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 linidn.
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.
26 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
27 supports it and a size is available (ie. the sending peer gave us one).
29 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
30 years of being deprecated). The replacements are the ACLs
31 acl_smtp_mime and acl_not_smtp_mime.
33 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
34 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
35 or even an in-clear connection were permitted. Now, if the host lookup
36 was dnssec and dane was requested then the host is only used if the
37 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
38 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
39 if one fails this test.
40 This means that a poorly-configured remote DNS will make it incommunicado;
41 but it protects against a DNS-interception attack on it.
43 JH/04 Bug 1810: make continued-use of an open smtp transport connection
44 non-noisy when a race steals the message being considered.
46 JH/05 If main configuration option tls_certificate is unset, generate a
47 self-signed certificate for inbound TLS connections.
49 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
50 in rewrites and routers.
52 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
53 and logged a warning sing 4.83; now they are a configuration file error.
55 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
56 (lacking @domain). Apply the same qualification processing as RCPT.
58 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
60 JH/10 Support ${sha256:} applied to a string (as well as the previous
63 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
64 a cutthrough deliver is pending, as we always want to make a connection.
65 This also avoids re-routing the message when later placing the cutthrough
66 connection after a verify cache hit.
67 Do not update it with the verify result either.
69 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
70 when routing results in more than one destination address.
72 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
73 signing (which inhibits the cutthrough capability). Previously only
74 the presence of an option was tested; now an expansion evaluating as
75 empty is permissible (obviously it should depend only on data available
76 when the cutthrough connection is made).
78 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
79 the relevant preceding SMTP command did not note the pipelining mode.
81 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
82 Previously they were not counted.
84 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
85 as one having no matching records. Previously we deferred the message
86 that needed the lookup.
88 JH/17 Fakereject: previously logged as a norml message arrival "<="; now
89 distinguished as "(=".
91 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
92 for missing MX records. Previously it only worked for missing A records.
94 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
96 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
97 after the data-go-ahead and data-ack. Patch from Jason Betts.
99 JH/21 Bug 1846: Send DMARC forensic reports for reject and quaratine results,
100 even for a "none" policy. Patch from Tony Meyer.
102 JH/22 Fix continued use of a connection for further deliveries. If a port was
103 specified by a router, it must also match for the delivery to be
106 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
107 When one of the recipients of a message was unsuitable for the connection
108 (has no matching addresses), we lost track of needing to mark it
109 deferred. As a result mail would be lost.
111 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
113 JH/25 Decoding ACL controls is now done using a binary search; the source code
114 takes up less space and should be simpler to maintain. Merge the ACL
115 condition decode tables also, with similar effect.
117 JH/26 Fix problem with one_time used on a redirect router which returned the
118 parent address unchanged. A retry would see the parent address marked as
119 delivered, so not attempt the (identical) child. As a result mail would
122 JH/27 Fix a possible security hole, wherein a process operating with the Exim
123 UID can gain a root shell. Credit to http://www.halfdog.net/ for
124 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
127 JH/28 Enable {spool,log} filesystem space and inode checks as default.
128 Main config options check_{log,spool}_{inodes,space} are now
129 100 inodes, 10MB unless set otherwise in the configuration.
131 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
132 Previously it only applied to the main-section connection policy
135 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
137 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
138 by me. Added RFC7919 DH primes as an alternative.
140 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
141 are not in the system include path.
143 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
144 GnuTLS, when a session startup failed (eg because the client disconnected)
145 Exim did stdio operations after fclose. This was exposed by a recent
146 change which nulled out the file handle after the fclose.
148 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
149 signed directly by the cert-signing cert, rather than an intermediate
150 OCSP-signing cert. This is the model used by LetsEncrypt.
152 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
154 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
155 an incoming connection.
157 HS/02 Bug 1802: Do not half-close the connection after sending a request
160 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
161 fallback to "prime256v1".
163 JH/34 SECURITY: Use proper copy of DATA command in error message.
164 Could leak key material. Remotely exploitable. CVE-2016-9963.
170 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
171 and 3.4.4 - once the server is enabled to respond to an OCSP request
172 it does even when not requested, resulting in a stapling non-aware
173 client dropping the TLS connection.
175 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
176 support variable-length bit vectors. No functional change.
178 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
179 The I= interface field on outgoing lines is now after the H= remote
180 host field, same as incoming lines. There is a separate
181 outgoing_interface log selector which allows you to disable the
184 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
185 If not running log_selector +smtp_connection the mainlog would be held
186 open indefinitely after a "too many connections" event, including to a
187 deleted file after a log rotate. Leave the per net connection logging
188 leaving it open for efficiency as that will be quickly detected by the
189 check on the next write.
191 HS/01 Bug 1671: Fix post transport crash.
192 Processing the wait-<transport> messages could crash the delivery
193 process if the message IDs didn't exist for some reason. When
194 using 'split_spool_directory=yes' the construction of the spool
195 file name failed already, exposing the same netto behaviour.
197 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
198 mime_regex ACL conditions.
200 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
201 to DSN fail messages (bounces): remote IP, remote greeting, remote response
202 to HELO, local diagnostic string.
204 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
205 log line to debug. Even when configured with a tls authenticator many
206 client connections are expected to not authenticate in this way, so
207 an authenticate fail is not an error.
209 HS/02 Add the Exim version string to the process info. This way exiwhat
210 gives some more detail about the running daemon.
212 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
213 matter for fast-change records such as DNSBLs.
215 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
216 retry record, even if constant. There may be multiple transports with
217 different interface settings and the retry behaviour needs to be kept
220 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
222 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
224 JH/10 Bug 840: fix log_defer_output option of pipe transport
226 JH/11 Bug 830: use same host for all RCPTS of a message, even under
227 hosts_randomize. This matters a lot when combined with mua_wrapper.
229 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
230 ${quote_pgsql:<string>} operator.
232 JH/13 Bug 1708: avoid misaligned access in cached lookup.
234 JH/14 Change header file name for freeradius-client. Relevant if compiling
235 with Radius support; from the Gentoo tree and checked under Fedora.
237 JH/15 Bug 1712: Introduce $prdr_requested flag variable
239 JH/16 Bug 1714: Permit an empty string as expansion result for transport
240 option transport_filter, meaning no filtering.
242 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
244 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
245 defaults to "*" (all hosts). The variable is now available when not built
246 with TLS, default unset, mainly to enable keeping the testuite sane.
247 If a server certificate is not supplied (via tls_certificate) an error is
248 logged, and clients will find TLS connections fail on startup. Presumably
249 they will retry in-clear.
250 Packagers of Exim are strongly encouraged to create a server certificate
251 at installation time.
253 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
254 with the $config_file variable.
256 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
257 in transport context, after the attempt, and per-recipient. The latter type
258 is per host attempted. The event data is the error message, and the errno
259 information encodes the lookup type (A vs. MX) used for the (first) host,
260 and the trailing two digits of the smtp 4xx reponse.
262 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
263 to write to mainlog (or rejectlog, paniclog) in the window between file
264 creation and permissions/ownership being changed. Particularly affects
265 installations where exicyclog is run as root, rather than exim user;
266 result is that the running daemon panics and dies.
268 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
270 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
271 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
272 "pri" and "weight". Note that the previous implicit priority given by the
273 list order is no longer honoured.
275 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
278 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
279 by defining SUPPORT_SOCKS.
281 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
282 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
283 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
284 variables are renamed to proxy_{local,external}_{address,port}.
286 JH/27 Move Internationalisation support from Experimental to mainline, enabled
287 for a build by defining SUPPORT_I18N
289 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
290 of the query string, and make ${quote_redis:} do that quoting.
292 JH/29 Move Events support from Experimental to mainline, enabled by default
293 and removable for a build by defining DISABLE_EVENT.
295 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
297 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
298 cached by the daemon.
300 JH/32 Move Redis support from Experimental to mainline, enabled for a build
301 by defining LOOKUP_REDIS. The libhiredis library is required.
303 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
304 keys are given for lookup.
306 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
307 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
308 only supported when built with TLS support. The PolarSSL SHA routines
309 are still used when the TLS library is too old for convenient support.
311 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
312 openssl_options), for security. OpenSSL forces this from version 1.1.0
313 server-side so match that on older versions.
315 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
316 allocation for $value could be released as the expansion processing
317 concluded, but leaving the global pointer active for it.
319 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
320 and to use the domains and local_parts ACL conditions.
322 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
323 incorrectly not doubled on cutthrough transmission, hence seen as a
324 body-termination at the receiving system - resulting in truncated mails.
325 Commonly the sender saw a TCP-level error, and retransmitted the message
326 via the normal store-and-forward channel. This could result in duplicates
327 received - but deduplicating mailstores were liable to retain only the
328 initial truncated version.
330 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
332 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
334 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
335 we're in there, support oversigning also; bug 1309.
337 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
339 HS/04 Add support for keep_environment and add_environment options.
341 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
342 either intentional arithmetic overflow during PRNG, or testing config-
345 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
346 delivery resulted in actual delivery. Cancel cutthrough before DATA
349 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
350 rejects a recipient: pass the reject to the originator.
352 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
353 Many were false-positives and ignorable, but it's worth fixing the
356 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
357 for the new environment-manipulation done at startup. Move the routines
358 from being local to tls.c to being global via the os.c file.
360 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
361 an extract embedded as result-arg for a map, the first arg for extract
362 is unavailable so we cannot tell if this is a numbered or keyed
363 extraction. Accept either.
369 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
372 JH/02 The smtp transport option "multi_domain" is now expanded.
374 JH/03 The smtp transport now requests PRDR by default, if the server offers
377 JH/04 Certificate name checking on server certificates, when exim is a client,
378 is now done by default. The transport option tls_verify_cert_hostnames
379 can be used to disable this per-host. The build option
380 EXPERIMENTAL_CERTNAMES is withdrawn.
382 JH/05 The value of the tls_verify_certificates smtp transport and main options
383 default to the word "system" to access the system default CA bundle.
384 For GnuTLS, only version 3.0.20 or later.
386 JH/06 Verification of the server certificate for a TLS connection is now tried
387 (but not required) by default. The verification status is now logged by
388 default, for both outbound TLS and client-certificate supplying inbound
391 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
394 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
395 Status Notification (bounce) messages are now MIME format per RFC 3464.
396 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
397 under the control of the dsn_advertise_hosts option, and routers may
398 have a dsn_lasthop option.
400 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
401 default, modifiable by a malware= option. The list separator for
402 the options can now be changed in the usual way. Bug 68.
404 JH/10 The smtp_receive_timeout main option is now expanded before use.
406 JH/11 The incoming_interface log option now also enables logging of the
407 local interface on delivery outgoing connections.
409 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
410 if the interface and destination host and port all match.
412 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
415 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
416 Patch from Andrew Lewis.
418 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
419 now supports optional time-restrictions, weighting, and priority
420 modifiers per server. Patch originally by <rommer@active.by>.
422 JH/16 The spamd_address main option now supports a mixed list of local
423 and remote servers. Remote servers can be IPv6 addresses, and
424 specify a port-range.
426 JH/17 Bug 68: The spamd_address main option now supports an optional
427 timeout value per server.
429 JH/18 Bug 1581: Router and transport options headers_add/remove can
430 now have the list separator specified.
432 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
435 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
438 JH/21 Support for the A6 type of dns record is withdrawn.
440 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
441 rather than the verbs used.
443 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
444 from 255 to 1024 chars.
446 JH/24 Verification callouts now attempt to use TLS by default.
448 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
449 are generic router options now. The defaults didn't change.
451 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
452 Original patch from Alexander Shikoff, worked over by JH.
454 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
455 files if ZCAT_COMMAND is not executable.
457 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
459 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
461 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
462 Normally benign, it bites when the pair was led to by a CNAME;
463 modern usage is to not canoicalize the domain to a CNAME target
464 (and we were inconsistent anyway for A-only vs AAAA+A).
466 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
468 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
469 when evaluating $sender_host_dnssec.
471 JH/31 Check the HELO verification lookup for DNSSEC, adding new
472 $sender_helo_dnssec variable.
474 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
476 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
478 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
480 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
481 documented as working, but never had. Support all but $spam_report.
483 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
484 added for tls authenticator.
486 HS/03 Add perl_taintmode main config option
492 TL/01 When running the test suite, the README says that variables such as
493 no_msglog_check are global and can be placed anywhere in a specific
494 test's script, however it was observed that placement needed to be near
495 the beginning for it to behave that way. Changed the runtest perl
496 script to read through the entire script once to detect and set these
497 variables, reset to the beginning of the script, and then run through
498 the script parsing/test process like normal.
500 TL/02 The BSD's have an arc4random API. One of the functions to induce
501 adding randomness was arc4random_stir(), but it has been removed in
502 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
503 function when detected.
505 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
506 cause callback expansion.
508 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
509 syntax errors in an expansion can be treated as a string instead of
510 logging or causing an error, due to the internal use of bool_lax
511 instead of bool when processing it.
513 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
514 server certificates when making smtp deliveries.
516 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
518 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
520 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
522 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
523 Merged patch from Sebastian Wiedenroth.
525 JH/05 Fix results-pipe from transport process. Several recipients, combined
526 with certificate use, exposed issues where response data items split
527 over buffer boundaries were not parsed properly. This eventually
528 resulted in duplicates being sent. This issue only became common enough
529 to notice due to the introduction of conection certificate information,
530 the item size being so much larger. Found and fixed by Wolfgang Breyha.
532 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
533 size buffer was used, resulting in syntax errors when an expansion
536 JH/07 Add support for directories of certificates when compiled with a GnuTLS
537 version 3.3.6 or later.
539 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
540 is EXPERIMENTAL_EVENT, the main-configuration and transport options
541 both become "event_action", the variables become $event_name, $event_data
542 and $event_defer_errno. There is a new variable $verify_mode, usable in
543 routers, transports and related events. The tls:cert event is now also
544 raised for inbound connections, if the main configuration event_action
547 TL/06 In test suite, disable OCSP for old versions of openssl which contained
548 early OCSP support, but no stapling (appears to be less than 1.0.0).
550 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
551 server certificate names available under the smtp transport option
552 "tls_verify_cert_hostname" now do not permit multi-component wildcard
555 JH/10 Time-related extraction expansions from certificates now use the main
556 option "timezone" setting for output formatting, and are consistent
557 between OpenSSL and GnuTLS compilations. Bug 1541.
559 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
560 encoded parameter in the incoming message. Bug 1558.
562 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
563 include certificate info, eximon was claiming there were spoolfile
566 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
568 JH/14 Log delivery-related information more consistently, using the sequence
569 "H=<name> [<ip>]" wherever possible.
571 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
572 are problematic for Debian distribution, omit them from the release
575 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
577 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
580 JH/17 Fix a null-indirection in certextract expansions when a nondefault
581 output list separator was used.
586 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
587 checkers that were complaining about end of non-void function with no
590 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
591 This was a regression intruduced in 4.83 by another bugfix.
593 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
595 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
596 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
602 TF/01 Correctly close the server side of TLS when forking for delivery.
604 When a message was received over SMTP with TLS, Exim failed to clear up
605 the incoming connection properly after forking off the child process to
606 deliver the message. In some situations the subsequent outgoing
607 delivery connection happened to have the same fd number as the incoming
608 connection previously had. Exim would try to use TLS and fail, logging
609 a "Bad file descriptor" error.
611 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
612 utilities have not been installed.
614 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
615 temporary space as the ACL may create new global variables.
617 TL/01 LDAP support uses per connection or global context settings, depending
618 upon the detected version of the libraries at build time.
620 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
621 to extract and use the src ip:port in logging and expansions as if it
622 were a direct connection from the outside internet. PPv2 support was
623 updated based on HAProxy spec change in May 2014.
625 JH/02 Add ${listextract {number}{list}{success}{fail}}.
627 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
628 Properly escape header and check for NULL return.
630 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
633 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
635 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
636 characters in header names, implemented as a verify condition.
637 Contributed by Michael Fischer v. Mollard.
639 TL/05 Rename SPF condition results err_perm and err_temp to standardized
640 results permerror and temperror. Previous values are deprecated but
641 still accepted. In a future release, err_perm and err_temp will be
642 completely removed, which will be a backward incompatibility if the
643 ACL tests for either of these two old results. Patch contributed by
644 user bes-internal on the mailing list.
646 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
648 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
649 selectors, in both main and reject logs.
651 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
654 JH/07 Add malware type "sock" for talking to simple daemon.
656 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
658 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
659 routers/transports under cutthrough routing.
661 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
662 numbers. Touch up "bool" conditional to keep the same definition.
664 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
666 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
668 JH/12 Expand items in router/transport headers_add or headers_remove lists
669 individually rather than the list as a whole. Bug 1452.
671 Required for reasonable handling of multiple headers_ options when
672 they may be empty; requires that headers_remove items with embedded
673 colons must have them doubled (or the list-separator changed).
675 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
676 view the policy declared in the DMARC record. Currently, $dmarc_status
677 is a combined value of both the record presence and the result of the
680 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
682 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
683 dnslookup router and the smtp transport (applying to the forward
686 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
687 of ldap servers used for a specific lookup. Patch provided by Heiko
690 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
691 New variable $lookup_dnssec_authenticated for observability.
693 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
694 Patch submitted by Lars Timman.
696 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
698 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
699 Requires trusted mode and valid format message id, aborts otherwise.
700 Patch contributed by Heiko Schlichting.
702 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
703 certextract with support for various fields. Bug 1358.
705 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
706 is requested by default, modifiable by smtp transport option
709 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
710 operate on certificate variables to give certificate fingerprints
711 Also new ${sha256:cert_variable}.
713 JH/23 The PRDR feature is moved from being Experimental into the mainline.
715 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
716 Christian Aistleitner.
718 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
720 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
721 file. Patch from Wolfgang Breyha.
723 JH/25 Expand the coverage of the delivery $host and $host_address to
724 client authenticators run in verify callout. Bug 1476.
726 JH/26 Port service names are now accepted for tls_on_connect_ports, to
727 align with daemon_smtp_ports. Bug 72.
729 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
730 support and error reporting did not work properly.
732 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
733 and is readable. Patch from Andrew Colin Kissa.
735 TL/14 Enhance documentation of ${run expansion and how it parses the
736 commandline after expansion, particularly in the case when an
737 unquoted variable expansion results in an empty value.
739 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
741 PP/02 Fix internal collision of T_APL on systems which support RFC3123
742 by renaming away from it. Addresses GH issue 15, reported by
745 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
747 TL/15 SECURITY: prevent double expansion in math comparison functions
748 (can expand unsanitized data). Not remotely exploitable.
755 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
757 PP/02 Make -n do something, by making it not do something.
758 When combined with -bP, the name of an option is not output.
760 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
763 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
764 $sender_host_name and config options to manage this, and basic check
767 PP/05 DSCP support for outbound connections and control modifier for inbound.
769 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
770 (Only plugin which currently uses this is kerberos4, which nobody should
771 be using, but we should make it available and other future plugins might
772 conceivably use it, even though it would break NAT; stuff *should* be
773 using channel bindings instead).
775 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
776 name; added for Sendmail compatibility; requires admin caller.
777 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
778 just ignore it); requires trusted caller.
779 Also parse but ignore: -Ac -Am -X<logfile>
782 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
784 TL/02 Add +smtp_confirmation as a default logging option.
786 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
787 Patch by Magnus Holmgren from 2007-02-20.
789 TL/04 Bugzilla 1281 - Spec typo.
790 Bugzilla 1283 - Spec typo.
791 Bugzilla 1290 - Spec grammar fixes.
793 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
795 TL/06 Add Experimental DMARC support using libopendmarc libraries.
797 TL/07 Fix an out of order global option causing a segfault. Reported to dev
798 mailing list by by Dmitry Isaikin.
800 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
802 JH/02 Support "G" suffix to numbers in ${if comparisons.
804 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
806 NM/01 Bugzilla 1197 - Spec typo
807 Bugzilla 1196 - Spec examples corrections
809 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
811 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
812 gnutls_enable_pkcs11, but renamed to more accurately indicate its
815 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
816 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
818 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
819 "acl {{name}{arg}...}", and optional args on acl condition
822 JH/05 Permit multiple router/transport headers_add/remove lines.
824 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
826 JH/07 Avoid using a waiting database for a single-message-only transport.
827 Performance patch from Paul Fisher. Bugzilla 1262.
829 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
832 JH/09 Add $headers_added variable, with content from use of ACL modifier
833 add_header (but not yet added to the message). Bugzilla 199.
835 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
836 Pulled from Bugzilla 817 by Wolfgang Breyha.
838 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
840 (nb: this is the same fix as in Exim 4.80.1)
842 JH/11 Add A= logging on delivery lines, and a client_set_id option on
845 JH/12 Add optional authenticated_sender logging to A= and a log_selector
848 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
850 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
851 advertise SMTP AUTH mechanism to us, instead of a generic
852 protocol violation error. Also, make Exim more robust to bad
853 data from the Dovecot auth socket.
855 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
857 When a queue runner is handling a message, Exim first routes the
858 recipient addresses, during which it prunes them based on the retry
859 hints database. After that it attempts to deliver the message to
860 any remaining recipients. It then updates the hints database using
863 So if a recipient address works intermittently, it can get repeatedly
864 deferred at routing time. The retry hints record remains fresh so the
865 address never reaches the final cutoff time.
867 This is a fairly common occurrence when a user is bumping up against
868 their storage quota. Exim had some logic in its local delivery code
869 to deal with this. However it did not apply to per-recipient defers
870 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
872 This change adds a proper retry rule check during routing so that the
873 final cutoff time is checked against the message's age. We only do
874 this check if there is an address retry record and there is not a
875 domain retry record; this implies that previous attempts to handle
876 the address had the retry_use_local_parts option turned on. We use
877 this as an approximation for the destination being like a local
878 delivery, as in LMTP.
880 I suspect this new check makes the old local delivery cutoff check
881 redundant, but I have not verified this so I left the code in place.
883 TF/02 Correct gecos expansion when From: is a prefix of the username.
885 Test 0254 submits a message to Exim with the header
889 When I ran the test suite under the user fanf2, Exim expanded
890 the header to contain my full name, whereas it should have added
891 a Resent-Sender: header. It erroneously treats any prefix of the
892 username as equal to the username.
894 This change corrects that bug.
896 GF/01 DCC debug and logging tidyup
897 Error conditions log to paniclog rather than rejectlog.
898 Debug lines prefixed by "DCC: " to remove any ambiguity.
900 TF/03 Avoid unnecessary rebuilds of lookup-related code.
902 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
903 Bug spotted by Jeremy Harris; was flawed since initial commit.
904 Would have resulted in OCSP responses post-SNI triggering an Exim
905 NULL dereference and crash.
907 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
909 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
910 Bug detection, analysis and fix by Samuel Thibault.
911 Bugzilla 1331, Debian bug #698092.
913 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
915 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
916 Server implementation by Todd Lyons, client by JH.
917 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
918 config variable "prdr_enable" controls whether the server
919 advertises the facility. If the client requests PRDR a new
920 acl_data_smtp_prdr ACL is called once for each recipient, after
921 the body content is received and before the acl_smtp_data ACL.
922 The client is controlled by bolth of: a hosts_try_prdr option
923 on the smtp transport, and the server advertisement.
924 Default client logging of deliveries and rejections involving
925 PRDR are flagged with the string "PRDR".
927 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
928 fclose(). Diagnosis by Todd Lyons.
930 PP/17 Update configure.default to handle IPv6 localhost better.
931 Patch by Alain Williams (plus minor tweaks).
934 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
935 This is now consistent with GnuTLS, and is now documented: the
936 previous undocumented portable approach to treating the option as
937 unset was to force an expansion failure. That still works, and
938 an empty string is now equivalent.
940 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
941 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
942 not performing validation itself.
944 PP/20 Added force_command boolean option to pipe transport.
945 Patch from Nick Koston, of cPanel Inc.
947 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
950 TF/04 Added udpsend ACL modifer and hexquote expansion operator
952 PP/21 Fix eximon continuous updating with timestamped log-files.
953 Broken in a format-string cleanup in 4.80, missed when I repaired the
954 other false fix of the same issue.
955 Report and fix from Heiko Schlichting.
958 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
959 Report from Prashanth Katuri.
961 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
962 It's SecureTransport, so affects any MacOS clients which use the
963 system-integrated TLS libraries, including email clients.
965 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
966 using a MIME ACL for non-SMTP local injection.
967 Report and assistance in diagnosis by Warren Baker.
969 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
971 JH/16 Fix comparisons for 64b. Bugzilla 1385.
973 TL/09 Add expansion variable $authenticated_fail_id to keep track of
974 last id that failed so it may be referenced in subsequent ACL's.
976 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
979 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
980 ldap library initialization, allowing self-signed CA's to be
981 used. Also properly sets require_cert option later in code by
982 using NULL (global ldap config) instead of ldap handle (per
983 session). Bug diagnosis and testing by alxgomz.
985 TL/12 Enhanced documentation in the ratelimit.pl script provided in
986 the src/util/ subdirectory.
988 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
989 renamed to Transport Post Delivery Action by Jeremy Harris, as
992 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
993 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
994 redis_servers = needs to be configured which will be used by the redis
995 lookup. Patch from Warren Baker, of The Packet Hub.
997 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
999 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1000 hostname or reverse DNS when processing a host list. Used suggestions
1001 from multiple comments on this bug.
1003 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1005 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1006 Missed a few lines, added it to make the runtest require no keyboard
1009 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1010 contains upper case chars. Make router use caseful_local_part.
1012 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1013 support when GnuTLS has been built with p11-kit.
1019 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1021 This, or similar/improved, will also be change PP/11 of 4.82.
1027 PP/01 Handle short writes when writing local log-files.
1028 In practice, only affects FreeBSD (8 onwards).
1029 Bugzilla 1053, with thanks to Dmitry Isaikin.
1031 NM/01 Bugzilla 949 - Documentation tweak
1033 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1036 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1038 PP/02 Implemented gsasl authenticator.
1040 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1042 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1043 `pkg-config foo` for cflags/libs.
1045 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1046 with rest of GSASL and with heimdal_gssapi.
1048 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1049 `pkg-config foo` for cflags/libs for the TLS implementation.
1051 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1052 properties get this fed in as external SSF. A number of robustness
1053 and debugging improvements to the cyrus_sasl authenticator.
1055 PP/08 cyrus_sasl server now expands the server_realm option.
1057 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1058 Patch by Jeremy Harris.
1060 PP/10 Added dbmjz lookup type.
1062 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1064 PP/12 MAIL args handles TAB as well as SP, for better interop with
1065 non-compliant senders.
1066 Analysis and variant patch by Todd Lyons.
1068 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1069 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1070 Patch from Dirk Mueller <dmueller@suse.com>
1072 PP/13 tls_peerdn now print-escaped for spool files.
1073 Observed some $tls_peerdn in wild which contained \n, which resulted
1074 in spool file corruption.
1076 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1077 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1078 or write after TLS renegotiation, which otherwise led to messages
1081 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1082 as a tracking header (ie: a signed header comes before the signature).
1083 Patch from Wolfgang Breyha.
1085 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1086 comma-sep list; embedded commas doubled.
1088 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1090 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1092 Report and patch from Dmitry Banschikov.
1094 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
1095 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1096 failures. We appear to now support TLS1.1+ with Exim.
1098 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1099 lets Exim select keys and certificates based upon TLS SNI from client.
1100 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1101 before an outbound SMTP session. New log_selector, +tls_sni.
1103 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1104 NULL dereference. Report and patch from Alun Jones.
1106 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1107 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1108 Not seeing resolver debug output on NetBSD, but suspect this is a
1109 resolver implementation change.
1111 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1112 Left warnings. Added "eximon gdb" invocation mode.
1114 PP/21 Defaulting "accept_8bitmime" to true, not false.
1116 PP/22 Added -bw for inetd wait mode support.
1118 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1119 locate the relevant includes and libraries. Made this the default.
1121 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1122 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1124 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1125 This may cause build issues on older platforms.
1127 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1128 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1129 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1130 Added SNI support via GnuTLS too.
1131 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1133 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1135 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1136 Applied second patch from Janne, implementing suggestion to default
1137 multiple-strings-in-record handling to match SPF spec.
1139 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1141 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1142 read-only, out of scope).
1143 Patch from Wolfgang Breyha, report from Stuart Northfield.
1145 PP/29 Fix three issues highlighted by clang analyser static analysis.
1146 Only crash-plausible issue would require the Cambridge-specific
1147 iplookup router and a misconfiguration.
1148 Report from Marcin Mirosław.
1150 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1152 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1153 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1154 As part of this, removing so much warning spew let me fix some minor
1155 real issues in debug logging.
1157 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1158 assignment on my part. Fixed.
1160 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1161 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1162 Janne Snabb (who went above and beyond: thank you).
1164 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1165 string otherwise requires a connection and a bunch more work and it's
1166 relatively easy to get wrong. Should also expose TLS library linkage
1169 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1170 64-bit ${eval} (JH/03).
1172 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1173 GNU libc to support some of the 64-bit stuff, should not lead to
1174 conflicts. Defined before os.h is pulled in, so if a given platform
1175 needs to override this, it can.
1177 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1178 protection layer was required, which is not implemented.
1179 Bugzilla 1254, patch from Wolfgang Breyha.
1181 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1182 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1183 tls_dhparam take prime identifiers. Also unbreak combination of
1184 OpenSSL+DH_params+TLSSNI.
1186 PP/39 Disable SSLv2 by default in OpenSSL support.
1192 PP/01 Solaris build fix for Oracle's LDAP libraries.
1193 Bugzilla 1109, patch from Stephen Usher.
1195 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1197 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1200 TF/02 Fix a couple more cases where we did not log the error message
1201 when unlink() failed. See also change 4.74-TF/03.
1203 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1204 lock up or crash if it happened to be inside a call to libc when it
1205 got a SIGUSR1 from exiwhat.
1207 The SIGUSR1 handler appends the current process status to the process
1208 log which is later printed by exiwhat. It used to use the general
1209 purpose logging code to do this, but several functions it calls are
1210 not safe for signals.
1212 The new output code in the SIGUSR1 handler is specific to the process
1213 log, and simple enough that it's easy to inspect for signal safety.
1214 Removing some special cases also simplifies the general logging code.
1215 Removing the spurious timestamps from the process log simplifies
1218 TF/04 Improved ratelimit ACL condition.
1220 The /noupdate option has been deprecated in favour of /readonly which
1221 has clearer semantics. The /leaky, /strict, and /readonly update modes
1222 are mutually exclusive. The update mode is no longer included in the
1223 database key; it just determines when the database is updated. (This
1224 means that when you upgrde Exim will forget old rate measurements.)
1226 Exim now checks that the per_* options are used with an update mode that
1227 makes sense for the current ACL. For example, when Exim is processing a
1228 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1229 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1230 must specify per_mail/readonly. If you omit the update mode it defaults to
1231 /leaky where that makes sense (as before) or /readonly where required.
1233 The /noupdate option is now undocumented but still supported for
1234 backwards compatibility. It is equivalent to /readonly except that in
1235 ACLs where /readonly is required you may specify /leaky/noupdate or
1236 /strict/noupdate which are treated the same as /readonly.
1238 A useful new feature is the /count= option. This is a generalization
1239 of the per_byte option, so that you can measure the throughput of other
1240 aggregate values. For example, the per_byte option is now equivalent
1241 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1243 The per_rcpt option has been generalized using the /count= mechanism
1244 (though it's more complicated than the per_byte equivalence). When it is
1245 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1246 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1247 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1248 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1249 Note that using per_rcpt with a non-readonly update mode in more than
1250 one ACL will cause the recipients to be double-counted. (The per_mail
1251 and per_byte options don't have this problem.)
1253 The handling of very low rates has changed slightly. If the computed rate
1254 is less than the event's count (usually one) then this event is the first
1255 after a long gap. In this case the rate is set to the same as this event's
1256 count, so that the first message of a spam run is counted properly.
1258 The major new feature is a mechanism for counting the rate of unique
1259 events. The new per_addr option counts the number of different
1260 recipients that someone has sent messages to in the last time period. It
1261 behaves like per_rcpt if all the recipient addresses are different, but
1262 duplicate recipient addresses do not increase the measured rate. Like
1263 the /count= option this is a general mechanism, so the per_addr option
1264 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1265 example, measure the rate that a client uses different sender addresses
1266 with the options per_mail/unique=$sender_address. There are further
1267 details in the main documentation.
1269 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1271 TF/06 Removed a few PCRE remnants.
1273 TF/07 Automatically extract Exim's version number from tags in the git
1274 repository when doing development or release builds.
1276 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1277 Bugzilla 879. Patch from Paul Fisher.
1279 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1280 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1283 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1285 PP/05 Variable $av_failed, true if the AV scanner deferred.
1286 Bugzilla 1078. Patch from John Horne.
1288 PP/06 Stop make process more reliably on build failure.
1289 Bugzilla 1087. Patch from Heiko Schlittermann.
1291 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1292 Bugzilla 1089. Patch from Heiko Schlittermann.
1294 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1295 Bugzilla 1131. Patch from Holger Weiß.
1297 PP/09 Handle IPv6 addresses with SPF.
1298 Bugzilla 860. Patch from Wolfgang Breyha.
1300 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1302 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1305 PP/11 match_* no longer expand right-hand-side by default.
1306 New compile-time build option, EXPAND_LISTMATCH_RHS.
1307 New expansion conditions, "inlist", "inlisti".
1309 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1311 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1313 PP/14 fix log_write() format string regression from TF/03.
1314 Bugzilla 1152. Patch from Dmitry Isaikin.
1320 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1322 PP/02 Harmonised TLS library version reporting; only show if debugging.
1323 Layout now matches that introduced for other libraries in 4.74 PP/03.
1325 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1327 PP/04 New "dns_use_edns0" global option.
1329 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1332 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1333 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1335 TK/01 Updated PolarSSL code to 0.14.2.
1336 Bugzilla 1097. Patch from Andreas Metzler.
1338 PP/07 Catch divide-by-zero in ${eval:...}.
1339 Fixes bugzilla 1102.
1341 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1344 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1345 format-string attack -- SECURITY: remote arbitrary code execution.
1347 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1348 time unintentionally subject to list matching rules, letting the header
1349 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1350 arbitrary string expansion). This allowed for information disclosure.
1352 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1353 INT_MIN/-1 -- value coerced to INT_MAX.
1359 NM/01 Workround for PCRE version dependency in version reporting
1362 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1363 This fixes portability to compilers other than gcc, notably
1364 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1366 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1367 makefiles for portability to HP-UX and POSIX correctness.
1369 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1370 Also via indented variable definition in the Makefile.
1371 (Debugging by Oliver Heesakkers).
1373 PP/02 Restore caching of spamd results with expanded spamd_address.
1374 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1376 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1377 Improves build reliability. Fix from: Frank Elsner
1379 NM/02 Fix wide character breakage in the rfc2047 coding
1380 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1382 NM/03 Allow underscore in dnslist lookups
1383 Fixes bug 1026. Patch from Graeme Fowler
1385 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1386 Code patches from Adam Ciarcinski of NetBSD.
1388 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1391 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1392 is logged, to avoid truncation. Patch from John Horne.
1394 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1395 Patch from Jakob Hirsch.
1397 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1398 SQL string expansion failure details.
1399 Patch from Andrey Oktyabrski.
1401 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1402 Patch from Simon Arlott.
1404 PP/09 New lookups functionality failed to compile on old gcc which rejects
1405 extern declarations in function scope.
1406 Patch from Oliver Fleischmann
1408 PP/10 Use sig_atomic_t for flags set from signal handlers.
1409 Check getgroups() return and improve debugging.
1410 Fixed developed for diagnosis in bug 927 (which turned out to be
1413 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1414 Patch from Mark Zealey.
1416 PP/12 Bugzilla 1056: Improved spamd server selection.
1417 Patch from Mark Zealey.
1419 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1420 Based on patch from Heiko Schlittermann.
1422 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1423 Patch from Uwe Doering, sign-off by Michael Haardt.
1425 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1426 variable declaration deep within a block. Bug and patch from
1429 PP/15 lookups-Makefile IRIX compatibilty coercion.
1431 PP/16 Make DISABLE_DKIM build knob functional.
1433 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1434 Patch by Simon Arlott
1436 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1437 variable argument macros. Our copy now differs from upstream.
1443 TF/01 Failure to get a lock on a hints database can have serious
1444 consequences so log it to the panic log.
1446 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1447 controlled using the smtp_confirmation log selector.
1449 TF/03 Include the error message when we fail to unlink a spool file.
1451 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1452 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1453 for maintaining out-of-tree patches for some time.
1455 PP/01 Bugzilla 139: Documentation and portability issues.
1456 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1457 Handle per-OS dynamic-module compilation flags.
1459 PP/02 Let /dev/null have normal permissions.
1460 The 4.73 fixes were a little too stringent and complained about the
1461 permissions on /dev/null. Exempt it from some checks.
1462 Reported by Andreas M. Kirchwitz.
1464 PP/03 Report version information for many libraries, including
1465 Exim version information for dynamically loaded libraries. Created
1466 version.h, now support a version extension string for distributors
1467 who patch heavily. Dynamic module ABI change.
1469 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1470 privilege escalation vulnerability whereby the Exim run-time user
1471 can cause root to append content of the attacker's choosing to
1474 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1477 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1478 If dropping privileges for untrusted macros, we disabled normal logging
1479 on the basis that it would fail; for the Exim run-time user, this is not
1480 the case, and it resulted in successful deliveries going unlogged.
1481 Fixed. Reported by Andreas Metzler.
1487 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1488 only prepend for the Resent-* case. Fixes regression introduced in
1489 Exim 4.70 by NM/22 for Bugzilla 607.
1491 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1492 increasing numbers of administrators be bitten by this.
1494 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1496 PP/03 Bugzilla 994: added openssl_options main configuration option.
1498 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1500 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1502 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1504 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1505 without a peer certificate, leading to a segfault because of an
1506 assumption that peers always have certificates. Be a little more
1507 paranoid. Problem reported by Martin Tscholak.
1509 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1510 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1511 NB: ClamAV planning to remove STREAM in "middle of 2010".
1512 CL also introduces -bmalware, various -d+acl logging additions and
1513 more caution in buffer sizes.
1515 PP/09 Implemented reverse_ip expansion operator.
1517 PP/10 Bugzilla 937: provide a "debug" ACL control.
1519 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1521 PP/12 Bugzilla 973: Implement --version.
1523 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1525 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1527 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1529 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1530 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1531 ignore trailing whitespace.
1533 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1535 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1538 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1539 Notification from Dr Andrew Aitchison.
1541 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1542 ExtendedDetectionInfo response format.
1543 Notification from John Horne.
1545 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1548 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1549 XSL and documented dependency on system catalogs, with examples of how
1552 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1555 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1556 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1557 configuration file which is writeable by the Exim user or group.
1559 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1560 of configuration files to cover files specified with the -C option if
1561 they are going to be used with root privileges, not just the default
1564 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1565 option (effectively making it always true).
1567 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1568 files to be used while preserving root privileges.
1570 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1571 that rogue child processes cannot use them.
1573 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1574 run-time user, instead of root.
1576 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1577 Exim run-time user without dropping privileges.
1579 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1580 result string, instead of calling string_vformat() twice with the same
1583 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1584 for other users. Others should always drop root privileges if they use
1585 -C on the command line, even for a whitelisted configure file.
1587 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1589 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1595 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1596 $data_path, and $header_path variables; fixed documentation bugs and
1599 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1600 exipick to access non-standard spools, including the "frozen" queue
1603 NM/01 Bugzilla 965: Support mysql stored procedures.
1604 Patch from Alain Williams
1606 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1608 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1609 Patch from Andreas Metzler
1611 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1612 Patch from Kirill Miazine
1614 NM/05 Bugzilla 671: Added umask to procmail example.
1616 JJ/03 installed exipick 20100323.0, fixing doc bug
1618 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1619 directory. Notification and patch from Dan Rosenberg.
1621 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1623 TK/02 Improve log output when DKIM signing operation fails.
1625 MH/01 Treat the transport option dkim_domain as a colon separated
1626 list, not as a single string, and sign the message with each element,
1627 omitting multiple occurences of the same signer.
1629 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1630 Bugzilla 985, 986. Patch by Simon Arlott
1632 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1633 Patch by Simon Arlott
1635 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1636 MBX locking. Notification from Dan Rosenberg.
1642 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1644 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1646 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1648 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1650 NM/04 Bugzilla 915: Flag broken perl installation during build.
1656 TK/01 Added patch by Johannes Berg that expands the main option
1657 "spamd_address" if it starts with a dollar sign.
1659 TK/02 Write list of recipients to X-Envelope-Sender header when building
1660 the mbox-format spool file for content scanning (suggested by Jakob
1663 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1664 (http://www.dcc-servers.net/) support via dccifd. Activated by
1665 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1667 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1668 by Mark Daniel Reidel <mr@df.eu>.
1670 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1671 When building exim an external PCRE library is now needed -
1672 PCRE is a system library on the majority of modern systems.
1673 See entry on PCRE_LIBS in EDITME file.
1675 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1676 conversation. Added nologin parameter to request.
1677 Patch contributed by Kirill Miazine.
1679 TF/01 Do not log submission mode rewrites if they do not change the address.
1681 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1683 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1684 log files in place. Contributed by Roberto Lima.
1686 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1688 TF/03 Bugzilla 615: When checking the local_parts router precondition
1689 after a local_part_suffix or local_part_prefix option, Exim now
1690 does not use the address's named list lookup cache, since this
1691 contains cached lookups for the whole local part.
1693 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1694 Robert Millan. Documentation is in experimental-spec.txt.
1696 TF/04 Bugzilla 668: Fix parallel build (make -j).
1698 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1700 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1701 Patch provided by Jan Srzednicki.
1703 TF/05 Leading white space used to be stripped from $spam_report which
1704 wrecked the formatting. Now it is preserved.
1706 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1707 that they are available at delivery time.
1709 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1711 TF/08 TLS error reporting now respects the incoming_interface and
1712 incoming_port log selectors.
1714 TF/09 Produce a more useful error message if an SMTP transport's hosts
1715 setting expands to an empty string.
1717 NM/06 Bugzilla 744: EXPN did not work under TLS.
1718 Patch provided by Phil Pennock.
1720 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1721 Patch provided by Richard Godbee.
1723 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1724 acl_smtp_notquit, added index entry.
1726 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1727 Patch provided by Eugene Bujak.
1729 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1730 accept(). Patch provided by Maxim Dounin.
1732 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1733 Patch provided by Phil Pennock.
1735 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1737 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1738 Patch provided by Brad "anomie" Jorsch.
1740 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1741 Patch provided by Dean Brooks.
1743 TK/05 Add native DKIM support (does not depend on external libraries).
1745 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1746 Patch provided by Graeme Fowler.
1748 NM/16 Bugzilla 851: Documentation example syntax fix.
1750 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1752 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1755 NM/19 Bugzilla 745: TLS version reporting.
1756 Patch provided by Phil Pennock.
1758 NM/20 Bugzilla 167: bool: condition support.
1759 Patch provided by Phil Pennock.
1761 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1762 clients. Patch provided by Phil Pennock.
1764 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1765 Patch provided by Brad "anomie" Jorsch.
1767 NM/23 Bugzilla 687: Fix misparses in eximstats.
1768 Patch provided by Heiko Schlittermann.
1770 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1771 Patch provided by Heiko Schlittermann.
1773 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1774 plus update to original patch.
1776 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1778 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1779 Patch provided by David Brownlee.
1781 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1783 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1785 NM/30 Bugzilla 888: TLS documentation bugfixes.
1787 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1789 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1790 Unlike the original bugzilla I have changed all shell scripts in src tree.
1792 NM/33 Bugzilla 898: Transport filter timeout fix.
1793 Patch by Todd Rinaldo.
1795 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1796 Patch by Serge Demonchaux.
1798 NM/35 Bugzilla 39: Base64 decode bug fixes.
1799 Patch by Jakob Hirsch.
1801 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1803 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1805 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1807 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1813 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1814 ALT-N's libdkim that I have put here:
1815 http://duncanthrax.net/exim-experimental/
1817 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1818 were called 'true' and it seems that C99 defines that as a reserved
1819 keyword to be used with 'bool' variable types. That means you could
1820 not include C99-style headers which use bools without triggering
1821 build errors in sieve.c.
1823 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1824 as mailq or other aliases. Changed the --help handling significantly
1825 to do whats expected. exim_usage() emits usage/help information.
1827 SC/01 Added the -bylocaldomain option to eximstats.
1829 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1831 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1833 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1839 PH/01 Another patch from the Sieve maintainer.
1841 PH/02 When an IPv6 address is converted to a string for single-key lookup
1842 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1843 dots are used instead of colons so that keys in lsearch files need not
1844 contain colons. This was done some time before quoting was made available
1845 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1846 (notated using the quote facility) so as to distinguish them from IPv4
1847 keys. This meant that lookups for IP addresses in host lists did not work
1848 for iplsearch lookups.
1850 This has been fixed by arranging for IPv6 addresses to be expressed with
1851 colons if the lookup type is iplsearch. This is not incompatible, because
1852 previously such lookups could never work.
1854 The situation is now rather anomolous, since one *can* have colons in
1855 ordinary lsearch keys. However, making the change in all cases is
1856 incompatible and would probably break a number of configurations.
1858 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1861 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1862 conversion specification without a maximum field width, thereby enabling
1863 a rogue spamd server to cause a buffer overflow. While nobody in their
1864 right mind would setup Exim to query an untrusted spamd server, an
1865 attacker that gains access to a server running spamd could potentially
1866 exploit this vulnerability to run arbitrary code as the Exim user.
1868 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1869 $primary_hostname instead of what libspf2 thinks the hosts name is.
1871 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1872 a directory entry by the name of the lookup key. Previously, if a
1873 symlink pointed to a non-existing file or a file in a directory that
1874 Exim lacked permissions to read, a lookup for a key matching that
1875 symlink would fail. Now it is enough that a matching directory entry
1876 exists, symlink or not. (Bugzilla 503.)
1878 PH/03 The body_linecount and body_zerocount variables are now exported in the
1881 PH/04 Added the $dnslist_matched variable.
1883 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1884 This means they are set thereafter only if the connection becomes
1887 PH/06 Added the client_condition to authenticators so that some can be skipped
1888 by clients under certain conditions.
1890 PH/07 The error message for a badly-placed control=no_multiline_responses left
1891 "_responses" off the end of the name.
1893 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1895 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1896 (without spaces) instead of just copying the configuration text.
1898 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1900 PH/11 Added $max_received_linelength.
1902 PH/12 Added +ignore_defer and +include_defer to host lists.
1904 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1905 way in which PCRE > 7.0 is built.
1907 PH/14 Implemented queue_only_load_latch.
1909 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1910 MAIL command. The effect was to mangle the value on 64-bit systems.
1912 PH/16 Another patch from the Sieve maintainer.
1914 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1916 PH/18 If a system quota error occurred while trying to create the file for
1917 a maildir delivery, the message "Mailbox is full" was not appended to the
1918 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1919 only to a quota excession during the actual writing of the file.
1921 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1922 characters?) which causes problems in log lines. The DN values are now
1923 passed through string_printing() before being added to log lines.
1925 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1926 and InterBase are left for another time.)
1928 PH/21 Added message_body_newlines option.
1930 PH/22 Guard against possible overflow in moan_check_errorcopy().
1932 PH/23 POSIX allows open() to be a macro; guard against that.
1934 PH/24 If the recipient of an error message contained an @ in the local part
1935 (suitably quoted, of course), incorrect values were put in $domain and
1936 $local_part during the evaluation of errors_copy.
1942 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1943 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1946 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1947 issue a MAIL command.
1949 PH/02 In an ACL statement such as
1951 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1953 if a client was not listed at all, or was listed with a value other than
1954 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1955 the condition was not true (as it should be), so access was not denied.
1956 The bug was that the ! inversion was incorrectly passed on to the second
1957 item. This has been fixed.
1959 PH/03 Added additional dnslists conditions == and =& which are different from
1960 = and & when the dns lookup returns more than one IP address.
1962 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1963 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1965 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1966 FSYNC, which compiles an option called disable_fsync that allows for
1967 bypassing fsync(). The documentation is heavily laced with warnings.
1969 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1971 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1972 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1973 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1974 including adding "make clean"; (3) Added -fPIC when compiling the test
1975 dynamically loaded module, to get rid of a warning.
1977 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1978 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1979 = 0s. The bug is otherwise harmless.
1981 PH/07 There was a bug in the dovecot authenticator such that the value of
1982 $auth1 could be overwritten, and so not correctly preserved, after a
1983 successful authentication. This usually meant that the value preserved by
1984 the server_setid option was incorrect.
1986 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1988 PH/09 Installed PCRE release 7.0.
1990 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1991 run for batched SMTP input. It is now run at the start of every message
1992 in the batch. While fixing this I discovered that the process information
1993 (output by running exiwhat) was not always getting set for -bs and -bS
1994 input. This is fixed, and it now also says "batched" for BSMTP.
1996 PH/11 Added control=no_pipelining.
1998 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1999 patch, slightly modified), and move the expansion of helo_data till after
2000 the connection is made in the smtp transport (so it can use these
2003 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2005 PH/14 Added log_selector = +pid.
2007 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2009 PH/16 Add ${if forany and ${if forall.
2011 PH/17 Added dsn_from option to vary the From: line in DSNs.
2013 PH/18 Flush SMTP output before performing a callout, unless control =
2014 no_callout_flush is set.
2016 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2017 was true (the default) a successful delivery failed to delete the retry
2018 item, thus causing premature timeout of the address. The bug is now
2021 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2023 PH/21 Long custom messages for fakedefer and fakereject are now split up
2024 into multiline reponses in the same way that messages for "deny" and
2025 other ACL rejections are.
2027 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2028 with slight modification.
2030 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2031 draft, changing the syntax and factoring some duplicate code".
2033 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2034 for deliveries of the second and subsequent messages over the same SMTP
2037 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2038 ${reduce, with only minor "tidies".
2040 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2042 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2043 expansion side effects.
2045 PH/27 When a message times out after an over-quota error from an Exim-imposed
2046 quota, the bounce message says "mailbox is full". This message was not
2047 being given when it was a system quota that was exceeded. It now should
2050 MH/03 Made $recipients available in local_scan(). local_scan() already has
2051 better access to the recipient list through recipients_list[], but
2052 $recipients can be useful in postmaster-provided expansion strings.
2054 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2055 in the case of a MAIL command with additional options following the
2056 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2057 were accidentally chopped off.
2059 PH/29 SMTP synchronization checks are implemented when a command is read -
2060 there is a check that no more input is waiting when there shouldn't be
2061 any. However, for some commands, a delay in an ACL can mean that it is
2062 some time before the response is written. In this time, more input might
2063 arrive, invalidly. So now there are extra checks after an ACL has run for
2064 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2065 pipelining has not been advertised.
2067 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2069 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2070 returned in the SMTP response when smtp_return_error_details was false.
2071 This has been fixed.
2073 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2074 instead of the C I/O that was originally supplied, because problems were
2075 reported on Solaris.
2077 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2078 Exim which did not show up earlier: it was assuming that a call to
2079 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2080 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2081 was a macro that became an assignment, so it seemed to work. This has
2082 changed to a proper function call with a void return, hence the compile
2083 error. Exim's code has been fixed.
2085 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2088 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2090 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2092 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2093 because it turns out to be too big or there is a timeout) there is no
2094 "Completed" line in the log. When some messages of this type were
2095 selected by exigrep, they were listed as "not completed". Others were
2096 picked up by some special patterns. I have improved the selection
2097 criteria to be more general.
2099 PH/38 The host_find_failed option in the manualroute router can now be set
2100 to "ignore", to completely ignore a host whose IP address cannot be
2101 found. If all hosts are ignored, the behaviour is controlled by the new
2102 host_all_ignored option.
2104 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2105 homing or because of multiple MX records with /mx) generated more than
2106 one IP address, and the following item turned out to be the local host,
2107 all the secondary addresses of the first item were incorrectly removed
2108 from the list, along with the local host and any following hosts (which
2109 is what is supposed to happen).
2111 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2112 whoever called Exim into the -H file. In the case of the daemon it was
2113 behaving confusingly. When first started, it used values for whoever
2114 started the daemon, but after a SIGHUP it used the Exim user (because it
2115 calls itself on a restart). I have changed the code so that it now always
2118 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2119 message are rejected with the same error (e.g. no authentication or bad
2120 sender address), and a DATA command is nevertheless sent (as can happen
2121 with PIPELINING or a stupid MUA), the error message that was given to the
2122 RCPT commands is included in the rejection of the DATA command. This is
2123 intended to be helpful for MUAs that show only the final error to their
2126 PH/42 Another patch from the Sieve maintainer.
2128 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2129 Eximstats - Fixed some broken HTML links and added missing column headers
2131 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2132 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2134 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2136 PH/43 Yet another patch from the Sieve maintainer.
2138 PH/44 I found a way to check for a TCP/IP connection going away before sending
2139 the response to the final '.' that terminates a message, but only in the
2140 case where the client has not sent further data following the '.'
2141 (unfortunately, this is allowed). However, in many cases there won't be
2142 any further data because there won't be any more messages to send. A call
2143 to select() can be used: if it shows that the input is "ready", there is
2144 either input waiting, or the socket has been closed. An attempt to read
2145 the next input character can distinguish the two cases. Previously, Exim
2146 would have sent an OK response which the client would never have see.
2147 This could lead to message repetition. This fix should cure that, at
2148 least in a lot of common cases.
2150 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2151 advertised in response to EHLO.
2157 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2158 fixed by 4.65/MH/01 (is this a record?) are fixed:
2160 (i) An empty string was always treated as zero by the numeric comparison
2161 operators. This behaviour has been restored.
2163 (ii) It is documented that the numeric comparison operators always treat
2164 their arguments as decimal numbers. This was broken in that numbers
2165 starting with 0 were being interpreted as octal.
2167 While fixing these problems I realized that there was another issue that
2168 hadn't been noticed. Values of message_size_limit (both the global option
2169 and the transport option) were treated as octal if they started with 0.
2170 The documentation was vague. These values are now always treated as
2171 decimal, and I will make that clear in the documentation.
2177 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2178 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2181 MH/01 Don't check that the operands of numeric comparison operators are
2182 integers when their expansion is in "skipping" mode (fixes bug
2183 introduced by 4.64-PH/07).
2185 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2186 child addresses, Exim now panics and dies. Previously, because the count
2187 is held in a short int, deliveries were likely to be lost. As such a
2188 large number of recipients for a single message is ridiculous
2189 (performance will be very, very poor), I have chosen to impose a limit
2190 rather than extend the field.
2196 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2197 leftover -K file (the existence of which was triggered by #402).
2198 While we were at it, introduced process PID as part of the -K
2199 filename. This should rule out race conditions when creating
2202 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2203 processing considerably. Previous code took too long for large mails,
2204 triggering a timeout which in turn triggers #401.
2206 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2207 in the DK code in transports.c. sendfile() is not really portable,
2208 hence the _LINUX specificness.
2210 TF/01 In the add_headers option to the mail command in an Exim filter,
2211 there was a bug that Exim would claim a syntax error in any
2212 header after the first one which had an odd number of characters
2215 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2216 callout verification, Exim cached a "reject" for the entire domain. This
2217 is correct for most verifications, but it is not correct for a recipient
2218 verification with use_sender or use_postmaster set, because in that case
2219 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2220 case of MAIL FROM:<> rejection from other early rejections (e.g.
2221 rejection of HELO). When verifying a recipient using a non-null MAIL
2222 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2223 Whatever the result of the callout, the value of the domain cache is
2224 left unchanged (for any other kind of callout, getting as far as trying
2225 RCPT means that the domain itself is ok).
2227 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2230 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2231 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2232 session, the daemon ignores ECONNECT errors and logs others; it now
2233 ignores EPIPE as well.
2235 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2236 (quoted-printable decoding).
2238 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2239 later the small subsequent patch to fix an introduced bug.
2241 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2243 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2245 PH/08 An error is now given if message_size_limit is specified negative.
2247 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2248 to be given (somewhat) arbitrary names.
2250 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2253 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2254 miscellaneous code fixes
2256 PH/10 Added the log_reject_target ACL modifier to specify where to log
2259 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2260 hostname. This is wrong, because it relates to the incoming message (and
2261 probably the interface on which it is arriving) and not to the outgoing
2262 callout (which could be using a different interface). This has been
2263 changed to use the value of the helo_data option from the smtp transport
2264 instead - this is what is used when a message is actually being sent. If
2265 there is no remote transport (possible with a router that sets up host
2266 addresses), $smtp_active_hostname is used.
2268 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2269 tweaks were necessary in order to get it to work (see also 21 below):
2270 (a) The code assumed that strncpy() returns a negative number on buffer
2271 overflow, which isn't the case. Replaced with Exim's string_format()
2273 (b) There were several signed/unsigned issues. I just did the minimum
2274 hacking in of casts. There is scope for a larger refactoring.
2275 (c) The code used strcasecmp() which is not a standard C function.
2276 Replaced with Exim's strcmpic() function.
2277 (d) The code set only $1; it now sets $auth1 as well.
2278 (e) A simple test gave the error "authentication client didn't specify
2279 service in request". It would seem that Dovecot has changed its
2280 interface. Fortunately there's a specification; I followed it and
2281 changed what the client sends and it appears to be working now.
2283 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2286 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2287 address A is aliased to B and C, where B exists and C does not. Without
2288 -v the output is "A verified" because verification stops after a
2289 successful redirection if more than one address is generated. However,
2290 with -v the child addresses are also verified. Exim was outputting "A
2291 failed to verify" and then showing the successful verification for C,
2292 with its parentage. It now outputs "B failed to verify", showing B's
2293 parentage before showing the successful verification of C.
2295 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2296 look up a TXT record in a specific list after matching in a combined
2299 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2300 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2301 they consult the DNS. I had assumed they would set it the way they
2302 wanted; and indeed my experiments on Linux seem to show that in some
2303 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2304 To be on the safe side, however, I have now made the interface to
2305 host_find_byname() similar to host_find_bydns(), with an argument
2306 containing the DNS resolver options. The host_find_byname() function now
2307 sets these options at its start, just as host_find_bydns() does. The smtp
2308 transport options dns_qualify_single and dns_search_parents are passed to
2309 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2310 of host_find_byname() use the default settings of RES_DEFNAMES
2311 (qualify_single) but not RES_DNSRCH (search_parents).
2313 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2314 spool_read_header() do less string testing, by means of a preliminary
2315 switch on the second character of optional "-foo" lines. (This is
2316 overdue, caused by the large number of possibilities that now exist.
2317 Originally there were few.) While I was there, I also converted the
2318 str(n)cmp tests so they don't re-test the leading "-" and the first
2319 character, in the hope this might squeeze out yet more improvement.
2321 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2322 flag allowing group syntax was set by the header_syntax check but not
2323 turned off, possible causing trouble later; (2) The flag was not being
2324 set at all for the header_verify test, causing "group"-style headers to
2325 be rejected. I have now set it in this case, and also caused header_
2326 verify to ignore an empty address taken from a group. While doing this, I
2327 came across some other cases where the code for allowing group syntax
2328 while scanning a header line wasn't quite right (mostly, not resetting
2329 the flag correctly in the right place). These bugs could have caused
2330 trouble for malformed header lines. I hope it is now all correct.
2332 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2333 with the "reply" argument non-NULL. The code, however (which originally
2334 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2335 but it didn't always do it. This confused somebody who was copying the
2336 code for some other use. I have removed all the tests.
2338 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2339 feature that was used to support insecure browsers during the U.S. crypto
2340 embargo. It requires special client support, and Exim is probably the
2341 only MTA that supported it -- and would never use it because real RSA is
2342 always available. This code has been removed, because it had the bad
2343 effect of slowing Exim down by computing (never used) parameters for the
2344 RSA_EXPORT functionality.
2346 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2347 authenticator to fail if there's a tab character in the incoming data
2348 (there should never be unless someone is messing about, as it's supposed
2349 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2350 if the connection is using TLS or if the remote IP is the same as the
2351 local IP, and the "valid-client-cert option" if a client certificate has
2354 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2355 authenticators. This can be used for authorization after authentication
2356 succeeds. (In the case of plaintext, it servers for both authentication
2359 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2360 if any retry times were supplied.
2362 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2363 connection, where there is no client IP address to check. In this
2364 situation, the verify now always succeeds.
2366 PH/25 Applied John Jetmore's -Mset patch.
2368 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2370 PH/27 In a string expansion for a processed (not raw) header when multiple
2371 headers of the same name were present, leading whitespace was being
2372 removed from all of them, but trailing whitespace was being removed only
2373 from the last one. Now trailing whitespace is removed from each header
2374 before concatenation. Completely empty headers in a concatenation (as
2375 before) are ignored.
2377 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2378 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2380 PH/29 [Removed. This was a change that I later backed out, and forgot to
2381 correct the ChangeLog entry (that I had efficiently created) before
2382 committing the later change.]
2384 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2385 address errors (4xx response to RCPT) over the same connection as other
2386 messages routed to the same hosts. Such deliveries are always "forced",
2387 so retry times are not inspected. This resulted in far too many retries
2388 for the affected addresses. The effect occurred only when there were more
2389 hosts than the hosts_max_try setting in the smtp transport when it had
2390 the 4xx errors. Those hosts that it had tried were not added to the list
2391 of hosts for which the message was waiting, so if all were tried, there
2392 was no problem. Two fixes have been applied:
2394 (i) If there are any address or message errors in an SMTP delivery, none
2395 of the hosts (tried or untried) are now added to the list of hosts
2396 for which the message is waiting, so the message should not be a
2397 candidate for sending over the same connection that was used for a
2398 successful delivery of some other message. This seems entirely
2399 reasonable: after all the message is NOT "waiting for some host".
2400 This is so "obvious" that I'm not sure why it wasn't done
2401 previously. Hope I haven't missed anything, but it can't do any
2402 harm, as the worst effect is to miss an optimization.
2404 (ii) If, despite (i), such a delivery is accidentally attempted, the
2405 routing retry time is respected, so at least it doesn't keep
2406 hammering the server.
2408 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2409 in ${readsocket because some servers need this prod.
2411 PH/32 Added some extra debug output when updating a wait-xxx database.
2413 PH/33 The hint "could be header name not terminated by colon", which has been
2414 given for certain expansion errors for a long time, was not being given
2415 for the ${if def:h_colon_omitted{... case.
2417 PH/34 The spec says: "With one important exception, whenever a domain list is
2418 being scanned, $domain contains the subject domain." There was at least
2419 one case where this was not true.
2421 PH/35 The error "getsockname() failed: connection reset by peer" was being
2422 written to the panic log as well as the main log, but it isn't really
2423 panic-worthy as it just means the connection died rather early on. I have
2424 removed the panic log writing for the ECONNRESET error when getsockname()
2427 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2428 runs only) independently of the message's sender address. This meant
2429 that, if the 4xx error was in fact related to the sender, a different
2430 message to the same recipient with a different sender could confuse
2431 things. In particualar, this can happen when sending to a greylisting
2432 server, but other circumstances could also provoke similar problems.
2433 I have changed the default so that the retry time for these errors is now
2434 based a combination of the sender and recipient addresses. This change
2435 can be overridden by setting address_retry_include_sender=false in the
2438 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2439 remote server are returned as part of bounce messages. This was not
2440 happening for LMTP over a pipe (the lmtp transport), but now it is the
2441 same for both kinds of LMTP.
2443 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2444 in header lines that were in fact CNAMEs. This is no longer the case.
2446 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2447 and queue runs started by the daemon processed all messages. This has
2448 been fixed so that -R and -S can now usefully be given with -q<time>.
2450 PH/40 Import PCRE release 6.7 (fixes some bugs).
2452 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2454 PH/42 Give an error if -q is specified more than once.
2456 PH/43 Renamed the variables $interface_address and $interface_port as
2457 $received_ip_address and $received_port, to make it clear that these
2458 values apply to message reception, and not to the outgoing interface when
2459 a message is delivered. (The old names remain recognized, of course.)
2461 PH/44 There was no timeout on the connect() call when using a Unix domain
2462 socket in the ${readsocket expansion. There now is.
2464 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2465 be meaningful with "accept".
2467 SC/01 Eximstats V1.43
2468 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2470 SC/02 Eximstats V1.44
2471 Use a glob alias rather than an array ref in the generated
2472 parser. This improves both readability and performance.
2474 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2475 Collect SpamAssassin and rejection statistics.
2476 Don't display local sender or destination tables unless
2477 there is data to show.
2478 Added average volumes into the top table text output.
2480 SC/04 Eximstats V1.46
2481 Collect data on the number of addresses (recipients)
2482 as well as the number of messages.
2484 SC/05 Eximstats V1.47
2485 Added 'Message too big' to the list of mail rejection
2486 reasons (thanks to Marco Gaiarin).
2488 SC/06 Eximstats V1.48
2489 Mainlog lines which have GMT offsets and are too short to
2490 have a flag are now skipped.
2492 SC/07 Eximstats V1.49 (Alain Williams)
2493 Added the -emptyok flag.
2495 SC/08 Eximstats V1.50
2496 Fixes for obtaining the IP address from reject messages.
2498 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2499 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2500 whitesspace changes from 4.64-PH/27
2502 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2505 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2506 are found, allow negative numbers in numeric criteria)
2508 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2510 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2511 to match changes made in 4.64-PH/43
2513 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2515 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2516 "domain" is actually an IP address, and force a failure. This locks out
2517 those revolvers/nameservers that support "A-for-A" lookups, in
2518 contravention of the specifications.
2520 PH/48 When a host name was looked up from an IP address, and the subsequent
2521 forward lookup of the name timed out, the host name was left in
2522 $sender_host_name, contrary to the specification.
2524 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2525 restricted to single-key lookups, Exim was not diagnosing an error if
2526 * or *@ was used with a query-style lookup.
2528 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2530 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2531 long ago, but noone interested enough thought of it. Let's just say that
2532 the "1.1" means that there are some new functions that weren't there at
2533 some point in the past.
2535 PH/51 Error processing for expansion failure of helo_data from an smtp
2536 transport during callout processing was broken.
2538 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2539 tested/used via the -bh/-bhc/-bs options.
2541 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2542 bug, fixed in subsequent PCRE releases).
2544 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2545 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2547 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2553 SC/01 Use a glob alias rather than an array ref in eximstats generated
2554 parser. This improves both readability and performance.
2556 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2557 Don't display local sender or destination tables in eximstats unless
2558 there is data to show.
2559 Added average volumes into the eximstats top table text output.
2561 SC/03 Collect data on the number of addresses (recipients) as well
2562 as the number of messages in eximstats.
2564 TF/01 Correct an error in the documentation for the redirect router. Exim
2565 does (usually) call initgroups() when daemonizing.
2567 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2568 with consistent privilege compared to when running as a daemon.
2570 TF/03 Note in the spec that $authenticated_id is not set for local
2571 submissions from trusted users.
2573 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2574 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2576 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2577 by adding some example configuration directives to the default
2578 configuration file. A little bit of work is required to uncomment the
2579 directives and define how usernames and passwords are checked, but
2580 there is now a framework to start from.
2582 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2583 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2584 without this. I don't know how relevant this is to other LDAP libraries.
2586 PH/02 Add the verb name to the "unknown ACL verb" error.
2588 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2590 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2592 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2593 directory not expanded when it should be if an expanded home directory
2594 was set for the address (which is overridden by the transport).
2596 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2599 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2600 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2601 because it is too late at that time, and has no effect.
2603 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2604 security issue with \' (bugzilla #107). I could not use the
2605 PQescapeStringConn() function, because it needs a PGconn value as one of
2608 PH/08 When testing addresses using -bt, indicate those final addresses that
2609 are duplicates that would not cause an additional delivery. At least one
2610 person was confused, thinking that -bt output corresponded to deliveries.
2611 (Suppressing duplicates isn't a good idea as you lose the information
2612 about possibly different redirections that led to the duplicates.)
2614 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2615 systems where poll() doesn't work, in particular OS X.
2617 PH/10 Added more information to debugging output for retry time not reached.
2619 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2620 operations in malware.c.
2622 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2625 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2626 syslog_duplication set false, log lines that would normally be written
2627 both the the main log and to the reject log were not written to syslog at
2630 PH/14 In the default configuration, change the use of "message" in ACL warn
2631 statements to "add_header".
2633 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2634 not followed by a command (e.g. "seen endif").
2636 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2637 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2640 PH/17 Added extra conditions to the default value of delay_warning_condition
2644 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2645 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2646 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2649 The Auto-Submitted: and various List- headers are standardised, whereas I
2650 don't think Precedence: ever was.
2652 PH/18 Refactored debugging code in route_finduser() to show more information,
2653 in particular, the error code if getpwnam() issues one.
2655 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2656 This is apparently needed in addition to the PH/07 change above to avoid
2657 any possible encoding problems.
2659 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2660 but not after initializing Perl.
2662 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2663 output them only if debugging. By default they are written stderr,
2664 apparently, which is not desirable.
2666 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2669 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2672 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2674 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2675 authenticated or an ident call has been made. Suppress the default
2676 values for $authenticated_id and $authenticated_sender (but permit -oMai
2677 and -oMas) when testing with -bh.
2679 PH/24 Re-jigged the order of the tests in the default configuration so that the
2680 tests for valid domains and recipients precede the DNS black list and CSA
2681 tests, on the grounds that those ones are more expensive.
2683 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2684 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2685 This bug exists in every version of Exim that I still have, right back to
2688 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2689 However, an attempt to turn on case-sensitivity in a regex key by
2690 including (?-i) didn't work because the subject string was already
2691 lowercased, and the effects were non-intuitive. It turns out that a
2692 one-line patch can be used to allow (?-i) to work as expected.
2698 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2699 other effects) broke the use of negated acl sub-conditions.
2701 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2704 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2705 "Deny" causes Exim to reject the incoming connection with a 554 error.
2706 Unfortunately, if there is a major crisis, such as a disk failure,
2707 tcp-wrappers gives "deny", whereas what one would like would be some
2708 kind of temporary error. A kludge has been added to help with this.
2709 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2710 554 error is used if errno is still zero or contains ENOENT (which occurs
2711 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2714 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2716 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2717 errors. Otherwise a message that provokes a temporary error (when other
2718 messages do not) can cause a whole host to time out.
2720 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2721 addresses were routed directly to files or pipes from a redirect router.
2722 File deliveries just didn't batch; pipe deliveries might have suffered
2725 PH/06 A failure to get a lock for a hints database would erroneously always say
2726 "Failed to get write lock", even when it was really a read lock.
2728 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2729 of 0600. This has been changed to use the value of the lockfile_mode
2730 option (which defaults to 0600).
2732 PH/08 Applied small patch from the Sieve maintainer.
2734 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2735 folder from quota calculations, a direct delivery into this folder messed
2736 up the contents of the maildirsize file. This was because the regex was
2737 used only to exclude .Trash (or whatever) when the size of the mailbox
2738 was calculated. There was no check that a delivery was happening into an
2739 excluded directory. This bug has been fixed by ignoring all quota
2740 processing for deliveries into excluded directories.
2742 PH/10 Added the maildirfolder_create_regex option to appendfile.
2748 PH/01 The code for finding all the local interface addresses on a FreeBSD
2749 system running IPv6 was broken. This may well have applied to all BSD
2750 systems, as well as to others that have similar system calls. The broken
2751 code found IPv4 interfaces correctly, but gave incorrect values for the
2752 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2753 that it would not match correctly against @[] and not recognize the IPv6
2756 PH/02 The ipliteral router was not recognizing addresses of the form user@
2757 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2759 PH/03 Added disable_ipv6.
2761 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2762 decoded form, because it is most often used to construct To: headers
2763 lines in autoreplies, and the decoded form may well be syntactically
2764 invalid. However, $reply_address has leading white space removed, and all
2765 newlines turned into spaces so that the autoreply transport does not
2768 PH/05 If group was specified without a user on a router, and no group or user
2769 was specified on a transport, the group from the router was ignored.
2771 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2772 for visible compile-time settings that can be used to change these
2773 numbers, for those that want even more. Backwards compatibility with old
2774 spool files has been maintained. However, going back to a previous Exim
2775 release will lost any variables that are in spool files.
2777 PH/07 Two small changes when running in the test harness: increase delay when
2778 passing a TCP/IP connection to a new process, in case the original
2779 process has to generate a bounce, and remove special handling of
2780 127.0.0.2 (sic), which is no longer necessary.
2782 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2783 be the same on different OS.
2785 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2788 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2789 whether --show-vars was specified or not
2791 JJ/02 exipick: Added support for new ACL variable spool format introduced
2794 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2795 syntactically invalid From: or Reply-to: line, and a filter used this to
2796 generate an autoreply, and therefore failed to obtain an address for the
2797 autoreply, Exim could try to deliver to a non-existent relative file
2798 name, causing unrelated and misleading errors. What now happens is that
2799 it logs this as a hard delivery error, but does not attempt to create a
2802 PH/11 The exinext utility has a -C option for testing purposes, but although
2803 the given file was scanned by exinext itself; it wasn't being passed on
2804 when Exim was called.
2806 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2807 an end-of-file indication when reading a command response.
2809 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2810 compiled. In many other places in Exim, IPv6 addresses are always
2811 recognized, so I have changed this. It also means that IPv4 domain
2812 literals of the form [IPV4:n.n.n.n] are now always recognized.
2814 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2815 used if the router is not running as root, for example, when verifying at
2816 ACL time, or when using -bh. The debugging output from this situation was
2817 non-existent - all you got was a failure to exec. I have made two
2820 (a) Failures to set uid/gid, the current directory, or a process leader
2821 in a subprocess such as that created by queryprogram now generate
2822 suitable debugging ouput when -d is set.
2824 (b) The queryprogram router detects when it is not running as root,
2825 outputs suitable debugging information if -d is set, and then runs
2826 the subprocess without attempting to change uid/gid.
2828 PH/15 Minor change to Makefile for building test_host (undocumented testing
2831 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2832 additional section of a DNS packet that returns MX or SRV records.
2833 Instead, it always explicitly searches for A/AAAA records. This avoids
2834 major problems that occur when a DNS server includes only records of one
2835 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2836 fixed another bug: if SRV records were looked up and the corresponding
2837 address records were *not* found in the additional section, the port
2838 values from the SRV records were lost.
2840 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2841 using the correct key (the original address) when searching the retry
2842 rules in order to find which one to use for generating the retry hint.
2844 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2845 adding the default one. Similarly, if it contains a Reply-To: header, the
2846 errors_reply_to option, if set, is not used.
2848 PH/19 When calculating a retry time, Exim used to measure the "time since
2849 failure" by looking at the "first failed" field in the retry record. Now
2850 it does not use this if it is later than than the arrival time of the
2851 message. Instead it uses the arrival time. This makes for better
2852 behaviour in cases where some deliveries succeed, thus re-setting the
2853 "first failed" field. An example is a quota failure for a huge message
2854 when small messages continue to be delivered. Without this change, the
2855 "time since failure" will always be short, possible causing more frequent
2856 delivery attempts for the huge message than are intended.
2857 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2859 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2860 $1, $2, $3) because the numerical variables can be reset during some
2861 expansion items (e.g. "match"), thereby losing the authentication data.
2863 PH/21 Make -bV show the size of off_t variables so that the test suite can
2864 decide whether to run tests for quotas > 2G.
2866 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2867 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2868 filecount value is greater than 2G or if a quota value is greater than 2G
2869 on a system where the size of off_t is not greater than 4, a panic error
2872 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2873 never match. The debug and -bh output now contains an explicit error
2874 message indicating a malformed IPv4 address or mask.
2876 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2877 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2878 PH/23 above applies.
2880 PH/25 Do not write to syslog when running in the test harness. The only
2881 occasion when this arises is a failure to open the main or panic logs
2882 (for which there is an explicit test).
2884 PH/26 Added the /no_tell option to "control=freeze".
2886 PH/27 If a host name lookup failed very early in a connection, for example, if
2887 the IP address matched host_lookup and the reverse lookup yielded a name
2888 that did not have a forward lookup, an error message of the form "no IP
2889 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2890 could be logged. Now it outputs the IP address instead of "NULL".
2892 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2893 allows the sender and the authenticated sender to be set when
2894 submitting a message from within Exim. Since child_open_exim() is
2895 documented for local_scan(), the new function should be too.
2897 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2898 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2899 results in an empty string is now treated as unset.
2901 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2903 PH/31 Added sender_verify_fail logging option.
2905 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2906 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2907 tidied the source and removed it altogether.
2909 PH/33 When a queue run was abandoned because the load average was too high, a
2910 log line was always written; now it is written only if the queue_run log
2911 selector is set. In addition, the log line for abandonment now contains
2912 information about the queue run such as the pid. This is always present
2913 in "start" and "stop" lines but was omitted from the "abandon" line.
2915 PH/34 Omit spaces between a header name and the colon in the error message that
2916 is given when verify = headers_syntax fails (if there are lots of them,
2917 the message gets confusing).
2919 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2920 names, as there are now some PTR records that contain slashes. This check
2921 is only to protect against broken name servers that fall over on strange
2922 characters, so the fact that it applies to all lookups doesn't matter.
2924 PH/36 Now that the new test suite is complete, we can remove some of the
2925 special code in Exim that was needed for the old test suite. For example,
2926 sorting DNS records because real resolvers return them in an arbitrary
2927 order. The new test suite's fake resolver always returns records in the
2930 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2931 bounces) except when queue_only is set, to avoid logging races between
2932 the different processes.
2934 PH/38 Panic-die if .include specifies a non-absolute path.
2936 PH/39 A tweak to the "H" retry rule from its user.
2938 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2939 a label. They prevented compilation on older perls.
2941 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2942 a warning to be raised on newish perls.
2944 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2945 on queue. Changes to match documented behaviour of showing count of
2946 messages matching specified criteria.
2948 PH/40 Changed the default ident timeout from 30s to 5s.
2950 PH/41 Added support for the use of login_cap features, on those BSD systems
2951 that have them, for controlling the resources used by pipe deliveries.
2953 PH/42 The content-scanning code uses fopen() to create files in which to put
2954 message data. Previously it was not paying any attention to the mode of
2955 the files. Exim runs with umask(0) because the rest of the code creates
2956 files with open(), and sets the required mode explicitly. Thus, these
2957 files were ending up world-writeable. This was not a big issue, because,
2958 being within the spool directory, they were not world-accessible. I have
2959 created a function called modefopen, which takes an additional mode
2960 argument. It sets umask(777), creates the file, chmods it to the required
2961 mode, then resets the umask. All the relevant calls to fopen() in the
2962 content scanning code have been changed to use this function.
2964 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2965 to 24 hours. This avoids potential overflow problems when processing G
2966 and H retry rules. I suspect nobody ever tinkers with this value.
2968 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2970 PH/45 When the plaintext authenticator is running as a client, the server's
2971 challenges are checked to ensure they are valid base64 strings. By
2972 default, the authentication attempt is cancelled if an invalid string is
2973 received. Setting client_ignore_invalid_base64 true ignores these errors.
2974 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2975 they are received. Thus, the responses can be made to depend on the
2976 challenges. If an invalid string is ignored, an empty string is placed in
2979 PH/46 Messages that are created by the autoreply transport now contains a
2980 References: header, in accordance with RFCs 2822 and 3834.
2982 PH/47 Added authenticated_sender_force to the smtp transport.
2984 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2986 PH/49 Installed latest patch from the Sieve maintainer.
2988 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2989 was also set, the appendfile transport was unnecessarily scanning a
2990 directory of message files (e.g. for maildir delivery) to find the count
2991 of files (along with the size), even though it did not need this
2992 information. It now does the scan only if it needs to find either the
2993 size of the count of files.
2995 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2997 PH/52 Two bugs concerned with error handling when the smtp transport is
3000 (i) Exim was not creating retry information for temporary errors given
3001 for individual recipients after the DATA command when the smtp transport
3002 was used in LMTP mode. This meant that they could be retried too
3003 frequently, and not timed out correctly.
3005 (ii) Exim was setting the flag that allows error details to be returned
3006 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3007 recipients that were returned after the DATA command.
3009 PH/53 This is related to PH/52, but is more general: for any failing address,
3010 when detailed error information was permitted to be returned to the
3011 sender, but the error was temporary, then after the final timeout, only
3012 "retry timeout exceeded" was returned. Now it returns the full error as
3013 well as "retry timeout exceeded".
3015 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3016 do this, and (what is worse) MTAs that accept it.
3018 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3019 will now be deprecated.
3021 PH/56 New os.c-cygwin from the Cygwin maintainer.
3023 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3024 formats (previously only available in exim formats via -bpr, -bpru,
3025 and -bpra. Now also available in native and exiqgrep formats)
3027 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3028 with very large, slow to parse queues
3030 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3032 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3034 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3035 responses to authentication challenges, though it was showing the
3036 challenges; (ii) I've removed the CR characters from the debug output for
3039 PH/58 Allow for the insertion of a newline as well as a space when a string
3040 is turned into more than one encoded-word during RFC 2047 encoding. The
3041 Sieve code now uses this.
3043 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3044 data_4xx, lost_connection, tls_required.
3046 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3047 message was being sent as an SMTP response.
3049 PH/61 Add -l and -k options to exicyclog.
3051 PH/62 When verifying, if an address was redirected to one new address, so that
3052 verification continued, and the new address failed or deferred after
3053 having set something in $address_data, the value of $address_data was not
3054 passed back to the ACL. This was different to the case when no
3055 redirection occurred. The value is now passed back in both cases.
3057 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3058 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3059 use login_cap.h, so on its own it isn't the distinguishing feature. The
3060 new name refers directly to the setclassresources() function.
3062 PH/65 Added configuration files for NetBSD3.
3064 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3066 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3067 is preferred over IPv4.
3069 PH/68 The bounce_return_message and bounce_return_body options were not being
3070 honoured for bounces generated during the reception of non-SMTP messages.
3071 In particular, this applied to messages rejected by the ACL. This bug has
3072 been fixed. However, if bounce_return_message is true and bounce_return_
3073 body is false, the headers that are returned for a non-SMTP message
3074 include only those that have been read before the error was detected.
3075 (In the case of an ACL rejection, they have all been read.)
3077 PH/69 The HTML version of the specification is now built in a directory called
3078 spec_html instead of spec.html, because the latter looks like a path with
3079 a MIME-type, and this confuses some software.
3081 PH/70 Catch two compiler warnings in sieve.c.
3083 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3084 function verify_get_ident() calls ip_connect() to connect a socket, but
3085 if the "connect()" function timed out, ip_connect() used to close the
3086 socket. However, verify_get_ident() also closes the socket later, and in
3087 between Exim writes to the log, which may get opened at this point. When
3088 the socket was closed in ip_connect(), the log could get the same file
3089 descriptor number as the socket. This naturally causes chaos. The fix is
3090 not to close the socket in ip_connect(); the socket should be closed by
3091 the function that creates it. There was only one place in the code where
3092 this was missing, in the iplookup router, which I don't think anybody now
3093 uses, but I've fixed it anyway.
3095 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3096 well as to direct DNS lookups. Otherwise the handling of names in host
3097 lists is inconsistent and therefore confusing.
3103 PH/01 Two changes to the default runtime configuration:
3105 (1) Move the checks for relay_from_hosts and authenticated clients from
3106 after to before the (commented out) DNS black list checks.
3108 (2) Add control=submission to the relay_from_hosts and authenticated
3109 clients checks, on the grounds that messages accepted by these
3110 statements are most likely to be submissions.
3112 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3114 (1) Generate an error if the third argument for the ${prvs expansion is
3117 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3120 (3) Reset the variables that are obtained from the first argument of
3121 ${prvscheck and used in the second argument before leaving the code,
3122 because their memory is reclaimed, so using them afterwards may do
3125 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3126 one (it's much easier than Tom thought :-).
3128 (5) Because of (4), we can now allow for the use of $prvscheck_result
3129 inside the third argument.
3131 PH/03 For some reason, the default setting of PATH when running a command from
3132 a pipe transport was just "/usr/bin". I have changed it to
3135 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3136 anything to be listed in the output from -bV.
3138 PH/05 When a filter generated an autoreply, the entire To: header line was
3139 quoted in the delivery log line, like this:
3141 => >A.N.Other <ano@some.domain> <original@ddress> ...
3143 This has been changed so that it extracts the operative address. There
3144 may be more than one such address. If so, they are comma-separated, like
3147 => >ano@some.domain,ona@other.domain <original@ddress> ...
3149 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3150 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3151 not being looked up in the rDNS to get a host name, Exim was showing the
3152 IP address twice in Received: lines, even though the IP addresses were
3153 identical. For example:
3155 Received: from [1.2.3.4] (helo=[1.2.3.4])
3157 However, if the real host name was known, it was omitting the HELO data
3158 if it matched the actual IP address. This has been tidied up so that it
3159 doesn't show the same IP address twice.
3161 PH/07 When both +timestamp and +memory debugging was on, the value given by
3162 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3163 called by the debug printing, thereby overwriting the timestamp buffer.
3164 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3166 PH/08 When the original message was included in an autoreply transport, it
3167 always said "this is a copy of the message, including all the headers",
3168 even if body_only or headers_only was set. It now gives an appropriate
3171 PH/09 Applied a patch from the Sieve maintainer which:
3173 o fixes some comments
3174 o adds the (disabled) notify extension core
3175 o adds some debug output for the result of if/elsif tests
3176 o points to the current vacation draft in the documentation
3177 and documents the missing references header update
3181 o fixes a bug in processing the envelope test (when testing
3182 multiple envelope elements, the last element determinted the
3185 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3186 Electronic Mail") by including:
3188 Auto-submitted: auto-generated
3190 in the messages that it generates (bounce messages and others, such as
3191 warnings). In the case of bounce messages for non-SMTP mesages, there was
3192 also a typo: it was using "Auto_submitted" (underscore instead of
3193 hyphen). Since every message generated by Exim is necessarily in response
3194 to another message, thes have all been changed to:
3196 Auto-Submitted: auto-replied
3198 in accordance with these statements in the RFC:
3200 The auto-replied keyword:
3202 - SHOULD be used on messages sent in direct response to another
3203 message by an automatic process,
3205 - MUST NOT be used on manually-generated messages,
3207 - MAY be used on Delivery Status Notifications (DSNs) and Message
3208 Disposition Notifications (MDNs),
3210 - MUST NOT be used on messages generated by automatic or periodic
3211 processes, except for messages which are automatic responses to
3214 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3215 to the default Received: header definition.
3217 PH/12 Added log selector acl_warn_skipped (default on).
3219 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3220 variables because (a) they are in the wrong storage pool and (b) even if
3221 they were copied, it wouldn't work properly because of the caching.
3223 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3224 checking when decoding. Apparently there are clients that generate
3225 overlong encoded strings. Why am I not surprised?
3227 PH/15 If the first argument of "${if match_address" was not empty, but did not
3228 contain an "@" character, Exim crashed. Now it writes a panic log message
3229 and treats the condition as false.
3231 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3233 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3234 extension "envelope-auth". The code is finished and in agreement with
3235 other implementations, but there is no documentation so far and in fact,
3236 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3237 not changing the active code.
3239 Print executed "if" and "elsif" statements when debugging is used. This
3240 helps a great deal to understand what a filter does.
3242 Document more things not specified clearly in RFC3028. I had all this
3243 sorted out, when out of a sudden new issues came to my mind. Oops."
3245 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3248 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3249 canonical form (as documented). However, after a host name lookup from
3250 the IP address, check_host() was doing a simple string comparison with
3251 addresses acquired from the DNS when checking that the found name did
3252 have the original IP as one of its addresses. Since any found IPv6
3253 addresses are likely to be in abbreviated form, the comparison could
3254 fail. Luckily, there already exists a function for doing the comparison
3255 by converting both addresses to binary, so now that is used instead of
3256 the text comparison.
3258 PH/20 There was another similar case to PH/19, when a complete host name was
3259 given in a host list; looking up its IP address could give an abbreviated
3260 form, whereas the current host's name might or might not be abbreviated.
3261 The same fix has been applied.
3267 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3268 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3271 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3272 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3274 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3276 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3277 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3278 Also, for "Auto-Submitted", treat anything other than "no" as
3279 non-personal, in accordance with RFC 3834. (Previously it treated
3280 anything starting "auto-" as non-personal.)
3282 TF/01 The control=submission/name=... option had a problem with syntax
3283 errors if the name included a slash character. The /name= option
3284 now slurps the rest of the string, so it can include any characters
3285 but it must come last in the list of options (after /sender_retain
3288 PH/05 Some modifications to the interface to the fake nameserver for the new
3296 TK/01 Added the "success_on_redirect" address verification option. See
3297 NewStuff for rationale and an example.
3299 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3301 PH/02 Patch to exigrep to allow it to work on syslog lines.
3303 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3304 fread() to skip over the body file's header line, because in Cygwin the
3305 header line is locked and is inaccessible.
3307 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3308 co-exist for some time) to make it clear that it is the Exim ID that is
3309 referenced, not the Message-ID: header line.
3311 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3312 string_format() function, because snprintf() does not exist on all
3315 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3316 ${stat: expansion item.
3318 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3319 protocol synchronization error", to keep the pedants happy.
3321 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3322 well as for IRIX systems, when gcc is being used. See the host.c source
3325 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3327 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3330 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3331 transport and to the smtp transport in LMTP mode.
3333 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3335 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3336 run a filter in a subprocess. This could lead to confusion in subsequent
3337 lookups in the parent process. There should also be a search_tidyup() at
3338 the end of the subprocess.
3340 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3341 only if the host matched helo_try_verify_hosts, which caused the
3342 verification to occur when the EHLO/HELO command was issued. The ACL just
3343 tested the remembered result. Now, if a previous verification attempt has
3344 not happened, "verify = helo" does it there and then.
3346 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3348 TK/03 Fix log output including CR from clamd.
3350 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3351 exist provoked a memory error which could cause a segfault.
3353 PH/15 Installed PCRE 6.2
3355 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3357 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3358 of the problem. Specifically, suggested +O2 rather than +O1 for the
3361 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3363 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3364 record, the port was not correctly logged when the outgoing_port log
3365 selector was set (it logged the transort's default port).
3367 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3368 fallback_hosts, and "hosts" in the smtp transport.
3370 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3371 host errors such as "Connection refused".
3373 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3374 authentication with radiusclient 0.4.9:
3376 - Error returned from rc_read_config was caught wrongly
3377 - Username/password not passed on to radius server due to wrong length.
3379 The presumption is that some radiusclient API changes for 4.51/PH/17
3380 were not taken care of correctly. The code is still untested by me (my
3381 Linux distribution still has 0.3.2 of radiusclient), but it was
3382 contributed by a Radius user.
3384 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3385 expanding the "port" option of the smtp transport.
3387 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3388 while reading a MIME header. Thanks to Tom Hughes for a patch.
3390 PH/24 Include config.h inside local_scan.h so that configuration settings are
3393 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3394 that in an ACL for RCPT (for example), you can examine exactly what was
3397 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3398 commands, but it was not correctly comparing the address with the actual
3399 client host address. Thus, it would show the EHLO address in Received:
3400 header lines when this was not necessary.
3402 PH/27 Added the % operator to ${eval:}.
3404 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3405 it should be ignoring failures (because with -C, for example, it has lost
3406 privilege). It wasn't ignoring creation failures other than "already
3409 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3412 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3413 because an input line was too long, either on its own, or by virtue of
3414 too many continuations, the temporary file was not being removed, and the
3415 return code was incorrect.
3417 PH/31 Missing "BOOL" in function definition in filtertest.c.
3419 PH/32 Applied Sieve patches from the maintainer.
3421 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3423 PH/33 Added "verify = not_blind".
3425 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3426 Local/Makefile (with some defaults set). These are used in built scripts
3427 such as exicyclog, but they have never been used in the exim_install
3428 script (though there are many overriding facilities there). I have
3429 arranged that the exim_install script now takes note of these two
3432 PH/35 Installed configuration files for Dragonfly.
3434 PH/36 When a locally submitted message by a trusted user did not contain a
3435 From: header, and the sender address was obtained from -f or from an SMTP
3436 MAIL command, and the trusted user did not use -F to supply a sender
3437 name, $originator_name was incorrectly used when constructing a From:
3438 header. Furthermore, $originator_name was used for submission mode
3439 messages from external hosts without From: headers in a similar way,
3440 which is clearly wrong.
3442 PH/37 Added control=suppress_local_fixups.
3444 PH/38 When log_selector = +received_sender was set, and the addition of the
3445 sender made the log line's construction buffer exactly full, or one byte
3446 less than full, an overflow happened when the terminating "\n" was
3449 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3450 when the result of a list match is failure because a DNS lookup failed.
3452 PH/40 RM_COMMAND is now used in the building process.
3454 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3455 the "build-* directories that it finds.
3457 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3458 address in a domain literal was a prefix of an interface address.
3460 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3461 when verifying a sender address, unless rewrite_headers is false.
3463 PH/44 Wrote a long comment about why errors_to addresses are verified as
3464 recipients, not senders.
3466 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3467 the ratelimit ACL was added.
3469 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3471 PH/46 Added extra information about PostgreSQL errors to the error string.
3473 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3474 avoiding the need to install special zones in a real server. This is
3475 backwards compatible; if it can't find the fake resolver, it drops back.
3476 Thus, both old and new test suites can be run.
3478 TF/02 Added util/ratelimit.pl
3480 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3481 clock is set back in time.
3483 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3484 Candler <B.Candler@pobox.com>.
3486 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3487 OK for addresses that are the result of redirections.
3489 PH/48 A number of further additions for the benefit of the new test suite,
3490 including a fake gethostbyname() that interfaces to the fake DNS resolver
3493 TF/06 The fix for widen_domains has also been applied to qualify_single and
3494 search_parents which are the other dnslookup options that can cause
3497 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3500 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3502 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3503 service" if it encountered an MX record with an empty target hostname.
3504 The message is now "an MX or SRV record indicated no SMTP service".
3506 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3507 if the DNS of the sending site is misconfigured. This is quite a
3508 common situation. This change restores the behaviour of treating a
3509 helo verification defer as a failure.
3511 PH/51 If self=fail was set on a router, the bounce message did not include the
3512 actual error message.
3518 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3520 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3521 command itself ended in error, the underlying message about the transport
3522 filter timeout was being overwritten with the pipe command error. Now the
3523 underlying error message should be appended to the second error message.
3525 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3527 PH/02 Reduce the amount of output that "make" produces by default. Full output
3528 can still be requested.
3530 PH/03 The warning log line about a condition test deferring for a "warn" verb
3531 was being output only once per connection, rather than after each
3532 occurrence (because it was using the same function as for successful
3533 "warn" verbs). This seems wrong, so I have changed it.
3535 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3536 it should not have, which might have caused a crash in the right
3537 circumstances, but probably never did.
3539 PH/04 Installed a modified version of Tony Finch's patch to make submission
3540 mode fix the return path as well as the Sender: header line, and to
3541 add a /name= option so that you can make the user's friendly name appear
3544 TF/03 Added the control = fakedefer ACL modifier.
3546 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3547 Mark Lowes for thorough testing.
3549 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3551 TK/03 Merged latest SRS patch from Miles Wilton.
3553 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3554 with the definition in sysexits.h (which is #included earlier).
3555 Fortunately, Exim does not actually use EX_OK. The code used to try to
3556 preserve the sysexits.h value, by assumimg that macro definitions were
3557 scanned for macro replacements. I have been disabused of this notion,
3558 so now the code just undefines EX_OK before #including unistd.h.
3560 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3561 in the smtp transport. When a block could not be written in a single
3562 write() function, the timeout was being re-applied to each part-write.
3563 This seems wrong - if the receiver was accepting one byte at a time it
3564 would take for ever. The timeout is now adjusted when this happens. It
3565 doesn't have to be particularly precise.
3567 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3568 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3570 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3571 without a domain if the check to <postmaster@domain> fails.
3573 SC/01 Eximstats: added -xls and the ability to specify output files
3574 (patch written by Frank Heydlauf).
3576 SC/02 Eximstats: use FileHandles for outputing results.
3578 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3580 SC/04 Eximstats: fixed display of large numbers with -nvr option
3582 SC/05 Eximstats: fixed merging of reports with empty tables.
3584 SC/06 Eximstats: added the -include_original_destination flag
3586 SC/07 Eximstats: removed tabs and trailing whitespace.
3588 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3590 TK/06 MBOX spool code: Add real "From " MBOX separator line
3591 so the .eml file is really in mbox format (even though
3592 most programs do not really care). Patch from Alex Miller.
3594 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3595 The latter is generated from $received_to and is only set if the
3596 message has one envelope recipient. SA can use these headers,
3597 obviously out-of-the-box. Patch from Alex Miller.
3599 PH/08 The ${def test on a variable was returning false if the variable's
3600 value was "0", contrary to what the specification has always said!
3601 The result should be true unless the variable is empty.
3603 PH/09 The syntax error of a character other than { following "${if
3604 def:variable_name" (after optional whitespace) was not being diagnosed.
3605 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3606 accidental colon was present, for example, could give incorrect results.
3608 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3609 result is used (not including appendfile, where other changes are about
3612 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3613 This involved changing a lot of size variables from int to off_t. It
3614 should work with maildirs and everything.
3616 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3617 spamd dying while we are connected to it.
3619 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3622 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3623 function for BDB 4.3.
3625 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3627 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3628 the sources for PCRE have been re-arranged and more files are now
3631 PH/15 The code I had for printing potentially long long variables in PH/11
3632 above was not the best (it lost precision). The length of off_t variables
3633 is now inspected at build time, and an appropriate printing format (%ld
3634 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3635 to be "long long int" or "long int". This is needed for the internal
3636 formatting function string_vformat().
3638 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3639 the configuration file to be ":syslog", then the script "guesses" where
3640 the logs files are, rather than using the compiled in default. In our
3641 case the guess is not the same as the compiled default, so the script
3642 suddenly stopped working when I started to use syslog. The patch checks
3643 to see if log_file_path is "". If so, it attempts to read it from exim
3644 with no configuration file to get the compiled in version, before it
3645 falls back to the previous guessing code."
3647 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3648 implementing BATV in an Exim configuration. See NewStuff for the gory
3651 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3652 Makefile that are specific to HP-UX.
3654 PH/18 If the "use_postmaster" option was set for a recipient callout together
3655 with the "random" option, the postmaster address was used as the MAIL
3656 FROM address for the random test, but not for the subsequent recipient
3657 test. It is now used for both.
3659 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3660 patch removes a few documentation additions to RFC 3028, because the
3661 latest draft now contains them. It adds the new en;ascii-case comparator
3662 and a new error check for 8bit text in MIME parts. Comparator and
3663 require names are now matched exactly. I enabled the subaddress
3664 extension, but it is not well tested yet (read: it works for me)."
3666 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3667 rework some of the code of TK/09 above to avoid the hardwired use of
3668 "%lld" and "long long". Replaced the call to snprintf() with a call to
3671 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3672 records point to non-existent hosts", "retry timeout exceeded", and
3673 "retry time not reached for any host after a long failure period".
3675 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3676 experimental DomainKeys support:
3678 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3679 (2) On an error such as an illegally used "control", the wrong name for
3680 the control was given.
3682 These problems did NOT occur unless DomainKeys support was compiled.
3684 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3686 PH/24 Added ${if match_ip condition.
3688 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3689 databases so that it will be absolutely obvious if a crash occurs in the
3690 DB library. This is a regular occurrence (often caused by mis-matched
3693 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3694 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3695 picked up on a user's system that detects such things. There doesn't seem
3696 to be a gcc warning option for this - only an attribute that has to be
3697 put on the function's prototype. It seems that in Fedora Core 4 they have
3698 set this on a number of new functions. No doubt there will be more in due
3701 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3702 specify a transport. However, if an address that was verified by such a
3703 router was the subject of a callout, Exim crashed because it tried to
3704 read the rcpt_include_affixes from the non-existent transport. Now it
3705 just assumes that the setting of that option is false. This bug was
3706 introduced by 4.51/PH/31.
3708 PH/28 Changed -d+all to exclude +memory, because that information is very
3709 rarely of interest, but it makes the output a lot bigger. People tend to
3710 do -d+all out of habit.
3712 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3713 code in os-type was giving problems when libc.so lives in lib64, like on
3716 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3717 aren't the modern standard, and it seems that some systems' include files
3718 don't always have them. Exim was already checking for some of the newer
3719 ones like T_AAAA, and defining it itself. I've added checks for all the
3720 record types that Exim uses.
3722 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3723 not automatically generating a new one, as it is supposed to. This
3724 prevented TLS from working. If the file did exist, but contained invalid
3725 data, a new version was generated, as expected. It was only the case of a
3726 non-existent file that was broken.
3728 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3729 with a change in libdomainkeys > 0.64.
3731 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3732 from DNS. If the selector record carries the flag, it now has
3733 precedence over the domain-wide flag.
3735 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3737 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3738 the use of a port name that isn't defined in /etc/services) occurred, the
3739 message was deferred as in a normal delivery, and thus remained on the
3740 spool, instead of being failed because of the mua_wrapper setting. This
3741 is now fixed, and I tidied up some of the mua_wrapper messages at the
3744 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3745 the messages in a hash of arrays rather than using individual hashes.
3746 This is a bit cleaner and results in dramatic memory savings, albeit
3747 at a slight CPU cost.
3749 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3750 as requested by Marc Sherman.
3752 SC/10 Eximstats: added histograms for user specified patterns as requested
3755 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3757 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3758 fopen() in the content-scanning modules that did not already have it.
3764 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3765 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3767 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3769 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3771 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3772 to test Sieve filters that use "vacation".
3774 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3775 that changes the way the GnuTLS parameters are stored in the cache file.
3776 The new format can be generated externally. For backward compatibility,
3777 if the data in the cache doesn't make sense, Exim assumes it has read an
3778 old-format file, and it generates new data and writes a new file. This
3779 means that you can't go back to an older release without removing the
3782 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3783 work if there are any delivery delays because "one_time" forces the
3784 parent to be marked "delivered", so its unseen clone is never tried
3785 again. For this reason, Exim now forbids the simultaneous setting of
3788 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3789 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3790 entry if you want to know the details. The fix, however, neglected to
3791 consider the case when local delivery batching is involved. The test for
3792 "previously delivered" was not happening when checking to see if an
3793 address could be batched with a previous (undelivered) one; under
3794 certain circumstances this could lead to multiple deliveries to the same
3797 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3798 in its include files, and this causes problems building Exim.
3800 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3801 header_syntax) but Exim was just ignoring anything given after a slash.
3802 In particular, this caused confusion with an attempt to use "verify =
3803 reverse_host_lookup/defer_ok". An error is now given when options are
3804 supplied for verify items that do not have them. (Maybe reverse_host_
3805 lookup should have a defer_ok option, but that's a different point.)
3807 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3808 defined by RFC 821) to 2048, because there were problems with some AUTH
3809 commands, and RFC 1869 says the size should be increased for extended
3810 SMTP commands that take arguments.
3812 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3815 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3816 "unknown" error; now it says that the functionality isn't in the binary.
3818 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3819 an address' error message when a string expansion fails (syntax or
3820 whatever). Otherwise the password may appear in the log. Following change
3821 PH/42 below, there is no longer a chance of it appearing in a bounce
3824 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3826 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3827 of its addresses was ever tried. (Bugzilla bug #2.)
3829 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3830 the result incorrectly in the debug output. (It correctly added a newline
3831 to what was transported.)
3833 TF/01 Added $received_time.
3835 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3836 commented out examples of how to interface to a virus scanner and to
3837 SpamAssassin. Also added commented examples of av_scanner and
3838 spamd_address settings.
3840 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3841 and controls are allowed in which ACLs. There were a couple of minor
3842 errors. Some of the entries in the conditions table (which is a table of
3843 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3844 negation of where the condition IS allowed.
3846 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3848 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3849 header file does not have a version number, so I've had to invent a new
3850 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3851 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3852 radiusclient), but it was contributed by a Radius user.
3854 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3855 files or directories, for OpenSSL.
3857 PH/19 When an Exim process that is running as root has to create an Exim log
3858 file, it does so in a subprocess that runs as exim:exim so as to get the
3859 ownership right at creation (otherwise, other Exim processes might see
3860 the file with the wrong ownership). There was no test for failure of this
3861 fork() call, which would lead to the process getting stuck as it waited
3862 for a non-existent subprocess. Forks do occasionally fail when resources
3863 run out. I reviewed all the other calls to fork(); they all seem to check
3866 PH/20 When checking for unexpected SMTP input at connect time (before writing
3867 the banner), Exim was not dealing correctly with a non-positive return
3868 from the read() function. If the client had disconnected by this time,
3869 the result was a log entry for a synchronization error with an empty
3870 string after "input=" when read() returned zero. If read() returned -1
3871 (an event I could not check), uninitialized data bytes were printed.
3872 There were reports of junk text (parts of files, etc) appearing after
3875 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3877 PH/22 Added support for macro redefinition, and (re)definition in between
3878 driver and ACL definitions.
3880 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3881 forgetting to use the resulting value; it was using the unexpanded value.
3883 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3884 hadn't been configured. The fix is from Juergen Kreileder, who
3885 understands it better than I do:
3887 "Here's what I see happening with three configured cyrus_sasl
3888 authenticators configured (plain, login, cram-md5):
3890 On startup auth_cyrus_sasl_init() gets called for each of these.
3891 This means three calls to sasl_listmech() without a specified mech_list.
3892 => SASL tests which mechs of all available mechs actually work
3893 => three warnings about OTP not working
3894 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3896 With the patch, sasl_listmech() also gets called three times. But now
3897 SASL's mech_list option is set to the server_mech specified in the the
3898 authenticator. Or in other words, the answer from sasl_listmech()
3899 gets limited to just the mech you're testing for (which is different
3901 => the return list contains just 'plain' or 'login', 'cram-md5' or
3902 nothing depending on the value of ob->server_mech.
3904 I've just tested the patch: Authentication still works fine,
3905 unavailable mechs specified in the exim configuration are still
3906 caught, and the auth.log warnings about OTP are gone."
3908 PH/25 When debugging is enabled, the contents of the command line are added
3909 to the debugging output, even when log_selector=+arguments is not
3912 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3913 answer is "GNU", and only if the return is "GNU/something" is the answer
3916 PH/27 $acl_verify_message is now set immediately after the failure of a
3917 verification in an ACL, and so is available in subsequent modifiers. In
3918 particular, the message can be preserved by coding like this:
3920 warn !verify = sender
3921 set acl_m0 = $acl_verify_message
3923 Previously, $acl_verify_message was set only while expanding "message"
3924 and "log_message" when a very denied access.
3926 PH/28 Modified OS/os.c-Linux with
3928 -#ifndef OS_LOAD_AVERAGE
3929 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3931 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3932 nomenclature these days.)
3934 PH/29 Installed patch from the Sieve maintainer that adds the options
3935 sieve_useraddress and sieve_subaddress to the redirect router.
3937 PH/30 In these circumstances:
3938 . Two addresses routed to the same list of hosts;
3939 . First host does not offer TLS;
3940 . First host accepts first address;
3941 . First host gives temporary error to second address;
3942 . Second host offers TLS and a TLS session is established;
3943 . Second host accepts second address.
3944 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3945 and peerdn, if requested) that were in fact used only for the second
3948 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3949 attention to any local part prefix or suffix that was matched by the
3950 router that accepted the address. It now behaves in the same way as it
3951 does for delivery: the affixes are removed from the local part unless
3952 rcpt_include_affixes is set on the transport.
3954 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3955 timeout during the DATA phase of an incoming message.
3957 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3958 applied a patch sanctioned by the Sieve maintainer.
3960 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3961 the uid or gid is negative. A case of a negative gid caused this to be
3962 noticed. The fix allows for either to be negative.
3964 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3965 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3968 PH/36 The change PH/12 above was broken. Fixed it.
3970 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3971 the grounds that routing the same address twice would always produce the
3972 same answer. This might have been true once, but it is certainly no
3973 longer true now. Routing a child address may depend on the previous
3974 routing that produced that child. Some complicated redirection strategies
3975 went wrong when messages had multiple recipients, and made Exim's
3976 behaviour dependent on the order in which the addresses were given.
3978 I have moved the duplicate checking until after the routing is complete.
3979 Exim scans the addresses that are assigned to local and remote
3980 transports, and removes any duplicates. This means that more work will be
3981 done, as duplicates will always all be routed, but duplicates are
3982 presumably rare, so I don't expect this is of any significance.
3984 For deliveries to pipes, files, and autoreplies, the duplicate checking
3985 still happens during the routing process, since they are not going to be
3988 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3989 It corrects a timeout issue with spamd. This is Ian's comment: "The
3990 background is that sometimes spamd either never reads data from a
3991 connection it has accepted, or it never writes response data. The exiscan
3992 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3993 blindly assumes that writes won't block so it may never time out."
3995 PH/39 Allow G after quota size as well as K and M.
3997 PH/40 The value set for $authenticated_id in an authenticator may not contain
3998 binary zeroes or newlines because the value is written to log lines and
3999 to spool files. There was no check on this. Now the value is run through
4000 the string_printing() function so that such characters are converted to
4001 printable escape sequences.
4003 PH/41 $message_linecount is a new variable that contains the total number of
4004 lines in the message. Compare $body_linecount, which is the count for the
4007 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4008 bounce and delay warning messages, except in certain special cases, which
4011 (a) An SMTP error message from a remote host;
4012 (b) A message specified in a :fail: redirection;
4013 (c) A message specified in a "fail" command in a system filter;
4014 (d) A message specified in a FAIL return from the queryprogram router;
4015 (e) A message specified by the cannot_route_message router option.
4017 In these cases only, Exim does include the error details in bounce and
4018 warning messages. There are also a few cases where bland messages such
4019 as "unrouteable address" or "local delivery error" are given.
4021 PH/43 $value is now also set for the "else" part of a ${run expansion.
4023 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4024 being worked on, but at least Exim now implements the latest version to
4027 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4028 process to complete was treated as a delivery failure, a timeout while
4029 writing the message to the pipe was logged, but erroneously treated as a
4030 successful delivery. Such timeouts include transport filter timeouts. For
4031 consistency with the overall process timeout, these timeouts are now
4032 treated as errors, giving rise to delivery failures by default. However,
4033 there is now a new Boolean option for the pipe transport called
4034 timeout_defer, which, if set TRUE, converts the failures into defers for
4035 both kinds of timeout. A transport filter timeout is now identified in
4038 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4039 systems where "make" and "gmake" are different, calling "gmake" at top
4040 level broke things. I've arranged for the value of $(MAKE) to be passed
4041 from the Makefile to this script so that it can call the same version of
4045 A note about Exim versions 4.44 and 4.50
4046 ----------------------------------------
4048 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4049 changes of various kinds. As a consequence, a big documentation update was
4050 needed. This delayed the release for rather longer than seemed good, especially
4051 in the light of a couple of (minor) security issues. Therefore, the changes
4052 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4053 release. So 4.44 and 4.50 are in effect two different branches that both start
4056 I have left the 4.50 change log unchanged; it contains all the changes since
4057 4.43. The change log for 4.44 is below; many of its items are identical to
4058 those for 4.50. This seems to be the most sensible way to preserve the
4059 historical information.
4065 1. Minor wording change to the doc/README.SIEVE file.
4067 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4068 computation of the current number of files was incorrect.
4070 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4071 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4074 4. Give more explanation in the error message when the command for a transport
4075 filter fails to execute.
4077 5. There are several places where Exim runs a non-Exim command in a
4078 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4079 was being done only for the command run by the queryprogram router. It is
4080 now done for all such subprocesses. The other cases are: ${run, transport
4081 filters, and the commands run by the lmtp and pipe transports.
4083 6. Added CONFIGURE_GROUP build-time option.
4085 7. Some older OS have a limit of 256 on the maximum number of file
4086 descriptors. Exim was using setrlimit() to set 1000 as a large value
4087 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4088 systems. I've change it so that if it can't get 1000, it tries for 256.
4090 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4091 was an oversight, and furthermore, ever since the addition of extra
4092 controls (e.g. 4.43/32), the checks on when to allow different forms of
4093 "control" were broken. There should now be diagnostics for all cases when a
4094 control that does not make sense is encountered.
4096 9. Added the /retain_sender option to "control=submission".
4098 10. $recipients is now available in the predata ACL (oversight).
4100 11. Tidy the search cache before the fork to do a delivery from a message
4101 received from the command line. Otherwise the child will trigger a lookup
4102 failure and thereby defer the delivery if it tries to use (for example) a
4103 cached ldap connection that the parent has called unbind on.
4105 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4106 of $address_data from the recipient verification was clobbered by the
4107 sender verification.
4109 13. The value of address_data from a sender verification is now available in
4110 $sender_address_data in subsequent conditions in the ACL statement.
4112 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4114 15. Added a new option "connect=<time>" to callout options, to set a different
4117 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4118 was its contents. (It was OK if the option was not defined at all.)
4120 17. A "Completed" log line is now written for messages that are removed from
4121 the spool by the -Mrm option.
4123 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4124 information about exactly what failed.
4126 19. Added -dd to debug only the daemon process.
4128 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4129 handles timeouts, both on the server side and network timeouts. Renamed the
4130 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4132 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4133 It is now set to "smtps".
4135 22. $host_address is now set to the target address during the checking of
4136 ignore_target_hosts.
4138 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4139 being passed; this would have caused $sender_host_name to have been used if
4140 matching the list had actually called for a host name (not very likely,
4141 since this list is usually IP addresses). A host name is now passed as
4144 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4145 code that specifies a non-restarting handler (typically sigaction() in
4146 modern systems) in an attempt to fix a rare and obscure crash bug.
4148 25. Narrowed the window for a race in the daemon that could cause it to ignore
4149 SIGCHLD signals. This is not a major problem, because they are used only to
4150 wake it up if nothing else does.
4152 26. A malformed maildirsize file could cause Exim to calculate negative values
4153 for the mailbox size or file count. Odd effects could occur as a result.
4154 The maildirsize information is now recalculated if the size or filecount
4157 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4158 support for a long time. Removed HAVE_SYS_VFS_H.
4160 28. Installed the latest version of exipick from John Jetmore.
4162 29. In an address list, if the pattern was not a regular expression, an empty
4163 subject address (from a bounce message) matched only if the pattern was an
4164 empty string. Non-empty patterns were not even tested. This was the wrong
4165 because it is perfectly reasonable to use an empty address as part of a
4166 database query. An empty address is now tested by patterns that are
4167 lookups. However, all the other forms of pattern expect the subject to
4168 contain a local part and a domain, and therefore, for them, an empty
4169 address still always fails if the pattern is not itself empty.
4171 30. Exim went into a mad DNS loop when attempting to do a callout where the
4172 host was specified on an smtp transport, and looking it up yielded more
4173 than one IP address.
4175 31. Re-factored the code for checking spool and log partition space into a
4176 function that finds that data and another that does the check. The former
4177 is then used to implement four new variables: $spool_space, $log_space,
4178 $spool_inodes, and $log_inodes.
4180 32. The RFC2047 encoding function was originally intended for short strings
4181 such as real names; it was not keeping to the 75-character limit for
4182 encoded words that the RFC imposes. It now respects the limit, and
4183 generates multiple encoded words if necessary. To be on the safe side, I
4184 have increased the buffer size for the ${rfc2047: expansion operator from
4187 33. It is now permitted to omit both strings after an "if" condition; if the
4188 condition is true, the result is "true". As before, when the second string
4189 is omitted, a false condition yields an empty string. This makes it less
4190 cumbersome to write custom ACL and router conditions.
4192 34. Failure to deliver a bounce message always caused it to be frozen, even if
4193 there was an errors_to setting on the router. The errors_to setting is now
4196 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4197 canonical form (fully expanded) before being placed in
4198 $sender_host_address.
4200 36. The table in the code that translates DNS record types into text (T_A to
4201 "A" for instance) was missing entries for NS and CNAME. It is just possible
4202 that this could have caused confusion if both these types were looked up
4203 for the same domain, because the text type is used as part of Exim's
4204 per-process caching. But the chance of anyone hitting this buglet seems
4207 37. The dnsdb lookup has been extended in a number of ways.
4209 (1) There is a new type, "zns", which walks up the domain tree until it
4210 finds some nameserver records. It should be used with care.
4212 (2) There is a new type, "mxh", which is like "mx" except that it returns
4213 just the host names, not the priorities.
4215 (3) It is now possible to give a list of domains (or IP addresses) to be
4216 looked up. The behaviour when one of the lookups defers can be
4217 controlled by a keyword.
4219 (4) It is now possible to specify the separator character for use when
4220 multiple records are returned.
4222 38. The dnslists ACL condition has been extended: it is now possible to supply
4223 a list of IP addresses and/or domains to be looked up in a particular DNS
4226 39. Added log_selector=+queue_time_overall.
4228 40. When running the queue in the test harness, wait just a tad after forking a
4229 delivery process, to get repeatability of debugging output.
4231 41. Include certificate and key file names in error message when GnuTLS fails
4232 to set them up, because the GnuTLS error message doesn't include the name
4233 of the failing file when there is a problem reading it.
4235 42. Allow both -bf and -bF in the same test run.
4237 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4239 44. The "Exiscan patch" is now merged into the mainline Exim source.
4241 45. Sometimes the final signoff response after QUIT could fail to get
4242 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4243 < 0 before doing a fflush(). This bug looks as though it goes back to the
4244 introduction of TLS in release 3.20, but "sometimes" must have been rare
4245 because the tests only now provoked it.
4247 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4248 (this can affect the format of dates).
4250 47. exim_tidydb, when checking for the continued existence of a message for
4251 which it has found a message-specific retry record, was not finding
4252 messages that were in split spool directories. Consequently, it was
4253 deleting retry records that should have stayed in existence.
4255 48. Steve fixed some bugs in eximstats.
4257 49. The SPA authentication driver was not abandoning authentication and moving
4258 on to the next authenticator when an expansion was forced to fail,
4259 contradicting the general specification for all authenticators. Instead it
4260 was generating a temporary error. It now behaves as specified.
4262 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4263 (the order specifies the preference for clients). The order is now AES256,
4264 AES128, 3DES, ARCFOUR128.
4266 51. Small patch to Sieve code - explicitly set From: when generating an
4269 52. Exim crashed if a remote delivery caused a very long error message to be
4270 recorded - for instance if somebody sent an entire SpamAssassin report back
4271 as a large number of 550 error lines. This bug was coincidentally fixed by
4272 increasing the size of one of Exim's internal buffers (big_buffer) that
4273 happened as part of the Exiscan merge. However, to be on the safe side, I
4274 have made the code more robust (and fixed the comments that describe what
4277 53. Now that there can be additional text after "Completed" in log lines (if
4278 the queue_time_overall log selector is set), a one-byte patch to exigrep
4279 was needed to allow it to recognize "Completed" as not the last thing in
4282 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4283 patch that reportedly fixes this has been added. I am not expert enough to
4284 create a test for it. This is what the patch creator wrote:
4286 "I found a little strange behaviour of ldap code when working with
4287 Windows 2003 AD Domain, where users was placed in more than one
4288 Organization Units. When I tried to give exim partial DN, the exit code
4289 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4290 But simultaneously result of request was absolutely normal ldap result,
4291 so I produce this patch..."
4293 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4294 REFERENCE, so I have modified the code to exclude the patch when that macro
4297 55. Some experimental protocols are using DNS PTR records for new purposes. The
4298 keys for these records are domain names, not reversed IP addresses. The
4299 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4300 leaves it alone. Component reversal etc. now happens only for IP addresses.
4303 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4305 57. Double the size of the debug message buffer (to 2048) so that more of very
4306 long debug lines gets shown.
4308 58. The exicyclog utility now does better if the number of log files to keep
4309 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4311 59. Two changes related to the smtp_active_hostname option:
4313 (1) $smtp_active_hostname is now available as a variable.
4314 (2) The default for smtp_banner uses $smtp_active_hostname instead
4315 of $primary_hostname.
4317 60. The host_aton() function is supposed to be passed a string that is known
4318 to be a valid IP address. However, in the case of IPv6 addresses, it was
4319 not checking this. This is a hostage to fortune. Exim now panics and dies
4320 if the condition is not met. A case was found where this could be provoked
4321 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4322 components; fortuitously, this particular loophole had already been fixed
4323 by change 4.50/55 above.
4325 If there are any other similar loopholes, the new check in host_aton()
4326 itself should stop them being exploited. The report I received stated that
4327 data on the command line could provoke the exploit when Exim was running as
4328 exim, but did not say which command line option was involved. All I could
4329 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4330 running as the user.
4333 61. There was a buffer overflow vulnerability in the SPA authentication code
4334 (which came originally from the Samba project). I have added a test to the
4335 spa_base64_to_bits() function which I hope fixes it.
4338 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4339 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4341 63. The daemon start-up calls getloadavg() while still root for those OS that
4342 need the first call to be done as root, but it missed one case: when
4343 deliver_queue_load_max is set with deliver_drop_privilege. This is
4344 necessary for the benefit of the queue runner, because there is no re-exec
4345 when deliver_drop_privilege is set.
4347 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4348 This has been fixed.
4350 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4351 was in use, was not putting the data itself into the right store pool;
4352 consequently, it could be overwritten for a subsequent message in the same
4353 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4356 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4358 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4359 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4360 the function were treating the return as a boolean value, which happened to
4361 work because 0=false and not-0=true, but is not correct code.
4363 68. The host_aton() function was not handling scoped IPv6 addresses (those
4364 with, for example, "%eth0" on the end) correctly.
4366 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4367 negated items (that is, ~something) in unsigned ints. Some compilers
4368 apparently mutter when there is no cast.
4370 70. If an address verification called from an ACL failed, and did not produce a
4371 user-specific message (i.e. there was only a "system" message), nothing was
4372 put in $acl_verify_message. In this situation, it now puts the system
4375 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4376 session; change 4.31/43 added the unwanted input to the log line - except
4377 that it did not do this in the start of session case. It now does.
4379 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4380 This is wrong and can cause the other end to generate a synchronization
4381 error if it is another Exim or anything else that does the synchronization
4382 check. A QUIT command is no longer sent after a timeout.
4384 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4385 during host lookups.
4387 74. The defer_ok option of callout verification was not working if it was used
4388 when verifying addresses in header lines, that is, for this case:
4390 verify = header_sender/callout=defer_ok
4392 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4393 those file descriptors could be used for SMTP connections. If anything
4394 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4395 could be sent to the SMTP client, causing chaos. The daemon now opens
4396 stdin, stdout, and stderr to /dev/null when it puts itself into the
4399 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4400 log by default. The user can override this with suitable Perl magic.
4402 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4403 the log message when discard triggers, was not working for the DATA ACL or
4404 for the non-SMTP ACL.
4406 78. Error message wording change in sieve.c.
4408 79. If smtp_accept_max_per_host was set, the number of connections could be
4409 restricted to fewer than expected, because the daemon was trying to set up
4410 a new connection before checking whether the processes handling previous
4411 connections had finished. The check for completed processes is now done
4412 earlier. On busy systems, this bug wouldn't be noticed because something
4413 else would have woken the daemon, and it would have reaped the completed
4416 80. If a message was submitted locally by a user whose login name contained one
4417 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4418 It caused a spool format error. I have fixed the spool reading code. A
4419 related problem was that the "from" clause in the Received: line became
4420 illegal because of the space(s). It is now covered by ${quote_local_part.
4422 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4425 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4427 83. Patch from Sieve maintainer to fix unterminated string problem in
4428 "vacation" handling.
4430 84. Some minor changes to the Linux configuration files to help with other
4431 OS variants using glibc.
4433 85. One more patch for Sieve to update vacation handling to latest spec.
4436 ----------------------------------------------------
4437 See the note above about the 4.44 and 4.50 releases.
4438 ----------------------------------------------------
4444 1. Change 4.43/35 introduced a bug that caused file counts to be
4445 incorrectly computed when quota_filecount was set in an appendfile
4448 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4449 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4452 3. Give more explanation in the error message when the command for a transport
4453 filter fails to execute.
4455 4. There are several places where Exim runs a non-Exim command in a
4456 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4457 was being done only for the command run by the queryprogram router. It is
4458 now done for all such subprocesses. The other cases are: ${run, transport
4459 filters, and the commands run by the lmtp and pipe transports.
4461 5. Some older OS have a limit of 256 on the maximum number of file
4462 descriptors. Exim was using setrlimit() to set 1000 as a large value
4463 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4464 systems. I've change it so that if it can't get 1000, it tries for 256.
4466 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4467 was an oversight, and furthermore, ever since the addition of extra
4468 controls (e.g. 4.43/32), the checks on when to allow different forms of
4469 "control" were broken. There should now be diagnostics for all cases when a
4470 control that does not make sense is encountered.
4472 7. $recipients is now available in the predata ACL (oversight).
4474 8. Tidy the search cache before the fork to do a delivery from a message
4475 received from the command line. Otherwise the child will trigger a lookup
4476 failure and thereby defer the delivery if it tries to use (for example) a
4477 cached ldap connection that the parent has called unbind on.
4479 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4480 of $address_data from the recipient verification was clobbered by the
4481 sender verification.
4483 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4484 was its contents. (It was OK if the option was not defined at all.)
4486 11. A "Completed" log line is now written for messages that are removed from
4487 the spool by the -Mrm option.
4489 12. $host_address is now set to the target address during the checking of
4490 ignore_target_hosts.
4492 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4493 being passed; this would have caused $sender_host_name to have been used if
4494 matching the list had actually called for a host name (not very likely,
4495 since this list is usually IP addresses). A host name is now passed as
4498 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4499 code that specifies a non-restarting handler (typically sigaction() in
4500 modern systems) in an attempt to fix a rare and obscure crash bug.
4502 15. Narrowed the window for a race in the daemon that could cause it to ignore
4503 SIGCHLD signals. This is not a major problem, because they are used only to
4504 wake it up if nothing else does.
4506 16. A malformed maildirsize file could cause Exim to calculate negative values
4507 for the mailbox size or file count. Odd effects could occur as a result.
4508 The maildirsize information is now recalculated if the size or filecount
4511 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4512 support for a long time. Removed HAVE_SYS_VFS_H.
4514 18. Updated exipick to current release from John Jetmore.
4516 19. Allow an empty sender to be matched against a lookup in an address list.
4517 Previously the only cases considered were a regular expression, or an
4520 20. Exim went into a mad DNS lookup loop when doing a callout where the
4521 host was specified on the transport, if the DNS lookup yielded more than
4524 21. The RFC2047 encoding function was originally intended for short strings
4525 such as real names; it was not keeping to the 75-character limit for
4526 encoded words that the RFC imposes. It now respects the limit, and
4527 generates multiple encoded words if necessary. To be on the safe side, I
4528 have increased the buffer size for the ${rfc2047: expansion operator from
4531 22. Failure to deliver a bounce message always caused it to be frozen, even if
4532 there was an errors_to setting on the router. The errors_to setting is now
4535 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4536 canonical form (fully expanded) before being placed in
4537 $sender_host_address.
4539 24. Updated eximstats to version 1.33
4541 25. Include certificate and key file names in error message when GnuTLS fails
4542 to set them up, because the GnuTLS error message doesn't include the name
4543 of the failing file when there is a problem reading it.
4545 26. Expand error message when OpenSSL has problems setting up cert/key files.
4548 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4549 (this can affect the format of dates).
4551 28. exim_tidydb, when checking for the continued existence of a message for
4552 which it has found a message-specific retry record, was not finding
4553 messages that were in split spool directories. Consequently, it was
4554 deleting retry records that should have stayed in existence.
4556 29. eximstats updated to version 1.35
4557 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4558 1.35 - bugfix such that pie charts by volume are generated correctly
4560 30. The SPA authentication driver was not abandoning authentication and moving
4561 on to the next authenticator when an expansion was forced to fail,
4562 contradicting the general specification for all authenticators. Instead it
4563 was generating a temporary error. It now behaves as specified.
4565 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4566 (the order specifies the preference for clients). The order is now AES256,
4567 AES128, 3DES, ARCFOUR128.
4569 31. Small patch to Sieve code - explicitly set From: when generating an
4572 32. Exim crashed if a remote delivery caused a very long error message to be
4573 recorded - for instance if somebody sent an entire SpamAssassin report back
4574 as a large number of 550 error lines. This bug was coincidentally fixed by
4575 increasing the size of one of Exim's internal buffers (big_buffer) that
4576 happened as part of the Exiscan merge. However, to be on the safe side, I
4577 have made the code more robust (and fixed the comments that describe what
4580 33. Some experimental protocols are using DNS PTR records for new purposes. The
4581 keys for these records are domain names, not reversed IP addresses. The
4582 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4583 leaves it alone. Component reversal etc. now happens only for IP addresses.
4586 34. The host_aton() function is supposed to be passed a string that is known
4587 to be a valid IP address. However, in the case of IPv6 addresses, it was
4588 not checking this. This is a hostage to fortune. Exim now panics and dies
4589 if the condition is not met. A case was found where this could be provoked
4590 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4591 components; fortuitously, this particular loophole had already been fixed
4592 by change 4.50/55 or 4.44/33 above.
4594 If there are any other similar loopholes, the new check in host_aton()
4595 itself should stop them being exploited. The report I received stated that
4596 data on the command line could provoke the exploit when Exim was running as
4597 exim, but did not say which command line option was involved. All I could
4598 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4599 running as the user.
4602 35. There was a buffer overflow vulnerability in the SPA authentication code
4603 (which came originally from the Samba project). I have added a test to the
4604 spa_base64_to_bits() function which I hope fixes it.
4607 36. The daemon start-up calls getloadavg() while still root for those OS that
4608 need the first call to be done as root, but it missed one case: when
4609 deliver_queue_load_max is set with deliver_drop_privilege. This is
4610 necessary for the benefit of the queue runner, because there is no re-exec
4611 when deliver_drop_privilege is set.
4613 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4614 was in use, was not putting the data itself into the right store pool;
4615 consequently, it could be overwritten for a subsequent message in the same
4616 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4619 38. Sometimes the final signoff response after QUIT could fail to get
4620 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4621 < 0 before doing a fflush(). This bug looks as though it goes back to the
4622 introduction of TLS in release 3.20, but "sometimes" must have been rare
4623 because the tests only now provoked it.
4629 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4630 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4631 Now it returns an integer. A number of calls were still expecting a T/F
4632 return. Fortuitously, in all cases, the tests worked in OK situations,
4633 which is the norm. However, things would have gone wrong on any write
4634 failures on the smtp file descriptor. This function is used when sending
4635 messages over SMTP and also when doing verify callouts.
4637 2. When Exim is called to do synchronous delivery of a locally submitted
4638 message (the -odf or -odi options), it no longer closes stderr before doing
4641 3. Implemented the mua_wrapper option.
4643 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4645 5. Implemented the functions header_remove(), header_testname(),
4646 header_add_at_position(), and receive_remove_recipient(), and exported them
4649 6. If an ACL "warn" statement specified the addition of headers, Exim already
4650 inserted X-ACL-Warn: at the start if there was no header name. However, it
4651 was not making this test for the second and subsequent header lines if
4652 there were newlines in the string. This meant that an invalid header could
4653 be inserted if Exim was badly configured.
4655 7. Allow an ACL "warn" statement to add header lines at the start or after all
4656 the Received: headers, as well as at the end.
4658 8. Added the rcpt_4xx retry error code.
4660 9. Added postmaster_mailfrom=xxx to callout verification option.
4662 10. Added mailfrom=xxxx to the callout verification option, for verify=
4665 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4666 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4668 12. Inserted some casts to stop certain compilers warning when using pointer
4669 differences as field lengths or precisions in printf-type calls (mostly
4670 affecting debugging statements).
4672 13. Added optional readline() support for -be (dynamically loaded).
4674 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4675 same clock tick as a message's arrival, so that its received time was the
4676 same as the "first fail" time on the retry record, and that message
4677 remained on the queue past the ultimate address timeout, every queue runner
4678 would try a delivery (because it was past the ultimate address timeout) but
4679 after another failure, the ultimate address timeout, which should have then
4680 bounced the address, did not kick in. This was a "< instead of <=" error;
4681 in most cases the first failure would have been in the next clock tick
4682 after the received time, and all would be well.
4684 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4685 being recognized when the domain list was tested by the match_domain
4686 condition in an expansion string.
4688 16. Added the ${str2b64: operator.
4690 17. Exim was always calling setrlimit() to set a large limit for the number of
4691 processes, without checking whether the existing limit was already
4692 adequate. (It did check for the limit on file descriptors.) Furthermore,
4693 errors from getrlimit() and setrlimit() were being ignored. Now they are
4694 logged to the main and panic logs, but Exim does carry on, to try to do its
4695 job under whatever limits there are.
4697 18. Imported PCRE 5.0.
4699 19. Trivial typo in log message " temporarily refused connection" (the leading
4702 20. If the log selector return_path_on_delivery was set and an address was
4703 redirected to /dev/null, the delivery process crashed because it assumed
4704 that a return path would always be set for a "successful" delivery. In this
4705 case, the whole delivery is bypassed as an optimization, and therefore no
4708 21. Internal re-arrangement: the function for sending a challenge and reading
4709 a response while authentication was assuming a zero-terminated challenge
4710 string. It's now changed to take a pointer and a length, to allow for
4711 binary data in such strings.
4713 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4715 23. Exim was not respecting finduser_retries when seeking the login of the
4716 uid under which it was called; it was always trying 10 times. (The default
4717 setting of finduser_retries is zero.) Also, it was sleeping after the final
4718 failure, which is pointless.
4720 24. Implemented tls_on_connect_ports.
4722 25. Implemented acl_smtp_predata.
4724 26. If the domain in control=submission is set empty, Exim assumes that the
4725 authenticated id is a complete email address when it generates From: or
4726 Sender: header lines.
4728 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4729 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4730 chown and chgrp in /bin and hostname in /usr/bin.
4732 28. Exim was keeping the "process log" file open after each use, just as it
4733 does for the main log. This opens the possibility of it remaining open for
4734 long periods when the USR1 signal hits a daemon. Occasional processlog
4735 errors were reported, that could have been caused by this. Anyway, it seems
4736 much more sensible not to leave this file open at all, so that is what now
4739 29. The long-running daemon process does not normally write to the log once it
4740 has entered its main loop, and it closes the log before doing so. This is
4741 so that log files can straightforwardly be renamed and moved. However,
4742 there are a couple of unusual error situations where the daemon does write
4743 log entries, and I had neglected to close the log afterwards.
4745 30. The text of an SMTP error response that was received during a remote
4746 delivery was being truncated at 512 bytes. This is too short for some of
4747 the long messages that one sometimes sees. I've increased the limit to
4750 31. It is now possible to make retry rules that apply only when a message has a
4751 specific sender, in particular, an empty sender.
4753 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4754 it possible to be selective about when SMTP synchronization is enforced.
4756 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4758 32. Implemented hosts_connection_nolog.
4760 33. Added an ACL for QUIT.
4762 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4765 35. Added mailbox_size and mailbox_filecount to appendfile.
4767 36. Added control = no_multiline_responses to ACLs.
4769 37. There was a bug in the logic of the code that waits for the clock to tick
4770 in the case where the clock went backwards by a substantial amount such
4771 that the microsecond fraction of "now" was more than the microsecond
4772 fraction of "then" (but the whole seconds number was less).
4774 38. Added support for the libradius Radius client library this is found on
4775 FreeBSD (previously only the radiusclient library was supported).
4781 1. When certain lookups returned multiple values in the form name=value, the
4782 quoting of the values was not always being done properly. Specifically:
4783 (a) If the value started with a double quote, but contained no whitespace,
4785 (b) If the value contained whitespace other than a space character (i.e.
4786 tabs or newlines or carriage returns) it was not quoted.
4787 This fix has been applied to the mysql and pgsql lookups by writing a
4788 separate quoting function and calling it from the lookup code. The fix
4789 should probably also be applied to nisplus, ibase and oracle lookups, but
4790 since I cannot test any of those, I have not disturbed their existing code.
4792 2. A hit in the callout cache for a specific address caused a log line with no
4793 reason for rejecting RCPT. Now it says "Previous (cached) callout
4794 verification failure".
4796 3. There was an off-by-one bug in the queryprogram router. An over-long
4797 return line was truncated at 256 instead of 255 characters, thereby
4798 overflowing its buffer with the terminating zero. As well as fixing this, I
4799 have increased the buffer size to 1024 (and made a note to document this).
4801 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4802 when Exim is waiting for an SMTP response from a remote server, Exim
4803 restarts its select() call on the socket, thereby resetting its timeout.
4804 This is not a problem when such interrupts are rare. Somebody set up a cron
4805 job to run exiwhat every 2 minutes, which is less than the normal select()
4806 timeout (5 or 10 minutes). This meant that the select() timeout never
4807 kicked in because it was always reset. I have fixed this by comparing the
4808 time when an interrupt arrives with the time at the start of the first call
4809 to select(). If more time than the timeout has elapsed, the interrupt is
4810 treated as a timeout.
4812 5. Some internal re-factoring in preparation for the addition of Sieve
4813 extensions (by MH). In particular, the "personal" test is moved to a
4814 separate function, and given an option for scanning Cc: and Bcc: (which is
4815 not set for Exim filters).
4817 6. When Exim created an email address using the login of the caller as the
4818 local part (e.g. when creating a From: or Sender: header line), it was not
4819 quoting the local part when it contained special characters such as @.
4821 7. Installed new OpenBSD configuration files.
4823 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4824 try to make them clearer.
4826 9. Callout options, other than the timeout value, were being ignored when
4827 verifying sender addresses in header lines. For example, when using
4829 verify = header_sender/callout=no_cache
4831 the cache was (incorrectly) being used.
4833 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4834 only the Cygwin environment.
4836 11. When return_path_on_delivery was set as a log selector, if different remote
4837 addresses in the same message used different return paths and parallel
4838 remote delivery occurred, the wrong values would sometimes be logged.
4839 (Whenever a remote delivery process finished, the return path value from
4840 the most recently started remote delivery process was logged.)
4842 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4843 lines when AUTH and/or TLS are in use. This is the "received protocol"
4844 field. Exim used to use "asmtp" for authenticated SMTP, without any
4845 indication (in the protocol name) for TLS use. Now it follows the RFC and
4846 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4847 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4848 names appear in log lines as well as in Received: header lines.
4850 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4851 extensions, and comparison tests, and to fix some bugs.
4853 14. Changes to the "personal" filter test:
4855 (1) The test was buggy in that it was just doing the equivalent of
4856 "contains" tests on header lines. For example, if a user's address was
4857 anne@some.where, the "personal" test would incorrectly be true for
4859 To: susanne@some.where
4861 This test is now done by extracting each address from the header in turn,
4862 and checking the entire address. Other tests that are part of "personal"
4863 are now done using regular expressions (for example, to check local parts
4864 of addresses in From: header lines).
4866 (2) The list of non-personal local parts in From: addresses has been
4867 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4868 taken from the Sieve specification recommendations.
4870 (3) If the message contains any header line starting with "List-" it is
4871 treated as non-personal.
4873 (4) The test for "circular" in the Subject: header line has been removed
4874 because it now seems ill-conceived.
4876 15. Minor typos in src/EDITME comments corrected.
4878 16. Installed latest exipick from John Jetmore.
4880 17. If headers_add on a router specified a text string that was too long for
4881 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4882 of string_sprintf() is now avoided.
4884 18. $message_body_size was not set (it was always zero) when running the DATA
4885 ACL and the local_scan() function.
4887 19. For the "mail" command in an Exim filter, no default was being set for
4888 the once_repeat time, causing a random time value to be used if "once" was
4889 specified. (If the value happened to be <= 0, no repeat happened.) The
4890 default is now 0s, meaning "never repeat". The "vacation" command was OK
4891 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4892 (I found it when inspecting the code).
4894 20. There is now an overall timeout for performing a callout verification. It
4895 defaults to 4 times the callout timeout, which applies to individual SMTP
4896 commands during the callout. The overall timeout applies when there is more
4897 than one host that can be tried. The timeout is checked before trying the
4898 next host. This prevents very long delays if there are a large number of
4899 hosts and all are timing out (e.g. when the network connections are timing
4900 out). The value of the overall timeout can be changed by specifying an
4901 additional sub-option for "callout", called "maxwait". For example:
4903 verify = sender/callout=5s,maxwait=20s
4905 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4906 to the end before writing, but this should make it even safer).
4908 22. Exim was forgetting that it had advertised PIPELINING for the second and
4909 subsequent messages on an SMTP connection. It was also not resetting its
4910 memory on STARTTLS and an internal HELO.
4912 23. When Exim logs an SMTP synchronization error within a session, it now
4913 records whether PIPELINING has been advertised or not.
4915 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4916 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4917 rather than long int.
4919 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4921 26. Added the never_mail option to autoreply.
4927 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4928 crash if the getsockname() call failed; this can happen if a connection is
4929 closed very soon after it is established. The problem was simply in the
4930 order in which certain operations were done, causing Exim to try to write
4931 to the SMTP stream before it had set up the file descriptor. The bug has
4932 been fixed by making things happen in the correct order.
4938 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4939 before the connection was closed, thus losing the rejection response.
4941 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4942 some early Solaris releases, but causes trouble in current releases where
4943 socklen_t is defined.
4945 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4948 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4951 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4952 that releasing the top of it at the end releases what was used for sub-
4953 expansions (unless the block got too big). However, discard this block if
4954 the first thing is a variable or header, so that we can use its block when
4955 it is dynamic (useful for very large $message_headers, for example).
4957 6. Lookups now cache *every* query, not just the most recent. A new, separate
4958 store pool is used for this. It can be recovered when all lookup caches are
4959 flushed. Lookups now release memory at the end of their result strings.
4960 This has involved some general refactoring of the lookup sources.
4962 7. Some code has been added to the store_xxx() functions to reduce the amount
4963 of flapping under certain conditions.
4965 8. log_incoming_interface used to affect only the <= reception log lines. Now
4966 it causes the local interface and port to be added to several more SMTP log
4967 lines, for example "SMTP connection from", and rejection lines.
4969 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4971 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4973 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4974 could be overwritten at the end of the current message (or the start of a
4975 new message if it was set in a HELO ACL). The value is now preserved for
4976 the duration of the SMTP connection.
4978 12. If a transport had a headers_rewrite setting, and a matching header line
4979 contained an unqualified address, that address was qualified, even if it
4980 did not match any rewriting rules. The underlying bug was that the values
4981 of the flags that permit the existence of unqualified sender and recipient
4982 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4983 non-local messages, and by -bnq for local messages) were not being
4984 preserved with the message after it was received.
4986 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4987 "next input=" as part of the text comprising the host identity instead of
4988 the correct text. The code was using the same buffer for two different
4989 strings. However, depending on which order the printing function evaluated
4990 its arguments, the bug did not always show up. Under Linux, for example, my
4991 test suite worked just fine.
4993 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4994 doesn't work with some older versions of Perl. It has been changed to "my",
4995 which in any case is probably the better facility to use.
4997 15. A really picky compiler found some instances of statements for creating
4998 error messages that either had too many or two few arguments for the format
5001 16. The size of the buffer for calls to the DNS resolver has been increased
5002 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5003 for addresses that have a lot of PTR records. This alleviates a problem; it
5004 does not fully solve it.
5006 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5007 buffer now truncates the list and logs the incident, which is the same
5008 action as happens when Exim is looking up a host name and its aliases.
5009 Previously in this situation something unpredictable would happen;
5010 sometimes it was "internal error: store_reset failed".
5012 18. If a server dropped the connection unexpectedly when an Exim client was
5013 using GnuTLS and trying to read a response, the client delivery process
5014 crashed while trying to generate an error log message.
5016 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5017 string, for example:
5019 warn message = H1: something\nH2: something
5021 the text was added as a single header line from Exim's point of view
5022 though it ended up OK in the delivered message. However, searching for the
5023 second and subsequent header lines using $h_h2: did not work. This has been
5024 fixed. Similarly, if a system filter added multiple headers in this way,
5025 the routers could not see them.
5027 20. Expanded the error message when iplsearch is called with an invalid key to
5028 suggest using net-iplsearch in a host list.
5030 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5031 ACLs are no longer actually imposed (and a message to that effect is
5034 22. If a "gecos" field in a passwd entry contained escaped characters, in
5035 particular, if it contained a \" sequence, Exim got it wrong when building
5036 a From: or a Sender: header from that name. A second bug also caused
5037 incorrect handling when an unquoted " was present following a character
5038 that needed quoting.
5040 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5041 was not being matched caselessly.
5043 24. Arranged for all hyphens in the exim.8 source to be escaped with
5046 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5047 properly. Recipient callout cache records were still being keyed to include
5048 the sender, even when use_sender was set false. This led to far more
5049 callouts that were necessary. The sender is no longer included in the key
5050 when use_sender is false.
5052 26. Added "control = submission" modifier to ACLs.
5054 27. Added the ${base62d: operator to decode base 62 numbers.
5056 28. dnsdb lookups can now access SRV records.
5058 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5059 the configuration file.
5061 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5062 (-v) mode. This makes the output for a verbose queue run more intelligible.
5064 31. Added a use_postmaster feature to recipient callouts.
5066 32. Added the $body_zerocount variable, containing the number of binary zero
5067 bytes in the message body.
5069 33. The time of last modification of the "new" subdirectory is now used as the
5070 "mailbox time last read" when there is a quota error for a maildir
5073 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5075 35. Added +ignore_unknown as a special item in host lists.
5077 36. Code for decoding IPv6 addresses in host lists is now included, even if
5078 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5079 address was recognized as an IP address, but was then not correctly decoded
5080 into binary, causing unexpected and incorrect effects when compared with
5087 1. Very minor rewording of debugging text in manualroute to say "list of
5088 hosts" instead of "hostlist".
5090 2. If verify=header_syntax was set, and a header line with an unqualified
5091 address (no domain) and a large number of spaces between the end of the
5092 name and the colon was received, the reception process suffered a buffer
5093 overflow, and (when I tested it) crashed. This was caused by some obsolete
5094 code that should have been removed. The fix is to remove it!
5096 3. When running in the test harness, delay a bit after writing a bounce
5097 message to get a bit more predictability in the log output.
5099 4. Added a call to search_tidyup() just before forking a reception process. In
5100 theory, someone could use a lookup in the expansion of smtp_accept_max_
5101 per_host which, without the tidyup, could leave open a database connection.
5103 5. Added the variables $recipient_data and $sender_data which get set from a
5104 lookup success in an ACL "recipients" or "senders" condition, or a router
5105 "senders" option, similar to $domain_data and $local_part_data.
5107 6. Moved the writing of debug_print from before to after the "senders" test
5110 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5111 problems for message scanning, either using a data ACL, or using
5112 local_scan() because the Received: header was not generated till after they
5113 were called (in order to set the time as the time of reception completion).
5114 I have revised the way this works. The header is now generated after the
5115 body is received, but before the ACL or local_scan() are called. After they
5116 are run, the timestamp in the header is updated.
5122 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5123 before starting a queue runner without re-exec. This happened only when
5124 deliver_drop_privilege was set or when the Exim user was set to root. The
5125 effect of the bug was that timeouts during subsequent deliveries caused
5126 crashes instead of being properly handled. The handler is now left at its
5127 default (and expected) setting.
5129 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5130 message, again when deliver_drop_privilege is set or Exim is run as root.
5131 The bug described in (1) was not present in this case, but the tidying up
5132 of the other signals was missing. I have made the two cases consistent.
5134 3. The ignore_target_hosts setting on a manualroute router was being ignored
5135 for hosts that were looked up using the /MX notation.
5137 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5140 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5141 operated on the sender address. After changing the $sender_address to <>
5142 for the sender address verify, Exim was re-instated it as the original
5143 (before rewriting) address, but remembering that it had rewritten it, so it
5144 wasn't rewriting it again. This bug also had the effect of breaking the
5145 sender address verification caching when the sender address was rewritten.
5147 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5148 This has been changed so that if the ip literal address matches
5149 ignore_target_hosts, the router declines.
5151 7. Added expansion conditions match_domain, match_address, and match_local_
5152 part (NOT match_host).
5154 8. The placeholder for the Received: header didn't have a length field set.
5156 9. Added code to Exim itself and to exim_lock to test for a specific race
5157 condition that could lead to file corruption when using MBX delivery. The
5158 issue is with the lockfile that is created in /tmp. If this file is removed
5159 after a process has opened it but before that process has acquired a lock,
5160 there is the potential for a second process to recreate the file and also
5161 acquire a lock. This could lead to two Exim processes writing to the file
5162 at the same time. The added code performs the same test as UW imapd; it
5163 checks after acquiring the lock that its file descriptor still refers to
5164 the same named file.
5166 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5167 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5170 11. Added the smtp_active_hostname option. If used, this will typically be made
5171 to depend on the incoming interface address. Because $interface_address is
5172 not set up until the daemon has forked a reception process, error responses
5173 that can happen earlier (such as "too many connections") no longer contain
5176 12. If an expansion in a condition on a "warn" statement fails because a lookup
5177 defers, the "warn" statement is abandoned, and the next ACL statement is
5178 processed. Previously this caused the whole ACL to be aborted.
5180 13. Added the iplsearch lookup type.
5182 14. Added ident_timeout as a log selector.
5184 15. Added tls_certificate_verified as a log selector.
5186 16. Added a global option tls_require_ciphers (compare the smtp transport
5187 option of the same name). This controls incoming TLS connections.
5189 17. I finally figured out how to make tls_require_ciphers do a similar thing
5190 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5191 before starting the TLS session.
5193 18. Tabs are now shown as \t in -bP output.
5195 19. If the log selector return_path_on_delivery was set, Exim crashed when
5196 bouncing a message because it had too many Received: header lines.
5198 20. If two routers both had headers_remove settings, and the first one included
5199 a superfluous trailing colon, the final name in the first list and the
5200 first name in the second list were incorrectly joined into one item (with a
5201 colon in the middle).
5207 1. Added -C and -D options to the exinext utility, mainly to make it easier
5208 to include in the automated testing, but these could be helpful when
5209 multiple configurations are in use.
5211 2. The exinext utility was not formatting the output nicely when there was
5212 an alternate port involved in the retry record key, nor when there was a
5213 message id as well (for retries that were specific to a specific message
5214 and a specific host). It was also confused by IPv6 addresses, because of
5215 the additional colons they contain. I have fixed the IPv4 problem, and
5216 patched it up to do a reasonable job for IPv6.
5218 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5219 delivery, the log line now contains "pipelined" if PIPELINING was used.
5221 4. An SMTP transport process used to panic and die if the bind() call to set
5222 an explicit outgoing interface failed. This has been changed; it is now
5223 treated in the same way as a connect() failure.
5225 5. A reference to $sender_host_name in the part of a conditional expansion
5226 that was being skipped was still causing a DNS lookup. This no longer
5229 6. The def: expansion condition was not recognizing references to header lines
5230 that used bh_ and bheader_.
5232 7. Added the _cache feature to named lists.
5234 8. The code for checking quota_filecount in the appendfile transport was
5235 allowing one more file than it should have been.
5237 9. For compatibility with Sendmail, the command line option
5245 and sets the incoming protocol and host name (for trusted callers). The
5246 host name and its colon can be omitted when only the protocol is to be set.
5247 Note the Exim already has two private options, -pd and -ps, that refer to
5248 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5249 "s", but I don't think that's a major issue.
5251 10. A number of refactoring changes to the code, none of which should affect
5254 (a) The number of logging options was getting close to filling up the
5255 32-bit word that was used as a bit map. I have split them into two classes:
5256 those that are passed in the argument to log_write(), and those that are
5257 only ever tested independently outside of that function. These are now in
5258 separate 32-bit words, so there is plenty of room for expansion again.
5259 There is no change in the user interface or the logging behaviour.
5261 (b) When building, for example, log lines, the code previously used a
5262 macro that called string_cat() twice, in order to add two strings. This is
5263 not really sufficiently general. Furthermore, there was one instance where
5264 it was actually wrong because one of the argument was used twice, and in
5265 one call a function was used. (As it happened, calling the function twice
5266 did not affect the overall behaviour.) The macro has been replaced by a
5267 function that can join an arbitrary number of extra strings onto a growing
5270 (c) The code for expansion conditions now uses a table and a binary chop
5271 instead of a serial search (which was left over from when there were very
5272 few conditions). Also, it now recognizes conditions like "pam" even when
5273 the relevant support is not compiled in: a suitably worded error message is
5274 given if an attempt is made to use such a condition.
5276 11. Added ${time_interval:xxxxx}.
5278 12. A bug was causing one of the ddress fields not to be passed back correctly
5279 from remote delivery subprocesses. The field in question was not being
5280 subsequently used, so this caused to problems in practice.
5282 13. Added new log selectors queue_time and deliver_time.
5284 14. Might have fixed a bug in maildirsizefile handling that threw up
5285 "unexpected character" debug warnings, and recalculated the data
5286 unnecessarily. In any case, I expanded the warning message to give more
5289 15. Added the message "Restricted characters in address" to the statements in
5290 the default ACL that block characters like @ and % in local parts.
5292 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5293 Three changes have been made:
5295 (a) There was a serious bug; a negative response to MAIL caused the whole
5296 recipient domain to be cached as invalid, thereby blocking all messages
5297 to all local parts at the same domain, from all senders. This bug has
5298 been fixed. The domain is no longer cached after a negative response to
5299 MAIL if the sender used is not empty.
5301 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5304 (c) A new callout option, "use_sender" has been added for people who want
5305 the modified behaviour.
5311 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5314 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5315 indeed breaks things for older releases.
5317 3. Added additional logging to the case where there is a problem reading data
5318 from a filter that is running in a subprocess using a pipe, in order to
5319 try to track down a specific problem.
5321 4. Testing facility fudge: when running in the test harness and attempting
5322 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5323 getting "No route to host". Convert this to a timeout.
5325 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5328 6. Some OS don't have socklen_t but use size_t instead. This affects the
5329 fifth argument of getsockopt() amongst other things. This is now
5330 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5331 can be set for individual OS. I have set it for SunOS5, OSF1, and
5332 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5333 some earlier ones do not.
5335 7. Change 4.30/15 was not doing the test caselessly.
5337 8. The standard form for an IPv6 address literal was being rejected by address
5338 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5339 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5340 this, as well as the form without the "IPv6" on the front (but only when
5341 address literals are enabled, of course).
5343 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5345 10. Exim crashed if a message with an empty sender address specified by -f
5346 encountered a router with an errors_to setting. This could be provoked only
5347 by a command such as
5351 where an empty string was supplied; "<>" did not hit this bug.
5353 11. Installed PCRE release 4.5.
5355 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5356 remained set. It is now erased.
5358 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5359 times from message ids (which are base 36 rather than the normal 62).
5361 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5362 were being counted as actual protocol errors, and logged if the log
5363 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5364 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5365 and DATA following a set of rejected RCPTs do not count as protocol errors.
5366 In other words, Exim assumes they were pipelined, though this may not
5367 actually be the case. Of course, in all cases the client gets an
5368 appropriate error code.
5370 15. If a lookup fails in an ACL condition, a message about the failure may
5371 be available; it is used if testing the ACL cannot continue, because most
5372 such messages specify what the cause of the deferral is. However, some
5373 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5374 that caused an old message to be retained and used if a later statement
5375 caused a defer, replacing the real cause of the deferral.
5377 16. If an IP address had so many PTR records that the DNS lookup buffer
5378 was not large enough to hold them, Exim could crash while trying to process
5379 the truncated data. It now detects and logs this case.
5381 17. Further to 4.21/58, another change has been made: if (and only if) the
5382 first line of a message (the first header line) ends with CRLF, a bare LF
5383 in a subsequent header line has a space inserted after it, so as not to
5384 terminate the header.
5386 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5387 unnecessarily, used atoi() instead of strtol(), and didn't check the
5388 termination when getting file sizes from file names by regex.
5390 19. Completely re-implemented the support for maildirsize files, in the light
5391 of a number of problems with the previous contributed implementation
5392 (4.30/29). In particular:
5394 . If the quota is zero, the maildirsize file is maintained, but no quota is
5397 . If the maildir directory does not exist, it is created before any attempt
5398 to write a maildirsize file.
5400 . The quota value in the file is just a cache; if the quota is changed in
5401 the transport, the new value overrides.
5403 . A regular expression is available for excluding directories from the
5406 20. The autoreply transport checks the characters in options that define the
5407 message's headers; it allows continued headers, but it was checking with
5408 isspace() after an embedded newline instead of explicitly looking for a
5411 21. If all the "regular" hosts to which an address was routed had passed their
5412 expiry times, and had not reached their retry times, the address was
5413 bounced, even if fallback hosts were defined. Now Exim should go on to try
5416 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5417 equivalent code in the SMTP transport. Some hosts send humungous responses
5418 to HELO/EHLO, more than 1024 it seems.
5420 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5421 gives compiler warnings in some environments. I've now done it "properly",
5424 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5425 (because of problems with the built-in one) was declared to return uschar *
5426 instead of char *, causing compiler failure.
5428 25. Fixed a file descriptor leak when processing alias/forward files.
5430 26. Fixed a minor format string issue in dbfn.c.
5432 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5434 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5435 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5438 29. When a sender address is verified, it is cached, to save repeating the test
5439 when there is more than one recipient in a message. However, when the
5440 verification involves a callout, it is possible for different callout
5441 options to be set for different recipients. It is too complicated to keep
5442 track of this in the cache, so now Exim always runs a verification when a
5443 callout is required, relying on the callout cache for the optimization.
5444 The overhead is duplication of the address routing, but this should not be
5447 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5448 to be verified with callout=postmaster, and the main callout worked but the
5449 postmaster check failed, the verification correctly failed. However, if a
5450 subsequent RCPT command asked for sender verification *without* the
5451 postmaster check, incorrect caching caused this verification also to fail,
5454 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5455 it was not caching the DNS options (qualify_single, search_parents) that
5456 were used when the lookup failed. A subsequent lookup with different
5457 options therefore always gave the same answer, though there were cases
5458 where it should not have. (Example: a "domains = !$mx_any" option on a
5459 dnslookup router: the "domains" option is always processed without any
5460 widening, but the router might have qualify_single set.) Now Exim uses the
5461 cached value only when the same options are set.
5463 32. Added John Jetmore's "exipick" utility to the distribution.
5465 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5466 than a timeout (e.g. a certificate is required, and is not provided), an
5467 Exim server now closes the connection immediately. Previously it waited for
5468 the client to close - but if the client is SSL, it seems that they each
5469 wait for each other, leading to a delay before one of them times out.
5471 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5472 maintained 0.8.x compatibility because I don't think many are using it, and
5473 it is clearly obsolete.
5475 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5478 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5479 client certificate was expired. A simple patch fixes this, though I don't
5480 understand the full logic of why the verify callback is called multiple
5483 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5484 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5485 which causes problems with some clients (such as the Certicom SSL Plus
5486 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5487 disables the coutermeasure allowing Eudora to connect."
5489 38. Exim was not checking that a write() to a log file succeeded. This could
5490 lead to Bad Things if a log got too big, in particular if it hit a file
5491 size limit. Exim now panics and dies if it cannot write to a log file, just
5492 as it does if it cannot open a log file.
5494 39. Modified OS/Makefile-Linux so that it now contains
5496 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5498 The two -D definitions ensure that Exim is compiled with large file
5499 support, which makes it possible to handle log files that are bigger than
5502 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5503 instance) a domain was checked against a named list that involved a lookup,
5504 causing $domain_data to be set, then another domain was checked against the
5505 same list, then the first domain was re-checked, the value of $domain_data
5506 after the final check could be wrong. In particular, if the second check
5507 failed, it could be set empty. This bug probably also applied to
5510 41. The strip_trailing_dot option was not being applied to the address given
5511 with the -f command-line option.
5513 42. The code for reading a message's header from the spool was incrementing
5514 $received_count, but never initializing it. This meant that the value was
5515 incorrect (doubled) while delivering a message in the same process in which
5516 it was received. In the most common configuration of Exim, this never
5517 happens - a fresh exec is done - but it can happen when
5518 deliver_drop_privilege is set.
5520 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5521 it now includes up to 150 characters of the unexpected data in the log
5524 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5525 and building data strings. The size of both of these buffers was 10 000
5526 bytes - far larger than anybody would *ever* want, thought I. Needless to
5527 say, somebody hit the limit. I have increased the maximum line length to
5528 20 000 and the maximum data length of concatenated lines to 100 000. I have
5529 also fixed two bugs, because there was no checking on these buffers. Tsk,
5530 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5531 buffer is too small.
5533 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5534 lsearch lookups. Now it does.
5536 46. When parsing a route_list item in a manualroute router, a fixed-length
5537 buffer was used for the list of hosts. I made this 1024 bytes long,
5538 thinking that nobody would ever have a list of hosts that long. Wrong.
5539 Somebody had a whole pile of complicated expansion conditions, and the
5540 string was silently truncated, leading to an expansion error. It turns out
5541 that it is easier to change to an unlimited length (owing to other changes
5542 that have happened since this code was originally written) than to build
5543 structure for giving a limitation error. The length of the item that
5544 expands into the list of hosts is now unlimited.
5546 47. The lsearch lookup could not handle data where the length of text line was
5547 more than 4095 characters. Such lines were truncated, leading to shortened
5548 data being returned. It should now handle lines of any length.
5550 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5551 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5554 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5555 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5556 no longer happens in comment lines. A list of replacements is now placed
5557 at the head of all of the source files, except those whose only change is
5558 to replace PERL_COMMAND in the very first #! line.
5560 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5561 messages on the queue, with a bottom-up merge sort, using code contributed
5562 by Michael Haardt. This should make operations like -bp somewhat faster on
5563 large queues. It won't affect queue runners, except when queue_run_in_order
5566 51. Installed eximstats 1.31 in the distribution.
5568 52. Added support for SRV lookups to the dnslookup router.
5570 53. If an ACL referred to $message_body or $message_body_end, the value was not
5571 reset for any messages that followed in the same SMTP session.
5573 54. The store-handling optimization for building very long strings was not
5574 differentiating between the different store pools. I don't think this
5575 actually made any difference in practice, but I've tidied it.
5577 55. While running the routers to verify a sender address, $sender_address
5578 was still set to the sender address. This is wrong, because when routing to
5579 send a bounce to the sender, it would be empty. Therefore, I have changed
5580 it so that, while verifying a sender address, $sender_address is set to <>.
5581 (There is no change to what happens when verifying a recipient address.)
5583 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5584 target A or AAAA records (if not already returned) without resetting the
5585 qualify_single or search_parents options of the DNS resolver. These are
5586 inappropriate in this case because the targets of MX and SRV records must
5587 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5588 target that, when qualified, matched something in the local domain. These
5589 two options are now turned off when doing these lookups.
5591 57. It seems that at least some releases of Reiserfs (which does not have the
5592 concept of a fixed number of inodes) returns zero and not -1 for the
5593 number of available inodes. This interacted badly with check_spool_inodes,
5594 which assumed that -1 was the "no such thing" setting. What I have done is
5595 to check that the total number of inodes is greater than zero before doing
5596 the test of how many are available.
5598 58. When a "warn" ACL statement has a log_message modifier, the message is
5599 remembered, and not repeated. This is to avoid a lot of repetition when a
5600 message has many recipients that cause the same warning to be written.
5601 Howewer, Exim was preserving the list of already written lines for an
5602 entire SMTP session, which doesn't seem right. The memory is now reset if a
5603 new message is started.
5605 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5606 debugging output unless log_rewrite was also set.
5608 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5609 of (int)(handle) when we know that handle contains (void *)(-1).
5611 61. The Exim daemon panic-logs an error return when it closes the incoming
5612 connection. However "connection reset by peer" seems to be common, and
5613 isn't really an error worthy of noting specially, so that particular error
5616 62. When Exim is trying to find all the local interfaces, it used to panic and
5617 die if the ioctl to get the interface flags failed. However, it seems that
5618 on at least one OS (Solaris 9) it is possible to have an interface that is
5619 included in the list of interfaces, but for which you get a failure error
5620 for this call. This happens when the interface is not "plumbed" into a
5621 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5622 failure of the "get flags" call assumes that the interface is down.
5624 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5625 makes life easier for people who are doing arithmetic on fields extracted
5626 from dates, where you often get leading zeros that should not be
5627 interpreted as octal.
5629 64. Added qualify_domain to the redirect router, to override the global
5632 65. If a pathologically long header line contained very many addresses (the
5633 report of this problem mentioned 10 000) and each of them was rewritten,
5634 Exim could use up a very large amount of memory. (It kept on making new
5635 copies of the header line as it rewrote, and never released the old ones.)
5636 At the expense of a bit more processing, the header rewriting function has
5637 been changed so that it no longer eats memory in this way.
5639 66. The generation of the Received: header has been moved from the time that a
5640 message starts to be received, to the time that it finishes. The timestamp
5641 in the Received: header should now be very close to that of the <= log
5642 line. There are two side-effects of this change:
5644 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5645 logged header lines no longer include the local Received: line, because
5646 it has not yet been created. The same applies to a copy of the message
5647 that is returned to a non-SMTP sender when a message is rejected.
5649 (b) When a filter file is tested using -bf, no additional Received: header
5650 is added to the test message. After some thought, I decided that this
5653 This change does not affect the value of $received_for. It is still set
5654 after address rewriting, but before local_scan() is called.
5656 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5658 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5659 gave an unhelpful panic error message, and a defer error. I have managed to
5660 change this behaviour so that it now rejects any supplied certificate,
5661 which seems right, as the list of acceptable certificates is empty.
5663 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5664 gave an unhelpful defer error. I have not managed to make this reject any
5665 supplied certificates, but the error message it gives is "no certificate
5666 supplied", which is not helpful.
5668 70. exigrep's output now also includes lines that are not associated with any
5669 message, but which match the given pattern. Implemented by a patch from
5670 Martin Sluka, which also tidied up the Perl a bit.
5672 71. Recipient callout verification, like sender verification, was using <> in
5673 the MAIL FROM command. This isn't really the right thing, since the actual
5674 sender may affect whether the remote host accepts the recipient or not. I
5675 have changed it to use the actual sender in the callout; this means that
5676 the cache record is now keyed on a recipient/sender pair, not just the
5677 recipient address. There doesn't seem to be a real danger of callout loops,
5678 since a callout by the remote host to check the sender would use <>.
5679 [SEE ABOVE: changed after hitting problems.]
5681 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5682 temporary errors. However, in the case of such a code being given after
5683 the end of a data transmission (i.e. after ".") Exim was failing to write
5684 a retry record for the message. (Yes, there was some broken host that was
5685 actually sending 8xx at this point.)
5687 73. An unknown lookup type in a host list could cause Exim to panic-die when
5688 the list was checked. (An example that provoked this was putting <; in the
5689 middle of a list instead of at the start.) If this happened during a DATA
5690 ACL check, a -D file could be left lying around. This kind of configuration
5691 error no longer causes Exim to die; instead it causes a defer errror. The
5692 incident is still logged to the main and panic logs.
5694 74. Buglet left over from Exim 3 conversion. The message "too many messages
5695 in one connection" was written to the rejectlog but not the mainlog, except
5696 when address rewriting (yes!) was being logged.
5698 75. Added write_rejectlog option.
5700 76. When a system filter was run not as root (that is, when system_filter_user
5701 was set), the values of the $n variables were not being returned to the
5702 main process; thus, they were not subsequently available in the $sn
5705 77. Added +return_path_on_delivery log selector.
5707 78. A connection timeout was being treated differently from recipients deferred
5708 when testing hosts_max_try with a message that was older than the host's
5709 retry timeout. (The host should not be counted, thus allowing all hosts to
5710 be tried at least once before bouncing.) This may have been the cause of an
5711 occasionally reported bug whereby a message would remain on the queue
5712 longer than the retry timeout, but would be bounced if a delivery was
5713 forced. I say "may" because I never totally pinned down the problem;
5714 setting up timeout/retry tests is difficult. See also the next item.
5716 79. The ultimate address timeout was not being applied to errors that involved
5717 a combination of host plus message (for example, a timeout on a MAIL
5718 command). When an address resolved to a number of possible hosts, and they
5719 were not all tried for each delivery (e.g. because of hosts_max_try), a
5720 message could remain on the queue longer than the retry timeout.
5722 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5725 81. Fixed an obscure SMTP outgoing bug which required at least the following
5726 conditions: (a) there was another message waiting for the same server;
5727 (b) the server returned 5xx to all RCPT commands in the first message so
5728 that the message was not completed; (c) the server dropped the connection
5729 or gave a negative response to the RSET that Exim sends to abort the
5730 transaction. The observed case was a dropped connection after DATA that had
5731 been sent in pipelining mode. That is, the server had advertised PIPELINING
5732 but was not implementing it correctly. The effect of the bug was incorrect
5733 behaviour, such as trying another host, and this could lead to a crash.
5739 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5740 and daemon.c were passed as pointers to ints; they should have been
5741 pointers to socklen_t variables (which are typically unsigned ints).
5743 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5746 3. Fixed a really odd bug that affected only the testing scheme; patching a
5747 certain fixed string in the binary changed the value of another string that
5748 happened to be identical to the end of the original first string.
5750 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5751 name", it returns that address as the IP address. On some operating
5752 systems (e.g. Solaris), it also passes back the IP address string as the
5753 "host name". However, on others (e.g. Linux), it passes back an empty
5754 string. Exim wasn't checking for this, and was changing the host name to an
5755 empty string, assuming it had been canonicized.
5757 5. Although rare, it is permitted to have more than one PTR record for a given
5758 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5759 all the names associated with an address, because they do in Solaris.
5760 However, it seems that they do not in Linux for data that comes from the
5761 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5762 I found this out when I moved to a new Linux workstation and tried to run
5763 the Exim test suite.
5765 To get round this problem I have changed the code so that it now does its
5766 own call to the DNS to look up PTR records when searching for a host name.
5767 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5768 addresses that are only in /etc/hosts are still found.
5770 This behaviour is, however, controlled by an option called host_lookup_
5771 order, which defaults to "bydns:byaddr". If people want to use the other
5772 order, or indeed, just use one or the other means of lookup, they can
5773 specify it in this variable.
5775 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5776 some operating systems, this comes back from gethostbyaddr() as an empty
5777 string, and this is what Exim used to test for. However, it seems that in
5778 other systems, "." is yielded. Exim now tests for this case too.
5780 7. The values of check_spool_space and check_log_space are now held internally
5781 as a number of kilobytes instead of an absolute number of bytes. If a
5782 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5783 kilobyte. This means that much larger values can be stored.
5785 8. Exim monitor: an attempt to get the action menu when not actually pointing
5786 at a message produces an empty menu entitled "No message selected". This
5787 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5788 no entries in it ("Shell widget menu has zero width and/or height"). So I
5789 have added a single, blank menu entry in this case.
5791 9. Added ${quote_local_part.
5793 10. MIME decoding is now applied to the contents of Subject: header lines when
5796 11. Now that a reference to $sender_host_address automatically causes a reverse
5797 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5798 host lookup before query-style lookups in lists that might use this
5799 variable. This has therefore been abolished, and the "net-" prefix is no
5800 longer necessary for query-style lookups.
5802 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5803 have been a typo for LFLAGS, so it has been changed.
5805 13. The install script calls Exim with "-C /dev/null" in order to find the
5806 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5807 to be output. Howeve, since Exim outputs its version number before the
5808 error, it didn't break the script. It just looked ugly. I fixed this by
5809 always allowing "-C /dev/null" if the caller is root.
5811 14. Ignore overlarge ACL variable number when reading spool file - insurance
5812 against a later release with more variables having written the file.
5814 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5815 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5816 this, as well as the form without the "IPv6" on the front.
5818 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5819 OS/Makefile-Darwin file.
5821 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5822 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5824 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5825 variable. However, some OS use ints for inodes. I've added cast to long int
5826 to get rid of the compiler warning.
5828 19. I had forgotten to lock out "/../" in configuration file names when
5829 ALT_CONFIG_PREFIX was set.
5831 20. Routers used for verification do not need to specify transports. However,
5832 if such a router generated a host list, and callout was configured, Exim
5833 crashed, because it could not find a port number from the (non-existent)
5834 transport. It now assumes port 25 in this circumstance.
5836 21. Added the -t option to exigrep.
5838 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5839 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5840 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5841 them is set, LOOKUP_LSEARCH is forced.
5843 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5845 24. Added sender and host information to the "rejected by local_scan()" log
5846 line; previously there was no indication of these.
5848 25. Added .include_if_exists.
5850 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5851 Linux. It turns out that not all file systems support this. Apparently some
5852 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5853 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5854 not supported on the file descriptor, is now ignored when Exim is trying to
5855 sync a directory. This applies only to Linux.
5857 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5859 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5860 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5863 29. Added support for maildirsize files from supplied patch (modified a bit).
5865 30. The use of :fail: followed by an empty string could lead Exim to respond to
5866 sender verification failures with (e.g.):
5868 550 Verification failed for <xxx>
5869 550 Sender verify failed
5871 where the first response line was missing the '-' that indicates it is not
5872 the final line of the response.
5874 31. The loop for finding the name of the user that called Exim had a hardwired
5875 limit of 10; it now uses the value of finduser_retries, which is used for
5876 all other user lookups.
5878 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5881 33. Exim was neglecting to zero errno before one call of strtol() when
5882 expanding a string and expecting an integer value. On some systems this
5883 resulted in spurious "integer overflow" errors. Also, it was casting the
5884 result into an int without checking.
5886 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5887 did not work. The code looks as if it has *never* worked, though it appears
5888 to have been documented since at least releast 1.62. I have made it work.
5890 35. The "timeout_DNS" error in retry rules, also documented since at least
5891 1.62, also never worked. As it isn't clear exactly what this means, and
5892 clearly it isn't a major issue, I have abolished the feature by treating it
5893 as "timeout", and writing a warning to the main and panic logs.
5895 36. The display of retry rules for -brt wasn't always showing the error code
5898 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5899 timeout_connect_A, timeout_connect_MX.
5901 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5902 to the empty sender.
5904 39. The daemon was not analysing the content of -oX till after it had closed
5905 stderr and disconnected from the controlling terminal. This meant that any
5906 syntax errors were only noted on the panic log, and the return code from
5907 the command was 0. By re-arranging the code a little, I've made the
5908 decoding happen first, so such errors now appear on stderr, and the return
5909 code is 1. However, the actual setting up of the sockets still happens in
5910 the disconnected process, so errors there are still only recorded on the
5913 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5914 connections (as happens on some IP stacks) was logged at start up time as
5915 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5916 it from "IPv6 and IPv4", which means that two separate sockets are being
5919 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5920 says whether AF_INET or AF_INET6 was passed as an argument.
5922 42. Exiwhat output was messed up when time zones were included in log
5925 43. Exiwhat now gives more information about the daemon's listening ports,
5926 and whether -tls-on-connect was used.
5928 44. The "port" option of the smtp transport is now expanded.
5930 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5931 silently ignored. Now an error message is written to the main and panic
5934 46. There's a new ACL modifier called "logwrite" which writes to a log file
5935 as soon as it is encountered.
5937 47. Added $local_user_uid and $local_user_gid at routing time.
5939 48. Exim crashed when trying to verify a sender address that was being
5942 49. Exim was recognizing only a space character after ".include". It now also
5943 recognizes a tab character.
5945 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5946 extracting the relevant information from the specification. The man page no
5947 longer contains scrambled data for the -d option, and I've added a section
5948 at the front about calling Exim under different names.
5950 51. Added "extra_headers" argument to the "mail" command in filter files.
5952 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5955 53. Installed eximstats 1.29.
5957 54. Added transport_filter_timeout as a generic transport option.
5959 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5960 Cc: header lines. This was required by RFC 822, but it not required by RFC
5963 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5964 incoming messages that did not have them. Now it does so only if the
5965 message originates locally, that is, if there is no associated remote host
5966 address. When Resent- header lines are present, this applies to the Resent-
5967 lines rather than the non-Resent- lines.
5969 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5970 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5972 58. Messages for configuration errors now include the name of the main
5973 configuration file - useful now that there may be more than one file in a
5974 list (.included file names were always shown).
5976 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5977 for those rare installations that do not start the daemon as root or run it
5978 setuid root. I've cut out the call to initgroups() if the daemon is not
5981 60. The Exim user and group can now be bound into the binary as text strings
5982 that are looked up at the start of Exim's processing.
5984 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5986 62. Added $mailstore_basename variable.
5988 63. Installed patch to sieve.c from Michael Haardt.
5990 64. When Exim failed to open the panic log after failing to open the main log,
5991 the original message it was trying to log was written to stderr and debug
5992 output, but if they were not available (the usual case in production), it
5993 was lost. Now it is written to syslog before the two lines that record the
5994 failures to open the logs.
5996 65. Users' Exim filters run in subprocesses under the user's uid. It is
5997 possible for a "deliver" command or an alias in a "personal" command to
5998 provoke an address rewrite. If logging of address rewriting is configured,
5999 this fails because the process is not running as root or exim. There may be
6000 a better way of dealing with this, but for the moment (because 4.30 needs
6001 to be released), I have disabled address rewrite logging when running a
6002 filter in a non-root, non-exim process.
6008 1. The buildconfig auxiliary program wasn't quoting the value set for
6009 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6010 not defined. This bug was masked in 4.22 by the effect that was fixed in
6013 2. Some messages that were rejected after a message id was allocated were
6014 shown as "incomplete" by exigrep. It no longer does this for messages that
6015 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6017 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6018 have allow_domain_literals set, the ID did not get logged in the <= line.
6019 Domain literals are now always recognized in Message-ID: header lines.
6021 4. The first argument for a ${extract expansion item is the key name or field
6022 number. Leading and trailing spaces in this item were not being ignored,
6023 causing some misleading effects.
6025 5. When deliver_drop_privilege was set, single queue runner processes started
6026 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6027 same command in the process it spins off) were not dropping privilege.
6029 6. When the daemon running as "exim" started a queue runner, it always
6030 re-executed Exim in the spun-off process. This is a waste of effort when
6031 deliver_drop_privilege is set. The new process now just calls the
6032 queue-runner function directly.
6038 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6041 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6042 ignored. Though the use of -oP was forcing the writing of a pid file, it
6043 was always written to the default place.
6045 3. If the message "no IP address found for host xxxx" is generated during
6046 incoming verification, it is now followed by identification of the incoming
6047 connection (so you can more easily find what provoked it).
6049 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6051 5. Added some features to "harden" Exim a bit more against certain attacks:
6053 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6054 be put in Local/Makefile. This is like the never_users runtime option,
6055 but it cannot be overridden. The default setting is "root".
6057 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6058 prefix string with which any file named in a -C command line option
6061 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6062 is retained for -C and -D only if the caller of Exim is root. Without
6063 it, the exim user may also use -C and -D and retain privilege.
6065 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6066 command line option is disabled.
6068 6. Macro names set by the -D option must start with an upper case letter, just
6069 like macro names defined in the configuration file.
6071 7. Added "dereference=" facility to LDAP.
6073 8. Two instances of the typo "uknown" in the source files are fixed.
6075 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6076 the Configure-Makefile script screwed up while processing it.
6078 10. Incorporated PCRE 4.4.
6080 11. The SMTP synchronization check was not operating right at the start of an
6081 SMTP session. For example, it could not catch a HELO sent before the client
6082 waited for the greeting. There is now a check for outstanding input at the
6083 point when the greeting is written. Because of the duplex, asynchronous
6084 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6085 way, but not yet received, when the check is performed.
6087 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6088 on TCP/IP sockets, because this apparently causes some broken clients to
6091 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6092 unchanged) from the Cygwin maintainer.
6094 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6095 was supported instead of testing for mbx. Effectively a typo.
6097 15. The spa authenticator server code was not checking that the input it
6098 received was valid base64.
6100 16. The debug output line for the "set" modifier in ACLs was not showing the
6101 name of the variable that was being set.
6103 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6105 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6106 lookup on its own. Something else was needed to trigger the lookup. For
6107 example, a match in host_lookup or the need for a host name in a host list.
6108 Now, if $sender_host_name is referenced and the host name has not yet been
6109 looked up, a lookup is performed. If the lookup fails, the variable remains
6110 empty, and $host_lookup_failed is set to "1".
6112 19. Added "eqi" as a case-independent comparison operator.
6114 20. The saslauthd authentication condition could segfault if neither service
6115 nor realm was specified.
6117 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6118 error message that was logged was misleading, and incoming SMTP
6119 connections were dropped. The message is now more accurate, and temporary
6120 errors are given to SMTP connections.
6122 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6123 (except RSET) with a 421 error, until QUIT is received. However, it was
6124 failing to send a response to QUIT.
6126 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6127 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6128 ACL is now run after the helo_try_verify_hosts code.
6130 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6131 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6132 case-independent, so other case variants are also recognized). Apparently
6133 some systems use these upper case variants.
6135 25. If more than two messages were waiting for the same host, and a transport
6136 filter was specified for the transport, Exim sent two messages over the
6137 same TCP/IP connection, and then failed with "socket operation on non-
6138 socket" when it tried to send the third.
6140 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6142 27. The extern definition of crypt16() in expand.c was not being excluded when
6143 the OS had its own crypt16() function.
6145 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6146 as a preferred synonym for return_size_limit, both as an option and as an
6149 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6151 30. Changed the default configuration ACL to relax the local part checking rule
6152 for addresses that are not in any local domains. For these addresses,
6153 slashes and pipe symbols are allowed within local parts, but the sequence
6154 /../ is explicitly forbidden.
6156 31. SPA server authentication was not clearing the challenge buffer before
6159 32. log_message in a "warn" ACL statement was writing to the reject log as
6160 well as to the main log, which contradicts the documentation and doesn't
6161 seem right (because no rejection is happening). So I have stopped it.
6163 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6164 However, I am unable to do any testing of this.
6166 34. Fixed an infelicity in the appendfile transport. When checking directories
6167 for a mailbox, to see if any needed to be created, it was accidentally
6168 using path names with one or more superfluous leading slashes; tracing
6169 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6171 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6172 contents are added to the log line that is written for every discarded
6173 recipient. (Previously a log_message setting was ignored.)
6175 36. The ${quote: operator now quotes the string if it is empty.
6177 37. The install script runs exim in order to find its version number. If for
6178 some reason other than non-existence or emptiness, which it checks, it
6179 could not run './exim', it was installing it with an empty version number,
6180 i.e. as "exim-". This error state is now caught, and the installation is
6183 38. An argument was missing from the function that creates an error message
6184 when Exim fails to connect to the socket for saslauthd authentication.
6185 This could cause Exim to crash, or give a corrupted message.
6187 39. Added isip, isip4, and isip6 to ${if conditions.
6189 40. The ACL variables $acl_xx are now saved with the message, and can be
6190 accessed later in routers, transports, and filters.
6192 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6193 strings in the file are not string-expanded.
6195 42. If a MAIL command specified a SIZE value that was too large to fit into an
6196 int variable, the check against message_size_limit failed. Such values are
6197 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6198 day this will have to be increased, but I don't think I want to be around
6199 when emails are that large.
6206 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6207 iconv() is not standard in FreeBSD.
6209 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6210 IPv6 enabled. The observed symptom was a segmentation fault on return from
6211 the function os_common_find_running_interfaces() in src/os.c.
6213 3. In the check_special_case() function in daemon.c I had used "errno" as an
6214 argument name, which causes warnings on some systems. This was basically a
6215 typo, since it was named "eno" in the comments!
6217 4. The code that waits for the clock to tick (at a resolution of some fraction
6218 of a second) so as to ensure message-id uniqueness was always waiting for
6219 at least one whole tick, when it could have waited for less. [This is
6220 almost certainly not relevant at current processor speeds, where it is
6221 unlikely to ever wait at all. But we try to future-proof.]
6223 5. The function that sleeps for a time interval that includes fractions of a
6224 second contained a race. It did not block SIGALRM between setting the
6225 timer, and suspending (a couple of lines later). If the interval was short
6226 and the sigsuspend() was delayed until after it had expired, the suspension
6227 never ended. On busy systems this could lead to processes getting stuck for
6230 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6231 process, before it forks any delivery processes. The open lookup caching
6232 mechanism meant that the open file or database connection was passed into
6233 the delivery process. The problem was that delivery processes always tidy
6234 up cached lookup data. This could cause a problem for the next delivery
6235 process started by the queue runner, because the external queue runner
6236 process does not know about the closure. So the next delivery process
6237 still has data in the lookup cache. In the case of a file lookup, there was
6238 no problem because closing a file descriptor in a subprocess doesn't affect
6239 the parent. However, if the lookup was caching a connection to a database,
6240 the connection was closed, and the second delivery process was likely to
6241 see errors such as "PGSQL: query failed: server closed the connection
6242 unexpectedly". The problem has been fixed by closing all cached lookups
6243 in a queue runner before running a delivery process.
6245 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6246 seem to have the "const" qualifier which it has on other OS. I've
6249 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6250 *of the same type* (client or server) with the same public name that an
6251 error should be diagnosed.
6253 9. When Exim looked up a host name for an IP address, but failed to find the
6254 original IP address when looking up the host name (a safety check), it
6255 output the message "<ip address> does not match any IP for NULL", which was
6256 confusing, to say the least. The bug was that the host name should have
6257 appeared instead of "NULL".
6259 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6260 user that is built into the binary, and the -C or -D options is used, root
6261 privilege is dropped before the configuration file is read. In addition,
6262 logging is switched to stderr instead of the normal log files. If the
6263 configuration then re-defines the Exim user, the unprivileged environment
6264 is probably not what is expected, so Exim logs a panic warning message (but
6267 However, if deliver_drop_privilege is set, the unprivileged state may well
6268 be exactly what is intended, so the warning has been cut out in that case,
6269 and Exim is allowed to try to write to its normal log files.
6275 1. smtp_return_error_details was not giving details for temporary sender
6276 or receiver verification errors.
6278 2. Diagnose a configuration error if two authenticators have the same public
6281 3. Exim used not to create the message log file for a message until the first
6282 delivery attempt. This could be confusing when incoming messages were held
6283 for policy or load reasons. The message log file is now created at the time
6284 the message is received, and an initial "Received" line is written to it.
6286 4. The automatically generated man page for command line options had a minor
6287 bug that caused no ill effects; however, a more serious problem was that
6288 the procedure for building the man page automatically didn't always
6289 operate. Consequently, release 4.20 contains an out-of-date version. This
6290 shouldn't happen again.
6292 5. When building Exim with embedded Perl support, the script that builds the
6293 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6294 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6296 6. The freeze_tell option was not being used for messages that were frozen on
6297 arrival, either by an ACL or by local_scan().
6299 7. Added the smtp_incomplete_transaction log selector.
6301 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6302 was accepting AUTH without a new EHLO.
6304 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6305 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6308 10. Logging of TCP/IP connections (when configured) now happens in the main
6309 daemon process instead of the child process, so that the TCP/IP connection
6310 count is more accurate (but it can never be perfect).
6312 11. The use of "drop" in a nested ACL was not being handled correctly in the
6313 outer ACL. Now, if condition failure induced by the nested "drop" causes
6314 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6315 or "require"), the connection is dropped.
6317 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6318 that yield "discard" can now be used with an "accept" or a "discard" verb,
6319 but an error is generated for any others (because I can't see a useful way
6320 to define what should happen).
6322 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6323 are now processed in the same way as lines in the Exim configuration file.
6324 In particular, continuation lines are supported.
6326 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6328 15. Added -ti meaning -t -i.
6330 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6331 domains, and warn by logging if others are found.
6333 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6334 returned from GIFCONF when Exim is trying to find the list of interfaces on
6335 a host. The code in os.c has been modified to copy each ifreq to an aligned
6336 structure in all cases.
6338 Also, in some cases, the returned ifreq's were being copied to a 'struct
6339 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6340 means the last couple of bytes of an IPv6 address could be chopped if the
6341 ifreq contained only a normal sockaddr (14 bytes storage).
6343 18. Named domain lists were not supported in the hosts_treat_as_local option.
6344 An entry such as +xxxx was not recognized, and was treated as a literal
6347 19. Ensure that header lines added by a DATA ACL are included in the reject log
6348 if the ACL subsequently rejects the message.
6350 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6351 MD5 (which is deprecated).
6353 21. When testing a filter file using -bf, Exim was writing a message when it
6354 took the sender from a "From " line in the message, but it was not doing so
6355 when it took $return_path from a Return-Path: header line. It now does.
6357 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6358 with a valid header line field name (a series of printing characters
6359 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6361 23. Changed "disc" in the source to "disk" to conform to the documentation and
6362 the book and for uniformity.
6364 24. Ignore Sendmail's -Ooption=value command line item.
6366 25. When execve() failed while trying to run a command in a pipe transport,
6367 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6368 could be confused with a return value of 69 from the command itself. This
6369 has been changed to 127, the value the shell returns if it is asked to run
6370 a non-existent command. The wording for the related log line suggests a
6371 non-existent command as the problem.
6373 26. If received_header_text expands to an empty string, do not add a Received:
6374 header line to the message. (Well, it adds a token one on the spool, but
6375 marks it "old" so that it doesn't get used or transmitted.)
6377 27. Installed eximstats 1.28 (addition of -nt option).
6379 28. There was no check for failure on the call to getsockname() in the daemon
6380 code. This can fail if there is a shortage of resources on the system, with
6381 ENOMEM, for example. A temporary error is now given on failure.
6383 29. Contrary to the C standard, it seems that in some environments, the
6384 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6385 program. Exim now does this explicitly; it affects the formatting of
6386 timestamps using strftime().
6388 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6389 complaints. I've now initialized all the variables, to avoid this.
6391 32. Header lines added by a system filter were not being "seen" during
6392 transport-time rewrites.
6394 33. The info_callback() function passed to OpenSSL is set up with type void
6395 (*)(SSL *, int, int), as described somewhere. However, when calling the
6396 function (actually a macro) that sets it up, the type void(*)() is
6397 expected. I've put in a cast to prevent warnings from picky compilers.
6399 34. If a DNS black list lookup found a CNAME record, but there were no A
6400 records associated with the domain it pointed at, Exim crashed.
6402 35. If a DNS black list lookup returned more than one A record, Exim ignored
6403 all but the first. It now scans all returned addresses if a particular IP
6404 value is being sought. In this situation, the contents of the
6405 $dnslist_value variable are a list of all the addresses, separated by a
6408 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6409 to accept a host name and all its aliases if the forward lookup for any of
6410 them yielded the IP address of the incoming connection. Now it accepts only
6411 those names whose forward lookup yields the correct IP address. Any other
6412 names are discarded. This closes a loophole whereby a rogue DNS
6413 administrator could create reverse DNS records to break through a
6414 wildcarded host restriction in an ACL.
6416 37. If a user filter or a system filter that ran in a subprocess used any of
6417 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6418 the wrong values were passed to the pipe command ($thisaddress had the
6419 value of $0, $0 had the value of $1, etc). This bug was introduced by
6420 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6422 38. Improved the line breaking for long SMTP error messages from ACLs.
6423 Previously, if there was no break point between 40 and 75 characters, Exim
6424 left the rest of the message alone. Two changes have been made: (a) I've
6425 reduced the minimum length to 35 characters; (b) if it can't find a break
6426 point between 35 and 75 characters, it looks ahead and uses the first one
6427 that it finds. This may give the occasional overlong line, but at least the
6428 remaining text gets split now.
6430 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6431 file descriptors might be low, and that setting 1000 would always raise it.
6432 It turns out that in some environments, the limit is already over 1000 and
6433 that lowering it causes trouble. So now Exim takes care not to decrease it.
6435 40. When delivering a message, the value of $return_path is set to $sender_
6436 address at the start of routing (routers may change the value). By an
6437 oversight, this default was not being set up when an address was tested by
6438 -bt or -bv, which affected the outcome if any router or filter referred to
6441 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6442 log only when "message" or "log_message" are set. However, if one of the
6443 conditions was an address verification, or a call to a nested ACL, the
6444 messages generated by the underlying test were being passed through. This
6445 no longer happens. The underlying message is available in $acl_verify_
6446 message for both "message" and "log_message" expansions, so it can be
6447 passed through if needed.
6449 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6450 new expansion $bh_ to give the encoded byte string without charset
6451 translation. Translation happens only if iconv() is available; HAVE_ICONV
6452 indicates this at build time. HEADERS_CHARSET gives the charset to
6453 translate to; headers_charset can change it in the configuration, and
6454 "headers charset" can change it in an individual filter file.
6456 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6457 that creates RFC 2047 encoded "words" labels them as that charset instead
6458 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6459 expansion operator; (ii) when Exim creates a From: line for a local
6460 message; (iii) when a header line is rewritten to include a "phrase" part.
6462 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6463 buggy, causing it to skip the first lines of messages whose message ID
6464 ended in 'D'. This would not have bitten before Exim release 4.14, because
6465 message IDs were unlikely to end in 'D' before then. The effect was to have
6466 incorrect size information for certain domains.
6468 45. #include "config.h" was missing at the start of the crypt16.c module. This
6469 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6472 46. If there was a timeout during a "random" callout check, Exim treated it as
6473 a failure of the random address, and carried on sending RSET and the real
6474 address. If the delay was just some slowness somewhere, the response to the
6475 original RCPT would be taken as a response to RSET and so on, causing
6476 mayhem of various kinds.
6478 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6479 when I implemented it. It didn't allow for the fact that some option values
6480 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6481 the right test for positive values.
6483 48. Domain names in DNS records are case-independent. Exim always looks them up
6484 in lower case. Some resolvers return domain names in exactly the case they
6485 appear in the zone file, that is, they may contain uppercase letters. Not
6486 all resolvers do this - some return always lower case. Exim was treating a
6487 change of case by a resolver as a change of domain, similar to a widening
6488 of a domain abbreviation. This triggered its re-routing code and so it was
6489 trying to route what was effectively the same domain again. This normally
6490 caused routing to fail (because the router wouldn't handle the domain
6491 twice). Now Exim checks for this case specially, and just changes the
6492 casing of the domain that it ultimately uses when it transmits the message
6495 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6498 50. If a filter generated a file delivery with a non-absolute name (possible if
6499 no home directory exists for the router), the forbid_file option was not
6502 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6503 the existing equality matching.
6505 52. Exim was using ints instead of ino_t variables in some places where it was
6506 dealing with inode numbers.
6508 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6509 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6510 called TMPDIR, and if it finds it is different, it changes its value.
6512 54. The smtp_printf() function is now made available to local_scan() so
6513 additional output lines can be written before returning. There is also an
6514 smtp_fflush() function to enable the detection of a dropped connection.
6515 The variables smtp_input and smtp_batched_input are exported to
6518 55. Changed the default runtime configuration: the message "Unknown user"
6519 has been removed from the ACL, and instead placed on the localuser router,
6520 using the cannot_route_message feature. This means that any verification
6521 failures that generate their own messages won't get overridden. Similarly,
6522 the "Unrouteable address" message that was in the ACL for unverifiable
6523 relay addresses has also been removed.
6525 56. Added hosts_avoid_esmtp to the smtp transport.
6527 57. The exicyclog script was not checking for the esoteric option
6528 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6529 will work only if exicyclog is run under the appropriate euid.
6531 58. Following a discussion on the list, the rules by which Exim recognises line
6532 endings on incoming messages have been changed. The -dropcr and drop_cr
6533 options are now no-ops, retained only for backwards compatibility. The
6534 following line terminators are recognized: LF CRLF CR. However, special
6535 processing applies to CR:
6537 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6538 nor a local message in the state where . is a terminator.
6540 (ii) If a bare CR is encountered in a header line, an extra space is added
6541 after the line terminator so as not to end the header. The reasoning
6542 behind this is that bare CRs in header lines are most likely either
6543 to be mistakes, or people trying to play silly games.
6545 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6546 was being incorrectly given as 18 bytes larger than it should have been.
6547 This is a VOB (very old bug).
6549 60. This may never have affected anything current, but just in case it has:
6550 When the local host is found other than at the start of a list of hosts,
6551 the local host, those with the same MX, and any that follow, are discarded.
6552 When the list in question was part of a longer list of hosts, the following
6553 hosts (not currently being processed) were also being discarded. This no
6554 longer happens. I'm not sure if this situation could ever has previously
6557 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6560 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6561 header. This is something that is recommended in a new Internet Draft, and
6562 is something that is documented as being done by Sendmail. There are two
6563 possible values. For messages generated by the autoreply transport, Exim
6566 Auto-Submitted: auto-replied
6568 whereas for all other generated messages (e.g. bounces) it adds
6570 Auto-Submitted: auto-generated
6572 63. The "personal" condition in filters now includes a test for the
6573 Auto-Submitted: header. If it contains the string "auto-" the message it
6574 not considered personal.
6576 64. Added rcpt_include_affixes as a generic transport option.
6578 65. Added queue_only_override (default true).
6580 66. Added the syslog_duplication option.
6582 67. If what should have been the first header line of a message consisted of
6583 a space followed by a colon, Exim was mis-interpreting it as a header line.
6584 It isn't of course - it is syntactically invalid and should therefore be
6585 treated as the start of the message body. The misbehaviour could have
6586 caused a number of strange effects, including loss of data in subsequent
6587 header lines, and spool format errors.
6589 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6590 client host had authenticated. This control can now be exercised by an ACL
6591 for more flexibility.
6593 69. By default, callouts do not happen when testing with -bh. There is now a
6594 variant, -bhc, which does actually run the callout code, including
6595 consulting and updating the callout cache.
6597 70. Added support for saslauthd authentication, courtesy of Alexander
6600 71. If statvfs() failed on the spool or log directories while checking their
6601 size for availability, Exim confusingly gave the error "space shortage".
6602 Furthermore, in debugging mode it crashed with a floating point exception.
6603 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6604 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6605 serious problem, Exim now writes to the main and panic logs when this
6606 happens, with details of the failure. It then refuses to accept the
6607 incoming message, giving the message "spool directory problem" or "log
6608 directory problem" with a 421 code for SMTP messages.
6610 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6611 0, 1, and 2 exist, because some OS complain for execs without them (see
6612 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6613 descriptors. However, the code omitted to check that the open succeeded,
6614 causing mysterious errors if for some reason the permissions on /dev/null
6615 got screwed. Now Exim writes a message to the main and panic logs, and
6616 bombs out if it can't open /dev/null.
6618 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6619 interact so that it is all more flexible. It is supposed to remain
6620 backwards compatible. Also added extra_local_interfaces.
6622 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6623 to bomb out with an assertion failure - to the client this appears as a
6624 connection drop. This problem occurs in the part of the code that was taken
6625 from the Samba project. Fortunately, the assertion is in a very simple
6626 function, so I have fixed this by reproducing the function inline in the
6627 one place where it is called, and arranging for authentication to fail
6628 instead of killing the process with assert().
6630 75. The SPA client code was not working when the server requested OEM rather
6631 than Unicode encoding.
6633 76. Added code to make require_files with a specific uid setting more usable in
6634 the case where statting the file as root fails - usually a non-root-mounted
6635 NFS file system. When this happens and the failure is EACCES, Exim now
6636 forks a subprocess and does the per-uid checking as the relevant uid.
6638 77. Added process_log_path.
6640 78. If log_file_path was not explicitly set, a setting of check_log_space or
6641 check_log_inodes was ignored.
6643 79. If a space check for the spool or log partitions fails, the incident is now
6644 logged. Of course, in the latter case the data may get lost...
6646 80. Added the %p formatting code to string_format() so that it can be used to
6647 print addresses in debug_print(). Adjusted all the address printing in the
6648 debugging in store.c to use %p rather than %d.
6650 81. There was a concern that a line of code in smtp_in.c could overflow a
6651 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6652 initially expressed, the concern was not well-founded, because trailing
6653 spaces are removed early. However, if the trailing spaces were followed by
6654 a NULL, they did not get removed, so the overflow was possible. Two fixes
6657 (a) I re-wrote the offending code in a cleaner fashion.
6658 (b) If an incoming SMTP command contains a NULL character, it is rejected
6661 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6662 runs initgroups(), so that if the Exim user is in any additional groups,
6663 they will be used during message reception.
6669 The change log for 4.20 and earlier releases has been archived.