; (c) the server dropped the connection
5759 or gave a negative response to the RSET that Exim sends to abort the
5760 transaction. The observed case was a dropped connection after DATA that had
5761 been sent in pipelining mode. That is, the server had advertised PIPELINING
5762 but was not implementing it correctly. The effect of the bug was incorrect
5763 behaviour, such as trying another host, and this could lead to a crash.
5769 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5770 and daemon.c were passed as pointers to ints; they should have been
5771 pointers to socklen_t variables (which are typically unsigned ints).
5773 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5776 3. Fixed a really odd bug that affected only the testing scheme; patching a
5777 certain fixed string in the binary changed the value of another string that
5778 happened to be identical to the end of the original first string.
5780 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5781 name", it returns that address as the IP address. On some operating
5782 systems (e.g. Solaris), it also passes back the IP address string as the
5783 "host name". However, on others (e.g. Linux), it passes back an empty
5784 string. Exim wasn't checking for this, and was changing the host name to an
5785 empty string, assuming it had been canonicalized.
5787 5. Although rare, it is permitted to have more than one PTR record for a given
5788 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5789 all the names associated with an address, because they do in Solaris.
5790 However, it seems that they do not in Linux for data that comes from the
5791 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5792 I found this out when I moved to a new Linux workstation and tried to run
5793 the Exim test suite.
5795 To get round this problem I have changed the code so that it now does its
5796 own call to the DNS to look up PTR records when searching for a host name.
5797 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5798 addresses that are only in /etc/hosts are still found.
5800 This behaviour is, however, controlled by an option called host_lookup_
5801 order, which defaults to "bydns:byaddr". If people want to use the other
5802 order, or indeed, just use one or the other means of lookup, they can
5803 specify it in this variable.
5805 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5806 some operating systems, this comes back from gethostbyaddr() as an empty
5807 string, and this is what Exim used to test for. However, it seems that in
5808 other systems, "." is yielded. Exim now tests for this case too.
5810 7. The values of check_spool_space and check_log_space are now held internally
5811 as a number of kilobytes instead of an absolute number of bytes. If a
5812 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5813 kilobyte. This means that much larger values can be stored.
5815 8. Exim monitor: an attempt to get the action menu when not actually pointing
5816 at a message produces an empty menu entitled "No message selected". This
5817 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5818 no entries in it ("Shell widget menu has zero width and/or height"). So I
5819 have added a single, blank menu entry in this case.
5821 9. Added ${quote_local_part.
5823 10. MIME decoding is now applied to the contents of Subject: header lines when
5826 11. Now that a reference to $sender_host_address automatically causes a reverse
5827 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5828 host lookup before query-style lookups in lists that might use this
5829 variable. This has therefore been abolished, and the "net-" prefix is no
5830 longer necessary for query-style lookups.
5832 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5833 have been a typo for LFLAGS, so it has been changed.
5835 13. The install script calls Exim with "-C /dev/null" in order to find the
5836 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5837 to be output. However, since Exim outputs its version number before the
5838 error, it didn't break the script. It just looked ugly. I fixed this by
5839 always allowing "-C /dev/null" if the caller is root.
5841 14. Ignore overlarge ACL variable number when reading spool file - insurance
5842 against a later release with more variables having written the file.
5844 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5845 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5846 this, as well as the form without the "IPv6" on the front.
5848 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5849 OS/Makefile-Darwin file.
5851 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5852 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5854 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5855 variable. However, some OS use ints for inodes. I've added cast to long int
5856 to get rid of the compiler warning.
5858 19. I had forgotten to lock out "/../" in configuration file names when
5859 ALT_CONFIG_PREFIX was set.
5861 20. Routers used for verification do not need to specify transports. However,
5862 if such a router generated a host list, and callout was configured, Exim
5863 crashed, because it could not find a port number from the (non-existent)
5864 transport. It now assumes port 25 in this circumstance.
5866 21. Added the -t option to exigrep.
5868 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5869 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5870 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5871 them is set, LOOKUP_LSEARCH is forced.
5873 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5875 24. Added sender and host information to the "rejected by local_scan()" log
5876 line; previously there was no indication of these.
5878 25. Added .include_if_exists.
5880 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5881 Linux. It turns out that not all file systems support this. Apparently some
5882 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5883 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5884 not supported on the file descriptor, is now ignored when Exim is trying to
5885 sync a directory. This applies only to Linux.
5887 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5889 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5890 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5893 29. Added support for maildirsize files from supplied patch (modified a bit).
5895 30. The use of :fail: followed by an empty string could lead Exim to respond to
5896 sender verification failures with (e.g.):
5898 550 Verification failed for <xxx>
5899 550 Sender verify failed
5901 where the first response line was missing the '-' that indicates it is not
5902 the final line of the response.
5904 31. The loop for finding the name of the user that called Exim had a hardwired
5905 limit of 10; it now uses the value of finduser_retries, which is used for
5906 all other user lookups.
5908 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5911 33. Exim was neglecting to zero errno before one call of strtol() when
5912 expanding a string and expecting an integer value. On some systems this
5913 resulted in spurious "integer overflow" errors. Also, it was casting the
5914 result into an int without checking.
5916 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5917 did not work. The code looks as if it has *never* worked, though it appears
5918 to have been documented since at least release 1.62. I have made it work.
5920 35. The "timeout_DNS" error in retry rules, also documented since at least
5921 1.62, also never worked. As it isn't clear exactly what this means, and
5922 clearly it isn't a major issue, I have abolished the feature by treating it
5923 as "timeout", and writing a warning to the main and panic logs.
5925 36. The display of retry rules for -brt wasn't always showing the error code
5928 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5929 timeout_connect_A, timeout_connect_MX.
5931 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5932 to the empty sender.
5934 39. The daemon was not analysing the content of -oX till after it had closed
5935 stderr and disconnected from the controlling terminal. This meant that any
5936 syntax errors were only noted on the panic log, and the return code from
5937 the command was 0. By re-arranging the code a little, I've made the
5938 decoding happen first, so such errors now appear on stderr, and the return
5939 code is 1. However, the actual setting up of the sockets still happens in
5940 the disconnected process, so errors there are still only recorded on the
5943 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5944 connections (as happens on some IP stacks) was logged at start up time as
5945 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5946 it from "IPv6 and IPv4", which means that two separate sockets are being
5949 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5950 says whether AF_INET or AF_INET6 was passed as an argument.
5952 42. Exiwhat output was messed up when time zones were included in log
5955 43. Exiwhat now gives more information about the daemon's listening ports,
5956 and whether -tls-on-connect was used.
5958 44. The "port" option of the smtp transport is now expanded.
5960 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5961 silently ignored. Now an error message is written to the main and panic
5964 46. There's a new ACL modifier called "logwrite" which writes to a log file
5965 as soon as it is encountered.
5967 47. Added $local_user_uid and $local_user_gid at routing time.
5969 48. Exim crashed when trying to verify a sender address that was being
5972 49. Exim was recognizing only a space character after ".include". It now also
5973 recognizes a tab character.
5975 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5976 extracting the relevant information from the specification. The man page no
5977 longer contains scrambled data for the -d option, and I've added a section
5978 at the front about calling Exim under different names.
5980 51. Added "extra_headers" argument to the "mail" command in filter files.
5982 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5985 53. Installed eximstats 1.29.
5987 54. Added transport_filter_timeout as a generic transport option.
5989 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5990 Cc: header lines. This was required by RFC 822, but it not required by RFC
5993 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5994 incoming messages that did not have them. Now it does so only if the
5995 message originates locally, that is, if there is no associated remote host
5996 address. When Resent- header lines are present, this applies to the Resent-
5997 lines rather than the non-Resent- lines.
5999 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6000 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6002 58. Messages for configuration errors now include the name of the main
6003 configuration file - useful now that there may be more than one file in a
6004 list (.included file names were always shown).
6006 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6007 for those rare installations that do not start the daemon as root or run it
6008 setuid root. I've cut out the call to initgroups() if the daemon is not
6011 60. The Exim user and group can now be bound into the binary as text strings
6012 that are looked up at the start of Exim's processing.
6014 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6016 62. Added $mailstore_basename variable.
6018 63. Installed patch to sieve.c from Michael Haardt.
6020 64. When Exim failed to open the panic log after failing to open the main log,
6021 the original message it was trying to log was written to stderr and debug
6022 output, but if they were not available (the usual case in production), it
6023 was lost. Now it is written to syslog before the two lines that record the
6024 failures to open the logs.
6026 65. Users' Exim filters run in subprocesses under the user's uid. It is
6027 possible for a "deliver" command or an alias in a "personal" command to
6028 provoke an address rewrite. If logging of address rewriting is configured,
6029 this fails because the process is not running as root or exim. There may be
6030 a better way of dealing with this, but for the moment (because 4.30 needs
6031 to be released), I have disabled address rewrite logging when running a
6032 filter in a non-root, non-exim process.
6038 1. The buildconfig auxiliary program wasn't quoting the value set for
6039 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6040 not defined. This bug was masked in 4.22 by the effect that was fixed in
6043 2. Some messages that were rejected after a message id was allocated were
6044 shown as "incomplete" by exigrep. It no longer does this for messages that
6045 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6047 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6048 have allow_domain_literals set, the ID did not get logged in the <= line.
6049 Domain literals are now always recognized in Message-ID: header lines.
6051 4. The first argument for a ${extract expansion item is the key name or field
6052 number. Leading and trailing spaces in this item were not being ignored,
6053 causing some misleading effects.
6055 5. When deliver_drop_privilege was set, single queue runner processes started
6056 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6057 same command in the process it spins off) were not dropping privilege.
6059 6. When the daemon running as "exim" started a queue runner, it always
6060 re-executed Exim in the spun-off process. This is a waste of effort when
6061 deliver_drop_privilege is set. The new process now just calls the
6062 queue-runner function directly.
6068 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6071 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6072 ignored. Though the use of -oP was forcing the writing of a pid file, it
6073 was always written to the default place.
6075 3. If the message "no IP address found for host xxxx" is generated during
6076 incoming verification, it is now followed by identification of the incoming
6077 connection (so you can more easily find what provoked it).
6079 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6081 5. Added some features to "harden" Exim a bit more against certain attacks:
6083 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6084 be put in Local/Makefile. This is like the never_users runtime option,
6085 but it cannot be overridden. The default setting is "root".
6087 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6088 prefix string with which any file named in a -C command line option
6091 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6092 is retained for -C and -D only if the caller of Exim is root. Without
6093 it, the exim user may also use -C and -D and retain privilege.
6095 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6096 command line option is disabled.
6098 6. Macro names set by the -D option must start with an upper case letter, just
6099 like macro names defined in the configuration file.
6101 7. Added "dereference=" facility to LDAP.
6103 8. Two instances of the typo "uknown" in the source files are fixed.
6105 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6106 the Configure-Makefile script screwed up while processing it.
6108 10. Incorporated PCRE 4.4.
6110 11. The SMTP synchronization check was not operating right at the start of an
6111 SMTP session. For example, it could not catch a HELO sent before the client
6112 waited for the greeting. There is now a check for outstanding input at the
6113 point when the greeting is written. Because of the duplex, asynchronous
6114 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6115 way, but not yet received, when the check is performed.
6117 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6118 on TCP/IP sockets, because this apparently causes some broken clients to
6121 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6122 unchanged) from the Cygwin maintainer.
6124 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6125 was supported instead of testing for mbx. Effectively a typo.
6127 15. The spa authenticator server code was not checking that the input it
6128 received was valid base64.
6130 16. The debug output line for the "set" modifier in ACLs was not showing the
6131 name of the variable that was being set.
6133 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6135 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6136 lookup on its own. Something else was needed to trigger the lookup. For
6137 example, a match in host_lookup or the need for a host name in a host list.
6138 Now, if $sender_host_name is referenced and the host name has not yet been
6139 looked up, a lookup is performed. If the lookup fails, the variable remains
6140 empty, and $host_lookup_failed is set to "1".
6142 19. Added "eqi" as a case-independent comparison operator.
6144 20. The saslauthd authentication condition could segfault if neither service
6145 nor realm was specified.
6147 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6148 error message that was logged was misleading, and incoming SMTP
6149 connections were dropped. The message is now more accurate, and temporary
6150 errors are given to SMTP connections.
6152 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6153 (except RSET) with a 421 error, until QUIT is received. However, it was
6154 failing to send a response to QUIT.
6156 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6157 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6158 ACL is now run after the helo_try_verify_hosts code.
6160 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6161 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6162 case-independent, so other case variants are also recognized). Apparently
6163 some systems use these upper case variants.
6165 25. If more than two messages were waiting for the same host, and a transport
6166 filter was specified for the transport, Exim sent two messages over the
6167 same TCP/IP connection, and then failed with "socket operation on non-
6168 socket" when it tried to send the third.
6170 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6172 27. The extern definition of crypt16() in expand.c was not being excluded when
6173 the OS had its own crypt16() function.
6175 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6176 as a preferred synonym for return_size_limit, both as an option and as an
6179 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6181 30. Changed the default configuration ACL to relax the local part checking rule
6182 for addresses that are not in any local domains. For these addresses,
6183 slashes and pipe symbols are allowed within local parts, but the sequence
6184 /../ is explicitly forbidden.
6186 31. SPA server authentication was not clearing the challenge buffer before
6189 32. log_message in a "warn" ACL statement was writing to the reject log as
6190 well as to the main log, which contradicts the documentation and doesn't
6191 seem right (because no rejection is happening). So I have stopped it.
6193 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6194 However, I am unable to do any testing of this.
6196 34. Fixed an infelicity in the appendfile transport. When checking directories
6197 for a mailbox, to see if any needed to be created, it was accidentally
6198 using path names with one or more superfluous leading slashes; tracing
6199 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6201 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6202 contents are added to the log line that is written for every discarded
6203 recipient. (Previously a log_message setting was ignored.)
6205 36. The ${quote: operator now quotes the string if it is empty.
6207 37. The install script runs exim in order to find its version number. If for
6208 some reason other than non-existence or emptiness, which it checks, it
6209 could not run './exim', it was installing it with an empty version number,
6210 i.e. as "exim-". This error state is now caught, and the installation is
6213 38. An argument was missing from the function that creates an error message
6214 when Exim fails to connect to the socket for saslauthd authentication.
6215 This could cause Exim to crash, or give a corrupted message.
6217 39. Added isip, isip4, and isip6 to ${if conditions.
6219 40. The ACL variables $acl_xx are now saved with the message, and can be
6220 accessed later in routers, transports, and filters.
6222 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6223 strings in the file are not string-expanded.
6225 42. If a MAIL command specified a SIZE value that was too large to fit into an
6226 int variable, the check against message_size_limit failed. Such values are
6227 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6228 day this will have to be increased, but I don't think I want to be around
6229 when emails are that large.
6236 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6237 iconv() is not standard in FreeBSD.
6239 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6240 IPv6 enabled. The observed symptom was a segmentation fault on return from
6241 the function os_common_find_running_interfaces() in src/os.c.
6243 3. In the check_special_case() function in daemon.c I had used "errno" as an
6244 argument name, which causes warnings on some systems. This was basically a
6245 typo, since it was named "eno" in the comments!
6247 4. The code that waits for the clock to tick (at a resolution of some fraction
6248 of a second) so as to ensure message-id uniqueness was always waiting for
6249 at least one whole tick, when it could have waited for less. [This is
6250 almost certainly not relevant at current processor speeds, where it is
6251 unlikely to ever wait at all. But we try to future-proof.]
6253 5. The function that sleeps for a time interval that includes fractions of a
6254 second contained a race. It did not block SIGALRM between setting the
6255 timer, and suspending (a couple of lines later). If the interval was short
6256 and the sigsuspend() was delayed until after it had expired, the suspension
6257 never ended. On busy systems this could lead to processes getting stuck for
6260 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6261 process, before it forks any delivery processes. The open lookup caching
6262 mechanism meant that the open file or database connection was passed into
6263 the delivery process. The problem was that delivery processes always tidy
6264 up cached lookup data. This could cause a problem for the next delivery
6265 process started by the queue runner, because the external queue runner
6266 process does not know about the closure. So the next delivery process
6267 still has data in the lookup cache. In the case of a file lookup, there was
6268 no problem because closing a file descriptor in a subprocess doesn't affect
6269 the parent. However, if the lookup was caching a connection to a database,
6270 the connection was closed, and the second delivery process was likely to
6271 see errors such as "PGSQL: query failed: server closed the connection
6272 unexpectedly". The problem has been fixed by closing all cached lookups
6273 in a queue runner before running a delivery process.
6275 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6276 seem to have the "const" qualifier which it has on other OS. I've
6279 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6280 *of the same type* (client or server) with the same public name that an
6281 error should be diagnosed.
6283 9. When Exim looked up a host name for an IP address, but failed to find the
6284 original IP address when looking up the host name (a safety check), it
6285 output the message "<ip address> does not match any IP for NULL", which was
6286 confusing, to say the least. The bug was that the host name should have
6287 appeared instead of "NULL".
6289 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6290 user that is built into the binary, and the -C or -D options is used, root
6291 privilege is dropped before the configuration file is read. In addition,
6292 logging is switched to stderr instead of the normal log files. If the
6293 configuration then re-defines the Exim user, the unprivileged environment
6294 is probably not what is expected, so Exim logs a panic warning message (but
6297 However, if deliver_drop_privilege is set, the unprivileged state may well
6298 be exactly what is intended, so the warning has been cut out in that case,
6299 and Exim is allowed to try to write to its normal log files.
6305 1. smtp_return_error_details was not giving details for temporary sender
6306 or receiver verification errors.
6308 2. Diagnose a configuration error if two authenticators have the same public
6311 3. Exim used not to create the message log file for a message until the first
6312 delivery attempt. This could be confusing when incoming messages were held
6313 for policy or load reasons. The message log file is now created at the time
6314 the message is received, and an initial "Received" line is written to it.
6316 4. The automatically generated man page for command line options had a minor
6317 bug that caused no ill effects; however, a more serious problem was that
6318 the procedure for building the man page automatically didn't always
6319 operate. Consequently, release 4.20 contains an out-of-date version. This
6320 shouldn't happen again.
6322 5. When building Exim with embedded Perl support, the script that builds the
6323 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6324 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6326 6. The freeze_tell option was not being used for messages that were frozen on
6327 arrival, either by an ACL or by local_scan().
6329 7. Added the smtp_incomplete_transaction log selector.
6331 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6332 was accepting AUTH without a new EHLO.
6334 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6335 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6338 10. Logging of TCP/IP connections (when configured) now happens in the main
6339 daemon process instead of the child process, so that the TCP/IP connection
6340 count is more accurate (but it can never be perfect).
6342 11. The use of "drop" in a nested ACL was not being handled correctly in the
6343 outer ACL. Now, if condition failure induced by the nested "drop" causes
6344 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6345 or "require"), the connection is dropped.
6347 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6348 that yield "discard" can now be used with an "accept" or a "discard" verb,
6349 but an error is generated for any others (because I can't see a useful way
6350 to define what should happen).
6352 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6353 are now processed in the same way as lines in the Exim configuration file.
6354 In particular, continuation lines are supported.
6356 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6358 15. Added -ti meaning -t -i.
6360 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6361 domains, and warn by logging if others are found.
6363 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6364 returned from GIFCONF when Exim is trying to find the list of interfaces on
6365 a host. The code in os.c has been modified to copy each ifreq to an aligned
6366 structure in all cases.
6368 Also, in some cases, the returned ifreq's were being copied to a 'struct
6369 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6370 means the last couple of bytes of an IPv6 address could be chopped if the
6371 ifreq contained only a normal sockaddr (14 bytes storage).
6373 18. Named domain lists were not supported in the hosts_treat_as_local option.
6374 An entry such as +xxxx was not recognized, and was treated as a literal
6377 19. Ensure that header lines added by a DATA ACL are included in the reject log
6378 if the ACL subsequently rejects the message.
6380 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6381 MD5 (which is deprecated).
6383 21. When testing a filter file using -bf, Exim was writing a message when it
6384 took the sender from a "From " line in the message, but it was not doing so
6385 when it took $return_path from a Return-Path: header line. It now does.
6387 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6388 with a valid header line field name (a series of printing characters
6389 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6391 23. Changed "disc" in the source to "disk" to conform to the documentation and
6392 the book and for uniformity.
6394 24. Ignore Sendmail's -Ooption=value command line item.
6396 25. When execve() failed while trying to run a command in a pipe transport,
6397 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6398 could be confused with a return value of 69 from the command itself. This
6399 has been changed to 127, the value the shell returns if it is asked to run
6400 a non-existent command. The wording for the related log line suggests a
6401 non-existent command as the problem.
6403 26. If received_header_text expands to an empty string, do not add a Received:
6404 header line to the message. (Well, it adds a token one on the spool, but
6405 marks it "old" so that it doesn't get used or transmitted.)
6407 27. Installed eximstats 1.28 (addition of -nt option).
6409 28. There was no check for failure on the call to getsockname() in the daemon
6410 code. This can fail if there is a shortage of resources on the system, with
6411 ENOMEM, for example. A temporary error is now given on failure.
6413 29. Contrary to the C standard, it seems that in some environments, the
6414 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6415 program. Exim now does this explicitly; it affects the formatting of
6416 timestamps using strftime().
6418 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6419 complaints. I've now initialized all the variables, to avoid this.
6421 32. Header lines added by a system filter were not being "seen" during
6422 transport-time rewrites.
6424 33. The info_callback() function passed to OpenSSL is set up with type void
6425 (*)(SSL *, int, int), as described somewhere. However, when calling the
6426 function (actually a macro) that sets it up, the type void(*)() is
6427 expected. I've put in a cast to prevent warnings from picky compilers.
6429 34. If a DNS black list lookup found a CNAME record, but there were no A
6430 records associated with the domain it pointed at, Exim crashed.
6432 35. If a DNS black list lookup returned more than one A record, Exim ignored
6433 all but the first. It now scans all returned addresses if a particular IP
6434 value is being sought. In this situation, the contents of the
6435 $dnslist_value variable are a list of all the addresses, separated by a
6438 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6439 to accept a host name and all its aliases if the forward lookup for any of
6440 them yielded the IP address of the incoming connection. Now it accepts only
6441 those names whose forward lookup yields the correct IP address. Any other
6442 names are discarded. This closes a loophole whereby a rogue DNS
6443 administrator could create reverse DNS records to break through a
6444 wildcarded host restriction in an ACL.
6446 37. If a user filter or a system filter that ran in a subprocess used any of
6447 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6448 the wrong values were passed to the pipe command ($thisaddress had the
6449 value of $0, $0 had the value of $1, etc). This bug was introduced by
6450 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6452 38. Improved the line breaking for long SMTP error messages from ACLs.
6453 Previously, if there was no break point between 40 and 75 characters, Exim
6454 left the rest of the message alone. Two changes have been made: (a) I've
6455 reduced the minimum length to 35 characters; (b) if it can't find a break
6456 point between 35 and 75 characters, it looks ahead and uses the first one
6457 that it finds. This may give the occasional overlong line, but at least the
6458 remaining text gets split now.
6460 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6461 file descriptors might be low, and that setting 1000 would always raise it.
6462 It turns out that in some environments, the limit is already over 1000 and
6463 that lowering it causes trouble. So now Exim takes care not to decrease it.
6465 40. When delivering a message, the value of $return_path is set to $sender_
6466 address at the start of routing (routers may change the value). By an
6467 oversight, this default was not being set up when an address was tested by
6468 -bt or -bv, which affected the outcome if any router or filter referred to
6471 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6472 log only when "message" or "log_message" are set. However, if one of the
6473 conditions was an address verification, or a call to a nested ACL, the
6474 messages generated by the underlying test were being passed through. This
6475 no longer happens. The underlying message is available in $acl_verify_
6476 message for both "message" and "log_message" expansions, so it can be
6477 passed through if needed.
6479 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6480 new expansion $bh_ to give the encoded byte string without charset
6481 translation. Translation happens only if iconv() is available; HAVE_ICONV
6482 indicates this at build time. HEADERS_CHARSET gives the charset to
6483 translate to; headers_charset can change it in the configuration, and
6484 "headers charset" can change it in an individual filter file.
6486 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6487 that creates RFC 2047 encoded "words" labels them as that charset instead
6488 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6489 expansion operator; (ii) when Exim creates a From: line for a local
6490 message; (iii) when a header line is rewritten to include a "phrase" part.
6492 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6493 buggy, causing it to skip the first lines of messages whose message ID
6494 ended in 'D'. This would not have bitten before Exim release 4.14, because
6495 message IDs were unlikely to end in 'D' before then. The effect was to have
6496 incorrect size information for certain domains.
6498 45. #include "config.h" was missing at the start of the crypt16.c module. This
6499 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6502 46. If there was a timeout during a "random" callout check, Exim treated it as
6503 a failure of the random address, and carried on sending RSET and the real
6504 address. If the delay was just some slowness somewhere, the response to the
6505 original RCPT would be taken as a response to RSET and so on, causing
6506 mayhem of various kinds.
6508 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6509 when I implemented it. It didn't allow for the fact that some option values
6510 may legitimately be negative (e.g. size_addition), and it didn't even do
6511 the right test for positive values.
6513 48. Domain names in DNS records are case-independent. Exim always looks them up
6514 in lower case. Some resolvers return domain names in exactly the case they
6515 appear in the zone file, that is, they may contain uppercase letters. Not
6516 all resolvers do this - some return always lower case. Exim was treating a
6517 change of case by a resolver as a change of domain, similar to a widening
6518 of a domain abbreviation. This triggered its re-routing code and so it was
6519 trying to route what was effectively the same domain again. This normally
6520 caused routing to fail (because the router wouldn't handle the domain
6521 twice). Now Exim checks for this case specially, and just changes the
6522 casing of the domain that it ultimately uses when it transmits the message
6525 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6528 50. If a filter generated a file delivery with a non-absolute name (possible if
6529 no home directory exists for the router), the forbid_file option was not
6532 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6533 the existing equality matching.
6535 52. Exim was using ints instead of ino_t variables in some places where it was
6536 dealing with inode numbers.
6538 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6539 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6540 called TMPDIR, and if it finds it is different, it changes its value.
6542 54. The smtp_printf() function is now made available to local_scan() so
6543 additional output lines can be written before returning. There is also an
6544 smtp_fflush() function to enable the detection of a dropped connection.
6545 The variables smtp_input and smtp_batched_input are exported to
6548 55. Changed the default runtime configuration: the message "Unknown user"
6549 has been removed from the ACL, and instead placed on the localuser router,
6550 using the cannot_route_message feature. This means that any verification
6551 failures that generate their own messages won't get overridden. Similarly,
6552 the "Unrouteable address" message that was in the ACL for unverifiable
6553 relay addresses has also been removed.
6555 56. Added hosts_avoid_esmtp to the smtp transport.
6557 57. The exicyclog script was not checking for the esoteric option
6558 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6559 will work only if exicyclog is run under the appropriate euid.
6561 58. Following a discussion on the list, the rules by which Exim recognises line
6562 endings on incoming messages have been changed. The -dropcr and drop_cr
6563 options are now no-ops, retained only for backwards compatibility. The
6564 following line terminators are recognized: LF CRLF CR. However, special
6565 processing applies to CR:
6567 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6568 nor a local message in the state where . is a terminator.
6570 (ii) If a bare CR is encountered in a header line, an extra space is added
6571 after the line terminator so as not to end the header. The reasoning
6572 behind this is that bare CRs in header lines are most likely either
6573 to be mistakes, or people trying to play silly games.
6575 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6576 was being incorrectly given as 18 bytes larger than it should have been.
6577 This is a VOB (very old bug).
6579 60. This may never have affected anything current, but just in case it has:
6580 When the local host is found other than at the start of a list of hosts,
6581 the local host, those with the same MX, and any that follow, are discarded.
6582 When the list in question was part of a longer list of hosts, the following
6583 hosts (not currently being processed) were also being discarded. This no
6584 longer happens. I'm not sure if this situation could ever has previously
6587 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6590 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6591 header. This is something that is recommended in a new Internet Draft, and
6592 is something that is documented as being done by Sendmail. There are two
6593 possible values. For messages generated by the autoreply transport, Exim
6596 Auto-Submitted: auto-replied
6598 whereas for all other generated messages (e.g. bounces) it adds
6600 Auto-Submitted: auto-generated
6602 63. The "personal" condition in filters now includes a test for the
6603 Auto-Submitted: header. If it contains the string "auto-" the message it
6604 not considered personal.
6606 64. Added rcpt_include_affixes as a generic transport option.
6608 65. Added queue_only_override (default true).
6610 66. Added the syslog_duplication option.
6612 67. If what should have been the first header line of a message consisted of
6613 a space followed by a colon, Exim was mis-interpreting it as a header line.
6614 It isn't of course - it is syntactically invalid and should therefore be
6615 treated as the start of the message body. The misbehaviour could have
6616 caused a number of strange effects, including loss of data in subsequent
6617 header lines, and spool format errors.
6619 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6620 client host had authenticated. This control can now be exercised by an ACL
6621 for more flexibility.
6623 69. By default, callouts do not happen when testing with -bh. There is now a
6624 variant, -bhc, which does actually run the callout code, including
6625 consulting and updating the callout cache.
6627 70. Added support for saslauthd authentication, courtesy of Alexander
6630 71. If statvfs() failed on the spool or log directories while checking their
6631 size for availability, Exim confusingly gave the error "space shortage".
6632 Furthermore, in debugging mode it crashed with a floating point exception.
6633 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6634 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6635 serious problem, Exim now writes to the main and panic logs when this
6636 happens, with details of the failure. It then refuses to accept the
6637 incoming message, giving the message "spool directory problem" or "log
6638 directory problem" with a 421 code for SMTP messages.
6640 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6641 0, 1, and 2 exist, because some OS complain for execs without them (see
6642 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6643 descriptors. However, the code omitted to check that the open succeeded,
6644 causing mysterious errors if for some reason the permissions on /dev/null
6645 got screwed. Now Exim writes a message to the main and panic logs, and
6646 bombs out if it can't open /dev/null.
6648 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6649 interact so that it is all more flexible. It is supposed to remain
6650 backwards compatible. Also added extra_local_interfaces.
6652 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6653 to bomb out with an assertion failure - to the client this appears as a
6654 connection drop. This problem occurs in the part of the code that was taken
6655 from the Samba project. Fortunately, the assertion is in a very simple
6656 function, so I have fixed this by reproducing the function inline in the
6657 one place where it is called, and arranging for authentication to fail
6658 instead of killing the process with assert().
6660 75. The SPA client code was not working when the server requested OEM rather
6661 than Unicode encoding.
6663 76. Added code to make require_files with a specific uid setting more usable in
6664 the case where statting the file as root fails - usually a non-root-mounted
6665 NFS file system. When this happens and the failure is EACCES, Exim now
6666 forks a subprocess and does the per-uid checking as the relevant uid.
6668 77. Added process_log_path.
6670 78. If log_file_path was not explicitly set, a setting of check_log_space or
6671 check_log_inodes was ignored.
6673 79. If a space check for the spool or log partitions fails, the incident is now
6674 logged. Of course, in the latter case the data may get lost...
6676 80. Added the %p formatting code to string_format() so that it can be used to
6677 print addresses in debug_print(). Adjusted all the address printing in the
6678 debugging in store.c to use %p rather than %d.
6680 81. There was a concern that a line of code in smtp_in.c could overflow a
6681 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6682 initially expressed, the concern was not well-founded, because trailing
6683 spaces are removed early. However, if the trailing spaces were followed by
6684 a NULL, they did not get removed, so the overflow was possible. Two fixes
6687 (a) I re-wrote the offending code in a cleaner fashion.
6688 (b) If an incoming SMTP command contains a NULL character, it is rejected
6691 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6692 runs initgroups(), so that if the Exim user is in any additional groups,
6693 they will be used during message reception.
6699 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob