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.
35 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
36 supports it and a size is available (ie. the sending peer gave us one).
38 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
39 years of being deprecated). The replacements are the ACLs
40 acl_smtp_mime and acl_not_smtp_mime.
42 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
43 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
44 or even an in-clear connection were permitted. Now, if the host lookup
45 was dnssec and dane was requested then the host is only used if the
46 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
47 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
48 if one fails this test.
49 This means that a poorly-configured remote DNS will make it incommunicado;
50 but it protects against a DNS-interception attack on it.
52 JH/04 Bug 1810: make continued-use of an open smtp transport connection
53 non-noisy when a race steals the message being considered.
55 JH/05 If main configuration option tls_certificate is unset, generate a
56 self-signed certificate for inbound TLS connections.
58 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
59 in rewrites and routers.
61 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
62 and logged a warning sing 4.83; now they are a configuration file error.
64 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
65 (lacking @domain). Apply the same qualification processing as RCPT.
67 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
69 JH/10 Support ${sha256:} applied to a string (as well as the previous
72 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
73 a cutthrough deliver is pending, as we always want to make a connection.
74 This also avoids re-routing the message when later placing the cutthrough
75 connection after a verify cache hit.
76 Do not update it with the verify result either.
78 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
79 when routing results in more than one destination address.
81 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
82 signing (which inhibits the cutthrough capability). Previously only
83 the presence of an option was tested; now an expansion evaluating as
84 empty is permissible (obviously it should depend only on data available
85 when the cutthrough connection is made).
87 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
88 the relevant preceding SMTP command did not note the pipelining mode.
90 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
91 Previously they were not counted.
93 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
94 as one having no matching records. Previously we deferred the message
95 that needed the lookup.
97 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
98 distinguished as "(=".
100 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
101 for missing MX records. Previously it only worked for missing A records.
103 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
105 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
106 after the data-go-ahead and data-ack. Patch from Jason Betts.
108 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
109 even for a "none" policy. Patch from Tony Meyer.
111 JH/22 Fix continued use of a connection for further deliveries. If a port was
112 specified by a router, it must also match for the delivery to be
115 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
116 When one of the recipients of a message was unsuitable for the connection
117 (has no matching addresses), we lost track of needing to mark it
118 deferred. As a result mail would be lost.
120 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
122 JH/25 Decoding ACL controls is now done using a binary search; the source code
123 takes up less space and should be simpler to maintain. Merge the ACL
124 condition decode tables also, with similar effect.
126 JH/26 Fix problem with one_time used on a redirect router which returned the
127 parent address unchanged. A retry would see the parent address marked as
128 delivered, so not attempt the (identical) child. As a result mail would
131 JH/27 Fix a possible security hole, wherein a process operating with the Exim
132 UID can gain a root shell. Credit to http://www.halfdog.net/ for
133 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
136 JH/28 Enable {spool,log} filesystem space and inode checks as default.
137 Main config options check_{log,spool}_{inodes,space} are now
138 100 inodes, 10MB unless set otherwise in the configuration.
140 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
141 Previously it only applied to the main-section connection policy
144 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
146 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
147 by me. Added RFC7919 DH primes as an alternative.
149 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
150 are not in the system include path.
152 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
153 GnuTLS, when a session startup failed (eg because the client disconnected)
154 Exim did stdio operations after fclose. This was exposed by a recent
155 change which nulled out the file handle after the fclose.
157 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
158 signed directly by the cert-signing cert, rather than an intermediate
159 OCSP-signing cert. This is the model used by LetsEncrypt.
161 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
163 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
164 an incoming connection.
166 HS/02 Bug 1802: Do not half-close the connection after sending a request
169 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
170 fallback to "prime256v1".
172 JH/34 SECURITY: Use proper copy of DATA command in error message.
173 Could leak key material. Remotely exploitable. CVE-2016-9963.
179 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
180 and 3.4.4 - once the server is enabled to respond to an OCSP request
181 it does even when not requested, resulting in a stapling non-aware
182 client dropping the TLS connection.
184 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
185 support variable-length bit vectors. No functional change.
187 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
188 The I= interface field on outgoing lines is now after the H= remote
189 host field, same as incoming lines. There is a separate
190 outgoing_interface log selector which allows you to disable the
193 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
194 If not running log_selector +smtp_connection the mainlog would be held
195 open indefinitely after a "too many connections" event, including to a
196 deleted file after a log rotate. Leave the per net connection logging
197 leaving it open for efficiency as that will be quickly detected by the
198 check on the next write.
200 HS/01 Bug 1671: Fix post transport crash.
201 Processing the wait-<transport> messages could crash the delivery
202 process if the message IDs didn't exist for some reason. When
203 using 'split_spool_directory=yes' the construction of the spool
204 file name failed already, exposing the same netto behaviour.
206 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
207 mime_regex ACL conditions.
209 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
210 to DSN fail messages (bounces): remote IP, remote greeting, remote response
211 to HELO, local diagnostic string.
213 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
214 log line to debug. Even when configured with a tls authenticator many
215 client connections are expected to not authenticate in this way, so
216 an authenticate fail is not an error.
218 HS/02 Add the Exim version string to the process info. This way exiwhat
219 gives some more detail about the running daemon.
221 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
222 matter for fast-change records such as DNSBLs.
224 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
225 retry record, even if constant. There may be multiple transports with
226 different interface settings and the retry behaviour needs to be kept
229 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
231 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
233 JH/10 Bug 840: fix log_defer_output option of pipe transport
235 JH/11 Bug 830: use same host for all RCPTS of a message, even under
236 hosts_randomize. This matters a lot when combined with mua_wrapper.
238 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
239 ${quote_pgsql:<string>} operator.
241 JH/13 Bug 1708: avoid misaligned access in cached lookup.
243 JH/14 Change header file name for freeradius-client. Relevant if compiling
244 with Radius support; from the Gentoo tree and checked under Fedora.
246 JH/15 Bug 1712: Introduce $prdr_requested flag variable
248 JH/16 Bug 1714: Permit an empty string as expansion result for transport
249 option transport_filter, meaning no filtering.
251 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
253 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
254 defaults to "*" (all hosts). The variable is now available when not built
255 with TLS, default unset, mainly to enable keeping the testsuite sane.
256 If a server certificate is not supplied (via tls_certificate) an error is
257 logged, and clients will find TLS connections fail on startup. Presumably
258 they will retry in-clear.
259 Packagers of Exim are strongly encouraged to create a server certificate
260 at installation time.
262 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
263 with the $config_file variable.
265 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
266 in transport context, after the attempt, and per-recipient. The latter type
267 is per host attempted. The event data is the error message, and the errno
268 information encodes the lookup type (A vs. MX) used for the (first) host,
269 and the trailing two digits of the smtp 4xx response.
271 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
272 to write to mainlog (or rejectlog, paniclog) in the window between file
273 creation and permissions/ownership being changed. Particularly affects
274 installations where exicyclog is run as root, rather than exim user;
275 result is that the running daemon panics and dies.
277 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
279 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
280 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
281 "pri" and "weight". Note that the previous implicit priority given by the
282 list order is no longer honoured.
284 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
287 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
288 by defining SUPPORT_SOCKS.
290 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
291 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
292 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
293 variables are renamed to proxy_{local,external}_{address,port}.
295 JH/27 Move Internationalisation support from Experimental to mainline, enabled
296 for a build by defining SUPPORT_I18N
298 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
299 of the query string, and make ${quote_redis:} do that quoting.
301 JH/29 Move Events support from Experimental to mainline, enabled by default
302 and removable for a build by defining DISABLE_EVENT.
304 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
306 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
307 cached by the daemon.
309 JH/32 Move Redis support from Experimental to mainline, enabled for a build
310 by defining LOOKUP_REDIS. The libhiredis library is required.
312 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
313 keys are given for lookup.
315 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
316 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
317 only supported when built with TLS support. The PolarSSL SHA routines
318 are still used when the TLS library is too old for convenient support.
320 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
321 openssl_options), for security. OpenSSL forces this from version 1.1.0
322 server-side so match that on older versions.
324 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
325 allocation for $value could be released as the expansion processing
326 concluded, but leaving the global pointer active for it.
328 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
329 and to use the domains and local_parts ACL conditions.
331 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
332 incorrectly not doubled on cutthrough transmission, hence seen as a
333 body-termination at the receiving system - resulting in truncated mails.
334 Commonly the sender saw a TCP-level error, and retransmitted the message
335 via the normal store-and-forward channel. This could result in duplicates
336 received - but deduplicating mailstores were liable to retain only the
337 initial truncated version.
339 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
341 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
343 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
344 we're in there, support oversigning also; bug 1309.
346 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
348 HS/04 Add support for keep_environment and add_environment options.
350 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
351 either intentional arithmetic overflow during PRNG, or testing config-
354 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
355 delivery resulted in actual delivery. Cancel cutthrough before DATA
358 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
359 rejects a recipient: pass the reject to the originator.
361 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
362 Many were false-positives and ignorable, but it's worth fixing the
365 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
366 for the new environment-manipulation done at startup. Move the routines
367 from being local to tls.c to being global via the os.c file.
369 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
370 an extract embedded as result-arg for a map, the first arg for extract
371 is unavailable so we cannot tell if this is a numbered or keyed
372 extraction. Accept either.
378 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
381 JH/02 The smtp transport option "multi_domain" is now expanded.
383 JH/03 The smtp transport now requests PRDR by default, if the server offers
386 JH/04 Certificate name checking on server certificates, when exim is a client,
387 is now done by default. The transport option tls_verify_cert_hostnames
388 can be used to disable this per-host. The build option
389 EXPERIMENTAL_CERTNAMES is withdrawn.
391 JH/05 The value of the tls_verify_certificates smtp transport and main options
392 default to the word "system" to access the system default CA bundle.
393 For GnuTLS, only version 3.0.20 or later.
395 JH/06 Verification of the server certificate for a TLS connection is now tried
396 (but not required) by default. The verification status is now logged by
397 default, for both outbound TLS and client-certificate supplying inbound
400 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
403 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
404 Status Notification (bounce) messages are now MIME format per RFC 3464.
405 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
406 under the control of the dsn_advertise_hosts option, and routers may
407 have a dsn_lasthop option.
409 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
410 default, modifiable by a malware= option. The list separator for
411 the options can now be changed in the usual way. Bug 68.
413 JH/10 The smtp_receive_timeout main option is now expanded before use.
415 JH/11 The incoming_interface log option now also enables logging of the
416 local interface on delivery outgoing connections.
418 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
419 if the interface and destination host and port all match.
421 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
424 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
425 Patch from Andrew Lewis.
427 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
428 now supports optional time-restrictions, weighting, and priority
429 modifiers per server. Patch originally by <rommer@active.by>.
431 JH/16 The spamd_address main option now supports a mixed list of local
432 and remote servers. Remote servers can be IPv6 addresses, and
433 specify a port-range.
435 JH/17 Bug 68: The spamd_address main option now supports an optional
436 timeout value per server.
438 JH/18 Bug 1581: Router and transport options headers_add/remove can
439 now have the list separator specified.
441 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
444 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
447 JH/21 Support for the A6 type of dns record is withdrawn.
449 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
450 rather than the verbs used.
452 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
453 from 255 to 1024 chars.
455 JH/24 Verification callouts now attempt to use TLS by default.
457 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
458 are generic router options now. The defaults didn't change.
460 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
461 Original patch from Alexander Shikoff, worked over by JH.
463 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
464 files if ZCAT_COMMAND is not executable.
466 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
468 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
470 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
471 Normally benign, it bites when the pair was led to by a CNAME;
472 modern usage is to not canonicalize the domain to a CNAME target
473 (and we were inconsistent anyway for A-only vs AAAA+A).
475 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
477 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
478 when evaluating $sender_host_dnssec.
480 JH/31 Check the HELO verification lookup for DNSSEC, adding new
481 $sender_helo_dnssec variable.
483 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
485 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
487 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
489 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
490 documented as working, but never had. Support all but $spam_report.
492 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
493 added for tls authenticator.
495 HS/03 Add perl_taintmode main config option
501 TL/01 When running the test suite, the README says that variables such as
502 no_msglog_check are global and can be placed anywhere in a specific
503 test's script, however it was observed that placement needed to be near
504 the beginning for it to behave that way. Changed the runtest perl
505 script to read through the entire script once to detect and set these
506 variables, reset to the beginning of the script, and then run through
507 the script parsing/test process like normal.
509 TL/02 The BSD's have an arc4random API. One of the functions to induce
510 adding randomness was arc4random_stir(), but it has been removed in
511 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
512 function when detected.
514 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
515 cause callback expansion.
517 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
518 syntax errors in an expansion can be treated as a string instead of
519 logging or causing an error, due to the internal use of bool_lax
520 instead of bool when processing it.
522 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
523 server certificates when making smtp deliveries.
525 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
527 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
529 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
531 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
532 Merged patch from Sebastian Wiedenroth.
534 JH/05 Fix results-pipe from transport process. Several recipients, combined
535 with certificate use, exposed issues where response data items split
536 over buffer boundaries were not parsed properly. This eventually
537 resulted in duplicates being sent. This issue only became common enough
538 to notice due to the introduction of connection certificate information,
539 the item size being so much larger. Found and fixed by Wolfgang Breyha.
541 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
542 size buffer was used, resulting in syntax errors when an expansion
545 JH/07 Add support for directories of certificates when compiled with a GnuTLS
546 version 3.3.6 or later.
548 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
549 is EXPERIMENTAL_EVENT, the main-configuration and transport options
550 both become "event_action", the variables become $event_name, $event_data
551 and $event_defer_errno. There is a new variable $verify_mode, usable in
552 routers, transports and related events. The tls:cert event is now also
553 raised for inbound connections, if the main configuration event_action
556 TL/06 In test suite, disable OCSP for old versions of openssl which contained
557 early OCSP support, but no stapling (appears to be less than 1.0.0).
559 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
560 server certificate names available under the smtp transport option
561 "tls_verify_cert_hostname" now do not permit multi-component wildcard
564 JH/10 Time-related extraction expansions from certificates now use the main
565 option "timezone" setting for output formatting, and are consistent
566 between OpenSSL and GnuTLS compilations. Bug 1541.
568 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
569 encoded parameter in the incoming message. Bug 1558.
571 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
572 include certificate info, eximon was claiming there were spoolfile
575 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
577 JH/14 Log delivery-related information more consistently, using the sequence
578 "H=<name> [<ip>]" wherever possible.
580 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
581 are problematic for Debian distribution, omit them from the release
584 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
586 JH/16 Fix string representation of time values on 64bit time_t architectures.
589 JH/17 Fix a null-indirection in certextract expansions when a nondefault
590 output list separator was used.
595 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
596 checkers that were complaining about end of non-void function with no
599 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
600 This was a regression introduced in 4.83 by another bugfix.
602 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
604 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
605 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
611 TF/01 Correctly close the server side of TLS when forking for delivery.
613 When a message was received over SMTP with TLS, Exim failed to clear up
614 the incoming connection properly after forking off the child process to
615 deliver the message. In some situations the subsequent outgoing
616 delivery connection happened to have the same fd number as the incoming
617 connection previously had. Exim would try to use TLS and fail, logging
618 a "Bad file descriptor" error.
620 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
621 utilities have not been installed.
623 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
624 temporary space as the ACL may create new global variables.
626 TL/01 LDAP support uses per connection or global context settings, depending
627 upon the detected version of the libraries at build time.
629 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
630 to extract and use the src ip:port in logging and expansions as if it
631 were a direct connection from the outside internet. PPv2 support was
632 updated based on HAProxy spec change in May 2014.
634 JH/02 Add ${listextract {number}{list}{success}{fail}}.
636 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
637 Properly escape header and check for NULL return.
639 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
642 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
644 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
645 characters in header names, implemented as a verify condition.
646 Contributed by Michael Fischer v. Mollard.
648 TL/05 Rename SPF condition results err_perm and err_temp to standardized
649 results permerror and temperror. Previous values are deprecated but
650 still accepted. In a future release, err_perm and err_temp will be
651 completely removed, which will be a backward incompatibility if the
652 ACL tests for either of these two old results. Patch contributed by
653 user bes-internal on the mailing list.
655 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
657 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
658 selectors, in both main and reject logs.
660 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
663 JH/07 Add malware type "sock" for talking to simple daemon.
665 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
667 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
668 routers/transports under cutthrough routing.
670 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
671 numbers. Touch up "bool" conditional to keep the same definition.
673 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
675 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
677 JH/12 Expand items in router/transport headers_add or headers_remove lists
678 individually rather than the list as a whole. Bug 1452.
680 Required for reasonable handling of multiple headers_ options when
681 they may be empty; requires that headers_remove items with embedded
682 colons must have them doubled (or the list-separator changed).
684 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
685 view the policy declared in the DMARC record. Currently, $dmarc_status
686 is a combined value of both the record presence and the result of the
689 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
691 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
692 dnslookup router and the smtp transport (applying to the forward
695 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
696 of ldap servers used for a specific lookup. Patch provided by Heiko
699 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
700 New variable $lookup_dnssec_authenticated for observability.
702 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
703 Patch submitted by Lars Timman.
705 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
707 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
708 Requires trusted mode and valid format message id, aborts otherwise.
709 Patch contributed by Heiko Schlichting.
711 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
712 certextract with support for various fields. Bug 1358.
714 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
715 is requested by default, modifiable by smtp transport option
718 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
719 operate on certificate variables to give certificate fingerprints
720 Also new ${sha256:cert_variable}.
722 JH/23 The PRDR feature is moved from being Experimental into the mainline.
724 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
725 Christian Aistleitner.
727 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
729 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
730 file. Patch from Wolfgang Breyha.
732 JH/25 Expand the coverage of the delivery $host and $host_address to
733 client authenticators run in verify callout. Bug 1476.
735 JH/26 Port service names are now accepted for tls_on_connect_ports, to
736 align with daemon_smtp_ports. Bug 72.
738 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
739 support and error reporting did not work properly.
741 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
742 and is readable. Patch from Andrew Colin Kissa.
744 TL/14 Enhance documentation of ${run expansion and how it parses the
745 commandline after expansion, particularly in the case when an
746 unquoted variable expansion results in an empty value.
748 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
750 PP/02 Fix internal collision of T_APL on systems which support RFC3123
751 by renaming away from it. Addresses GH issue 15, reported by
754 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
756 TL/15 SECURITY: prevent double expansion in math comparison functions
757 (can expand unsanitized data). Not remotely exploitable.
764 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
766 PP/02 Make -n do something, by making it not do something.
767 When combined with -bP, the name of an option is not output.
769 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
772 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
773 $sender_host_name and config options to manage this, and basic check
776 PP/05 DSCP support for outbound connections and control modifier for inbound.
778 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
779 (Only plugin which currently uses this is kerberos4, which nobody should
780 be using, but we should make it available and other future plugins might
781 conceivably use it, even though it would break NAT; stuff *should* be
782 using channel bindings instead).
784 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
785 name; added for Sendmail compatibility; requires admin caller.
786 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
787 just ignore it); requires trusted caller.
788 Also parse but ignore: -Ac -Am -X<logfile>
791 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
793 TL/02 Add +smtp_confirmation as a default logging option.
795 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
796 Patch by Magnus Holmgren from 2007-02-20.
798 TL/04 Bugzilla 1281 - Spec typo.
799 Bugzilla 1283 - Spec typo.
800 Bugzilla 1290 - Spec grammar fixes.
802 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
804 TL/06 Add Experimental DMARC support using libopendmarc libraries.
806 TL/07 Fix an out of order global option causing a segfault. Reported to dev
807 mailing list by by Dmitry Isaikin.
809 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
811 JH/02 Support "G" suffix to numbers in ${if comparisons.
813 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
815 NM/01 Bugzilla 1197 - Spec typo
816 Bugzilla 1196 - Spec examples corrections
818 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
820 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
821 gnutls_enable_pkcs11, but renamed to more accurately indicate its
824 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
825 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
827 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
828 "acl {{name}{arg}...}", and optional args on acl condition
831 JH/05 Permit multiple router/transport headers_add/remove lines.
833 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
835 JH/07 Avoid using a waiting database for a single-message-only transport.
836 Performance patch from Paul Fisher. Bugzilla 1262.
838 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
841 JH/09 Add $headers_added variable, with content from use of ACL modifier
842 add_header (but not yet added to the message). Bugzilla 199.
844 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
845 Pulled from Bugzilla 817 by Wolfgang Breyha.
847 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
849 (nb: this is the same fix as in Exim 4.80.1)
851 JH/11 Add A= logging on delivery lines, and a client_set_id option on
854 JH/12 Add optional authenticated_sender logging to A= and a log_selector
857 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
859 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
860 advertise SMTP AUTH mechanism to us, instead of a generic
861 protocol violation error. Also, make Exim more robust to bad
862 data from the Dovecot auth socket.
864 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
866 When a queue runner is handling a message, Exim first routes the
867 recipient addresses, during which it prunes them based on the retry
868 hints database. After that it attempts to deliver the message to
869 any remaining recipients. It then updates the hints database using
872 So if a recipient address works intermittently, it can get repeatedly
873 deferred at routing time. The retry hints record remains fresh so the
874 address never reaches the final cutoff time.
876 This is a fairly common occurrence when a user is bumping up against
877 their storage quota. Exim had some logic in its local delivery code
878 to deal with this. However it did not apply to per-recipient defers
879 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
881 This change adds a proper retry rule check during routing so that the
882 final cutoff time is checked against the message's age. We only do
883 this check if there is an address retry record and there is not a
884 domain retry record; this implies that previous attempts to handle
885 the address had the retry_use_local_parts option turned on. We use
886 this as an approximation for the destination being like a local
887 delivery, as in LMTP.
889 I suspect this new check makes the old local delivery cutoff check
890 redundant, but I have not verified this so I left the code in place.
892 TF/02 Correct gecos expansion when From: is a prefix of the username.
894 Test 0254 submits a message to Exim with the header
898 When I ran the test suite under the user fanf2, Exim expanded
899 the header to contain my full name, whereas it should have added
900 a Resent-Sender: header. It erroneously treats any prefix of the
901 username as equal to the username.
903 This change corrects that bug.
905 GF/01 DCC debug and logging tidyup
906 Error conditions log to paniclog rather than rejectlog.
907 Debug lines prefixed by "DCC: " to remove any ambiguity.
909 TF/03 Avoid unnecessary rebuilds of lookup-related code.
911 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
912 Bug spotted by Jeremy Harris; was flawed since initial commit.
913 Would have resulted in OCSP responses post-SNI triggering an Exim
914 NULL dereference and crash.
916 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
918 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
919 Bug detection, analysis and fix by Samuel Thibault.
920 Bugzilla 1331, Debian bug #698092.
922 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
924 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
925 Server implementation by Todd Lyons, client by JH.
926 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
927 config variable "prdr_enable" controls whether the server
928 advertises the facility. If the client requests PRDR a new
929 acl_data_smtp_prdr ACL is called once for each recipient, after
930 the body content is received and before the acl_smtp_data ACL.
931 The client is controlled by both of: a hosts_try_prdr option
932 on the smtp transport, and the server advertisement.
933 Default client logging of deliveries and rejections involving
934 PRDR are flagged with the string "PRDR".
936 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
937 fclose(). Diagnosis by Todd Lyons.
939 PP/17 Update configure.default to handle IPv6 localhost better.
940 Patch by Alain Williams (plus minor tweaks).
943 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
944 This is now consistent with GnuTLS, and is now documented: the
945 previous undocumented portable approach to treating the option as
946 unset was to force an expansion failure. That still works, and
947 an empty string is now equivalent.
949 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
950 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
951 not performing validation itself.
953 PP/20 Added force_command boolean option to pipe transport.
954 Patch from Nick Koston, of cPanel Inc.
956 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
959 TF/04 Added udpsend ACL modifier and hexquote expansion operator
961 PP/21 Fix eximon continuous updating with timestamped log-files.
962 Broken in a format-string cleanup in 4.80, missed when I repaired the
963 other false fix of the same issue.
964 Report and fix from Heiko Schlichting.
967 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
968 Report from Prashanth Katuri.
970 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
971 It's SecureTransport, so affects any MacOS clients which use the
972 system-integrated TLS libraries, including email clients.
974 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
975 using a MIME ACL for non-SMTP local injection.
976 Report and assistance in diagnosis by Warren Baker.
978 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
980 JH/16 Fix comparisons for 64b. Bugzilla 1385.
982 TL/09 Add expansion variable $authenticated_fail_id to keep track of
983 last id that failed so it may be referenced in subsequent ACL's.
985 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
988 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
989 ldap library initialization, allowing self-signed CA's to be
990 used. Also properly sets require_cert option later in code by
991 using NULL (global ldap config) instead of ldap handle (per
992 session). Bug diagnosis and testing by alxgomz.
994 TL/12 Enhanced documentation in the ratelimit.pl script provided in
995 the src/util/ subdirectory.
997 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
998 renamed to Transport Post Delivery Action by Jeremy Harris, as
1001 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1002 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1003 redis_servers = needs to be configured which will be used by the redis
1004 lookup. Patch from Warren Baker, of The Packet Hub.
1006 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1008 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1009 hostname or reverse DNS when processing a host list. Used suggestions
1010 from multiple comments on this bug.
1012 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1014 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1015 Missed a few lines, added it to make the runtest require no keyboard
1018 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1019 contains upper case chars. Make router use caseful_local_part.
1021 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1022 support when GnuTLS has been built with p11-kit.
1028 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1030 This, or similar/improved, will also be change PP/11 of 4.82.
1036 PP/01 Handle short writes when writing local log-files.
1037 In practice, only affects FreeBSD (8 onwards).
1038 Bugzilla 1053, with thanks to Dmitry Isaikin.
1040 NM/01 Bugzilla 949 - Documentation tweak
1042 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1045 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1047 PP/02 Implemented gsasl authenticator.
1049 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1051 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1052 `pkg-config foo` for cflags/libs.
1054 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1055 with rest of GSASL and with heimdal_gssapi.
1057 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1058 `pkg-config foo` for cflags/libs for the TLS implementation.
1060 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1061 properties get this fed in as external SSF. A number of robustness
1062 and debugging improvements to the cyrus_sasl authenticator.
1064 PP/08 cyrus_sasl server now expands the server_realm option.
1066 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1067 Patch by Jeremy Harris.
1069 PP/10 Added dbmjz lookup type.
1071 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1073 PP/12 MAIL args handles TAB as well as SP, for better interop with
1074 non-compliant senders.
1075 Analysis and variant patch by Todd Lyons.
1077 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1078 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1079 Patch from Dirk Mueller <dmueller@suse.com>
1081 PP/13 tls_peerdn now print-escaped for spool files.
1082 Observed some $tls_peerdn in wild which contained \n, which resulted
1083 in spool file corruption.
1085 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1086 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1087 or write after TLS renegotiation, which otherwise led to messages
1090 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1091 as a tracking header (ie: a signed header comes before the signature).
1092 Patch from Wolfgang Breyha.
1094 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1095 comma-sep list; embedded commas doubled.
1097 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1099 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1101 Report and patch from Dmitry Banschikov.
1103 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1104 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1105 failures. We appear to now support TLS1.1+ with Exim.
1107 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1108 lets Exim select keys and certificates based upon TLS SNI from client.
1109 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1110 before an outbound SMTP session. New log_selector, +tls_sni.
1112 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1113 NULL dereference. Report and patch from Alun Jones.
1115 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1116 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1117 Not seeing resolver debug output on NetBSD, but suspect this is a
1118 resolver implementation change.
1120 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1121 Left warnings. Added "eximon gdb" invocation mode.
1123 PP/21 Defaulting "accept_8bitmime" to true, not false.
1125 PP/22 Added -bw for inetd wait mode support.
1127 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1128 locate the relevant includes and libraries. Made this the default.
1130 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1131 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1133 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1134 This may cause build issues on older platforms.
1136 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1137 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1138 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1139 Added SNI support via GnuTLS too.
1140 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1142 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1144 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1145 Applied second patch from Janne, implementing suggestion to default
1146 multiple-strings-in-record handling to match SPF spec.
1148 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1150 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1151 read-only, out of scope).
1152 Patch from Wolfgang Breyha, report from Stuart Northfield.
1154 PP/29 Fix three issues highlighted by clang analyser static analysis.
1155 Only crash-plausible issue would require the Cambridge-specific
1156 iplookup router and a misconfiguration.
1157 Report from Marcin Mirosław.
1159 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1161 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1162 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1163 As part of this, removing so much warning spew let me fix some minor
1164 real issues in debug logging.
1166 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1167 assignment on my part. Fixed.
1169 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1170 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1171 Janne Snabb (who went above and beyond: thank you).
1173 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1174 string otherwise requires a connection and a bunch more work and it's
1175 relatively easy to get wrong. Should also expose TLS library linkage
1178 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1179 64-bit ${eval} (JH/03).
1181 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1182 GNU libc to support some of the 64-bit stuff, should not lead to
1183 conflicts. Defined before os.h is pulled in, so if a given platform
1184 needs to override this, it can.
1186 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1187 protection layer was required, which is not implemented.
1188 Bugzilla 1254, patch from Wolfgang Breyha.
1190 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1191 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1192 tls_dhparam take prime identifiers. Also unbreak combination of
1193 OpenSSL+DH_params+TLSSNI.
1195 PP/39 Disable SSLv2 by default in OpenSSL support.
1201 PP/01 Solaris build fix for Oracle's LDAP libraries.
1202 Bugzilla 1109, patch from Stephen Usher.
1204 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1206 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1209 TF/02 Fix a couple more cases where we did not log the error message
1210 when unlink() failed. See also change 4.74-TF/03.
1212 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1213 lock up or crash if it happened to be inside a call to libc when it
1214 got a SIGUSR1 from exiwhat.
1216 The SIGUSR1 handler appends the current process status to the process
1217 log which is later printed by exiwhat. It used to use the general
1218 purpose logging code to do this, but several functions it calls are
1219 not safe for signals.
1221 The new output code in the SIGUSR1 handler is specific to the process
1222 log, and simple enough that it's easy to inspect for signal safety.
1223 Removing some special cases also simplifies the general logging code.
1224 Removing the spurious timestamps from the process log simplifies
1227 TF/04 Improved ratelimit ACL condition.
1229 The /noupdate option has been deprecated in favour of /readonly which
1230 has clearer semantics. The /leaky, /strict, and /readonly update modes
1231 are mutually exclusive. The update mode is no longer included in the
1232 database key; it just determines when the database is updated. (This
1233 means that when you upgrade Exim will forget old rate measurements.)
1235 Exim now checks that the per_* options are used with an update mode that
1236 makes sense for the current ACL. For example, when Exim is processing a
1237 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1238 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1239 must specify per_mail/readonly. If you omit the update mode it defaults to
1240 /leaky where that makes sense (as before) or /readonly where required.
1242 The /noupdate option is now undocumented but still supported for
1243 backwards compatibility. It is equivalent to /readonly except that in
1244 ACLs where /readonly is required you may specify /leaky/noupdate or
1245 /strict/noupdate which are treated the same as /readonly.
1247 A useful new feature is the /count= option. This is a generalization
1248 of the per_byte option, so that you can measure the throughput of other
1249 aggregate values. For example, the per_byte option is now equivalent
1250 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1252 The per_rcpt option has been generalized using the /count= mechanism
1253 (though it's more complicated than the per_byte equivalence). When it is
1254 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1255 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1256 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1257 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1258 Note that using per_rcpt with a non-readonly update mode in more than
1259 one ACL will cause the recipients to be double-counted. (The per_mail
1260 and per_byte options don't have this problem.)
1262 The handling of very low rates has changed slightly. If the computed rate
1263 is less than the event's count (usually one) then this event is the first
1264 after a long gap. In this case the rate is set to the same as this event's
1265 count, so that the first message of a spam run is counted properly.
1267 The major new feature is a mechanism for counting the rate of unique
1268 events. The new per_addr option counts the number of different
1269 recipients that someone has sent messages to in the last time period. It
1270 behaves like per_rcpt if all the recipient addresses are different, but
1271 duplicate recipient addresses do not increase the measured rate. Like
1272 the /count= option this is a general mechanism, so the per_addr option
1273 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1274 example, measure the rate that a client uses different sender addresses
1275 with the options per_mail/unique=$sender_address. There are further
1276 details in the main documentation.
1278 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1280 TF/06 Removed a few PCRE remnants.
1282 TF/07 Automatically extract Exim's version number from tags in the git
1283 repository when doing development or release builds.
1285 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1286 Bugzilla 879. Patch from Paul Fisher.
1288 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1289 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1292 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1294 PP/05 Variable $av_failed, true if the AV scanner deferred.
1295 Bugzilla 1078. Patch from John Horne.
1297 PP/06 Stop make process more reliably on build failure.
1298 Bugzilla 1087. Patch from Heiko Schlittermann.
1300 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1301 Bugzilla 1089. Patch from Heiko Schlittermann.
1303 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1304 Bugzilla 1131. Patch from Holger Weiß.
1306 PP/09 Handle IPv6 addresses with SPF.
1307 Bugzilla 860. Patch from Wolfgang Breyha.
1309 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1311 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1314 PP/11 match_* no longer expand right-hand-side by default.
1315 New compile-time build option, EXPAND_LISTMATCH_RHS.
1316 New expansion conditions, "inlist", "inlisti".
1318 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1320 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1322 PP/14 fix log_write() format string regression from TF/03.
1323 Bugzilla 1152. Patch from Dmitry Isaikin.
1329 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1331 PP/02 Harmonised TLS library version reporting; only show if debugging.
1332 Layout now matches that introduced for other libraries in 4.74 PP/03.
1334 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1336 PP/04 New "dns_use_edns0" global option.
1338 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1341 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1342 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1344 TK/01 Updated PolarSSL code to 0.14.2.
1345 Bugzilla 1097. Patch from Andreas Metzler.
1347 PP/07 Catch divide-by-zero in ${eval:...}.
1348 Fixes bugzilla 1102.
1350 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1353 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1354 format-string attack -- SECURITY: remote arbitrary code execution.
1356 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1357 time unintentionally subject to list matching rules, letting the header
1358 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1359 arbitrary string expansion). This allowed for information disclosure.
1361 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1362 INT_MIN/-1 -- value coerced to INT_MAX.
1368 NM/01 Workaround for PCRE version dependency in version reporting
1371 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1372 This fixes portability to compilers other than gcc, notably
1373 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1375 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1376 makefiles for portability to HP-UX and POSIX correctness.
1378 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1379 Also via indented variable definition in the Makefile.
1380 (Debugging by Oliver Heesakkers).
1382 PP/02 Restore caching of spamd results with expanded spamd_address.
1383 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1385 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1386 Improves build reliability. Fix from: Frank Elsner
1388 NM/02 Fix wide character breakage in the rfc2047 coding
1389 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1391 NM/03 Allow underscore in dnslist lookups
1392 Fixes bug 1026. Patch from Graeme Fowler
1394 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1395 Code patches from Adam Ciarcinski of NetBSD.
1397 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1400 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1401 is logged, to avoid truncation. Patch from John Horne.
1403 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1404 Patch from Jakob Hirsch.
1406 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1407 SQL string expansion failure details.
1408 Patch from Andrey Oktyabrski.
1410 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1411 Patch from Simon Arlott.
1413 PP/09 New lookups functionality failed to compile on old gcc which rejects
1414 extern declarations in function scope.
1415 Patch from Oliver Fleischmann
1417 PP/10 Use sig_atomic_t for flags set from signal handlers.
1418 Check getgroups() return and improve debugging.
1419 Fixed developed for diagnosis in bug 927 (which turned out to be
1422 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1423 Patch from Mark Zealey.
1425 PP/12 Bugzilla 1056: Improved spamd server selection.
1426 Patch from Mark Zealey.
1428 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1429 Based on patch from Heiko Schlittermann.
1431 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1432 Patch from Uwe Doering, sign-off by Michael Haardt.
1434 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1435 variable declaration deep within a block. Bug and patch from
1438 PP/15 lookups-Makefile IRIX compatibility coercion.
1440 PP/16 Make DISABLE_DKIM build knob functional.
1442 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1443 Patch by Simon Arlott
1445 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1446 variable argument macros. Our copy now differs from upstream.
1452 TF/01 Failure to get a lock on a hints database can have serious
1453 consequences so log it to the panic log.
1455 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1456 controlled using the smtp_confirmation log selector.
1458 TF/03 Include the error message when we fail to unlink a spool file.
1460 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1461 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1462 for maintaining out-of-tree patches for some time.
1464 PP/01 Bugzilla 139: Documentation and portability issues.
1465 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1466 Handle per-OS dynamic-module compilation flags.
1468 PP/02 Let /dev/null have normal permissions.
1469 The 4.73 fixes were a little too stringent and complained about the
1470 permissions on /dev/null. Exempt it from some checks.
1471 Reported by Andreas M. Kirchwitz.
1473 PP/03 Report version information for many libraries, including
1474 Exim version information for dynamically loaded libraries. Created
1475 version.h, now support a version extension string for distributors
1476 who patch heavily. Dynamic module ABI change.
1478 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1479 privilege escalation vulnerability whereby the Exim run-time user
1480 can cause root to append content of the attacker's choosing to
1483 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1486 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1487 If dropping privileges for untrusted macros, we disabled normal logging
1488 on the basis that it would fail; for the Exim run-time user, this is not
1489 the case, and it resulted in successful deliveries going unlogged.
1490 Fixed. Reported by Andreas Metzler.
1496 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1497 only prepend for the Resent-* case. Fixes regression introduced in
1498 Exim 4.70 by NM/22 for Bugzilla 607.
1500 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1501 increasing numbers of administrators be bitten by this.
1503 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1505 PP/03 Bugzilla 994: added openssl_options main configuration option.
1507 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1509 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1511 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1513 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1514 without a peer certificate, leading to a segfault because of an
1515 assumption that peers always have certificates. Be a little more
1516 paranoid. Problem reported by Martin Tscholak.
1518 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1519 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1520 NB: ClamAV planning to remove STREAM in "middle of 2010".
1521 CL also introduces -bmalware, various -d+acl logging additions and
1522 more caution in buffer sizes.
1524 PP/09 Implemented reverse_ip expansion operator.
1526 PP/10 Bugzilla 937: provide a "debug" ACL control.
1528 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1530 PP/12 Bugzilla 973: Implement --version.
1532 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1534 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1536 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1538 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1539 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1540 ignore trailing whitespace.
1542 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1544 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1547 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1548 Notification from Dr Andrew Aitchison.
1550 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1551 ExtendedDetectionInfo response format.
1552 Notification from John Horne.
1554 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1557 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1558 XSL and documented dependency on system catalogs, with examples of how
1561 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1564 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1565 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1566 configuration file which is writeable by the Exim user or group.
1568 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1569 of configuration files to cover files specified with the -C option if
1570 they are going to be used with root privileges, not just the default
1573 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1574 option (effectively making it always true).
1576 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1577 files to be used while preserving root privileges.
1579 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1580 that rogue child processes cannot use them.
1582 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1583 run-time user, instead of root.
1585 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1586 Exim run-time user without dropping privileges.
1588 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1589 result string, instead of calling string_vformat() twice with the same
1592 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1593 for other users. Others should always drop root privileges if they use
1594 -C on the command line, even for a whitelisted configure file.
1596 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1598 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1604 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1605 $data_path, and $header_path variables; fixed documentation bugs and
1608 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1609 exipick to access non-standard spools, including the "frozen" queue
1612 NM/01 Bugzilla 965: Support mysql stored procedures.
1613 Patch from Alain Williams
1615 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1617 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1618 Patch from Andreas Metzler
1620 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1621 Patch from Kirill Miazine
1623 NM/05 Bugzilla 671: Added umask to procmail example.
1625 JJ/03 installed exipick 20100323.0, fixing doc bug
1627 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1628 directory. Notification and patch from Dan Rosenberg.
1630 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1632 TK/02 Improve log output when DKIM signing operation fails.
1634 MH/01 Treat the transport option dkim_domain as a colon separated
1635 list, not as a single string, and sign the message with each element,
1636 omitting multiple occurences of the same signer.
1638 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1639 Bugzilla 985, 986. Patch by Simon Arlott
1641 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1642 Patch by Simon Arlott
1644 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1645 MBX locking. Notification from Dan Rosenberg.
1651 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1653 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1655 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1657 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1659 NM/04 Bugzilla 915: Flag broken perl installation during build.
1665 TK/01 Added patch by Johannes Berg that expands the main option
1666 "spamd_address" if it starts with a dollar sign.
1668 TK/02 Write list of recipients to X-Envelope-Sender header when building
1669 the mbox-format spool file for content scanning (suggested by Jakob
1672 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1673 (http://www.dcc-servers.net/) support via dccifd. Activated by
1674 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1676 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1677 by Mark Daniel Reidel <mr@df.eu>.
1679 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1680 When building exim an external PCRE library is now needed -
1681 PCRE is a system library on the majority of modern systems.
1682 See entry on PCRE_LIBS in EDITME file.
1684 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1685 conversation. Added nologin parameter to request.
1686 Patch contributed by Kirill Miazine.
1688 TF/01 Do not log submission mode rewrites if they do not change the address.
1690 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1692 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1693 log files in place. Contributed by Roberto Lima.
1695 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1697 TF/03 Bugzilla 615: When checking the local_parts router precondition
1698 after a local_part_suffix or local_part_prefix option, Exim now
1699 does not use the address's named list lookup cache, since this
1700 contains cached lookups for the whole local part.
1702 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1703 Robert Millan. Documentation is in experimental-spec.txt.
1705 TF/04 Bugzilla 668: Fix parallel build (make -j).
1707 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1709 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1710 Patch provided by Jan Srzednicki.
1712 TF/05 Leading white space used to be stripped from $spam_report which
1713 wrecked the formatting. Now it is preserved.
1715 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1716 that they are available at delivery time.
1718 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1720 TF/08 TLS error reporting now respects the incoming_interface and
1721 incoming_port log selectors.
1723 TF/09 Produce a more useful error message if an SMTP transport's hosts
1724 setting expands to an empty string.
1726 NM/06 Bugzilla 744: EXPN did not work under TLS.
1727 Patch provided by Phil Pennock.
1729 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1730 Patch provided by Richard Godbee.
1732 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1733 acl_smtp_notquit, added index entry.
1735 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1736 Patch provided by Eugene Bujak.
1738 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1739 accept(). Patch provided by Maxim Dounin.
1741 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1742 Patch provided by Phil Pennock.
1744 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1746 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1747 Patch provided by Brad "anomie" Jorsch.
1749 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1750 Patch provided by Dean Brooks.
1752 TK/05 Add native DKIM support (does not depend on external libraries).
1754 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1755 Patch provided by Graeme Fowler.
1757 NM/16 Bugzilla 851: Documentation example syntax fix.
1759 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1761 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1764 NM/19 Bugzilla 745: TLS version reporting.
1765 Patch provided by Phil Pennock.
1767 NM/20 Bugzilla 167: bool: condition support.
1768 Patch provided by Phil Pennock.
1770 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1771 clients. Patch provided by Phil Pennock.
1773 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1774 Patch provided by Brad "anomie" Jorsch.
1776 NM/23 Bugzilla 687: Fix misparses in eximstats.
1777 Patch provided by Heiko Schlittermann.
1779 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1780 Patch provided by Heiko Schlittermann.
1782 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1783 plus update to original patch.
1785 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1787 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1788 Patch provided by David Brownlee.
1790 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1792 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1794 NM/30 Bugzilla 888: TLS documentation bugfixes.
1796 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1798 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1799 Unlike the original bugzilla I have changed all shell scripts in src tree.
1801 NM/33 Bugzilla 898: Transport filter timeout fix.
1802 Patch by Todd Rinaldo.
1804 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1805 Patch by Serge Demonchaux.
1807 NM/35 Bugzilla 39: Base64 decode bug fixes.
1808 Patch by Jakob Hirsch.
1810 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1812 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1814 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1816 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1822 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1823 ALT-N's libdkim that I have put here:
1824 http://duncanthrax.net/exim-experimental/
1826 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1827 were called 'true' and it seems that C99 defines that as a reserved
1828 keyword to be used with 'bool' variable types. That means you could
1829 not include C99-style headers which use bools without triggering
1830 build errors in sieve.c.
1832 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1833 as mailq or other aliases. Changed the --help handling significantly
1834 to do whats expected. exim_usage() emits usage/help information.
1836 SC/01 Added the -bylocaldomain option to eximstats.
1838 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1840 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1842 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1848 PH/01 Another patch from the Sieve maintainer.
1850 PH/02 When an IPv6 address is converted to a string for single-key lookup
1851 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1852 dots are used instead of colons so that keys in lsearch files need not
1853 contain colons. This was done some time before quoting was made available
1854 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1855 (notated using the quote facility) so as to distinguish them from IPv4
1856 keys. This meant that lookups for IP addresses in host lists did not work
1857 for iplsearch lookups.
1859 This has been fixed by arranging for IPv6 addresses to be expressed with
1860 colons if the lookup type is iplsearch. This is not incompatible, because
1861 previously such lookups could never work.
1863 The situation is now rather anomalous, since one *can* have colons in
1864 ordinary lsearch keys. However, making the change in all cases is
1865 incompatible and would probably break a number of configurations.
1867 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1870 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1871 conversion specification without a maximum field width, thereby enabling
1872 a rogue spamd server to cause a buffer overflow. While nobody in their
1873 right mind would setup Exim to query an untrusted spamd server, an
1874 attacker that gains access to a server running spamd could potentially
1875 exploit this vulnerability to run arbitrary code as the Exim user.
1877 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1878 $primary_hostname instead of what libspf2 thinks the hosts name is.
1880 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1881 a directory entry by the name of the lookup key. Previously, if a
1882 symlink pointed to a non-existing file or a file in a directory that
1883 Exim lacked permissions to read, a lookup for a key matching that
1884 symlink would fail. Now it is enough that a matching directory entry
1885 exists, symlink or not. (Bugzilla 503.)
1887 PH/03 The body_linecount and body_zerocount variables are now exported in the
1890 PH/04 Added the $dnslist_matched variable.
1892 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1893 This means they are set thereafter only if the connection becomes
1896 PH/06 Added the client_condition to authenticators so that some can be skipped
1897 by clients under certain conditions.
1899 PH/07 The error message for a badly-placed control=no_multiline_responses left
1900 "_responses" off the end of the name.
1902 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1904 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1905 (without spaces) instead of just copying the configuration text.
1907 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1909 PH/11 Added $max_received_linelength.
1911 PH/12 Added +ignore_defer and +include_defer to host lists.
1913 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1914 way in which PCRE > 7.0 is built.
1916 PH/14 Implemented queue_only_load_latch.
1918 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1919 MAIL command. The effect was to mangle the value on 64-bit systems.
1921 PH/16 Another patch from the Sieve maintainer.
1923 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1925 PH/18 If a system quota error occurred while trying to create the file for
1926 a maildir delivery, the message "Mailbox is full" was not appended to the
1927 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1928 only to a quota excession during the actual writing of the file.
1930 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1931 characters?) which causes problems in log lines. The DN values are now
1932 passed through string_printing() before being added to log lines.
1934 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1935 and InterBase are left for another time.)
1937 PH/21 Added message_body_newlines option.
1939 PH/22 Guard against possible overflow in moan_check_errorcopy().
1941 PH/23 POSIX allows open() to be a macro; guard against that.
1943 PH/24 If the recipient of an error message contained an @ in the local part
1944 (suitably quoted, of course), incorrect values were put in $domain and
1945 $local_part during the evaluation of errors_copy.
1951 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1952 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1955 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1956 issue a MAIL command.
1958 PH/02 In an ACL statement such as
1960 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1962 if a client was not listed at all, or was listed with a value other than
1963 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1964 the condition was not true (as it should be), so access was not denied.
1965 The bug was that the ! inversion was incorrectly passed on to the second
1966 item. This has been fixed.
1968 PH/03 Added additional dnslists conditions == and =& which are different from
1969 = and & when the dns lookup returns more than one IP address.
1971 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1972 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1974 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1975 FSYNC, which compiles an option called disable_fsync that allows for
1976 bypassing fsync(). The documentation is heavily laced with warnings.
1978 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1980 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1981 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1982 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1983 including adding "make clean"; (3) Added -fPIC when compiling the test
1984 dynamically loaded module, to get rid of a warning.
1986 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1987 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1988 = 0s. The bug is otherwise harmless.
1990 PH/07 There was a bug in the dovecot authenticator such that the value of
1991 $auth1 could be overwritten, and so not correctly preserved, after a
1992 successful authentication. This usually meant that the value preserved by
1993 the server_setid option was incorrect.
1995 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1997 PH/09 Installed PCRE release 7.0.
1999 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2000 run for batched SMTP input. It is now run at the start of every message
2001 in the batch. While fixing this I discovered that the process information
2002 (output by running exiwhat) was not always getting set for -bs and -bS
2003 input. This is fixed, and it now also says "batched" for BSMTP.
2005 PH/11 Added control=no_pipelining.
2007 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2008 patch, slightly modified), and move the expansion of helo_data till after
2009 the connection is made in the smtp transport (so it can use these
2012 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2014 PH/14 Added log_selector = +pid.
2016 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2018 PH/16 Add ${if forany and ${if forall.
2020 PH/17 Added dsn_from option to vary the From: line in DSNs.
2022 PH/18 Flush SMTP output before performing a callout, unless control =
2023 no_callout_flush is set.
2025 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2026 was true (the default) a successful delivery failed to delete the retry
2027 item, thus causing premature timeout of the address. The bug is now
2030 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2032 PH/21 Long custom messages for fakedefer and fakereject are now split up
2033 into multiline responses in the same way that messages for "deny" and
2034 other ACL rejections are.
2036 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2037 with slight modification.
2039 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2040 draft, changing the syntax and factoring some duplicate code".
2042 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2043 for deliveries of the second and subsequent messages over the same SMTP
2046 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2047 ${reduce, with only minor "tidies".
2049 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2051 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2052 expansion side effects.
2054 PH/27 When a message times out after an over-quota error from an Exim-imposed
2055 quota, the bounce message says "mailbox is full". This message was not
2056 being given when it was a system quota that was exceeded. It now should
2059 MH/03 Made $recipients available in local_scan(). local_scan() already has
2060 better access to the recipient list through recipients_list[], but
2061 $recipients can be useful in postmaster-provided expansion strings.
2063 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2064 in the case of a MAIL command with additional options following the
2065 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2066 were accidentally chopped off.
2068 PH/29 SMTP synchronization checks are implemented when a command is read -
2069 there is a check that no more input is waiting when there shouldn't be
2070 any. However, for some commands, a delay in an ACL can mean that it is
2071 some time before the response is written. In this time, more input might
2072 arrive, invalidly. So now there are extra checks after an ACL has run for
2073 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2074 pipelining has not been advertised.
2076 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2078 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2079 returned in the SMTP response when smtp_return_error_details was false.
2080 This has been fixed.
2082 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2083 instead of the C I/O that was originally supplied, because problems were
2084 reported on Solaris.
2086 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2087 Exim which did not show up earlier: it was assuming that a call to
2088 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2089 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2090 was a macro that became an assignment, so it seemed to work. This has
2091 changed to a proper function call with a void return, hence the compile
2092 error. Exim's code has been fixed.
2094 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2097 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2099 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2101 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2102 because it turns out to be too big or there is a timeout) there is no
2103 "Completed" line in the log. When some messages of this type were
2104 selected by exigrep, they were listed as "not completed". Others were
2105 picked up by some special patterns. I have improved the selection
2106 criteria to be more general.
2108 PH/38 The host_find_failed option in the manualroute router can now be set
2109 to "ignore", to completely ignore a host whose IP address cannot be
2110 found. If all hosts are ignored, the behaviour is controlled by the new
2111 host_all_ignored option.
2113 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2114 homing or because of multiple MX records with /mx) generated more than
2115 one IP address, and the following item turned out to be the local host,
2116 all the secondary addresses of the first item were incorrectly removed
2117 from the list, along with the local host and any following hosts (which
2118 is what is supposed to happen).
2120 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2121 whoever called Exim into the -H file. In the case of the daemon it was
2122 behaving confusingly. When first started, it used values for whoever
2123 started the daemon, but after a SIGHUP it used the Exim user (because it
2124 calls itself on a restart). I have changed the code so that it now always
2127 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2128 message are rejected with the same error (e.g. no authentication or bad
2129 sender address), and a DATA command is nevertheless sent (as can happen
2130 with PIPELINING or a stupid MUA), the error message that was given to the
2131 RCPT commands is included in the rejection of the DATA command. This is
2132 intended to be helpful for MUAs that show only the final error to their
2135 PH/42 Another patch from the Sieve maintainer.
2137 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2138 Eximstats - Fixed some broken HTML links and added missing column headers
2140 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2141 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2143 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2145 PH/43 Yet another patch from the Sieve maintainer.
2147 PH/44 I found a way to check for a TCP/IP connection going away before sending
2148 the response to the final '.' that terminates a message, but only in the
2149 case where the client has not sent further data following the '.'
2150 (unfortunately, this is allowed). However, in many cases there won't be
2151 any further data because there won't be any more messages to send. A call
2152 to select() can be used: if it shows that the input is "ready", there is
2153 either input waiting, or the socket has been closed. An attempt to read
2154 the next input character can distinguish the two cases. Previously, Exim
2155 would have sent an OK response which the client would never have see.
2156 This could lead to message repetition. This fix should cure that, at
2157 least in a lot of common cases.
2159 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2160 advertised in response to EHLO.
2166 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2167 fixed by 4.65/MH/01 (is this a record?) are fixed:
2169 (i) An empty string was always treated as zero by the numeric comparison
2170 operators. This behaviour has been restored.
2172 (ii) It is documented that the numeric comparison operators always treat
2173 their arguments as decimal numbers. This was broken in that numbers
2174 starting with 0 were being interpreted as octal.
2176 While fixing these problems I realized that there was another issue that
2177 hadn't been noticed. Values of message_size_limit (both the global option
2178 and the transport option) were treated as octal if they started with 0.
2179 The documentation was vague. These values are now always treated as
2180 decimal, and I will make that clear in the documentation.
2186 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2187 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2190 MH/01 Don't check that the operands of numeric comparison operators are
2191 integers when their expansion is in "skipping" mode (fixes bug
2192 introduced by 4.64-PH/07).
2194 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2195 child addresses, Exim now panics and dies. Previously, because the count
2196 is held in a short int, deliveries were likely to be lost. As such a
2197 large number of recipients for a single message is ridiculous
2198 (performance will be very, very poor), I have chosen to impose a limit
2199 rather than extend the field.
2205 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2206 leftover -K file (the existence of which was triggered by #402).
2207 While we were at it, introduced process PID as part of the -K
2208 filename. This should rule out race conditions when creating
2211 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2212 processing considerably. Previous code took too long for large mails,
2213 triggering a timeout which in turn triggers #401.
2215 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2216 in the DK code in transports.c. sendfile() is not really portable,
2217 hence the _LINUX specificness.
2219 TF/01 In the add_headers option to the mail command in an Exim filter,
2220 there was a bug that Exim would claim a syntax error in any
2221 header after the first one which had an odd number of characters
2224 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2225 callout verification, Exim cached a "reject" for the entire domain. This
2226 is correct for most verifications, but it is not correct for a recipient
2227 verification with use_sender or use_postmaster set, because in that case
2228 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2229 case of MAIL FROM:<> rejection from other early rejections (e.g.
2230 rejection of HELO). When verifying a recipient using a non-null MAIL
2231 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2232 Whatever the result of the callout, the value of the domain cache is
2233 left unchanged (for any other kind of callout, getting as far as trying
2234 RCPT means that the domain itself is ok).
2236 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2239 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2240 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2241 session, the daemon ignores ECONNECT errors and logs others; it now
2242 ignores EPIPE as well.
2244 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2245 (quoted-printable decoding).
2247 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2248 later the small subsequent patch to fix an introduced bug.
2250 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2252 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2254 PH/08 An error is now given if message_size_limit is specified negative.
2256 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2257 to be given (somewhat) arbitrary names.
2259 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2262 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2263 miscellaneous code fixes
2265 PH/10 Added the log_reject_target ACL modifier to specify where to log
2268 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2269 hostname. This is wrong, because it relates to the incoming message (and
2270 probably the interface on which it is arriving) and not to the outgoing
2271 callout (which could be using a different interface). This has been
2272 changed to use the value of the helo_data option from the smtp transport
2273 instead - this is what is used when a message is actually being sent. If
2274 there is no remote transport (possible with a router that sets up host
2275 addresses), $smtp_active_hostname is used.
2277 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2278 tweaks were necessary in order to get it to work (see also 21 below):
2279 (a) The code assumed that strncpy() returns a negative number on buffer
2280 overflow, which isn't the case. Replaced with Exim's string_format()
2282 (b) There were several signed/unsigned issues. I just did the minimum
2283 hacking in of casts. There is scope for a larger refactoring.
2284 (c) The code used strcasecmp() which is not a standard C function.
2285 Replaced with Exim's strcmpic() function.
2286 (d) The code set only $1; it now sets $auth1 as well.
2287 (e) A simple test gave the error "authentication client didn't specify
2288 service in request". It would seem that Dovecot has changed its
2289 interface. Fortunately there's a specification; I followed it and
2290 changed what the client sends and it appears to be working now.
2292 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2295 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2296 address A is aliased to B and C, where B exists and C does not. Without
2297 -v the output is "A verified" because verification stops after a
2298 successful redirection if more than one address is generated. However,
2299 with -v the child addresses are also verified. Exim was outputting "A
2300 failed to verify" and then showing the successful verification for C,
2301 with its parentage. It now outputs "B failed to verify", showing B's
2302 parentage before showing the successful verification of C.
2304 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2305 look up a TXT record in a specific list after matching in a combined
2308 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2309 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2310 they consult the DNS. I had assumed they would set it the way they
2311 wanted; and indeed my experiments on Linux seem to show that in some
2312 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2313 To be on the safe side, however, I have now made the interface to
2314 host_find_byname() similar to host_find_bydns(), with an argument
2315 containing the DNS resolver options. The host_find_byname() function now
2316 sets these options at its start, just as host_find_bydns() does. The smtp
2317 transport options dns_qualify_single and dns_search_parents are passed to
2318 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2319 of host_find_byname() use the default settings of RES_DEFNAMES
2320 (qualify_single) but not RES_DNSRCH (search_parents).
2322 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2323 spool_read_header() do less string testing, by means of a preliminary
2324 switch on the second character of optional "-foo" lines. (This is
2325 overdue, caused by the large number of possibilities that now exist.
2326 Originally there were few.) While I was there, I also converted the
2327 str(n)cmp tests so they don't re-test the leading "-" and the first
2328 character, in the hope this might squeeze out yet more improvement.
2330 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2331 flag allowing group syntax was set by the header_syntax check but not
2332 turned off, possible causing trouble later; (2) The flag was not being
2333 set at all for the header_verify test, causing "group"-style headers to
2334 be rejected. I have now set it in this case, and also caused header_
2335 verify to ignore an empty address taken from a group. While doing this, I
2336 came across some other cases where the code for allowing group syntax
2337 while scanning a header line wasn't quite right (mostly, not resetting
2338 the flag correctly in the right place). These bugs could have caused
2339 trouble for malformed header lines. I hope it is now all correct.
2341 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2342 with the "reply" argument non-NULL. The code, however (which originally
2343 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2344 but it didn't always do it. This confused somebody who was copying the
2345 code for some other use. I have removed all the tests.
2347 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2348 feature that was used to support insecure browsers during the U.S. crypto
2349 embargo. It requires special client support, and Exim is probably the
2350 only MTA that supported it -- and would never use it because real RSA is
2351 always available. This code has been removed, because it had the bad
2352 effect of slowing Exim down by computing (never used) parameters for the
2353 RSA_EXPORT functionality.
2355 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2356 authenticator to fail if there's a tab character in the incoming data
2357 (there should never be unless someone is messing about, as it's supposed
2358 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2359 if the connection is using TLS or if the remote IP is the same as the
2360 local IP, and the "valid-client-cert option" if a client certificate has
2363 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2364 authenticators. This can be used for authorization after authentication
2365 succeeds. (In the case of plaintext, it servers for both authentication
2368 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2369 if any retry times were supplied.
2371 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2372 connection, where there is no client IP address to check. In this
2373 situation, the verify now always succeeds.
2375 PH/25 Applied John Jetmore's -Mset patch.
2377 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2379 PH/27 In a string expansion for a processed (not raw) header when multiple
2380 headers of the same name were present, leading whitespace was being
2381 removed from all of them, but trailing whitespace was being removed only
2382 from the last one. Now trailing whitespace is removed from each header
2383 before concatenation. Completely empty headers in a concatenation (as
2384 before) are ignored.
2386 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2387 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2389 PH/29 [Removed. This was a change that I later backed out, and forgot to
2390 correct the ChangeLog entry (that I had efficiently created) before
2391 committing the later change.]
2393 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2394 address errors (4xx response to RCPT) over the same connection as other
2395 messages routed to the same hosts. Such deliveries are always "forced",
2396 so retry times are not inspected. This resulted in far too many retries
2397 for the affected addresses. The effect occurred only when there were more
2398 hosts than the hosts_max_try setting in the smtp transport when it had
2399 the 4xx errors. Those hosts that it had tried were not added to the list
2400 of hosts for which the message was waiting, so if all were tried, there
2401 was no problem. Two fixes have been applied:
2403 (i) If there are any address or message errors in an SMTP delivery, none
2404 of the hosts (tried or untried) are now added to the list of hosts
2405 for which the message is waiting, so the message should not be a
2406 candidate for sending over the same connection that was used for a
2407 successful delivery of some other message. This seems entirely
2408 reasonable: after all the message is NOT "waiting for some host".
2409 This is so "obvious" that I'm not sure why it wasn't done
2410 previously. Hope I haven't missed anything, but it can't do any
2411 harm, as the worst effect is to miss an optimization.
2413 (ii) If, despite (i), such a delivery is accidentally attempted, the
2414 routing retry time is respected, so at least it doesn't keep
2415 hammering the server.
2417 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2418 in ${readsocket because some servers need this prod.
2420 PH/32 Added some extra debug output when updating a wait-xxx database.
2422 PH/33 The hint "could be header name not terminated by colon", which has been
2423 given for certain expansion errors for a long time, was not being given
2424 for the ${if def:h_colon_omitted{... case.
2426 PH/34 The spec says: "With one important exception, whenever a domain list is
2427 being scanned, $domain contains the subject domain." There was at least
2428 one case where this was not true.
2430 PH/35 The error "getsockname() failed: connection reset by peer" was being
2431 written to the panic log as well as the main log, but it isn't really
2432 panic-worthy as it just means the connection died rather early on. I have
2433 removed the panic log writing for the ECONNRESET error when getsockname()
2436 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2437 runs only) independently of the message's sender address. This meant
2438 that, if the 4xx error was in fact related to the sender, a different
2439 message to the same recipient with a different sender could confuse
2440 things. In particular, this can happen when sending to a greylisting
2441 server, but other circumstances could also provoke similar problems.
2442 I have changed the default so that the retry time for these errors is now
2443 based a combination of the sender and recipient addresses. This change
2444 can be overridden by setting address_retry_include_sender=false in the
2447 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2448 remote server are returned as part of bounce messages. This was not
2449 happening for LMTP over a pipe (the lmtp transport), but now it is the
2450 same for both kinds of LMTP.
2452 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2453 in header lines that were in fact CNAMEs. This is no longer the case.
2455 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2456 and queue runs started by the daemon processed all messages. This has
2457 been fixed so that -R and -S can now usefully be given with -q<time>.
2459 PH/40 Import PCRE release 6.7 (fixes some bugs).
2461 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2463 PH/42 Give an error if -q is specified more than once.
2465 PH/43 Renamed the variables $interface_address and $interface_port as
2466 $received_ip_address and $received_port, to make it clear that these
2467 values apply to message reception, and not to the outgoing interface when
2468 a message is delivered. (The old names remain recognized, of course.)
2470 PH/44 There was no timeout on the connect() call when using a Unix domain
2471 socket in the ${readsocket expansion. There now is.
2473 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2474 be meaningful with "accept".
2476 SC/01 Eximstats V1.43
2477 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2479 SC/02 Eximstats V1.44
2480 Use a glob alias rather than an array ref in the generated
2481 parser. This improves both readability and performance.
2483 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2484 Collect SpamAssassin and rejection statistics.
2485 Don't display local sender or destination tables unless
2486 there is data to show.
2487 Added average volumes into the top table text output.
2489 SC/04 Eximstats V1.46
2490 Collect data on the number of addresses (recipients)
2491 as well as the number of messages.
2493 SC/05 Eximstats V1.47
2494 Added 'Message too big' to the list of mail rejection
2495 reasons (thanks to Marco Gaiarin).
2497 SC/06 Eximstats V1.48
2498 Mainlog lines which have GMT offsets and are too short to
2499 have a flag are now skipped.
2501 SC/07 Eximstats V1.49 (Alain Williams)
2502 Added the -emptyok flag.
2504 SC/08 Eximstats V1.50
2505 Fixes for obtaining the IP address from reject messages.
2507 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2508 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2509 whitespace changes from 4.64-PH/27
2511 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2514 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2515 are found, allow negative numbers in numeric criteria)
2517 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2519 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2520 to match changes made in 4.64-PH/43
2522 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2524 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2525 "domain" is actually an IP address, and force a failure. This locks out
2526 those revolvers/nameservers that support "A-for-A" lookups, in
2527 contravention of the specifications.
2529 PH/48 When a host name was looked up from an IP address, and the subsequent
2530 forward lookup of the name timed out, the host name was left in
2531 $sender_host_name, contrary to the specification.
2533 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2534 restricted to single-key lookups, Exim was not diagnosing an error if
2535 * or *@ was used with a query-style lookup.
2537 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2539 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2540 long ago, but noone interested enough thought of it. Let's just say that
2541 the "1.1" means that there are some new functions that weren't there at
2542 some point in the past.
2544 PH/51 Error processing for expansion failure of helo_data from an smtp
2545 transport during callout processing was broken.
2547 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2548 tested/used via the -bh/-bhc/-bs options.
2550 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2551 bug, fixed in subsequent PCRE releases).
2553 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2554 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2556 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2562 SC/01 Use a glob alias rather than an array ref in eximstats generated
2563 parser. This improves both readability and performance.
2565 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2566 Don't display local sender or destination tables in eximstats unless
2567 there is data to show.
2568 Added average volumes into the eximstats top table text output.
2570 SC/03 Collect data on the number of addresses (recipients) as well
2571 as the number of messages in eximstats.
2573 TF/01 Correct an error in the documentation for the redirect router. Exim
2574 does (usually) call initgroups() when daemonizing.
2576 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2577 with consistent privilege compared to when running as a daemon.
2579 TF/03 Note in the spec that $authenticated_id is not set for local
2580 submissions from trusted users.
2582 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2583 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2585 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2586 by adding some example configuration directives to the default
2587 configuration file. A little bit of work is required to uncomment the
2588 directives and define how usernames and passwords are checked, but
2589 there is now a framework to start from.
2591 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2592 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2593 without this. I don't know how relevant this is to other LDAP libraries.
2595 PH/02 Add the verb name to the "unknown ACL verb" error.
2597 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2599 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2601 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2602 directory not expanded when it should be if an expanded home directory
2603 was set for the address (which is overridden by the transport).
2605 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2608 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2609 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2610 because it is too late at that time, and has no effect.
2612 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2613 security issue with \' (bugzilla #107). I could not use the
2614 PQescapeStringConn() function, because it needs a PGconn value as one of
2617 PH/08 When testing addresses using -bt, indicate those final addresses that
2618 are duplicates that would not cause an additional delivery. At least one
2619 person was confused, thinking that -bt output corresponded to deliveries.
2620 (Suppressing duplicates isn't a good idea as you lose the information
2621 about possibly different redirections that led to the duplicates.)
2623 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2624 systems where poll() doesn't work, in particular OS X.
2626 PH/10 Added more information to debugging output for retry time not reached.
2628 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2629 operations in malware.c.
2631 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2634 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2635 syslog_duplication set false, log lines that would normally be written
2636 both the the main log and to the reject log were not written to syslog at
2639 PH/14 In the default configuration, change the use of "message" in ACL warn
2640 statements to "add_header".
2642 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2643 not followed by a command (e.g. "seen endif").
2645 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2646 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2649 PH/17 Added extra conditions to the default value of delay_warning_condition
2653 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2654 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2655 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2658 The Auto-Submitted: and various List- headers are standardised, whereas I
2659 don't think Precedence: ever was.
2661 PH/18 Refactored debugging code in route_finduser() to show more information,
2662 in particular, the error code if getpwnam() issues one.
2664 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2665 This is apparently needed in addition to the PH/07 change above to avoid
2666 any possible encoding problems.
2668 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2669 but not after initializing Perl.
2671 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2672 output them only if debugging. By default they are written stderr,
2673 apparently, which is not desirable.
2675 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2678 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2681 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2683 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2684 authenticated or an ident call has been made. Suppress the default
2685 values for $authenticated_id and $authenticated_sender (but permit -oMai
2686 and -oMas) when testing with -bh.
2688 PH/24 Re-jigged the order of the tests in the default configuration so that the
2689 tests for valid domains and recipients precede the DNS black list and CSA
2690 tests, on the grounds that those ones are more expensive.
2692 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2693 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2694 This bug exists in every version of Exim that I still have, right back to
2697 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2698 However, an attempt to turn on case-sensitivity in a regex key by
2699 including (?-i) didn't work because the subject string was already
2700 lowercased, and the effects were non-intuitive. It turns out that a
2701 one-line patch can be used to allow (?-i) to work as expected.
2707 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2708 other effects) broke the use of negated acl sub-conditions.
2710 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2713 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2714 "Deny" causes Exim to reject the incoming connection with a 554 error.
2715 Unfortunately, if there is a major crisis, such as a disk failure,
2716 tcp-wrappers gives "deny", whereas what one would like would be some
2717 kind of temporary error. A kludge has been added to help with this.
2718 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2719 554 error is used if errno is still zero or contains ENOENT (which occurs
2720 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2723 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2725 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2726 errors. Otherwise a message that provokes a temporary error (when other
2727 messages do not) can cause a whole host to time out.
2729 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2730 addresses were routed directly to files or pipes from a redirect router.
2731 File deliveries just didn't batch; pipe deliveries might have suffered
2734 PH/06 A failure to get a lock for a hints database would erroneously always say
2735 "Failed to get write lock", even when it was really a read lock.
2737 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2738 of 0600. This has been changed to use the value of the lockfile_mode
2739 option (which defaults to 0600).
2741 PH/08 Applied small patch from the Sieve maintainer.
2743 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2744 folder from quota calculations, a direct delivery into this folder messed
2745 up the contents of the maildirsize file. This was because the regex was
2746 used only to exclude .Trash (or whatever) when the size of the mailbox
2747 was calculated. There was no check that a delivery was happening into an
2748 excluded directory. This bug has been fixed by ignoring all quota
2749 processing for deliveries into excluded directories.
2751 PH/10 Added the maildirfolder_create_regex option to appendfile.
2757 PH/01 The code for finding all the local interface addresses on a FreeBSD
2758 system running IPv6 was broken. This may well have applied to all BSD
2759 systems, as well as to others that have similar system calls. The broken
2760 code found IPv4 interfaces correctly, but gave incorrect values for the
2761 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2762 that it would not match correctly against @[] and not recognize the IPv6
2765 PH/02 The ipliteral router was not recognizing addresses of the form user@
2766 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2768 PH/03 Added disable_ipv6.
2770 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2771 decoded form, because it is most often used to construct To: headers
2772 lines in autoreplies, and the decoded form may well be syntactically
2773 invalid. However, $reply_address has leading white space removed, and all
2774 newlines turned into spaces so that the autoreply transport does not
2777 PH/05 If group was specified without a user on a router, and no group or user
2778 was specified on a transport, the group from the router was ignored.
2780 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2781 for visible compile-time settings that can be used to change these
2782 numbers, for those that want even more. Backwards compatibility with old
2783 spool files has been maintained. However, going back to a previous Exim
2784 release will lost any variables that are in spool files.
2786 PH/07 Two small changes when running in the test harness: increase delay when
2787 passing a TCP/IP connection to a new process, in case the original
2788 process has to generate a bounce, and remove special handling of
2789 127.0.0.2 (sic), which is no longer necessary.
2791 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2792 be the same on different OS.
2794 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2797 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2798 whether --show-vars was specified or not
2800 JJ/02 exipick: Added support for new ACL variable spool format introduced
2803 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2804 syntactically invalid From: or Reply-to: line, and a filter used this to
2805 generate an autoreply, and therefore failed to obtain an address for the
2806 autoreply, Exim could try to deliver to a non-existent relative file
2807 name, causing unrelated and misleading errors. What now happens is that
2808 it logs this as a hard delivery error, but does not attempt to create a
2811 PH/11 The exinext utility has a -C option for testing purposes, but although
2812 the given file was scanned by exinext itself; it wasn't being passed on
2813 when Exim was called.
2815 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2816 an end-of-file indication when reading a command response.
2818 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2819 compiled. In many other places in Exim, IPv6 addresses are always
2820 recognized, so I have changed this. It also means that IPv4 domain
2821 literals of the form [IPV4:n.n.n.n] are now always recognized.
2823 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2824 used if the router is not running as root, for example, when verifying at
2825 ACL time, or when using -bh. The debugging output from this situation was
2826 non-existent - all you got was a failure to exec. I have made two
2829 (a) Failures to set uid/gid, the current directory, or a process leader
2830 in a subprocess such as that created by queryprogram now generate
2831 suitable debugging output when -d is set.
2833 (b) The queryprogram router detects when it is not running as root,
2834 outputs suitable debugging information if -d is set, and then runs
2835 the subprocess without attempting to change uid/gid.
2837 PH/15 Minor change to Makefile for building test_host (undocumented testing
2840 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2841 additional section of a DNS packet that returns MX or SRV records.
2842 Instead, it always explicitly searches for A/AAAA records. This avoids
2843 major problems that occur when a DNS server includes only records of one
2844 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2845 fixed another bug: if SRV records were looked up and the corresponding
2846 address records were *not* found in the additional section, the port
2847 values from the SRV records were lost.
2849 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2850 using the correct key (the original address) when searching the retry
2851 rules in order to find which one to use for generating the retry hint.
2853 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2854 adding the default one. Similarly, if it contains a Reply-To: header, the
2855 errors_reply_to option, if set, is not used.
2857 PH/19 When calculating a retry time, Exim used to measure the "time since
2858 failure" by looking at the "first failed" field in the retry record. Now
2859 it does not use this if it is later than than the arrival time of the
2860 message. Instead it uses the arrival time. This makes for better
2861 behaviour in cases where some deliveries succeed, thus re-setting the
2862 "first failed" field. An example is a quota failure for a huge message
2863 when small messages continue to be delivered. Without this change, the
2864 "time since failure" will always be short, possible causing more frequent
2865 delivery attempts for the huge message than are intended.
2866 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2868 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2869 $1, $2, $3) because the numerical variables can be reset during some
2870 expansion items (e.g. "match"), thereby losing the authentication data.
2872 PH/21 Make -bV show the size of off_t variables so that the test suite can
2873 decide whether to run tests for quotas > 2G.
2875 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2876 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2877 filecount value is greater than 2G or if a quota value is greater than 2G
2878 on a system where the size of off_t is not greater than 4, a panic error
2881 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2882 never match. The debug and -bh output now contains an explicit error
2883 message indicating a malformed IPv4 address or mask.
2885 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2886 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2887 PH/23 above applies.
2889 PH/25 Do not write to syslog when running in the test harness. The only
2890 occasion when this arises is a failure to open the main or panic logs
2891 (for which there is an explicit test).
2893 PH/26 Added the /no_tell option to "control=freeze".
2895 PH/27 If a host name lookup failed very early in a connection, for example, if
2896 the IP address matched host_lookup and the reverse lookup yielded a name
2897 that did not have a forward lookup, an error message of the form "no IP
2898 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2899 could be logged. Now it outputs the IP address instead of "NULL".
2901 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2902 allows the sender and the authenticated sender to be set when
2903 submitting a message from within Exim. Since child_open_exim() is
2904 documented for local_scan(), the new function should be too.
2906 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2907 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2908 results in an empty string is now treated as unset.
2910 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2912 PH/31 Added sender_verify_fail logging option.
2914 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2915 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2916 tidied the source and removed it altogether.
2918 PH/33 When a queue run was abandoned because the load average was too high, a
2919 log line was always written; now it is written only if the queue_run log
2920 selector is set. In addition, the log line for abandonment now contains
2921 information about the queue run such as the pid. This is always present
2922 in "start" and "stop" lines but was omitted from the "abandon" line.
2924 PH/34 Omit spaces between a header name and the colon in the error message that
2925 is given when verify = headers_syntax fails (if there are lots of them,
2926 the message gets confusing).
2928 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2929 names, as there are now some PTR records that contain slashes. This check
2930 is only to protect against broken name servers that fall over on strange
2931 characters, so the fact that it applies to all lookups doesn't matter.
2933 PH/36 Now that the new test suite is complete, we can remove some of the
2934 special code in Exim that was needed for the old test suite. For example,
2935 sorting DNS records because real resolvers return them in an arbitrary
2936 order. The new test suite's fake resolver always returns records in the
2939 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2940 bounces) except when queue_only is set, to avoid logging races between
2941 the different processes.
2943 PH/38 Panic-die if .include specifies a non-absolute path.
2945 PH/39 A tweak to the "H" retry rule from its user.
2947 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2948 a label. They prevented compilation on older perls.
2950 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2951 a warning to be raised on newish perls.
2953 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2954 on queue. Changes to match documented behaviour of showing count of
2955 messages matching specified criteria.
2957 PH/40 Changed the default ident timeout from 30s to 5s.
2959 PH/41 Added support for the use of login_cap features, on those BSD systems
2960 that have them, for controlling the resources used by pipe deliveries.
2962 PH/42 The content-scanning code uses fopen() to create files in which to put
2963 message data. Previously it was not paying any attention to the mode of
2964 the files. Exim runs with umask(0) because the rest of the code creates
2965 files with open(), and sets the required mode explicitly. Thus, these
2966 files were ending up world-writeable. This was not a big issue, because,
2967 being within the spool directory, they were not world-accessible. I have
2968 created a function called modefopen, which takes an additional mode
2969 argument. It sets umask(777), creates the file, chmods it to the required
2970 mode, then resets the umask. All the relevant calls to fopen() in the
2971 content scanning code have been changed to use this function.
2973 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2974 to 24 hours. This avoids potential overflow problems when processing G
2975 and H retry rules. I suspect nobody ever tinkers with this value.
2977 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2979 PH/45 When the plaintext authenticator is running as a client, the server's
2980 challenges are checked to ensure they are valid base64 strings. By
2981 default, the authentication attempt is cancelled if an invalid string is
2982 received. Setting client_ignore_invalid_base64 true ignores these errors.
2983 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2984 they are received. Thus, the responses can be made to depend on the
2985 challenges. If an invalid string is ignored, an empty string is placed in
2988 PH/46 Messages that are created by the autoreply transport now contains a
2989 References: header, in accordance with RFCs 2822 and 3834.
2991 PH/47 Added authenticated_sender_force to the smtp transport.
2993 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2995 PH/49 Installed latest patch from the Sieve maintainer.
2997 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2998 was also set, the appendfile transport was unnecessarily scanning a
2999 directory of message files (e.g. for maildir delivery) to find the count
3000 of files (along with the size), even though it did not need this
3001 information. It now does the scan only if it needs to find either the
3002 size of the count of files.
3004 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3006 PH/52 Two bugs concerned with error handling when the smtp transport is
3009 (i) Exim was not creating retry information for temporary errors given
3010 for individual recipients after the DATA command when the smtp transport
3011 was used in LMTP mode. This meant that they could be retried too
3012 frequently, and not timed out correctly.
3014 (ii) Exim was setting the flag that allows error details to be returned
3015 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3016 recipients that were returned after the DATA command.
3018 PH/53 This is related to PH/52, but is more general: for any failing address,
3019 when detailed error information was permitted to be returned to the
3020 sender, but the error was temporary, then after the final timeout, only
3021 "retry timeout exceeded" was returned. Now it returns the full error as
3022 well as "retry timeout exceeded".
3024 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3025 do this, and (what is worse) MTAs that accept it.
3027 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3028 will now be deprecated.
3030 PH/56 New os.c-cygwin from the Cygwin maintainer.
3032 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3033 formats (previously only available in exim formats via -bpr, -bpru,
3034 and -bpra. Now also available in native and exiqgrep formats)
3036 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3037 with very large, slow to parse queues
3039 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3041 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3043 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3044 responses to authentication challenges, though it was showing the
3045 challenges; (ii) I've removed the CR characters from the debug output for
3048 PH/58 Allow for the insertion of a newline as well as a space when a string
3049 is turned into more than one encoded-word during RFC 2047 encoding. The
3050 Sieve code now uses this.
3052 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3053 data_4xx, lost_connection, tls_required.
3055 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3056 message was being sent as an SMTP response.
3058 PH/61 Add -l and -k options to exicyclog.
3060 PH/62 When verifying, if an address was redirected to one new address, so that
3061 verification continued, and the new address failed or deferred after
3062 having set something in $address_data, the value of $address_data was not
3063 passed back to the ACL. This was different to the case when no
3064 redirection occurred. The value is now passed back in both cases.
3066 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3067 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3068 use login_cap.h, so on its own it isn't the distinguishing feature. The
3069 new name refers directly to the setclassresources() function.
3071 PH/65 Added configuration files for NetBSD3.
3073 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3075 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3076 is preferred over IPv4.
3078 PH/68 The bounce_return_message and bounce_return_body options were not being
3079 honoured for bounces generated during the reception of non-SMTP messages.
3080 In particular, this applied to messages rejected by the ACL. This bug has
3081 been fixed. However, if bounce_return_message is true and bounce_return_
3082 body is false, the headers that are returned for a non-SMTP message
3083 include only those that have been read before the error was detected.
3084 (In the case of an ACL rejection, they have all been read.)
3086 PH/69 The HTML version of the specification is now built in a directory called
3087 spec_html instead of spec.html, because the latter looks like a path with
3088 a MIME-type, and this confuses some software.
3090 PH/70 Catch two compiler warnings in sieve.c.
3092 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3093 function verify_get_ident() calls ip_connect() to connect a socket, but
3094 if the "connect()" function timed out, ip_connect() used to close the
3095 socket. However, verify_get_ident() also closes the socket later, and in
3096 between Exim writes to the log, which may get opened at this point. When
3097 the socket was closed in ip_connect(), the log could get the same file
3098 descriptor number as the socket. This naturally causes chaos. The fix is
3099 not to close the socket in ip_connect(); the socket should be closed by
3100 the function that creates it. There was only one place in the code where
3101 this was missing, in the iplookup router, which I don't think anybody now
3102 uses, but I've fixed it anyway.
3104 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3105 well as to direct DNS lookups. Otherwise the handling of names in host
3106 lists is inconsistent and therefore confusing.
3112 PH/01 Two changes to the default runtime configuration:
3114 (1) Move the checks for relay_from_hosts and authenticated clients from
3115 after to before the (commented out) DNS black list checks.
3117 (2) Add control=submission to the relay_from_hosts and authenticated
3118 clients checks, on the grounds that messages accepted by these
3119 statements are most likely to be submissions.
3121 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3123 (1) Generate an error if the third argument for the ${prvs expansion is
3126 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3129 (3) Reset the variables that are obtained from the first argument of
3130 ${prvscheck and used in the second argument before leaving the code,
3131 because their memory is reclaimed, so using them afterwards may do
3134 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3135 one (it's much easier than Tom thought :-).
3137 (5) Because of (4), we can now allow for the use of $prvscheck_result
3138 inside the third argument.
3140 PH/03 For some reason, the default setting of PATH when running a command from
3141 a pipe transport was just "/usr/bin". I have changed it to
3144 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3145 anything to be listed in the output from -bV.
3147 PH/05 When a filter generated an autoreply, the entire To: header line was
3148 quoted in the delivery log line, like this:
3150 => >A.N.Other <ano@some.domain> <original@ddress> ...
3152 This has been changed so that it extracts the operative address. There
3153 may be more than one such address. If so, they are comma-separated, like
3156 => >ano@some.domain,ona@other.domain <original@ddress> ...
3158 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3159 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3160 not being looked up in the rDNS to get a host name, Exim was showing the
3161 IP address twice in Received: lines, even though the IP addresses were
3162 identical. For example:
3164 Received: from [1.2.3.4] (helo=[1.2.3.4])
3166 However, if the real host name was known, it was omitting the HELO data
3167 if it matched the actual IP address. This has been tidied up so that it
3168 doesn't show the same IP address twice.
3170 PH/07 When both +timestamp and +memory debugging was on, the value given by
3171 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3172 called by the debug printing, thereby overwriting the timestamp buffer.
3173 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3175 PH/08 When the original message was included in an autoreply transport, it
3176 always said "this is a copy of the message, including all the headers",
3177 even if body_only or headers_only was set. It now gives an appropriate
3180 PH/09 Applied a patch from the Sieve maintainer which:
3182 o fixes some comments
3183 o adds the (disabled) notify extension core
3184 o adds some debug output for the result of if/elsif tests
3185 o points to the current vacation draft in the documentation
3186 and documents the missing references header update
3190 o fixes a bug in processing the envelope test (when testing
3191 multiple envelope elements, the last element determined the
3194 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3195 Electronic Mail") by including:
3197 Auto-submitted: auto-generated
3199 in the messages that it generates (bounce messages and others, such as
3200 warnings). In the case of bounce messages for non-SMTP messages, there was
3201 also a typo: it was using "Auto_submitted" (underscore instead of
3202 hyphen). Since every message generated by Exim is necessarily in response
3203 to another message, thes have all been changed to:
3205 Auto-Submitted: auto-replied
3207 in accordance with these statements in the RFC:
3209 The auto-replied keyword:
3211 - SHOULD be used on messages sent in direct response to another
3212 message by an automatic process,
3214 - MUST NOT be used on manually-generated messages,
3216 - MAY be used on Delivery Status Notifications (DSNs) and Message
3217 Disposition Notifications (MDNs),
3219 - MUST NOT be used on messages generated by automatic or periodic
3220 processes, except for messages which are automatic responses to
3223 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3224 to the default Received: header definition.
3226 PH/12 Added log selector acl_warn_skipped (default on).
3228 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3229 variables because (a) they are in the wrong storage pool and (b) even if
3230 they were copied, it wouldn't work properly because of the caching.
3232 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3233 checking when decoding. Apparently there are clients that generate
3234 overlong encoded strings. Why am I not surprised?
3236 PH/15 If the first argument of "${if match_address" was not empty, but did not
3237 contain an "@" character, Exim crashed. Now it writes a panic log message
3238 and treats the condition as false.
3240 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3242 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3243 extension "envelope-auth". The code is finished and in agreement with
3244 other implementations, but there is no documentation so far and in fact,
3245 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3246 not changing the active code.
3248 Print executed "if" and "elsif" statements when debugging is used. This
3249 helps a great deal to understand what a filter does.
3251 Document more things not specified clearly in RFC3028. I had all this
3252 sorted out, when out of a sudden new issues came to my mind. Oops."
3254 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3257 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3258 canonical form (as documented). However, after a host name lookup from
3259 the IP address, check_host() was doing a simple string comparison with
3260 addresses acquired from the DNS when checking that the found name did
3261 have the original IP as one of its addresses. Since any found IPv6
3262 addresses are likely to be in abbreviated form, the comparison could
3263 fail. Luckily, there already exists a function for doing the comparison
3264 by converting both addresses to binary, so now that is used instead of
3265 the text comparison.
3267 PH/20 There was another similar case to PH/19, when a complete host name was
3268 given in a host list; looking up its IP address could give an abbreviated
3269 form, whereas the current host's name might or might not be abbreviated.
3270 The same fix has been applied.
3276 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3277 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3280 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3281 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3283 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3285 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3286 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3287 Also, for "Auto-Submitted", treat anything other than "no" as
3288 non-personal, in accordance with RFC 3834. (Previously it treated
3289 anything starting "auto-" as non-personal.)
3291 TF/01 The control=submission/name=... option had a problem with syntax
3292 errors if the name included a slash character. The /name= option
3293 now slurps the rest of the string, so it can include any characters
3294 but it must come last in the list of options (after /sender_retain
3297 PH/05 Some modifications to the interface to the fake nameserver for the new
3305 TK/01 Added the "success_on_redirect" address verification option. See
3306 NewStuff for rationale and an example.
3308 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3310 PH/02 Patch to exigrep to allow it to work on syslog lines.
3312 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3313 fread() to skip over the body file's header line, because in Cygwin the
3314 header line is locked and is inaccessible.
3316 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3317 co-exist for some time) to make it clear that it is the Exim ID that is
3318 referenced, not the Message-ID: header line.
3320 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3321 string_format() function, because snprintf() does not exist on all
3324 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3325 ${stat: expansion item.
3327 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3328 protocol synchronization error", to keep the pedants happy.
3330 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3331 well as for IRIX systems, when gcc is being used. See the host.c source
3334 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3336 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3339 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3340 transport and to the smtp transport in LMTP mode.
3342 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3344 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3345 run a filter in a subprocess. This could lead to confusion in subsequent
3346 lookups in the parent process. There should also be a search_tidyup() at
3347 the end of the subprocess.
3349 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3350 only if the host matched helo_try_verify_hosts, which caused the
3351 verification to occur when the EHLO/HELO command was issued. The ACL just
3352 tested the remembered result. Now, if a previous verification attempt has
3353 not happened, "verify = helo" does it there and then.
3355 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3357 TK/03 Fix log output including CR from clamd.
3359 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3360 exist provoked a memory error which could cause a segfault.
3362 PH/15 Installed PCRE 6.2
3364 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3366 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3367 of the problem. Specifically, suggested +O2 rather than +O1 for the
3370 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3372 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3373 record, the port was not correctly logged when the outgoing_port log
3374 selector was set (it logged the transort's default port).
3376 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3377 fallback_hosts, and "hosts" in the smtp transport.
3379 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3380 host errors such as "Connection refused".
3382 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3383 authentication with radiusclient 0.4.9:
3385 - Error returned from rc_read_config was caught wrongly
3386 - Username/password not passed on to radius server due to wrong length.
3388 The presumption is that some radiusclient API changes for 4.51/PH/17
3389 were not taken care of correctly. The code is still untested by me (my
3390 Linux distribution still has 0.3.2 of radiusclient), but it was
3391 contributed by a Radius user.
3393 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3394 expanding the "port" option of the smtp transport.
3396 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3397 while reading a MIME header. Thanks to Tom Hughes for a patch.
3399 PH/24 Include config.h inside local_scan.h so that configuration settings are
3402 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3403 that in an ACL for RCPT (for example), you can examine exactly what was
3406 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3407 commands, but it was not correctly comparing the address with the actual
3408 client host address. Thus, it would show the EHLO address in Received:
3409 header lines when this was not necessary.
3411 PH/27 Added the % operator to ${eval:}.
3413 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3414 it should be ignoring failures (because with -C, for example, it has lost
3415 privilege). It wasn't ignoring creation failures other than "already
3418 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3421 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3422 because an input line was too long, either on its own, or by virtue of
3423 too many continuations, the temporary file was not being removed, and the
3424 return code was incorrect.
3426 PH/31 Missing "BOOL" in function definition in filtertest.c.
3428 PH/32 Applied Sieve patches from the maintainer.
3430 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3432 PH/33 Added "verify = not_blind".
3434 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3435 Local/Makefile (with some defaults set). These are used in built scripts
3436 such as exicyclog, but they have never been used in the exim_install
3437 script (though there are many overriding facilities there). I have
3438 arranged that the exim_install script now takes note of these two
3441 PH/35 Installed configuration files for Dragonfly.
3443 PH/36 When a locally submitted message by a trusted user did not contain a
3444 From: header, and the sender address was obtained from -f or from an SMTP
3445 MAIL command, and the trusted user did not use -F to supply a sender
3446 name, $originator_name was incorrectly used when constructing a From:
3447 header. Furthermore, $originator_name was used for submission mode
3448 messages from external hosts without From: headers in a similar way,
3449 which is clearly wrong.
3451 PH/37 Added control=suppress_local_fixups.
3453 PH/38 When log_selector = +received_sender was set, and the addition of the
3454 sender made the log line's construction buffer exactly full, or one byte
3455 less than full, an overflow happened when the terminating "\n" was
3458 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3459 when the result of a list match is failure because a DNS lookup failed.
3461 PH/40 RM_COMMAND is now used in the building process.
3463 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3464 the "build-* directories that it finds.
3466 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3467 address in a domain literal was a prefix of an interface address.
3469 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3470 when verifying a sender address, unless rewrite_headers is false.
3472 PH/44 Wrote a long comment about why errors_to addresses are verified as
3473 recipients, not senders.
3475 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3476 the ratelimit ACL was added.
3478 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3480 PH/46 Added extra information about PostgreSQL errors to the error string.
3482 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3483 avoiding the need to install special zones in a real server. This is
3484 backwards compatible; if it can't find the fake resolver, it drops back.
3485 Thus, both old and new test suites can be run.
3487 TF/02 Added util/ratelimit.pl
3489 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3490 clock is set back in time.
3492 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3493 Candler <B.Candler@pobox.com>.
3495 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3496 OK for addresses that are the result of redirections.
3498 PH/48 A number of further additions for the benefit of the new test suite,
3499 including a fake gethostbyname() that interfaces to the fake DNS resolver
3502 TF/06 The fix for widen_domains has also been applied to qualify_single and
3503 search_parents which are the other dnslookup options that can cause
3506 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3509 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3511 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3512 service" if it encountered an MX record with an empty target hostname.
3513 The message is now "an MX or SRV record indicated no SMTP service".
3515 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3516 if the DNS of the sending site is misconfigured. This is quite a
3517 common situation. This change restores the behaviour of treating a
3518 helo verification defer as a failure.
3520 PH/51 If self=fail was set on a router, the bounce message did not include the
3521 actual error message.
3527 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3529 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3530 command itself ended in error, the underlying message about the transport
3531 filter timeout was being overwritten with the pipe command error. Now the
3532 underlying error message should be appended to the second error message.
3534 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3536 PH/02 Reduce the amount of output that "make" produces by default. Full output
3537 can still be requested.
3539 PH/03 The warning log line about a condition test deferring for a "warn" verb
3540 was being output only once per connection, rather than after each
3541 occurrence (because it was using the same function as for successful
3542 "warn" verbs). This seems wrong, so I have changed it.
3544 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3545 it should not have, which might have caused a crash in the right
3546 circumstances, but probably never did.
3548 PH/04 Installed a modified version of Tony Finch's patch to make submission
3549 mode fix the return path as well as the Sender: header line, and to
3550 add a /name= option so that you can make the user's friendly name appear
3553 TF/03 Added the control = fakedefer ACL modifier.
3555 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3556 Mark Lowes for thorough testing.
3558 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3560 TK/03 Merged latest SRS patch from Miles Wilton.
3562 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3563 with the definition in sysexits.h (which is #included earlier).
3564 Fortunately, Exim does not actually use EX_OK. The code used to try to
3565 preserve the sysexits.h value, by assuming that macro definitions were
3566 scanned for macro replacements. I have been disabused of this notion,
3567 so now the code just undefines EX_OK before #including unistd.h.
3569 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3570 in the smtp transport. When a block could not be written in a single
3571 write() function, the timeout was being re-applied to each part-write.
3572 This seems wrong - if the receiver was accepting one byte at a time it
3573 would take for ever. The timeout is now adjusted when this happens. It
3574 doesn't have to be particularly precise.
3576 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3577 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3579 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3580 without a domain if the check to <postmaster@domain> fails.
3582 SC/01 Eximstats: added -xls and the ability to specify output files
3583 (patch written by Frank Heydlauf).
3585 SC/02 Eximstats: use FileHandles for outputting results.
3587 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3589 SC/04 Eximstats: fixed display of large numbers with -nvr option
3591 SC/05 Eximstats: fixed merging of reports with empty tables.
3593 SC/06 Eximstats: added the -include_original_destination flag
3595 SC/07 Eximstats: removed tabs and trailing whitespace.
3597 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3599 TK/06 MBOX spool code: Add real "From " MBOX separator line
3600 so the .eml file is really in mbox format (even though
3601 most programs do not really care). Patch from Alex Miller.
3603 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3604 The latter is generated from $received_to and is only set if the
3605 message has one envelope recipient. SA can use these headers,
3606 obviously out-of-the-box. Patch from Alex Miller.
3608 PH/08 The ${def test on a variable was returning false if the variable's
3609 value was "0", contrary to what the specification has always said!
3610 The result should be true unless the variable is empty.
3612 PH/09 The syntax error of a character other than { following "${if
3613 def:variable_name" (after optional whitespace) was not being diagnosed.
3614 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3615 accidental colon was present, for example, could give incorrect results.
3617 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3618 result is used (not including appendfile, where other changes are about
3621 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3622 This involved changing a lot of size variables from int to off_t. It
3623 should work with maildirs and everything.
3625 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3626 spamd dying while we are connected to it.
3628 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3631 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3632 function for BDB 4.3.
3634 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3636 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3637 the sources for PCRE have been re-arranged and more files are now
3640 PH/15 The code I had for printing potentially long long variables in PH/11
3641 above was not the best (it lost precision). The length of off_t variables
3642 is now inspected at build time, and an appropriate printing format (%ld
3643 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3644 to be "long long int" or "long int". This is needed for the internal
3645 formatting function string_vformat().
3647 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3648 the configuration file to be ":syslog", then the script "guesses" where
3649 the logs files are, rather than using the compiled in default. In our
3650 case the guess is not the same as the compiled default, so the script
3651 suddenly stopped working when I started to use syslog. The patch checks
3652 to see if log_file_path is "". If so, it attempts to read it from exim
3653 with no configuration file to get the compiled in version, before it
3654 falls back to the previous guessing code."
3656 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3657 implementing BATV in an Exim configuration. See NewStuff for the gory
3660 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3661 Makefile that are specific to HP-UX.
3663 PH/18 If the "use_postmaster" option was set for a recipient callout together
3664 with the "random" option, the postmaster address was used as the MAIL
3665 FROM address for the random test, but not for the subsequent recipient
3666 test. It is now used for both.
3668 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3669 patch removes a few documentation additions to RFC 3028, because the
3670 latest draft now contains them. It adds the new en;ascii-case comparator
3671 and a new error check for 8bit text in MIME parts. Comparator and
3672 require names are now matched exactly. I enabled the subaddress
3673 extension, but it is not well tested yet (read: it works for me)."
3675 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3676 rework some of the code of TK/09 above to avoid the hardwired use of
3677 "%lld" and "long long". Replaced the call to snprintf() with a call to
3680 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3681 records point to non-existent hosts", "retry timeout exceeded", and
3682 "retry time not reached for any host after a long failure period".
3684 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3685 experimental DomainKeys support:
3687 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3688 (2) On an error such as an illegally used "control", the wrong name for
3689 the control was given.
3691 These problems did NOT occur unless DomainKeys support was compiled.
3693 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3695 PH/24 Added ${if match_ip condition.
3697 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3698 databases so that it will be absolutely obvious if a crash occurs in the
3699 DB library. This is a regular occurrence (often caused by mis-matched
3702 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3703 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3704 picked up on a user's system that detects such things. There doesn't seem
3705 to be a gcc warning option for this - only an attribute that has to be
3706 put on the function's prototype. It seems that in Fedora Core 4 they have
3707 set this on a number of new functions. No doubt there will be more in due
3710 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3711 specify a transport. However, if an address that was verified by such a
3712 router was the subject of a callout, Exim crashed because it tried to
3713 read the rcpt_include_affixes from the non-existent transport. Now it
3714 just assumes that the setting of that option is false. This bug was
3715 introduced by 4.51/PH/31.
3717 PH/28 Changed -d+all to exclude +memory, because that information is very
3718 rarely of interest, but it makes the output a lot bigger. People tend to
3719 do -d+all out of habit.
3721 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3722 code in os-type was giving problems when libc.so lives in lib64, like on
3725 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3726 aren't the modern standard, and it seems that some systems' include files
3727 don't always have them. Exim was already checking for some of the newer
3728 ones like T_AAAA, and defining it itself. I've added checks for all the
3729 record types that Exim uses.
3731 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3732 not automatically generating a new one, as it is supposed to. This
3733 prevented TLS from working. If the file did exist, but contained invalid
3734 data, a new version was generated, as expected. It was only the case of a
3735 non-existent file that was broken.
3737 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3738 with a change in libdomainkeys > 0.64.
3740 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3741 from DNS. If the selector record carries the flag, it now has
3742 precedence over the domain-wide flag.
3744 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3746 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3747 the use of a port name that isn't defined in /etc/services) occurred, the
3748 message was deferred as in a normal delivery, and thus remained on the
3749 spool, instead of being failed because of the mua_wrapper setting. This
3750 is now fixed, and I tidied up some of the mua_wrapper messages at the
3753 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3754 the messages in a hash of arrays rather than using individual hashes.
3755 This is a bit cleaner and results in dramatic memory savings, albeit
3756 at a slight CPU cost.
3758 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3759 as requested by Marc Sherman.
3761 SC/10 Eximstats: added histograms for user specified patterns as requested
3764 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3766 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3767 fopen() in the content-scanning modules that did not already have it.
3773 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3774 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3776 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3778 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3780 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3781 to test Sieve filters that use "vacation".
3783 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3784 that changes the way the GnuTLS parameters are stored in the cache file.
3785 The new format can be generated externally. For backward compatibility,
3786 if the data in the cache doesn't make sense, Exim assumes it has read an
3787 old-format file, and it generates new data and writes a new file. This
3788 means that you can't go back to an older release without removing the
3791 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3792 work if there are any delivery delays because "one_time" forces the
3793 parent to be marked "delivered", so its unseen clone is never tried
3794 again. For this reason, Exim now forbids the simultaneous setting of
3797 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3798 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3799 entry if you want to know the details. The fix, however, neglected to
3800 consider the case when local delivery batching is involved. The test for
3801 "previously delivered" was not happening when checking to see if an
3802 address could be batched with a previous (undelivered) one; under
3803 certain circumstances this could lead to multiple deliveries to the same
3806 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3807 in its include files, and this causes problems building Exim.
3809 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3810 header_syntax) but Exim was just ignoring anything given after a slash.
3811 In particular, this caused confusion with an attempt to use "verify =
3812 reverse_host_lookup/defer_ok". An error is now given when options are
3813 supplied for verify items that do not have them. (Maybe reverse_host_
3814 lookup should have a defer_ok option, but that's a different point.)
3816 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3817 defined by RFC 821) to 2048, because there were problems with some AUTH
3818 commands, and RFC 1869 says the size should be increased for extended
3819 SMTP commands that take arguments.
3821 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3824 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3825 "unknown" error; now it says that the functionality isn't in the binary.
3827 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3828 an address' error message when a string expansion fails (syntax or
3829 whatever). Otherwise the password may appear in the log. Following change
3830 PH/42 below, there is no longer a chance of it appearing in a bounce
3833 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3835 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3836 of its addresses was ever tried. (Bugzilla bug #2.)
3838 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3839 the result incorrectly in the debug output. (It correctly added a newline
3840 to what was transported.)
3842 TF/01 Added $received_time.
3844 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3845 commented out examples of how to interface to a virus scanner and to
3846 SpamAssassin. Also added commented examples of av_scanner and
3847 spamd_address settings.
3849 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3850 and controls are allowed in which ACLs. There were a couple of minor
3851 errors. Some of the entries in the conditions table (which is a table of
3852 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3853 negation of where the condition IS allowed.
3855 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3857 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3858 header file does not have a version number, so I've had to invent a new
3859 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3860 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3861 radiusclient), but it was contributed by a Radius user.
3863 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3864 files or directories, for OpenSSL.
3866 PH/19 When an Exim process that is running as root has to create an Exim log
3867 file, it does so in a subprocess that runs as exim:exim so as to get the
3868 ownership right at creation (otherwise, other Exim processes might see
3869 the file with the wrong ownership). There was no test for failure of this
3870 fork() call, which would lead to the process getting stuck as it waited
3871 for a non-existent subprocess. Forks do occasionally fail when resources
3872 run out. I reviewed all the other calls to fork(); they all seem to check
3875 PH/20 When checking for unexpected SMTP input at connect time (before writing
3876 the banner), Exim was not dealing correctly with a non-positive return
3877 from the read() function. If the client had disconnected by this time,
3878 the result was a log entry for a synchronization error with an empty
3879 string after "input=" when read() returned zero. If read() returned -1
3880 (an event I could not check), uninitialized data bytes were printed.
3881 There were reports of junk text (parts of files, etc) appearing after
3884 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3886 PH/22 Added support for macro redefinition, and (re)definition in between
3887 driver and ACL definitions.
3889 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3890 forgetting to use the resulting value; it was using the unexpanded value.
3892 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3893 hadn't been configured. The fix is from Juergen Kreileder, who
3894 understands it better than I do:
3896 "Here's what I see happening with three configured cyrus_sasl
3897 authenticators configured (plain, login, cram-md5):
3899 On startup auth_cyrus_sasl_init() gets called for each of these.
3900 This means three calls to sasl_listmech() without a specified mech_list.
3901 => SASL tests which mechs of all available mechs actually work
3902 => three warnings about OTP not working
3903 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3905 With the patch, sasl_listmech() also gets called three times. But now
3906 SASL's mech_list option is set to the server_mech specified in the the
3907 authenticator. Or in other words, the answer from sasl_listmech()
3908 gets limited to just the mech you're testing for (which is different
3910 => the return list contains just 'plain' or 'login', 'cram-md5' or
3911 nothing depending on the value of ob->server_mech.
3913 I've just tested the patch: Authentication still works fine,
3914 unavailable mechs specified in the exim configuration are still
3915 caught, and the auth.log warnings about OTP are gone."
3917 PH/25 When debugging is enabled, the contents of the command line are added
3918 to the debugging output, even when log_selector=+arguments is not
3921 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3922 answer is "GNU", and only if the return is "GNU/something" is the answer
3925 PH/27 $acl_verify_message is now set immediately after the failure of a
3926 verification in an ACL, and so is available in subsequent modifiers. In
3927 particular, the message can be preserved by coding like this:
3929 warn !verify = sender
3930 set acl_m0 = $acl_verify_message
3932 Previously, $acl_verify_message was set only while expanding "message"
3933 and "log_message" when a very denied access.
3935 PH/28 Modified OS/os.c-Linux with
3937 -#ifndef OS_LOAD_AVERAGE
3938 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3940 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3941 nomenclature these days.)
3943 PH/29 Installed patch from the Sieve maintainer that adds the options
3944 sieve_useraddress and sieve_subaddress to the redirect router.
3946 PH/30 In these circumstances:
3947 . Two addresses routed to the same list of hosts;
3948 . First host does not offer TLS;
3949 . First host accepts first address;
3950 . First host gives temporary error to second address;
3951 . Second host offers TLS and a TLS session is established;
3952 . Second host accepts second address.
3953 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3954 and peerdn, if requested) that were in fact used only for the second
3957 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3958 attention to any local part prefix or suffix that was matched by the
3959 router that accepted the address. It now behaves in the same way as it
3960 does for delivery: the affixes are removed from the local part unless
3961 rcpt_include_affixes is set on the transport.
3963 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3964 timeout during the DATA phase of an incoming message.
3966 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3967 applied a patch sanctioned by the Sieve maintainer.
3969 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3970 the uid or gid is negative. A case of a negative gid caused this to be
3971 noticed. The fix allows for either to be negative.
3973 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3974 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3977 PH/36 The change PH/12 above was broken. Fixed it.
3979 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3980 the grounds that routing the same address twice would always produce the
3981 same answer. This might have been true once, but it is certainly no
3982 longer true now. Routing a child address may depend on the previous
3983 routing that produced that child. Some complicated redirection strategies
3984 went wrong when messages had multiple recipients, and made Exim's
3985 behaviour dependent on the order in which the addresses were given.
3987 I have moved the duplicate checking until after the routing is complete.
3988 Exim scans the addresses that are assigned to local and remote
3989 transports, and removes any duplicates. This means that more work will be
3990 done, as duplicates will always all be routed, but duplicates are
3991 presumably rare, so I don't expect this is of any significance.
3993 For deliveries to pipes, files, and autoreplies, the duplicate checking
3994 still happens during the routing process, since they are not going to be
3997 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3998 It corrects a timeout issue with spamd. This is Ian's comment: "The
3999 background is that sometimes spamd either never reads data from a
4000 connection it has accepted, or it never writes response data. The exiscan
4001 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4002 blindly assumes that writes won't block so it may never time out."
4004 PH/39 Allow G after quota size as well as K and M.
4006 PH/40 The value set for $authenticated_id in an authenticator may not contain
4007 binary zeroes or newlines because the value is written to log lines and
4008 to spool files. There was no check on this. Now the value is run through
4009 the string_printing() function so that such characters are converted to
4010 printable escape sequences.
4012 PH/41 $message_linecount is a new variable that contains the total number of
4013 lines in the message. Compare $body_linecount, which is the count for the
4016 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4017 bounce and delay warning messages, except in certain special cases, which
4020 (a) An SMTP error message from a remote host;
4021 (b) A message specified in a :fail: redirection;
4022 (c) A message specified in a "fail" command in a system filter;
4023 (d) A message specified in a FAIL return from the queryprogram router;
4024 (e) A message specified by the cannot_route_message router option.
4026 In these cases only, Exim does include the error details in bounce and
4027 warning messages. There are also a few cases where bland messages such
4028 as "unrouteable address" or "local delivery error" are given.
4030 PH/43 $value is now also set for the "else" part of a ${run expansion.
4032 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4033 being worked on, but at least Exim now implements the latest version to
4036 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4037 process to complete was treated as a delivery failure, a timeout while
4038 writing the message to the pipe was logged, but erroneously treated as a
4039 successful delivery. Such timeouts include transport filter timeouts. For
4040 consistency with the overall process timeout, these timeouts are now
4041 treated as errors, giving rise to delivery failures by default. However,
4042 there is now a new Boolean option for the pipe transport called
4043 timeout_defer, which, if set TRUE, converts the failures into defers for
4044 both kinds of timeout. A transport filter timeout is now identified in
4047 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4048 systems where "make" and "gmake" are different, calling "gmake" at top
4049 level broke things. I've arranged for the value of $(MAKE) to be passed
4050 from the Makefile to this script so that it can call the same version of
4054 A note about Exim versions 4.44 and 4.50
4055 ----------------------------------------
4057 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4058 changes of various kinds. As a consequence, a big documentation update was
4059 needed. This delayed the release for rather longer than seemed good, especially
4060 in the light of a couple of (minor) security issues. Therefore, the changes
4061 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4062 release. So 4.44 and 4.50 are in effect two different branches that both start
4065 I have left the 4.50 change log unchanged; it contains all the changes since
4066 4.43. The change log for 4.44 is below; many of its items are identical to
4067 those for 4.50. This seems to be the most sensible way to preserve the
4068 historical information.
4074 1. Minor wording change to the doc/README.SIEVE file.
4076 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4077 computation of the current number of files was incorrect.
4079 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4080 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4083 4. Give more explanation in the error message when the command for a transport
4084 filter fails to execute.
4086 5. There are several places where Exim runs a non-Exim command in a
4087 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4088 was being done only for the command run by the queryprogram router. It is
4089 now done for all such subprocesses. The other cases are: ${run, transport
4090 filters, and the commands run by the lmtp and pipe transports.
4092 6. Added CONFIGURE_GROUP build-time option.
4094 7. Some older OS have a limit of 256 on the maximum number of file
4095 descriptors. Exim was using setrlimit() to set 1000 as a large value
4096 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4097 systems. I've change it so that if it can't get 1000, it tries for 256.
4099 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4100 was an oversight, and furthermore, ever since the addition of extra
4101 controls (e.g. 4.43/32), the checks on when to allow different forms of
4102 "control" were broken. There should now be diagnostics for all cases when a
4103 control that does not make sense is encountered.
4105 9. Added the /retain_sender option to "control=submission".
4107 10. $recipients is now available in the predata ACL (oversight).
4109 11. Tidy the search cache before the fork to do a delivery from a message
4110 received from the command line. Otherwise the child will trigger a lookup
4111 failure and thereby defer the delivery if it tries to use (for example) a
4112 cached ldap connection that the parent has called unbind on.
4114 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4115 of $address_data from the recipient verification was clobbered by the
4116 sender verification.
4118 13. The value of address_data from a sender verification is now available in
4119 $sender_address_data in subsequent conditions in the ACL statement.
4121 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4123 15. Added a new option "connect=<time>" to callout options, to set a different
4126 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4127 was its contents. (It was OK if the option was not defined at all.)
4129 17. A "Completed" log line is now written for messages that are removed from
4130 the spool by the -Mrm option.
4132 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4133 information about exactly what failed.
4135 19. Added -dd to debug only the daemon process.
4137 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4138 handles timeouts, both on the server side and network timeouts. Renamed the
4139 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4141 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4142 It is now set to "smtps".
4144 22. $host_address is now set to the target address during the checking of
4145 ignore_target_hosts.
4147 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4148 being passed; this would have caused $sender_host_name to have been used if
4149 matching the list had actually called for a host name (not very likely,
4150 since this list is usually IP addresses). A host name is now passed as
4153 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4154 code that specifies a non-restarting handler (typically sigaction() in
4155 modern systems) in an attempt to fix a rare and obscure crash bug.
4157 25. Narrowed the window for a race in the daemon that could cause it to ignore
4158 SIGCHLD signals. This is not a major problem, because they are used only to
4159 wake it up if nothing else does.
4161 26. A malformed maildirsize file could cause Exim to calculate negative values
4162 for the mailbox size or file count. Odd effects could occur as a result.
4163 The maildirsize information is now recalculated if the size or filecount
4166 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4167 support for a long time. Removed HAVE_SYS_VFS_H.
4169 28. Installed the latest version of exipick from John Jetmore.
4171 29. In an address list, if the pattern was not a regular expression, an empty
4172 subject address (from a bounce message) matched only if the pattern was an
4173 empty string. Non-empty patterns were not even tested. This was the wrong
4174 because it is perfectly reasonable to use an empty address as part of a
4175 database query. An empty address is now tested by patterns that are
4176 lookups. However, all the other forms of pattern expect the subject to
4177 contain a local part and a domain, and therefore, for them, an empty
4178 address still always fails if the pattern is not itself empty.
4180 30. Exim went into a mad DNS loop when attempting to do a callout where the
4181 host was specified on an smtp transport, and looking it up yielded more
4182 than one IP address.
4184 31. Re-factored the code for checking spool and log partition space into a
4185 function that finds that data and another that does the check. The former
4186 is then used to implement four new variables: $spool_space, $log_space,
4187 $spool_inodes, and $log_inodes.
4189 32. The RFC2047 encoding function was originally intended for short strings
4190 such as real names; it was not keeping to the 75-character limit for
4191 encoded words that the RFC imposes. It now respects the limit, and
4192 generates multiple encoded words if necessary. To be on the safe side, I
4193 have increased the buffer size for the ${rfc2047: expansion operator from
4196 33. It is now permitted to omit both strings after an "if" condition; if the
4197 condition is true, the result is "true". As before, when the second string
4198 is omitted, a false condition yields an empty string. This makes it less
4199 cumbersome to write custom ACL and router conditions.
4201 34. Failure to deliver a bounce message always caused it to be frozen, even if
4202 there was an errors_to setting on the router. The errors_to setting is now
4205 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4206 canonical form (fully expanded) before being placed in
4207 $sender_host_address.
4209 36. The table in the code that translates DNS record types into text (T_A to
4210 "A" for instance) was missing entries for NS and CNAME. It is just possible
4211 that this could have caused confusion if both these types were looked up
4212 for the same domain, because the text type is used as part of Exim's
4213 per-process caching. But the chance of anyone hitting this buglet seems
4216 37. The dnsdb lookup has been extended in a number of ways.
4218 (1) There is a new type, "zns", which walks up the domain tree until it
4219 finds some nameserver records. It should be used with care.
4221 (2) There is a new type, "mxh", which is like "mx" except that it returns
4222 just the host names, not the priorities.
4224 (3) It is now possible to give a list of domains (or IP addresses) to be
4225 looked up. The behaviour when one of the lookups defers can be
4226 controlled by a keyword.
4228 (4) It is now possible to specify the separator character for use when
4229 multiple records are returned.
4231 38. The dnslists ACL condition has been extended: it is now possible to supply
4232 a list of IP addresses and/or domains to be looked up in a particular DNS
4235 39. Added log_selector=+queue_time_overall.
4237 40. When running the queue in the test harness, wait just a tad after forking a
4238 delivery process, to get repeatability of debugging output.
4240 41. Include certificate and key file names in error message when GnuTLS fails
4241 to set them up, because the GnuTLS error message doesn't include the name
4242 of the failing file when there is a problem reading it.
4244 42. Allow both -bf and -bF in the same test run.
4246 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4248 44. The "Exiscan patch" is now merged into the mainline Exim source.
4250 45. Sometimes the final signoff response after QUIT could fail to get
4251 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4252 < 0 before doing a fflush(). This bug looks as though it goes back to the
4253 introduction of TLS in release 3.20, but "sometimes" must have been rare
4254 because the tests only now provoked it.
4256 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4257 (this can affect the format of dates).
4259 47. exim_tidydb, when checking for the continued existence of a message for
4260 which it has found a message-specific retry record, was not finding
4261 messages that were in split spool directories. Consequently, it was
4262 deleting retry records that should have stayed in existence.
4264 48. Steve fixed some bugs in eximstats.
4266 49. The SPA authentication driver was not abandoning authentication and moving
4267 on to the next authenticator when an expansion was forced to fail,
4268 contradicting the general specification for all authenticators. Instead it
4269 was generating a temporary error. It now behaves as specified.
4271 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4272 (the order specifies the preference for clients). The order is now AES256,
4273 AES128, 3DES, ARCFOUR128.
4275 51. Small patch to Sieve code - explicitly set From: when generating an
4278 52. Exim crashed if a remote delivery caused a very long error message to be
4279 recorded - for instance if somebody sent an entire SpamAssassin report back
4280 as a large number of 550 error lines. This bug was coincidentally fixed by
4281 increasing the size of one of Exim's internal buffers (big_buffer) that
4282 happened as part of the Exiscan merge. However, to be on the safe side, I
4283 have made the code more robust (and fixed the comments that describe what
4286 53. Now that there can be additional text after "Completed" in log lines (if
4287 the queue_time_overall log selector is set), a one-byte patch to exigrep
4288 was needed to allow it to recognize "Completed" as not the last thing in
4291 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4292 patch that reportedly fixes this has been added. I am not expert enough to
4293 create a test for it. This is what the patch creator wrote:
4295 "I found a little strange behaviour of ldap code when working with
4296 Windows 2003 AD Domain, where users was placed in more than one
4297 Organization Units. When I tried to give exim partial DN, the exit code
4298 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4299 But simultaneously result of request was absolutely normal ldap result,
4300 so I produce this patch..."
4302 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4303 REFERENCE, so I have modified the code to exclude the patch when that macro
4306 55. Some experimental protocols are using DNS PTR records for new purposes. The
4307 keys for these records are domain names, not reversed IP addresses. The
4308 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4309 leaves it alone. Component reversal etc. now happens only for IP addresses.
4312 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4314 57. Double the size of the debug message buffer (to 2048) so that more of very
4315 long debug lines gets shown.
4317 58. The exicyclog utility now does better if the number of log files to keep
4318 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4320 59. Two changes related to the smtp_active_hostname option:
4322 (1) $smtp_active_hostname is now available as a variable.
4323 (2) The default for smtp_banner uses $smtp_active_hostname instead
4324 of $primary_hostname.
4326 60. The host_aton() function is supposed to be passed a string that is known
4327 to be a valid IP address. However, in the case of IPv6 addresses, it was
4328 not checking this. This is a hostage to fortune. Exim now panics and dies
4329 if the condition is not met. A case was found where this could be provoked
4330 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4331 components; fortuitously, this particular loophole had already been fixed
4332 by change 4.50/55 above.
4334 If there are any other similar loopholes, the new check in host_aton()
4335 itself should stop them being exploited. The report I received stated that
4336 data on the command line could provoke the exploit when Exim was running as
4337 exim, but did not say which command line option was involved. All I could
4338 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4339 running as the user.
4342 61. There was a buffer overflow vulnerability in the SPA authentication code
4343 (which came originally from the Samba project). I have added a test to the
4344 spa_base64_to_bits() function which I hope fixes it.
4347 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4348 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4350 63. The daemon start-up calls getloadavg() while still root for those OS that
4351 need the first call to be done as root, but it missed one case: when
4352 deliver_queue_load_max is set with deliver_drop_privilege. This is
4353 necessary for the benefit of the queue runner, because there is no re-exec
4354 when deliver_drop_privilege is set.
4356 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4357 This has been fixed.
4359 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4360 was in use, was not putting the data itself into the right store pool;
4361 consequently, it could be overwritten for a subsequent message in the same
4362 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4365 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4367 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4368 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4369 the function were treating the return as a boolean value, which happened to
4370 work because 0=false and not-0=true, but is not correct code.
4372 68. The host_aton() function was not handling scoped IPv6 addresses (those
4373 with, for example, "%eth0" on the end) correctly.
4375 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4376 negated items (that is, ~something) in unsigned ints. Some compilers
4377 apparently mutter when there is no cast.
4379 70. If an address verification called from an ACL failed, and did not produce a
4380 user-specific message (i.e. there was only a "system" message), nothing was
4381 put in $acl_verify_message. In this situation, it now puts the system
4384 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4385 session; change 4.31/43 added the unwanted input to the log line - except
4386 that it did not do this in the start of session case. It now does.
4388 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4389 This is wrong and can cause the other end to generate a synchronization
4390 error if it is another Exim or anything else that does the synchronization
4391 check. A QUIT command is no longer sent after a timeout.
4393 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4394 during host lookups.
4396 74. The defer_ok option of callout verification was not working if it was used
4397 when verifying addresses in header lines, that is, for this case:
4399 verify = header_sender/callout=defer_ok
4401 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4402 those file descriptors could be used for SMTP connections. If anything
4403 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4404 could be sent to the SMTP client, causing chaos. The daemon now opens
4405 stdin, stdout, and stderr to /dev/null when it puts itself into the
4408 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4409 log by default. The user can override this with suitable Perl magic.
4411 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4412 the log message when discard triggers, was not working for the DATA ACL or
4413 for the non-SMTP ACL.
4415 78. Error message wording change in sieve.c.
4417 79. If smtp_accept_max_per_host was set, the number of connections could be
4418 restricted to fewer than expected, because the daemon was trying to set up
4419 a new connection before checking whether the processes handling previous
4420 connections had finished. The check for completed processes is now done
4421 earlier. On busy systems, this bug wouldn't be noticed because something
4422 else would have woken the daemon, and it would have reaped the completed
4425 80. If a message was submitted locally by a user whose login name contained one
4426 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4427 It caused a spool format error. I have fixed the spool reading code. A
4428 related problem was that the "from" clause in the Received: line became
4429 illegal because of the space(s). It is now covered by ${quote_local_part.
4431 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4434 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4436 83. Patch from Sieve maintainer to fix unterminated string problem in
4437 "vacation" handling.
4439 84. Some minor changes to the Linux configuration files to help with other
4440 OS variants using glibc.
4442 85. One more patch for Sieve to update vacation handling to latest spec.
4445 ----------------------------------------------------
4446 See the note above about the 4.44 and 4.50 releases.
4447 ----------------------------------------------------
4453 1. Change 4.43/35 introduced a bug that caused file counts to be
4454 incorrectly computed when quota_filecount was set in an appendfile
4457 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4458 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4461 3. Give more explanation in the error message when the command for a transport
4462 filter fails to execute.
4464 4. There are several places where Exim runs a non-Exim command in a
4465 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4466 was being done only for the command run by the queryprogram router. It is
4467 now done for all such subprocesses. The other cases are: ${run, transport
4468 filters, and the commands run by the lmtp and pipe transports.
4470 5. Some older OS have a limit of 256 on the maximum number of file
4471 descriptors. Exim was using setrlimit() to set 1000 as a large value
4472 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4473 systems. I've change it so that if it can't get 1000, it tries for 256.
4475 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4476 was an oversight, and furthermore, ever since the addition of extra
4477 controls (e.g. 4.43/32), the checks on when to allow different forms of
4478 "control" were broken. There should now be diagnostics for all cases when a
4479 control that does not make sense is encountered.
4481 7. $recipients is now available in the predata ACL (oversight).
4483 8. Tidy the search cache before the fork to do a delivery from a message
4484 received from the command line. Otherwise the child will trigger a lookup
4485 failure and thereby defer the delivery if it tries to use (for example) a
4486 cached ldap connection that the parent has called unbind on.
4488 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4489 of $address_data from the recipient verification was clobbered by the
4490 sender verification.
4492 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4493 was its contents. (It was OK if the option was not defined at all.)
4495 11. A "Completed" log line is now written for messages that are removed from
4496 the spool by the -Mrm option.
4498 12. $host_address is now set to the target address during the checking of
4499 ignore_target_hosts.
4501 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4502 being passed; this would have caused $sender_host_name to have been used if
4503 matching the list had actually called for a host name (not very likely,
4504 since this list is usually IP addresses). A host name is now passed as
4507 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4508 code that specifies a non-restarting handler (typically sigaction() in
4509 modern systems) in an attempt to fix a rare and obscure crash bug.
4511 15. Narrowed the window for a race in the daemon that could cause it to ignore
4512 SIGCHLD signals. This is not a major problem, because they are used only to
4513 wake it up if nothing else does.
4515 16. A malformed maildirsize file could cause Exim to calculate negative values
4516 for the mailbox size or file count. Odd effects could occur as a result.
4517 The maildirsize information is now recalculated if the size or filecount
4520 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4521 support for a long time. Removed HAVE_SYS_VFS_H.
4523 18. Updated exipick to current release from John Jetmore.
4525 19. Allow an empty sender to be matched against a lookup in an address list.
4526 Previously the only cases considered were a regular expression, or an
4529 20. Exim went into a mad DNS lookup loop when doing a callout where the
4530 host was specified on the transport, if the DNS lookup yielded more than
4533 21. The RFC2047 encoding function was originally intended for short strings
4534 such as real names; it was not keeping to the 75-character limit for
4535 encoded words that the RFC imposes. It now respects the limit, and
4536 generates multiple encoded words if necessary. To be on the safe side, I
4537 have increased the buffer size for the ${rfc2047: expansion operator from
4540 22. Failure to deliver a bounce message always caused it to be frozen, even if
4541 there was an errors_to setting on the router. The errors_to setting is now
4544 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4545 canonical form (fully expanded) before being placed in
4546 $sender_host_address.
4548 24. Updated eximstats to version 1.33
4550 25. Include certificate and key file names in error message when GnuTLS fails
4551 to set them up, because the GnuTLS error message doesn't include the name
4552 of the failing file when there is a problem reading it.
4554 26. Expand error message when OpenSSL has problems setting up cert/key files.
4557 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4558 (this can affect the format of dates).
4560 28. exim_tidydb, when checking for the continued existence of a message for
4561 which it has found a message-specific retry record, was not finding
4562 messages that were in split spool directories. Consequently, it was
4563 deleting retry records that should have stayed in existence.
4565 29. eximstats updated to version 1.35
4566 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4567 1.35 - bugfix such that pie charts by volume are generated correctly
4569 30. The SPA authentication driver was not abandoning authentication and moving
4570 on to the next authenticator when an expansion was forced to fail,
4571 contradicting the general specification for all authenticators. Instead it
4572 was generating a temporary error. It now behaves as specified.
4574 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4575 (the order specifies the preference for clients). The order is now AES256,
4576 AES128, 3DES, ARCFOUR128.
4578 31. Small patch to Sieve code - explicitly set From: when generating an
4581 32. Exim crashed if a remote delivery caused a very long error message to be
4582 recorded - for instance if somebody sent an entire SpamAssassin report back
4583 as a large number of 550 error lines. This bug was coincidentally fixed by
4584 increasing the size of one of Exim's internal buffers (big_buffer) that
4585 happened as part of the Exiscan merge. However, to be on the safe side, I
4586 have made the code more robust (and fixed the comments that describe what
4589 33. Some experimental protocols are using DNS PTR records for new purposes. The
4590 keys for these records are domain names, not reversed IP addresses. The
4591 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4592 leaves it alone. Component reversal etc. now happens only for IP addresses.
4595 34. The host_aton() function is supposed to be passed a string that is known
4596 to be a valid IP address. However, in the case of IPv6 addresses, it was
4597 not checking this. This is a hostage to fortune. Exim now panics and dies
4598 if the condition is not met. A case was found where this could be provoked
4599 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4600 components; fortuitously, this particular loophole had already been fixed
4601 by change 4.50/55 or 4.44/33 above.
4603 If there are any other similar loopholes, the new check in host_aton()
4604 itself should stop them being exploited. The report I received stated that
4605 data on the command line could provoke the exploit when Exim was running as
4606 exim, but did not say which command line option was involved. All I could
4607 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4608 running as the user.
4611 35. There was a buffer overflow vulnerability in the SPA authentication code
4612 (which came originally from the Samba project). I have added a test to the
4613 spa_base64_to_bits() function which I hope fixes it.
4616 36. The daemon start-up calls getloadavg() while still root for those OS that
4617 need the first call to be done as root, but it missed one case: when
4618 deliver_queue_load_max is set with deliver_drop_privilege. This is
4619 necessary for the benefit of the queue runner, because there is no re-exec
4620 when deliver_drop_privilege is set.
4622 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4623 was in use, was not putting the data itself into the right store pool;
4624 consequently, it could be overwritten for a subsequent message in the same
4625 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4628 38. Sometimes the final signoff response after QUIT could fail to get
4629 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4630 < 0 before doing a fflush(). This bug looks as though it goes back to the
4631 introduction of TLS in release 3.20, but "sometimes" must have been rare
4632 because the tests only now provoked it.
4638 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4639 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4640 Now it returns an integer. A number of calls were still expecting a T/F
4641 return. Fortuitously, in all cases, the tests worked in OK situations,
4642 which is the norm. However, things would have gone wrong on any write
4643 failures on the smtp file descriptor. This function is used when sending
4644 messages over SMTP and also when doing verify callouts.
4646 2. When Exim is called to do synchronous delivery of a locally submitted
4647 message (the -odf or -odi options), it no longer closes stderr before doing
4650 3. Implemented the mua_wrapper option.
4652 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4654 5. Implemented the functions header_remove(), header_testname(),
4655 header_add_at_position(), and receive_remove_recipient(), and exported them
4658 6. If an ACL "warn" statement specified the addition of headers, Exim already
4659 inserted X-ACL-Warn: at the start if there was no header name. However, it
4660 was not making this test for the second and subsequent header lines if
4661 there were newlines in the string. This meant that an invalid header could
4662 be inserted if Exim was badly configured.
4664 7. Allow an ACL "warn" statement to add header lines at the start or after all
4665 the Received: headers, as well as at the end.
4667 8. Added the rcpt_4xx retry error code.
4669 9. Added postmaster_mailfrom=xxx to callout verification option.
4671 10. Added mailfrom=xxxx to the callout verification option, for verify=
4674 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4675 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4677 12. Inserted some casts to stop certain compilers warning when using pointer
4678 differences as field lengths or precisions in printf-type calls (mostly
4679 affecting debugging statements).
4681 13. Added optional readline() support for -be (dynamically loaded).
4683 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4684 same clock tick as a message's arrival, so that its received time was the
4685 same as the "first fail" time on the retry record, and that message
4686 remained on the queue past the ultimate address timeout, every queue runner
4687 would try a delivery (because it was past the ultimate address timeout) but
4688 after another failure, the ultimate address timeout, which should have then
4689 bounced the address, did not kick in. This was a "< instead of <=" error;
4690 in most cases the first failure would have been in the next clock tick
4691 after the received time, and all would be well.
4693 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4694 being recognized when the domain list was tested by the match_domain
4695 condition in an expansion string.
4697 16. Added the ${str2b64: operator.
4699 17. Exim was always calling setrlimit() to set a large limit for the number of
4700 processes, without checking whether the existing limit was already
4701 adequate. (It did check for the limit on file descriptors.) Furthermore,
4702 errors from getrlimit() and setrlimit() were being ignored. Now they are
4703 logged to the main and panic logs, but Exim does carry on, to try to do its
4704 job under whatever limits there are.
4706 18. Imported PCRE 5.0.
4708 19. Trivial typo in log message " temporarily refused connection" (the leading
4711 20. If the log selector return_path_on_delivery was set and an address was
4712 redirected to /dev/null, the delivery process crashed because it assumed
4713 that a return path would always be set for a "successful" delivery. In this
4714 case, the whole delivery is bypassed as an optimization, and therefore no
4717 21. Internal re-arrangement: the function for sending a challenge and reading
4718 a response while authentication was assuming a zero-terminated challenge
4719 string. It's now changed to take a pointer and a length, to allow for
4720 binary data in such strings.
4722 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4724 23. Exim was not respecting finduser_retries when seeking the login of the
4725 uid under which it was called; it was always trying 10 times. (The default
4726 setting of finduser_retries is zero.) Also, it was sleeping after the final
4727 failure, which is pointless.
4729 24. Implemented tls_on_connect_ports.
4731 25. Implemented acl_smtp_predata.
4733 26. If the domain in control=submission is set empty, Exim assumes that the
4734 authenticated id is a complete email address when it generates From: or
4735 Sender: header lines.
4737 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4738 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4739 chown and chgrp in /bin and hostname in /usr/bin.
4741 28. Exim was keeping the "process log" file open after each use, just as it
4742 does for the main log. This opens the possibility of it remaining open for
4743 long periods when the USR1 signal hits a daemon. Occasional processlog
4744 errors were reported, that could have been caused by this. Anyway, it seems
4745 much more sensible not to leave this file open at all, so that is what now
4748 29. The long-running daemon process does not normally write to the log once it
4749 has entered its main loop, and it closes the log before doing so. This is
4750 so that log files can straightforwardly be renamed and moved. However,
4751 there are a couple of unusual error situations where the daemon does write
4752 log entries, and I had neglected to close the log afterwards.
4754 30. The text of an SMTP error response that was received during a remote
4755 delivery was being truncated at 512 bytes. This is too short for some of
4756 the long messages that one sometimes sees. I've increased the limit to
4759 31. It is now possible to make retry rules that apply only when a message has a
4760 specific sender, in particular, an empty sender.
4762 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4763 it possible to be selective about when SMTP synchronization is enforced.
4765 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4767 32. Implemented hosts_connection_nolog.
4769 33. Added an ACL for QUIT.
4771 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4774 35. Added mailbox_size and mailbox_filecount to appendfile.
4776 36. Added control = no_multiline_responses to ACLs.
4778 37. There was a bug in the logic of the code that waits for the clock to tick
4779 in the case where the clock went backwards by a substantial amount such
4780 that the microsecond fraction of "now" was more than the microsecond
4781 fraction of "then" (but the whole seconds number was less).
4783 38. Added support for the libradius Radius client library this is found on
4784 FreeBSD (previously only the radiusclient library was supported).
4790 1. When certain lookups returned multiple values in the form name=value, the
4791 quoting of the values was not always being done properly. Specifically:
4792 (a) If the value started with a double quote, but contained no whitespace,
4794 (b) If the value contained whitespace other than a space character (i.e.
4795 tabs or newlines or carriage returns) it was not quoted.
4796 This fix has been applied to the mysql and pgsql lookups by writing a
4797 separate quoting function and calling it from the lookup code. The fix
4798 should probably also be applied to nisplus, ibase and oracle lookups, but
4799 since I cannot test any of those, I have not disturbed their existing code.
4801 2. A hit in the callout cache for a specific address caused a log line with no
4802 reason for rejecting RCPT. Now it says "Previous (cached) callout
4803 verification failure".
4805 3. There was an off-by-one bug in the queryprogram router. An over-long
4806 return line was truncated at 256 instead of 255 characters, thereby
4807 overflowing its buffer with the terminating zero. As well as fixing this, I
4808 have increased the buffer size to 1024 (and made a note to document this).
4810 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4811 when Exim is waiting for an SMTP response from a remote server, Exim
4812 restarts its select() call on the socket, thereby resetting its timeout.
4813 This is not a problem when such interrupts are rare. Somebody set up a cron
4814 job to run exiwhat every 2 minutes, which is less than the normal select()
4815 timeout (5 or 10 minutes). This meant that the select() timeout never
4816 kicked in because it was always reset. I have fixed this by comparing the
4817 time when an interrupt arrives with the time at the start of the first call
4818 to select(). If more time than the timeout has elapsed, the interrupt is
4819 treated as a timeout.
4821 5. Some internal re-factoring in preparation for the addition of Sieve
4822 extensions (by MH). In particular, the "personal" test is moved to a
4823 separate function, and given an option for scanning Cc: and Bcc: (which is
4824 not set for Exim filters).
4826 6. When Exim created an email address using the login of the caller as the
4827 local part (e.g. when creating a From: or Sender: header line), it was not
4828 quoting the local part when it contained special characters such as @.
4830 7. Installed new OpenBSD configuration files.
4832 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4833 try to make them clearer.
4835 9. Callout options, other than the timeout value, were being ignored when
4836 verifying sender addresses in header lines. For example, when using
4838 verify = header_sender/callout=no_cache
4840 the cache was (incorrectly) being used.
4842 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4843 only the Cygwin environment.
4845 11. When return_path_on_delivery was set as a log selector, if different remote
4846 addresses in the same message used different return paths and parallel
4847 remote delivery occurred, the wrong values would sometimes be logged.
4848 (Whenever a remote delivery process finished, the return path value from
4849 the most recently started remote delivery process was logged.)
4851 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4852 lines when AUTH and/or TLS are in use. This is the "received protocol"
4853 field. Exim used to use "asmtp" for authenticated SMTP, without any
4854 indication (in the protocol name) for TLS use. Now it follows the RFC and
4855 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4856 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4857 names appear in log lines as well as in Received: header lines.
4859 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4860 extensions, and comparison tests, and to fix some bugs.
4862 14. Changes to the "personal" filter test:
4864 (1) The test was buggy in that it was just doing the equivalent of
4865 "contains" tests on header lines. For example, if a user's address was
4866 anne@some.where, the "personal" test would incorrectly be true for
4868 To: susanne@some.where
4870 This test is now done by extracting each address from the header in turn,
4871 and checking the entire address. Other tests that are part of "personal"
4872 are now done using regular expressions (for example, to check local parts
4873 of addresses in From: header lines).
4875 (2) The list of non-personal local parts in From: addresses has been
4876 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4877 taken from the Sieve specification recommendations.
4879 (3) If the message contains any header line starting with "List-" it is
4880 treated as non-personal.
4882 (4) The test for "circular" in the Subject: header line has been removed
4883 because it now seems ill-conceived.
4885 15. Minor typos in src/EDITME comments corrected.
4887 16. Installed latest exipick from John Jetmore.
4889 17. If headers_add on a router specified a text string that was too long for
4890 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4891 of string_sprintf() is now avoided.
4893 18. $message_body_size was not set (it was always zero) when running the DATA
4894 ACL and the local_scan() function.
4896 19. For the "mail" command in an Exim filter, no default was being set for
4897 the once_repeat time, causing a random time value to be used if "once" was
4898 specified. (If the value happened to be <= 0, no repeat happened.) The
4899 default is now 0s, meaning "never repeat". The "vacation" command was OK
4900 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4901 (I found it when inspecting the code).
4903 20. There is now an overall timeout for performing a callout verification. It
4904 defaults to 4 times the callout timeout, which applies to individual SMTP
4905 commands during the callout. The overall timeout applies when there is more
4906 than one host that can be tried. The timeout is checked before trying the
4907 next host. This prevents very long delays if there are a large number of
4908 hosts and all are timing out (e.g. when the network connections are timing
4909 out). The value of the overall timeout can be changed by specifying an
4910 additional sub-option for "callout", called "maxwait". For example:
4912 verify = sender/callout=5s,maxwait=20s
4914 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4915 to the end before writing, but this should make it even safer).
4917 22. Exim was forgetting that it had advertised PIPELINING for the second and
4918 subsequent messages on an SMTP connection. It was also not resetting its
4919 memory on STARTTLS and an internal HELO.
4921 23. When Exim logs an SMTP synchronization error within a session, it now
4922 records whether PIPELINING has been advertised or not.
4924 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4925 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4926 rather than long int.
4928 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4930 26. Added the never_mail option to autoreply.
4936 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4937 crash if the getsockname() call failed; this can happen if a connection is
4938 closed very soon after it is established. The problem was simply in the
4939 order in which certain operations were done, causing Exim to try to write
4940 to the SMTP stream before it had set up the file descriptor. The bug has
4941 been fixed by making things happen in the correct order.
4947 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4948 before the connection was closed, thus losing the rejection response.
4950 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4951 some early Solaris releases, but causes trouble in current releases where
4952 socklen_t is defined.
4954 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4957 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4960 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4961 that releasing the top of it at the end releases what was used for sub-
4962 expansions (unless the block got too big). However, discard this block if
4963 the first thing is a variable or header, so that we can use its block when
4964 it is dynamic (useful for very large $message_headers, for example).
4966 6. Lookups now cache *every* query, not just the most recent. A new, separate
4967 store pool is used for this. It can be recovered when all lookup caches are
4968 flushed. Lookups now release memory at the end of their result strings.
4969 This has involved some general refactoring of the lookup sources.
4971 7. Some code has been added to the store_xxx() functions to reduce the amount
4972 of flapping under certain conditions.
4974 8. log_incoming_interface used to affect only the <= reception log lines. Now
4975 it causes the local interface and port to be added to several more SMTP log
4976 lines, for example "SMTP connection from", and rejection lines.
4978 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4980 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4982 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4983 could be overwritten at the end of the current message (or the start of a
4984 new message if it was set in a HELO ACL). The value is now preserved for
4985 the duration of the SMTP connection.
4987 12. If a transport had a headers_rewrite setting, and a matching header line
4988 contained an unqualified address, that address was qualified, even if it
4989 did not match any rewriting rules. The underlying bug was that the values
4990 of the flags that permit the existence of unqualified sender and recipient
4991 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4992 non-local messages, and by -bnq for local messages) were not being
4993 preserved with the message after it was received.
4995 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4996 "next input=" as part of the text comprising the host identity instead of
4997 the correct text. The code was using the same buffer for two different
4998 strings. However, depending on which order the printing function evaluated
4999 its arguments, the bug did not always show up. Under Linux, for example, my
5000 test suite worked just fine.
5002 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5003 doesn't work with some older versions of Perl. It has been changed to "my",
5004 which in any case is probably the better facility to use.
5006 15. A really picky compiler found some instances of statements for creating
5007 error messages that either had too many or two few arguments for the format
5010 16. The size of the buffer for calls to the DNS resolver has been increased
5011 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5012 for addresses that have a lot of PTR records. This alleviates a problem; it
5013 does not fully solve it.
5015 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5016 buffer now truncates the list and logs the incident, which is the same
5017 action as happens when Exim is looking up a host name and its aliases.
5018 Previously in this situation something unpredictable would happen;
5019 sometimes it was "internal error: store_reset failed".
5021 18. If a server dropped the connection unexpectedly when an Exim client was
5022 using GnuTLS and trying to read a response, the client delivery process
5023 crashed while trying to generate an error log message.
5025 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5026 string, for example:
5028 warn message = H1: something\nH2: something
5030 the text was added as a single header line from Exim's point of view
5031 though it ended up OK in the delivered message. However, searching for the
5032 second and subsequent header lines using $h_h2: did not work. This has been
5033 fixed. Similarly, if a system filter added multiple headers in this way,
5034 the routers could not see them.
5036 20. Expanded the error message when iplsearch is called with an invalid key to
5037 suggest using net-iplsearch in a host list.
5039 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5040 ACLs are no longer actually imposed (and a message to that effect is
5043 22. If a "gecos" field in a passwd entry contained escaped characters, in
5044 particular, if it contained a \" sequence, Exim got it wrong when building
5045 a From: or a Sender: header from that name. A second bug also caused
5046 incorrect handling when an unquoted " was present following a character
5047 that needed quoting.
5049 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5050 was not being matched caselessly.
5052 24. Arranged for all hyphens in the exim.8 source to be escaped with
5055 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5056 properly. Recipient callout cache records were still being keyed to include
5057 the sender, even when use_sender was set false. This led to far more
5058 callouts that were necessary. The sender is no longer included in the key
5059 when use_sender is false.
5061 26. Added "control = submission" modifier to ACLs.
5063 27. Added the ${base62d: operator to decode base 62 numbers.
5065 28. dnsdb lookups can now access SRV records.
5067 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5068 the configuration file.
5070 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5071 (-v) mode. This makes the output for a verbose queue run more intelligible.
5073 31. Added a use_postmaster feature to recipient callouts.
5075 32. Added the $body_zerocount variable, containing the number of binary zero
5076 bytes in the message body.
5078 33. The time of last modification of the "new" subdirectory is now used as the
5079 "mailbox time last read" when there is a quota error for a maildir
5082 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5084 35. Added +ignore_unknown as a special item in host lists.
5086 36. Code for decoding IPv6 addresses in host lists is now included, even if
5087 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5088 address was recognized as an IP address, but was then not correctly decoded
5089 into binary, causing unexpected and incorrect effects when compared with
5096 1. Very minor rewording of debugging text in manualroute to say "list of
5097 hosts" instead of "hostlist".
5099 2. If verify=header_syntax was set, and a header line with an unqualified
5100 address (no domain) and a large number of spaces between the end of the
5101 name and the colon was received, the reception process suffered a buffer
5102 overflow, and (when I tested it) crashed. This was caused by some obsolete
5103 code that should have been removed. The fix is to remove it!
5105 3. When running in the test harness, delay a bit after writing a bounce
5106 message to get a bit more predictability in the log output.
5108 4. Added a call to search_tidyup() just before forking a reception process. In
5109 theory, someone could use a lookup in the expansion of smtp_accept_max_
5110 per_host which, without the tidyup, could leave open a database connection.
5112 5. Added the variables $recipient_data and $sender_data which get set from a
5113 lookup success in an ACL "recipients" or "senders" condition, or a router
5114 "senders" option, similar to $domain_data and $local_part_data.
5116 6. Moved the writing of debug_print from before to after the "senders" test
5119 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5120 problems for message scanning, either using a data ACL, or using
5121 local_scan() because the Received: header was not generated till after they
5122 were called (in order to set the time as the time of reception completion).
5123 I have revised the way this works. The header is now generated after the
5124 body is received, but before the ACL or local_scan() are called. After they
5125 are run, the timestamp in the header is updated.
5131 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5132 before starting a queue runner without re-exec. This happened only when
5133 deliver_drop_privilege was set or when the Exim user was set to root. The
5134 effect of the bug was that timeouts during subsequent deliveries caused
5135 crashes instead of being properly handled. The handler is now left at its
5136 default (and expected) setting.
5138 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5139 message, again when deliver_drop_privilege is set or Exim is run as root.
5140 The bug described in (1) was not present in this case, but the tidying up
5141 of the other signals was missing. I have made the two cases consistent.
5143 3. The ignore_target_hosts setting on a manualroute router was being ignored
5144 for hosts that were looked up using the /MX notation.
5146 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5149 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5150 operated on the sender address. After changing the $sender_address to <>
5151 for the sender address verify, Exim was re-instated it as the original
5152 (before rewriting) address, but remembering that it had rewritten it, so it
5153 wasn't rewriting it again. This bug also had the effect of breaking the
5154 sender address verification caching when the sender address was rewritten.
5156 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5157 This has been changed so that if the ip literal address matches
5158 ignore_target_hosts, the router declines.
5160 7. Added expansion conditions match_domain, match_address, and match_local_
5161 part (NOT match_host).
5163 8. The placeholder for the Received: header didn't have a length field set.
5165 9. Added code to Exim itself and to exim_lock to test for a specific race
5166 condition that could lead to file corruption when using MBX delivery. The
5167 issue is with the lockfile that is created in /tmp. If this file is removed
5168 after a process has opened it but before that process has acquired a lock,
5169 there is the potential for a second process to recreate the file and also
5170 acquire a lock. This could lead to two Exim processes writing to the file
5171 at the same time. The added code performs the same test as UW imapd; it
5172 checks after acquiring the lock that its file descriptor still refers to
5173 the same named file.
5175 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5176 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5179 11. Added the smtp_active_hostname option. If used, this will typically be made
5180 to depend on the incoming interface address. Because $interface_address is
5181 not set up until the daemon has forked a reception process, error responses
5182 that can happen earlier (such as "too many connections") no longer contain
5185 12. If an expansion in a condition on a "warn" statement fails because a lookup
5186 defers, the "warn" statement is abandoned, and the next ACL statement is
5187 processed. Previously this caused the whole ACL to be aborted.
5189 13. Added the iplsearch lookup type.
5191 14. Added ident_timeout as a log selector.
5193 15. Added tls_certificate_verified as a log selector.
5195 16. Added a global option tls_require_ciphers (compare the smtp transport
5196 option of the same name). This controls incoming TLS connections.
5198 17. I finally figured out how to make tls_require_ciphers do a similar thing
5199 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5200 before starting the TLS session.
5202 18. Tabs are now shown as \t in -bP output.
5204 19. If the log selector return_path_on_delivery was set, Exim crashed when
5205 bouncing a message because it had too many Received: header lines.
5207 20. If two routers both had headers_remove settings, and the first one included
5208 a superfluous trailing colon, the final name in the first list and the
5209 first name in the second list were incorrectly joined into one item (with a
5210 colon in the middle).
5216 1. Added -C and -D options to the exinext utility, mainly to make it easier
5217 to include in the automated testing, but these could be helpful when
5218 multiple configurations are in use.
5220 2. The exinext utility was not formatting the output nicely when there was
5221 an alternate port involved in the retry record key, nor when there was a
5222 message id as well (for retries that were specific to a specific message
5223 and a specific host). It was also confused by IPv6 addresses, because of
5224 the additional colons they contain. I have fixed the IPv4 problem, and
5225 patched it up to do a reasonable job for IPv6.
5227 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5228 delivery, the log line now contains "pipelined" if PIPELINING was used.
5230 4. An SMTP transport process used to panic and die if the bind() call to set
5231 an explicit outgoing interface failed. This has been changed; it is now
5232 treated in the same way as a connect() failure.
5234 5. A reference to $sender_host_name in the part of a conditional expansion
5235 that was being skipped was still causing a DNS lookup. This no longer
5238 6. The def: expansion condition was not recognizing references to header lines
5239 that used bh_ and bheader_.
5241 7. Added the _cache feature to named lists.
5243 8. The code for checking quota_filecount in the appendfile transport was
5244 allowing one more file than it should have been.
5246 9. For compatibility with Sendmail, the command line option
5254 and sets the incoming protocol and host name (for trusted callers). The
5255 host name and its colon can be omitted when only the protocol is to be set.
5256 Note the Exim already has two private options, -pd and -ps, that refer to
5257 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5258 "s", but I don't think that's a major issue.
5260 10. A number of refactoring changes to the code, none of which should affect
5263 (a) The number of logging options was getting close to filling up the
5264 32-bit word that was used as a bit map. I have split them into two classes:
5265 those that are passed in the argument to log_write(), and those that are
5266 only ever tested independently outside of that function. These are now in
5267 separate 32-bit words, so there is plenty of room for expansion again.
5268 There is no change in the user interface or the logging behaviour.
5270 (b) When building, for example, log lines, the code previously used a
5271 macro that called string_cat() twice, in order to add two strings. This is
5272 not really sufficiently general. Furthermore, there was one instance where
5273 it was actually wrong because one of the argument was used twice, and in
5274 one call a function was used. (As it happened, calling the function twice
5275 did not affect the overall behaviour.) The macro has been replaced by a
5276 function that can join an arbitrary number of extra strings onto a growing
5279 (c) The code for expansion conditions now uses a table and a binary chop
5280 instead of a serial search (which was left over from when there were very
5281 few conditions). Also, it now recognizes conditions like "pam" even when
5282 the relevant support is not compiled in: a suitably worded error message is
5283 given if an attempt is made to use such a condition.
5285 11. Added ${time_interval:xxxxx}.
5287 12. A bug was causing one of the ddress fields not to be passed back correctly
5288 from remote delivery subprocesses. The field in question was not being
5289 subsequently used, so this caused to problems in practice.
5291 13. Added new log selectors queue_time and deliver_time.
5293 14. Might have fixed a bug in maildirsizefile handling that threw up
5294 "unexpected character" debug warnings, and recalculated the data
5295 unnecessarily. In any case, I expanded the warning message to give more
5298 15. Added the message "Restricted characters in address" to the statements in
5299 the default ACL that block characters like @ and % in local parts.
5301 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5302 Three changes have been made:
5304 (a) There was a serious bug; a negative response to MAIL caused the whole
5305 recipient domain to be cached as invalid, thereby blocking all messages
5306 to all local parts at the same domain, from all senders. This bug has
5307 been fixed. The domain is no longer cached after a negative response to
5308 MAIL if the sender used is not empty.
5310 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5313 (c) A new callout option, "use_sender" has been added for people who want
5314 the modified behaviour.
5320 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5323 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5324 indeed breaks things for older releases.
5326 3. Added additional logging to the case where there is a problem reading data
5327 from a filter that is running in a subprocess using a pipe, in order to
5328 try to track down a specific problem.
5330 4. Testing facility fudge: when running in the test harness and attempting
5331 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5332 getting "No route to host". Convert this to a timeout.
5334 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5337 6. Some OS don't have socklen_t but use size_t instead. This affects the
5338 fifth argument of getsockopt() amongst other things. This is now
5339 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5340 can be set for individual OS. I have set it for SunOS5, OSF1, and
5341 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5342 some earlier ones do not.
5344 7. Change 4.30/15 was not doing the test caselessly.
5346 8. The standard form for an IPv6 address literal was being rejected by address
5347 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5348 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5349 this, as well as the form without the "IPv6" on the front (but only when
5350 address literals are enabled, of course).
5352 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5354 10. Exim crashed if a message with an empty sender address specified by -f
5355 encountered a router with an errors_to setting. This could be provoked only
5356 by a command such as
5360 where an empty string was supplied; "<>" did not hit this bug.
5362 11. Installed PCRE release 4.5.
5364 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5365 remained set. It is now erased.
5367 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5368 times from message ids (which are base 36 rather than the normal 62).
5370 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5371 were being counted as actual protocol errors, and logged if the log
5372 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5373 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5374 and DATA following a set of rejected RCPTs do not count as protocol errors.
5375 In other words, Exim assumes they were pipelined, though this may not
5376 actually be the case. Of course, in all cases the client gets an
5377 appropriate error code.
5379 15. If a lookup fails in an ACL condition, a message about the failure may
5380 be available; it is used if testing the ACL cannot continue, because most
5381 such messages specify what the cause of the deferral is. However, some
5382 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5383 that caused an old message to be retained and used if a later statement
5384 caused a defer, replacing the real cause of the deferral.
5386 16. If an IP address had so many PTR records that the DNS lookup buffer
5387 was not large enough to hold them, Exim could crash while trying to process
5388 the truncated data. It now detects and logs this case.
5390 17. Further to 4.21/58, another change has been made: if (and only if) the
5391 first line of a message (the first header line) ends with CRLF, a bare LF
5392 in a subsequent header line has a space inserted after it, so as not to
5393 terminate the header.
5395 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5396 unnecessarily, used atoi() instead of strtol(), and didn't check the
5397 termination when getting file sizes from file names by regex.
5399 19. Completely re-implemented the support for maildirsize files, in the light
5400 of a number of problems with the previous contributed implementation
5401 (4.30/29). In particular:
5403 . If the quota is zero, the maildirsize file is maintained, but no quota is
5406 . If the maildir directory does not exist, it is created before any attempt
5407 to write a maildirsize file.
5409 . The quota value in the file is just a cache; if the quota is changed in
5410 the transport, the new value overrides.
5412 . A regular expression is available for excluding directories from the
5415 20. The autoreply transport checks the characters in options that define the
5416 message's headers; it allows continued headers, but it was checking with
5417 isspace() after an embedded newline instead of explicitly looking for a
5420 21. If all the "regular" hosts to which an address was routed had passed their
5421 expiry times, and had not reached their retry times, the address was
5422 bounced, even if fallback hosts were defined. Now Exim should go on to try
5425 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5426 equivalent code in the SMTP transport. Some hosts send humungous responses
5427 to HELO/EHLO, more than 1024 it seems.
5429 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5430 gives compiler warnings in some environments. I've now done it "properly",
5433 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5434 (because of problems with the built-in one) was declared to return uschar *
5435 instead of char *, causing compiler failure.
5437 25. Fixed a file descriptor leak when processing alias/forward files.
5439 26. Fixed a minor format string issue in dbfn.c.
5441 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5443 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5444 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5447 29. When a sender address is verified, it is cached, to save repeating the test
5448 when there is more than one recipient in a message. However, when the
5449 verification involves a callout, it is possible for different callout
5450 options to be set for different recipients. It is too complicated to keep
5451 track of this in the cache, so now Exim always runs a verification when a
5452 callout is required, relying on the callout cache for the optimization.
5453 The overhead is duplication of the address routing, but this should not be
5456 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5457 to be verified with callout=postmaster, and the main callout worked but the
5458 postmaster check failed, the verification correctly failed. However, if a
5459 subsequent RCPT command asked for sender verification *without* the
5460 postmaster check, incorrect caching caused this verification also to fail,
5463 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5464 it was not caching the DNS options (qualify_single, search_parents) that
5465 were used when the lookup failed. A subsequent lookup with different
5466 options therefore always gave the same answer, though there were cases
5467 where it should not have. (Example: a "domains = !$mx_any" option on a
5468 dnslookup router: the "domains" option is always processed without any
5469 widening, but the router might have qualify_single set.) Now Exim uses the
5470 cached value only when the same options are set.
5472 32. Added John Jetmore's "exipick" utility to the distribution.
5474 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5475 than a timeout (e.g. a certificate is required, and is not provided), an
5476 Exim server now closes the connection immediately. Previously it waited for
5477 the client to close - but if the client is SSL, it seems that they each
5478 wait for each other, leading to a delay before one of them times out.
5480 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5481 maintained 0.8.x compatibility because I don't think many are using it, and
5482 it is clearly obsolete.
5484 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5487 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5488 client certificate was expired. A simple patch fixes this, though I don't
5489 understand the full logic of why the verify callback is called multiple
5492 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5493 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5494 which causes problems with some clients (such as the Certicom SSL Plus
5495 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5496 disables the coutermeasure allowing Eudora to connect."
5498 38. Exim was not checking that a write() to a log file succeeded. This could
5499 lead to Bad Things if a log got too big, in particular if it hit a file
5500 size limit. Exim now panics and dies if it cannot write to a log file, just
5501 as it does if it cannot open a log file.
5503 39. Modified OS/Makefile-Linux so that it now contains
5505 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5507 The two -D definitions ensure that Exim is compiled with large file
5508 support, which makes it possible to handle log files that are bigger than
5511 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5512 instance) a domain was checked against a named list that involved a lookup,
5513 causing $domain_data to be set, then another domain was checked against the
5514 same list, then the first domain was re-checked, the value of $domain_data
5515 after the final check could be wrong. In particular, if the second check
5516 failed, it could be set empty. This bug probably also applied to
5519 41. The strip_trailing_dot option was not being applied to the address given
5520 with the -f command-line option.
5522 42. The code for reading a message's header from the spool was incrementing
5523 $received_count, but never initializing it. This meant that the value was
5524 incorrect (doubled) while delivering a message in the same process in which
5525 it was received. In the most common configuration of Exim, this never
5526 happens - a fresh exec is done - but it can happen when
5527 deliver_drop_privilege is set.
5529 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5530 it now includes up to 150 characters of the unexpected data in the log
5533 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5534 and building data strings. The size of both of these buffers was 10 000
5535 bytes - far larger than anybody would *ever* want, thought I. Needless to
5536 say, somebody hit the limit. I have increased the maximum line length to
5537 20 000 and the maximum data length of concatenated lines to 100 000. I have
5538 also fixed two bugs, because there was no checking on these buffers. Tsk,
5539 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5540 buffer is too small.
5542 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5543 lsearch lookups. Now it does.
5545 46. When parsing a route_list item in a manualroute router, a fixed-length
5546 buffer was used for the list of hosts. I made this 1024 bytes long,
5547 thinking that nobody would ever have a list of hosts that long. Wrong.
5548 Somebody had a whole pile of complicated expansion conditions, and the
5549 string was silently truncated, leading to an expansion error. It turns out
5550 that it is easier to change to an unlimited length (owing to other changes
5551 that have happened since this code was originally written) than to build
5552 structure for giving a limitation error. The length of the item that
5553 expands into the list of hosts is now unlimited.
5555 47. The lsearch lookup could not handle data where the length of text line was
5556 more than 4095 characters. Such lines were truncated, leading to shortened
5557 data being returned. It should now handle lines of any length.
5559 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5560 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5563 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5564 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5565 no longer happens in comment lines. A list of replacements is now placed
5566 at the head of all of the source files, except those whose only change is
5567 to replace PERL_COMMAND in the very first #! line.
5569 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5570 messages on the queue, with a bottom-up merge sort, using code contributed
5571 by Michael Haardt. This should make operations like -bp somewhat faster on
5572 large queues. It won't affect queue runners, except when queue_run_in_order
5575 51. Installed eximstats 1.31 in the distribution.
5577 52. Added support for SRV lookups to the dnslookup router.
5579 53. If an ACL referred to $message_body or $message_body_end, the value was not
5580 reset for any messages that followed in the same SMTP session.
5582 54. The store-handling optimization for building very long strings was not
5583 differentiating between the different store pools. I don't think this
5584 actually made any difference in practice, but I've tidied it.
5586 55. While running the routers to verify a sender address, $sender_address
5587 was still set to the sender address. This is wrong, because when routing to
5588 send a bounce to the sender, it would be empty. Therefore, I have changed
5589 it so that, while verifying a sender address, $sender_address is set to <>.
5590 (There is no change to what happens when verifying a recipient address.)
5592 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5593 target A or AAAA records (if not already returned) without resetting the
5594 qualify_single or search_parents options of the DNS resolver. These are
5595 inappropriate in this case because the targets of MX and SRV records must
5596 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5597 target that, when qualified, matched something in the local domain. These
5598 two options are now turned off when doing these lookups.
5600 57. It seems that at least some releases of Reiserfs (which does not have the
5601 concept of a fixed number of inodes) returns zero and not -1 for the
5602 number of available inodes. This interacted badly with check_spool_inodes,
5603 which assumed that -1 was the "no such thing" setting. What I have done is
5604 to check that the total number of inodes is greater than zero before doing
5605 the test of how many are available.
5607 58. When a "warn" ACL statement has a log_message modifier, the message is
5608 remembered, and not repeated. This is to avoid a lot of repetition when a
5609 message has many recipients that cause the same warning to be written.
5610 However, Exim was preserving the list of already written lines for an
5611 entire SMTP session, which doesn't seem right. The memory is now reset if a
5612 new message is started.
5614 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5615 debugging output unless log_rewrite was also set.
5617 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5618 of (int)(handle) when we know that handle contains (void *)(-1).
5620 61. The Exim daemon panic-logs an error return when it closes the incoming
5621 connection. However "connection reset by peer" seems to be common, and
5622 isn't really an error worthy of noting specially, so that particular error
5625 62. When Exim is trying to find all the local interfaces, it used to panic and
5626 die if the ioctl to get the interface flags failed. However, it seems that
5627 on at least one OS (Solaris 9) it is possible to have an interface that is
5628 included in the list of interfaces, but for which you get a failure error
5629 for this call. This happens when the interface is not "plumbed" into a
5630 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5631 failure of the "get flags" call assumes that the interface is down.
5633 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5634 makes life easier for people who are doing arithmetic on fields extracted
5635 from dates, where you often get leading zeros that should not be
5636 interpreted as octal.
5638 64. Added qualify_domain to the redirect router, to override the global
5641 65. If a pathologically long header line contained very many addresses (the
5642 report of this problem mentioned 10 000) and each of them was rewritten,
5643 Exim could use up a very large amount of memory. (It kept on making new
5644 copies of the header line as it rewrote, and never released the old ones.)
5645 At the expense of a bit more processing, the header rewriting function has
5646 been changed so that it no longer eats memory in this way.
5648 66. The generation of the Received: header has been moved from the time that a
5649 message starts to be received, to the time that it finishes. The timestamp
5650 in the Received: header should now be very close to that of the <= log
5651 line. There are two side-effects of this change:
5653 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5654 logged header lines no longer include the local Received: line, because
5655 it has not yet been created. The same applies to a copy of the message
5656 that is returned to a non-SMTP sender when a message is rejected.
5658 (b) When a filter file is tested using -bf, no additional Received: header
5659 is added to the test message. After some thought, I decided that this
5662 This change does not affect the value of $received_for. It is still set
5663 after address rewriting, but before local_scan() is called.
5665 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5667 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5668 gave an unhelpful panic error message, and a defer error. I have managed to
5669 change this behaviour so that it now rejects any supplied certificate,
5670 which seems right, as the list of acceptable certificates is empty.
5672 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5673 gave an unhelpful defer error. I have not managed to make this reject any
5674 supplied certificates, but the error message it gives is "no certificate
5675 supplied", which is not helpful.
5677 70. exigrep's output now also includes lines that are not associated with any
5678 message, but which match the given pattern. Implemented by a patch from
5679 Martin Sluka, which also tidied up the Perl a bit.
5681 71. Recipient callout verification, like sender verification, was using <> in
5682 the MAIL FROM command. This isn't really the right thing, since the actual
5683 sender may affect whether the remote host accepts the recipient or not. I
5684 have changed it to use the actual sender in the callout; this means that
5685 the cache record is now keyed on a recipient/sender pair, not just the
5686 recipient address. There doesn't seem to be a real danger of callout loops,
5687 since a callout by the remote host to check the sender would use <>.
5688 [SEE ABOVE: changed after hitting problems.]
5690 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5691 temporary errors. However, in the case of such a code being given after
5692 the end of a data transmission (i.e. after ".") Exim was failing to write
5693 a retry record for the message. (Yes, there was some broken host that was
5694 actually sending 8xx at this point.)
5696 73. An unknown lookup type in a host list could cause Exim to panic-die when
5697 the list was checked. (An example that provoked this was putting <; in the
5698 middle of a list instead of at the start.) If this happened during a DATA
5699 ACL check, a -D file could be left lying around. This kind of configuration
5700 error no longer causes Exim to die; instead it causes a defer error. The
5701 incident is still logged to the main and panic logs.
5703 74. Buglet left over from Exim 3 conversion. The message "too many messages
5704 in one connection" was written to the rejectlog but not the mainlog, except
5705 when address rewriting (yes!) was being logged.
5707 75. Added write_rejectlog option.
5709 76. When a system filter was run not as root (that is, when system_filter_user
5710 was set), the values of the $n variables were not being returned to the
5711 main process; thus, they were not subsequently available in the $sn
5714 77. Added +return_path_on_delivery log selector.
5716 78. A connection timeout was being treated differently from recipients deferred
5717 when testing hosts_max_try with a message that was older than the host's
5718 retry timeout. (The host should not be counted, thus allowing all hosts to
5719 be tried at least once before bouncing.) This may have been the cause of an
5720 occasionally reported bug whereby a message would remain on the queue
5721 longer than the retry timeout, but would be bounced if a delivery was
5722 forced. I say "may" because I never totally pinned down the problem;
5723 setting up timeout/retry tests is difficult. See also the next item.
5725 79. The ultimate address timeout was not being applied to errors that involved
5726 a combination of host plus message (for example, a timeout on a MAIL
5727 command). When an address resolved to a number of possible hosts, and they
5728 were not all tried for each delivery (e.g. because of hosts_max_try), a
5729 message could remain on the queue longer than the retry timeout.
5731 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5734 81. Fixed an obscure SMTP outgoing bug which required at least the following
5735 conditions: (a) there was another message waiting for the same server;
5736 (b) the server returned 5xx to all RCPT commands in the first message so
5737 that the message was not completed; (c) the server dropped the connection
5738 or gave a negative response to the RSET that Exim sends to abort the
5739 transaction. The observed case was a dropped connection after DATA that had
5740 been sent in pipelining mode. That is, the server had advertised PIPELINING
5741 but was not implementing it correctly. The effect of the bug was incorrect
5742 behaviour, such as trying another host, and this could lead to a crash.
5748 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5749 and daemon.c were passed as pointers to ints; they should have been
5750 pointers to socklen_t variables (which are typically unsigned ints).
5752 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5755 3. Fixed a really odd bug that affected only the testing scheme; patching a
5756 certain fixed string in the binary changed the value of another string that
5757 happened to be identical to the end of the original first string.
5759 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5760 name", it returns that address as the IP address. On some operating
5761 systems (e.g. Solaris), it also passes back the IP address string as the
5762 "host name". However, on others (e.g. Linux), it passes back an empty
5763 string. Exim wasn't checking for this, and was changing the host name to an
5764 empty string, assuming it had been canonicalized.
5766 5. Although rare, it is permitted to have more than one PTR record for a given
5767 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5768 all the names associated with an address, because they do in Solaris.
5769 However, it seems that they do not in Linux for data that comes from the
5770 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5771 I found this out when I moved to a new Linux workstation and tried to run
5772 the Exim test suite.
5774 To get round this problem I have changed the code so that it now does its
5775 own call to the DNS to look up PTR records when searching for a host name.
5776 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5777 addresses that are only in /etc/hosts are still found.
5779 This behaviour is, however, controlled by an option called host_lookup_
5780 order, which defaults to "bydns:byaddr". If people want to use the other
5781 order, or indeed, just use one or the other means of lookup, they can
5782 specify it in this variable.
5784 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5785 some operating systems, this comes back from gethostbyaddr() as an empty
5786 string, and this is what Exim used to test for. However, it seems that in
5787 other systems, "." is yielded. Exim now tests for this case too.
5789 7. The values of check_spool_space and check_log_space are now held internally
5790 as a number of kilobytes instead of an absolute number of bytes. If a
5791 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5792 kilobyte. This means that much larger values can be stored.
5794 8. Exim monitor: an attempt to get the action menu when not actually pointing
5795 at a message produces an empty menu entitled "No message selected". This
5796 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5797 no entries in it ("Shell widget menu has zero width and/or height"). So I
5798 have added a single, blank menu entry in this case.
5800 9. Added ${quote_local_part.
5802 10. MIME decoding is now applied to the contents of Subject: header lines when
5805 11. Now that a reference to $sender_host_address automatically causes a reverse
5806 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5807 host lookup before query-style lookups in lists that might use this
5808 variable. This has therefore been abolished, and the "net-" prefix is no
5809 longer necessary for query-style lookups.
5811 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5812 have been a typo for LFLAGS, so it has been changed.
5814 13. The install script calls Exim with "-C /dev/null" in order to find the
5815 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5816 to be output. However, since Exim outputs its version number before the
5817 error, it didn't break the script. It just looked ugly. I fixed this by
5818 always allowing "-C /dev/null" if the caller is root.
5820 14. Ignore overlarge ACL variable number when reading spool file - insurance
5821 against a later release with more variables having written the file.
5823 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5824 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5825 this, as well as the form without the "IPv6" on the front.
5827 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5828 OS/Makefile-Darwin file.
5830 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5831 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5833 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5834 variable. However, some OS use ints for inodes. I've added cast to long int
5835 to get rid of the compiler warning.
5837 19. I had forgotten to lock out "/../" in configuration file names when
5838 ALT_CONFIG_PREFIX was set.
5840 20. Routers used for verification do not need to specify transports. However,
5841 if such a router generated a host list, and callout was configured, Exim
5842 crashed, because it could not find a port number from the (non-existent)
5843 transport. It now assumes port 25 in this circumstance.
5845 21. Added the -t option to exigrep.
5847 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5848 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5849 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5850 them is set, LOOKUP_LSEARCH is forced.
5852 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5854 24. Added sender and host information to the "rejected by local_scan()" log
5855 line; previously there was no indication of these.
5857 25. Added .include_if_exists.
5859 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5860 Linux. It turns out that not all file systems support this. Apparently some
5861 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5862 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5863 not supported on the file descriptor, is now ignored when Exim is trying to
5864 sync a directory. This applies only to Linux.
5866 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5868 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5869 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5872 29. Added support for maildirsize files from supplied patch (modified a bit).
5874 30. The use of :fail: followed by an empty string could lead Exim to respond to
5875 sender verification failures with (e.g.):
5877 550 Verification failed for <xxx>
5878 550 Sender verify failed
5880 where the first response line was missing the '-' that indicates it is not
5881 the final line of the response.
5883 31. The loop for finding the name of the user that called Exim had a hardwired
5884 limit of 10; it now uses the value of finduser_retries, which is used for
5885 all other user lookups.
5887 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5890 33. Exim was neglecting to zero errno before one call of strtol() when
5891 expanding a string and expecting an integer value. On some systems this
5892 resulted in spurious "integer overflow" errors. Also, it was casting the
5893 result into an int without checking.
5895 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5896 did not work. The code looks as if it has *never* worked, though it appears
5897 to have been documented since at least release 1.62. I have made it work.
5899 35. The "timeout_DNS" error in retry rules, also documented since at least
5900 1.62, also never worked. As it isn't clear exactly what this means, and
5901 clearly it isn't a major issue, I have abolished the feature by treating it
5902 as "timeout", and writing a warning to the main and panic logs.
5904 36. The display of retry rules for -brt wasn't always showing the error code
5907 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5908 timeout_connect_A, timeout_connect_MX.
5910 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5911 to the empty sender.
5913 39. The daemon was not analysing the content of -oX till after it had closed
5914 stderr and disconnected from the controlling terminal. This meant that any
5915 syntax errors were only noted on the panic log, and the return code from
5916 the command was 0. By re-arranging the code a little, I've made the
5917 decoding happen first, so such errors now appear on stderr, and the return
5918 code is 1. However, the actual setting up of the sockets still happens in
5919 the disconnected process, so errors there are still only recorded on the
5922 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5923 connections (as happens on some IP stacks) was logged at start up time as
5924 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5925 it from "IPv6 and IPv4", which means that two separate sockets are being
5928 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5929 says whether AF_INET or AF_INET6 was passed as an argument.
5931 42. Exiwhat output was messed up when time zones were included in log
5934 43. Exiwhat now gives more information about the daemon's listening ports,
5935 and whether -tls-on-connect was used.
5937 44. The "port" option of the smtp transport is now expanded.
5939 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5940 silently ignored. Now an error message is written to the main and panic
5943 46. There's a new ACL modifier called "logwrite" which writes to a log file
5944 as soon as it is encountered.
5946 47. Added $local_user_uid and $local_user_gid at routing time.
5948 48. Exim crashed when trying to verify a sender address that was being
5951 49. Exim was recognizing only a space character after ".include". It now also
5952 recognizes a tab character.
5954 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5955 extracting the relevant information from the specification. The man page no
5956 longer contains scrambled data for the -d option, and I've added a section
5957 at the front about calling Exim under different names.
5959 51. Added "extra_headers" argument to the "mail" command in filter files.
5961 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5964 53. Installed eximstats 1.29.
5966 54. Added transport_filter_timeout as a generic transport option.
5968 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5969 Cc: header lines. This was required by RFC 822, but it not required by RFC
5972 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5973 incoming messages that did not have them. Now it does so only if the
5974 message originates locally, that is, if there is no associated remote host
5975 address. When Resent- header lines are present, this applies to the Resent-
5976 lines rather than the non-Resent- lines.
5978 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5979 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5981 58. Messages for configuration errors now include the name of the main
5982 configuration file - useful now that there may be more than one file in a
5983 list (.included file names were always shown).
5985 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5986 for those rare installations that do not start the daemon as root or run it
5987 setuid root. I've cut out the call to initgroups() if the daemon is not
5990 60. The Exim user and group can now be bound into the binary as text strings
5991 that are looked up at the start of Exim's processing.
5993 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5995 62. Added $mailstore_basename variable.
5997 63. Installed patch to sieve.c from Michael Haardt.
5999 64. When Exim failed to open the panic log after failing to open the main log,
6000 the original message it was trying to log was written to stderr and debug
6001 output, but if they were not available (the usual case in production), it
6002 was lost. Now it is written to syslog before the two lines that record the
6003 failures to open the logs.
6005 65. Users' Exim filters run in subprocesses under the user's uid. It is
6006 possible for a "deliver" command or an alias in a "personal" command to
6007 provoke an address rewrite. If logging of address rewriting is configured,
6008 this fails because the process is not running as root or exim. There may be
6009 a better way of dealing with this, but for the moment (because 4.30 needs
6010 to be released), I have disabled address rewrite logging when running a
6011 filter in a non-root, non-exim process.
6017 1. The buildconfig auxiliary program wasn't quoting the value set for
6018 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6019 not defined. This bug was masked in 4.22 by the effect that was fixed in
6022 2. Some messages that were rejected after a message id was allocated were
6023 shown as "incomplete" by exigrep. It no longer does this for messages that
6024 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6026 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6027 have allow_domain_literals set, the ID did not get logged in the <= line.
6028 Domain literals are now always recognized in Message-ID: header lines.
6030 4. The first argument for a ${extract expansion item is the key name or field
6031 number. Leading and trailing spaces in this item were not being ignored,
6032 causing some misleading effects.
6034 5. When deliver_drop_privilege was set, single queue runner processes started
6035 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6036 same command in the process it spins off) were not dropping privilege.
6038 6. When the daemon running as "exim" started a queue runner, it always
6039 re-executed Exim in the spun-off process. This is a waste of effort when
6040 deliver_drop_privilege is set. The new process now just calls the
6041 queue-runner function directly.
6047 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6050 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6051 ignored. Though the use of -oP was forcing the writing of a pid file, it
6052 was always written to the default place.
6054 3. If the message "no IP address found for host xxxx" is generated during
6055 incoming verification, it is now followed by identification of the incoming
6056 connection (so you can more easily find what provoked it).
6058 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6060 5. Added some features to "harden" Exim a bit more against certain attacks:
6062 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6063 be put in Local/Makefile. This is like the never_users runtime option,
6064 but it cannot be overridden. The default setting is "root".
6066 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6067 prefix string with which any file named in a -C command line option
6070 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6071 is retained for -C and -D only if the caller of Exim is root. Without
6072 it, the exim user may also use -C and -D and retain privilege.
6074 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6075 command line option is disabled.
6077 6. Macro names set by the -D option must start with an upper case letter, just
6078 like macro names defined in the configuration file.
6080 7. Added "dereference=" facility to LDAP.
6082 8. Two instances of the typo "uknown" in the source files are fixed.
6084 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6085 the Configure-Makefile script screwed up while processing it.
6087 10. Incorporated PCRE 4.4.
6089 11. The SMTP synchronization check was not operating right at the start of an
6090 SMTP session. For example, it could not catch a HELO sent before the client
6091 waited for the greeting. There is now a check for outstanding input at the
6092 point when the greeting is written. Because of the duplex, asynchronous
6093 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6094 way, but not yet received, when the check is performed.
6096 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6097 on TCP/IP sockets, because this apparently causes some broken clients to
6100 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6101 unchanged) from the Cygwin maintainer.
6103 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6104 was supported instead of testing for mbx. Effectively a typo.
6106 15. The spa authenticator server code was not checking that the input it
6107 received was valid base64.
6109 16. The debug output line for the "set" modifier in ACLs was not showing the
6110 name of the variable that was being set.
6112 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6114 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6115 lookup on its own. Something else was needed to trigger the lookup. For
6116 example, a match in host_lookup or the need for a host name in a host list.
6117 Now, if $sender_host_name is referenced and the host name has not yet been
6118 looked up, a lookup is performed. If the lookup fails, the variable remains
6119 empty, and $host_lookup_failed is set to "1".
6121 19. Added "eqi" as a case-independent comparison operator.
6123 20. The saslauthd authentication condition could segfault if neither service
6124 nor realm was specified.
6126 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6127 error message that was logged was misleading, and incoming SMTP
6128 connections were dropped. The message is now more accurate, and temporary
6129 errors are given to SMTP connections.
6131 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6132 (except RSET) with a 421 error, until QUIT is received. However, it was
6133 failing to send a response to QUIT.
6135 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6136 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6137 ACL is now run after the helo_try_verify_hosts code.
6139 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6140 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6141 case-independent, so other case variants are also recognized). Apparently
6142 some systems use these upper case variants.
6144 25. If more than two messages were waiting for the same host, and a transport
6145 filter was specified for the transport, Exim sent two messages over the
6146 same TCP/IP connection, and then failed with "socket operation on non-
6147 socket" when it tried to send the third.
6149 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6151 27. The extern definition of crypt16() in expand.c was not being excluded when
6152 the OS had its own crypt16() function.
6154 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6155 as a preferred synonym for return_size_limit, both as an option and as an
6158 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6160 30. Changed the default configuration ACL to relax the local part checking rule
6161 for addresses that are not in any local domains. For these addresses,
6162 slashes and pipe symbols are allowed within local parts, but the sequence
6163 /../ is explicitly forbidden.
6165 31. SPA server authentication was not clearing the challenge buffer before
6168 32. log_message in a "warn" ACL statement was writing to the reject log as
6169 well as to the main log, which contradicts the documentation and doesn't
6170 seem right (because no rejection is happening). So I have stopped it.
6172 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6173 However, I am unable to do any testing of this.
6175 34. Fixed an infelicity in the appendfile transport. When checking directories
6176 for a mailbox, to see if any needed to be created, it was accidentally
6177 using path names with one or more superfluous leading slashes; tracing
6178 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6180 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6181 contents are added to the log line that is written for every discarded
6182 recipient. (Previously a log_message setting was ignored.)
6184 36. The ${quote: operator now quotes the string if it is empty.
6186 37. The install script runs exim in order to find its version number. If for
6187 some reason other than non-existence or emptiness, which it checks, it
6188 could not run './exim', it was installing it with an empty version number,
6189 i.e. as "exim-". This error state is now caught, and the installation is
6192 38. An argument was missing from the function that creates an error message
6193 when Exim fails to connect to the socket for saslauthd authentication.
6194 This could cause Exim to crash, or give a corrupted message.
6196 39. Added isip, isip4, and isip6 to ${if conditions.
6198 40. The ACL variables $acl_xx are now saved with the message, and can be
6199 accessed later in routers, transports, and filters.
6201 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6202 strings in the file are not string-expanded.
6204 42. If a MAIL command specified a SIZE value that was too large to fit into an
6205 int variable, the check against message_size_limit failed. Such values are
6206 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6207 day this will have to be increased, but I don't think I want to be around
6208 when emails are that large.
6215 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6216 iconv() is not standard in FreeBSD.
6218 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6219 IPv6 enabled. The observed symptom was a segmentation fault on return from
6220 the function os_common_find_running_interfaces() in src/os.c.
6222 3. In the check_special_case() function in daemon.c I had used "errno" as an
6223 argument name, which causes warnings on some systems. This was basically a
6224 typo, since it was named "eno" in the comments!
6226 4. The code that waits for the clock to tick (at a resolution of some fraction
6227 of a second) so as to ensure message-id uniqueness was always waiting for
6228 at least one whole tick, when it could have waited for less. [This is
6229 almost certainly not relevant at current processor speeds, where it is
6230 unlikely to ever wait at all. But we try to future-proof.]
6232 5. The function that sleeps for a time interval that includes fractions of a
6233 second contained a race. It did not block SIGALRM between setting the
6234 timer, and suspending (a couple of lines later). If the interval was short
6235 and the sigsuspend() was delayed until after it had expired, the suspension
6236 never ended. On busy systems this could lead to processes getting stuck for
6239 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6240 process, before it forks any delivery processes. The open lookup caching
6241 mechanism meant that the open file or database connection was passed into
6242 the delivery process. The problem was that delivery processes always tidy
6243 up cached lookup data. This could cause a problem for the next delivery
6244 process started by the queue runner, because the external queue runner
6245 process does not know about the closure. So the next delivery process
6246 still has data in the lookup cache. In the case of a file lookup, there was
6247 no problem because closing a file descriptor in a subprocess doesn't affect
6248 the parent. However, if the lookup was caching a connection to a database,
6249 the connection was closed, and the second delivery process was likely to
6250 see errors such as "PGSQL: query failed: server closed the connection
6251 unexpectedly". The problem has been fixed by closing all cached lookups
6252 in a queue runner before running a delivery process.
6254 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6255 seem to have the "const" qualifier which it has on other OS. I've
6258 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6259 *of the same type* (client or server) with the same public name that an
6260 error should be diagnosed.
6262 9. When Exim looked up a host name for an IP address, but failed to find the
6263 original IP address when looking up the host name (a safety check), it
6264 output the message "<ip address> does not match any IP for NULL", which was
6265 confusing, to say the least. The bug was that the host name should have
6266 appeared instead of "NULL".
6268 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6269 user that is built into the binary, and the -C or -D options is used, root
6270 privilege is dropped before the configuration file is read. In addition,
6271 logging is switched to stderr instead of the normal log files. If the
6272 configuration then re-defines the Exim user, the unprivileged environment
6273 is probably not what is expected, so Exim logs a panic warning message (but
6276 However, if deliver_drop_privilege is set, the unprivileged state may well
6277 be exactly what is intended, so the warning has been cut out in that case,
6278 and Exim is allowed to try to write to its normal log files.
6284 1. smtp_return_error_details was not giving details for temporary sender
6285 or receiver verification errors.
6287 2. Diagnose a configuration error if two authenticators have the same public
6290 3. Exim used not to create the message log file for a message until the first
6291 delivery attempt. This could be confusing when incoming messages were held
6292 for policy or load reasons. The message log file is now created at the time
6293 the message is received, and an initial "Received" line is written to it.
6295 4. The automatically generated man page for command line options had a minor
6296 bug that caused no ill effects; however, a more serious problem was that
6297 the procedure for building the man page automatically didn't always
6298 operate. Consequently, release 4.20 contains an out-of-date version. This
6299 shouldn't happen again.
6301 5. When building Exim with embedded Perl support, the script that builds the
6302 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6303 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6305 6. The freeze_tell option was not being used for messages that were frozen on
6306 arrival, either by an ACL or by local_scan().
6308 7. Added the smtp_incomplete_transaction log selector.
6310 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6311 was accepting AUTH without a new EHLO.
6313 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6314 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6317 10. Logging of TCP/IP connections (when configured) now happens in the main
6318 daemon process instead of the child process, so that the TCP/IP connection
6319 count is more accurate (but it can never be perfect).
6321 11. The use of "drop" in a nested ACL was not being handled correctly in the
6322 outer ACL. Now, if condition failure induced by the nested "drop" causes
6323 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6324 or "require"), the connection is dropped.
6326 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6327 that yield "discard" can now be used with an "accept" or a "discard" verb,
6328 but an error is generated for any others (because I can't see a useful way
6329 to define what should happen).
6331 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6332 are now processed in the same way as lines in the Exim configuration file.
6333 In particular, continuation lines are supported.
6335 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6337 15. Added -ti meaning -t -i.
6339 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6340 domains, and warn by logging if others are found.
6342 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6343 returned from GIFCONF when Exim is trying to find the list of interfaces on
6344 a host. The code in os.c has been modified to copy each ifreq to an aligned
6345 structure in all cases.
6347 Also, in some cases, the returned ifreq's were being copied to a 'struct
6348 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6349 means the last couple of bytes of an IPv6 address could be chopped if the
6350 ifreq contained only a normal sockaddr (14 bytes storage).
6352 18. Named domain lists were not supported in the hosts_treat_as_local option.
6353 An entry such as +xxxx was not recognized, and was treated as a literal
6356 19. Ensure that header lines added by a DATA ACL are included in the reject log
6357 if the ACL subsequently rejects the message.
6359 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6360 MD5 (which is deprecated).
6362 21. When testing a filter file using -bf, Exim was writing a message when it
6363 took the sender from a "From " line in the message, but it was not doing so
6364 when it took $return_path from a Return-Path: header line. It now does.
6366 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6367 with a valid header line field name (a series of printing characters
6368 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6370 23. Changed "disc" in the source to "disk" to conform to the documentation and
6371 the book and for uniformity.
6373 24. Ignore Sendmail's -Ooption=value command line item.
6375 25. When execve() failed while trying to run a command in a pipe transport,
6376 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6377 could be confused with a return value of 69 from the command itself. This
6378 has been changed to 127, the value the shell returns if it is asked to run
6379 a non-existent command. The wording for the related log line suggests a
6380 non-existent command as the problem.
6382 26. If received_header_text expands to an empty string, do not add a Received:
6383 header line to the message. (Well, it adds a token one on the spool, but
6384 marks it "old" so that it doesn't get used or transmitted.)
6386 27. Installed eximstats 1.28 (addition of -nt option).
6388 28. There was no check for failure on the call to getsockname() in the daemon
6389 code. This can fail if there is a shortage of resources on the system, with
6390 ENOMEM, for example. A temporary error is now given on failure.
6392 29. Contrary to the C standard, it seems that in some environments, the
6393 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6394 program. Exim now does this explicitly; it affects the formatting of
6395 timestamps using strftime().
6397 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6398 complaints. I've now initialized all the variables, to avoid this.
6400 32. Header lines added by a system filter were not being "seen" during
6401 transport-time rewrites.
6403 33. The info_callback() function passed to OpenSSL is set up with type void
6404 (*)(SSL *, int, int), as described somewhere. However, when calling the
6405 function (actually a macro) that sets it up, the type void(*)() is
6406 expected. I've put in a cast to prevent warnings from picky compilers.
6408 34. If a DNS black list lookup found a CNAME record, but there were no A
6409 records associated with the domain it pointed at, Exim crashed.
6411 35. If a DNS black list lookup returned more than one A record, Exim ignored
6412 all but the first. It now scans all returned addresses if a particular IP
6413 value is being sought. In this situation, the contents of the
6414 $dnslist_value variable are a list of all the addresses, separated by a
6417 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6418 to accept a host name and all its aliases if the forward lookup for any of
6419 them yielded the IP address of the incoming connection. Now it accepts only
6420 those names whose forward lookup yields the correct IP address. Any other
6421 names are discarded. This closes a loophole whereby a rogue DNS
6422 administrator could create reverse DNS records to break through a
6423 wildcarded host restriction in an ACL.
6425 37. If a user filter or a system filter that ran in a subprocess used any of
6426 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6427 the wrong values were passed to the pipe command ($thisaddress had the
6428 value of $0, $0 had the value of $1, etc). This bug was introduced by
6429 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6431 38. Improved the line breaking for long SMTP error messages from ACLs.
6432 Previously, if there was no break point between 40 and 75 characters, Exim
6433 left the rest of the message alone. Two changes have been made: (a) I've
6434 reduced the minimum length to 35 characters; (b) if it can't find a break
6435 point between 35 and 75 characters, it looks ahead and uses the first one
6436 that it finds. This may give the occasional overlong line, but at least the
6437 remaining text gets split now.
6439 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6440 file descriptors might be low, and that setting 1000 would always raise it.
6441 It turns out that in some environments, the limit is already over 1000 and
6442 that lowering it causes trouble. So now Exim takes care not to decrease it.
6444 40. When delivering a message, the value of $return_path is set to $sender_
6445 address at the start of routing (routers may change the value). By an
6446 oversight, this default was not being set up when an address was tested by
6447 -bt or -bv, which affected the outcome if any router or filter referred to
6450 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6451 log only when "message" or "log_message" are set. However, if one of the
6452 conditions was an address verification, or a call to a nested ACL, the
6453 messages generated by the underlying test were being passed through. This
6454 no longer happens. The underlying message is available in $acl_verify_
6455 message for both "message" and "log_message" expansions, so it can be
6456 passed through if needed.
6458 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6459 new expansion $bh_ to give the encoded byte string without charset
6460 translation. Translation happens only if iconv() is available; HAVE_ICONV
6461 indicates this at build time. HEADERS_CHARSET gives the charset to
6462 translate to; headers_charset can change it in the configuration, and
6463 "headers charset" can change it in an individual filter file.
6465 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6466 that creates RFC 2047 encoded "words" labels them as that charset instead
6467 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6468 expansion operator; (ii) when Exim creates a From: line for a local
6469 message; (iii) when a header line is rewritten to include a "phrase" part.
6471 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6472 buggy, causing it to skip the first lines of messages whose message ID
6473 ended in 'D'. This would not have bitten before Exim release 4.14, because
6474 message IDs were unlikely to end in 'D' before then. The effect was to have
6475 incorrect size information for certain domains.
6477 45. #include "config.h" was missing at the start of the crypt16.c module. This
6478 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6481 46. If there was a timeout during a "random" callout check, Exim treated it as
6482 a failure of the random address, and carried on sending RSET and the real
6483 address. If the delay was just some slowness somewhere, the response to the
6484 original RCPT would be taken as a response to RSET and so on, causing
6485 mayhem of various kinds.
6487 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6488 when I implemented it. It didn't allow for the fact that some option values
6489 may legitimately be negative (e.g. size_addition), and it didn't even do
6490 the right test for positive values.
6492 48. Domain names in DNS records are case-independent. Exim always looks them up
6493 in lower case. Some resolvers return domain names in exactly the case they
6494 appear in the zone file, that is, they may contain uppercase letters. Not
6495 all resolvers do this - some return always lower case. Exim was treating a
6496 change of case by a resolver as a change of domain, similar to a widening
6497 of a domain abbreviation. This triggered its re-routing code and so it was
6498 trying to route what was effectively the same domain again. This normally
6499 caused routing to fail (because the router wouldn't handle the domain
6500 twice). Now Exim checks for this case specially, and just changes the
6501 casing of the domain that it ultimately uses when it transmits the message
6504 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6507 50. If a filter generated a file delivery with a non-absolute name (possible if
6508 no home directory exists for the router), the forbid_file option was not
6511 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6512 the existing equality matching.
6514 52. Exim was using ints instead of ino_t variables in some places where it was
6515 dealing with inode numbers.
6517 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6518 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6519 called TMPDIR, and if it finds it is different, it changes its value.
6521 54. The smtp_printf() function is now made available to local_scan() so
6522 additional output lines can be written before returning. There is also an
6523 smtp_fflush() function to enable the detection of a dropped connection.
6524 The variables smtp_input and smtp_batched_input are exported to
6527 55. Changed the default runtime configuration: the message "Unknown user"
6528 has been removed from the ACL, and instead placed on the localuser router,
6529 using the cannot_route_message feature. This means that any verification
6530 failures that generate their own messages won't get overridden. Similarly,
6531 the "Unrouteable address" message that was in the ACL for unverifiable
6532 relay addresses has also been removed.
6534 56. Added hosts_avoid_esmtp to the smtp transport.
6536 57. The exicyclog script was not checking for the esoteric option
6537 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6538 will work only if exicyclog is run under the appropriate euid.
6540 58. Following a discussion on the list, the rules by which Exim recognises line
6541 endings on incoming messages have been changed. The -dropcr and drop_cr
6542 options are now no-ops, retained only for backwards compatibility. The
6543 following line terminators are recognized: LF CRLF CR. However, special
6544 processing applies to CR:
6546 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6547 nor a local message in the state where . is a terminator.
6549 (ii) If a bare CR is encountered in a header line, an extra space is added
6550 after the line terminator so as not to end the header. The reasoning
6551 behind this is that bare CRs in header lines are most likely either
6552 to be mistakes, or people trying to play silly games.
6554 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6555 was being incorrectly given as 18 bytes larger than it should have been.
6556 This is a VOB (very old bug).
6558 60. This may never have affected anything current, but just in case it has:
6559 When the local host is found other than at the start of a list of hosts,
6560 the local host, those with the same MX, and any that follow, are discarded.
6561 When the list in question was part of a longer list of hosts, the following
6562 hosts (not currently being processed) were also being discarded. This no
6563 longer happens. I'm not sure if this situation could ever has previously
6566 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6569 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6570 header. This is something that is recommended in a new Internet Draft, and
6571 is something that is documented as being done by Sendmail. There are two
6572 possible values. For messages generated by the autoreply transport, Exim
6575 Auto-Submitted: auto-replied
6577 whereas for all other generated messages (e.g. bounces) it adds
6579 Auto-Submitted: auto-generated
6581 63. The "personal" condition in filters now includes a test for the
6582 Auto-Submitted: header. If it contains the string "auto-" the message it
6583 not considered personal.
6585 64. Added rcpt_include_affixes as a generic transport option.
6587 65. Added queue_only_override (default true).
6589 66. Added the syslog_duplication option.
6591 67. If what should have been the first header line of a message consisted of
6592 a space followed by a colon, Exim was mis-interpreting it as a header line.
6593 It isn't of course - it is syntactically invalid and should therefore be
6594 treated as the start of the message body. The misbehaviour could have
6595 caused a number of strange effects, including loss of data in subsequent
6596 header lines, and spool format errors.
6598 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6599 client host had authenticated. This control can now be exercised by an ACL
6600 for more flexibility.
6602 69. By default, callouts do not happen when testing with -bh. There is now a
6603 variant, -bhc, which does actually run the callout code, including
6604 consulting and updating the callout cache.
6606 70. Added support for saslauthd authentication, courtesy of Alexander
6609 71. If statvfs() failed on the spool or log directories while checking their
6610 size for availability, Exim confusingly gave the error "space shortage".
6611 Furthermore, in debugging mode it crashed with a floating point exception.
6612 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6613 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6614 serious problem, Exim now writes to the main and panic logs when this
6615 happens, with details of the failure. It then refuses to accept the
6616 incoming message, giving the message "spool directory problem" or "log
6617 directory problem" with a 421 code for SMTP messages.
6619 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6620 0, 1, and 2 exist, because some OS complain for execs without them (see
6621 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6622 descriptors. However, the code omitted to check that the open succeeded,
6623 causing mysterious errors if for some reason the permissions on /dev/null
6624 got screwed. Now Exim writes a message to the main and panic logs, and
6625 bombs out if it can't open /dev/null.
6627 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6628 interact so that it is all more flexible. It is supposed to remain
6629 backwards compatible. Also added extra_local_interfaces.
6631 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6632 to bomb out with an assertion failure - to the client this appears as a
6633 connection drop. This problem occurs in the part of the code that was taken
6634 from the Samba project. Fortunately, the assertion is in a very simple
6635 function, so I have fixed this by reproducing the function inline in the
6636 one place where it is called, and arranging for authentication to fail
6637 instead of killing the process with assert().
6639 75. The SPA client code was not working when the server requested OEM rather
6640 than Unicode encoding.
6642 76. Added code to make require_files with a specific uid setting more usable in
6643 the case where statting the file as root fails - usually a non-root-mounted
6644 NFS file system. When this happens and the failure is EACCES, Exim now
6645 forks a subprocess and does the per-uid checking as the relevant uid.
6647 77. Added process_log_path.
6649 78. If log_file_path was not explicitly set, a setting of check_log_space or
6650 check_log_inodes was ignored.
6652 79. If a space check for the spool or log partitions fails, the incident is now
6653 logged. Of course, in the latter case the data may get lost...
6655 80. Added the %p formatting code to string_format() so that it can be used to
6656 print addresses in debug_print(). Adjusted all the address printing in the
6657 debugging in store.c to use %p rather than %d.
6659 81. There was a concern that a line of code in smtp_in.c could overflow a
6660 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6661 initially expressed, the concern was not well-founded, because trailing
6662 spaces are removed early. However, if the trailing spaces were followed by
6663 a NULL, they did not get removed, so the overflow was possible. Two fixes
6666 (a) I re-wrote the offending code in a cleaner fashion.
6667 (b) If an incoming SMTP command contains a NULL character, it is rejected
6670 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6671 runs initgroups(), so that if the Exim user is in any additional groups,
6672 they will be used during message reception.
6678 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob