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.
23 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
24 supports it and a size is available (ie. the sending peer gave us one).
26 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
27 years of being deprecated). The replacements are the ACLs
28 acl_smtp_mime and acl_not_smtp_mime.
30 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
31 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
32 or even an in-clear connection were permitted. Now, if the host lookup
33 was dnssec and dane was requested then the host is only used if the
34 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
35 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
36 if one fails this test.
37 This means that a poorly-configured remote DNS will make it incommunicado;
38 but it protects against a DNS-interception attack on it.
40 JH/04 Bug 1810: make continued-use of an open smtp transport connection
41 non-noisy when a race steals the message being considered.
43 JH/05 If main configuration option tls_certificate is unset, generate a
44 self-signed certificate for inbound TLS connections.
46 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
47 in rewrites and routers.
49 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
50 and logged a warning sing 4.83; now they are a configuration file error.
52 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
53 (lacking @domain). Apply the same qualification processing as RCPT.
55 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
57 JH/10 Support ${sha256:} applied to a string (as well as the previous
60 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
61 a cutthrough deliver is pending, as we always want to make a connection.
62 This also avoids re-routing the message when later placing the cutthrough
63 connection after a verify cache hit.
64 Do not update it with the verify result either.
66 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
67 when routing results in more than one destination address.
69 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
70 signing (which inhibits the cutthrough capability). Previously only
71 the presence of an option was tested; now an expansion evaluating as
72 empty is permissible (obviously it should depend only on data available
73 when the cutthrough connection is made).
75 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
76 the relevant preceding SMTP command did not note the pipelining mode.
78 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
79 Previously they were not counted.
81 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
82 as one having no matching records. Previously we deferred the message
83 that needed the lookup.
85 JH/17 Fakereject: previously logged as a norml message arrival "<="; now
86 distinguished as "(=".
88 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
89 for missing MX records. Previously it only worked for missing A records.
91 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
93 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
94 after the data-go-ahead and data-ack. Patch from Jason Betts.
96 JH/21 Bug 1846: Send DMARC forensic reports for reject and quaratine results,
97 even for a "none" policy. Patch from Tony Meyer.
99 JH/22 Fix continued use of a connection for further deliveries. If a port was
100 specified by a router, it must also match for the delivery to be
103 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
104 When one of the recipients of a message was unsuitable for the connection
105 (has no matching addresses), we lost track of needing to mark it
106 deferred. As a result mail would be lost.
108 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
110 JH/25 Decoding ACL controls is now done using a binary search; the source code
111 takes up less space and should be simpler to maintain. Merge the ACL
112 condition decode tables also, with similar effect.
114 JH/26 Fix problem with one_time used on a redirect router which returned the
115 parent address unchanged. A retry would see the parent address marked as
116 delivered, so not attempt the (identical) child. As a result mail would
119 JH/27 Fix a possible security hole, wherein a process operating with the Exim
120 UID can gain a root shell. Credit to http://www.halfdog.net/ for
121 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
124 JH/28 Enable {spool,log} filesystem space and inode checks as default.
125 Main config options check_{log,spool}_{inodes,space} are now
126 100 inodes, 10MB unless set otherwise in the configuration.
128 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
129 Previously it only applied to the main-section connection policy
132 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
134 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
135 by me. Added RFC7919 DH primes as an alternative.
137 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
138 are not in the system include path.
140 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
141 GnuTLS, when a session startup failed (eg because the client disconnected)
142 Exim did stdio operations after fclose. This was exposed by a recent
143 change which nulled out the file handle after the fclose.
145 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
146 signed directly by the cert-signing cert, rather than an intermediate
147 OCSP-signing cert. This is the model used by LetsEncrypt.
149 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
151 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
152 an incoming connection.
154 HS/02 Bug 1802: Do not half-close the connection after sending a request
157 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
158 fallback to "prime256v1".
160 JH/34 SECURITY: Use proper copy of DATA command in error message.
161 Could leak key material. Remotely exploitable. CVE-2016-9963.
167 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
168 and 3.4.4 - once the server is enabled to respond to an OCSP request
169 it does even when not requested, resulting in a stapling non-aware
170 client dropping the TLS connection.
172 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
173 support variable-length bit vectors. No functional change.
175 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
176 The I= interface field on outgoing lines is now after the H= remote
177 host field, same as incoming lines. There is a separate
178 outgoing_interface log selector which allows you to disable the
181 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
182 If not running log_selector +smtp_connection the mainlog would be held
183 open indefinitely after a "too many connections" event, including to a
184 deleted file after a log rotate. Leave the per net connection logging
185 leaving it open for efficiency as that will be quickly detected by the
186 check on the next write.
188 HS/01 Bug 1671: Fix post transport crash.
189 Processing the wait-<transport> messages could crash the delivery
190 process if the message IDs didn't exist for some reason. When
191 using 'split_spool_directory=yes' the construction of the spool
192 file name failed already, exposing the same netto behaviour.
194 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
195 mime_regex ACL conditions.
197 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
198 to DSN fail messages (bounces): remote IP, remote greeting, remote response
199 to HELO, local diagnostic string.
201 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
202 log line to debug. Even when configured with a tls authenticator many
203 client connections are expected to not authenticate in this way, so
204 an authenticate fail is not an error.
206 HS/02 Add the Exim version string to the process info. This way exiwhat
207 gives some more detail about the running daemon.
209 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
210 matter for fast-change records such as DNSBLs.
212 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
213 retry record, even if constant. There may be multiple transports with
214 different interface settings and the retry behaviour needs to be kept
217 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
219 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
221 JH/10 Bug 840: fix log_defer_output option of pipe transport
223 JH/11 Bug 830: use same host for all RCPTS of a message, even under
224 hosts_randomize. This matters a lot when combined with mua_wrapper.
226 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
227 ${quote_pgsql:<string>} operator.
229 JH/13 Bug 1708: avoid misaligned access in cached lookup.
231 JH/14 Change header file name for freeradius-client. Relevant if compiling
232 with Radius support; from the Gentoo tree and checked under Fedora.
234 JH/15 Bug 1712: Introduce $prdr_requested flag variable
236 JH/16 Bug 1714: Permit an empty string as expansion result for transport
237 option transport_filter, meaning no filtering.
239 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
241 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
242 defaults to "*" (all hosts). The variable is now available when not built
243 with TLS, default unset, mainly to enable keeping the testuite sane.
244 If a server certificate is not supplied (via tls_certificate) an error is
245 logged, and clients will find TLS connections fail on startup. Presumably
246 they will retry in-clear.
247 Packagers of Exim are strongly encouraged to create a server certificate
248 at installation time.
250 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
251 with the $config_file variable.
253 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
254 in transport context, after the attempt, and per-recipient. The latter type
255 is per host attempted. The event data is the error message, and the errno
256 information encodes the lookup type (A vs. MX) used for the (first) host,
257 and the trailing two digits of the smtp 4xx reponse.
259 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
260 to write to mainlog (or rejectlog, paniclog) in the window between file
261 creation and permissions/ownership being changed. Particularly affects
262 installations where exicyclog is run as root, rather than exim user;
263 result is that the running daemon panics and dies.
265 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
267 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
268 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
269 "pri" and "weight". Note that the previous implicit priority given by the
270 list order is no longer honoured.
272 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
275 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
276 by defining SUPPORT_SOCKS.
278 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
279 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
280 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
281 variables are renamed to proxy_{local,external}_{address,port}.
283 JH/27 Move Internationalisation support from Experimental to mainline, enabled
284 for a build by defining SUPPORT_I18N
286 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
287 of the query string, and make ${quote_redis:} do that quoting.
289 JH/29 Move Events support from Experimental to mainline, enabled by default
290 and removable for a build by defining DISABLE_EVENT.
292 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
294 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
295 cached by the daemon.
297 JH/32 Move Redis support from Experimental to mainline, enabled for a build
298 by defining LOOKUP_REDIS. The libhiredis library is required.
300 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
301 keys are given for lookup.
303 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
304 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
305 only supported when built with TLS support. The PolarSSL SHA routines
306 are still used when the TLS library is too old for convenient support.
308 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
309 openssl_options), for security. OpenSSL forces this from version 1.1.0
310 server-side so match that on older versions.
312 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
313 allocation for $value could be released as the expansion processing
314 concluded, but leaving the global pointer active for it.
316 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
317 and to use the domains and local_parts ACL conditions.
319 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
320 incorrectly not doubled on cutthrough transmission, hence seen as a
321 body-termination at the receiving system - resulting in truncated mails.
322 Commonly the sender saw a TCP-level error, and retransmitted the message
323 via the normal store-and-forward channel. This could result in duplicates
324 received - but deduplicating mailstores were liable to retain only the
325 initial truncated version.
327 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
329 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
331 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
332 we're in there, support oversigning also; bug 1309.
334 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
336 HS/04 Add support for keep_environment and add_environment options.
338 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
339 either intentional arithmetic overflow during PRNG, or testing config-
342 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
343 delivery resulted in actual delivery. Cancel cutthrough before DATA
346 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
347 rejects a recipient: pass the reject to the originator.
349 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
350 Many were false-positives and ignorable, but it's worth fixing the
353 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
354 for the new environment-manipulation done at startup. Move the routines
355 from being local to tls.c to being global via the os.c file.
357 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
358 an extract embedded as result-arg for a map, the first arg for extract
359 is unavailable so we cannot tell if this is a numbered or keyed
360 extraction. Accept either.
366 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
369 JH/02 The smtp transport option "multi_domain" is now expanded.
371 JH/03 The smtp transport now requests PRDR by default, if the server offers
374 JH/04 Certificate name checking on server certificates, when exim is a client,
375 is now done by default. The transport option tls_verify_cert_hostnames
376 can be used to disable this per-host. The build option
377 EXPERIMENTAL_CERTNAMES is withdrawn.
379 JH/05 The value of the tls_verify_certificates smtp transport and main options
380 default to the word "system" to access the system default CA bundle.
381 For GnuTLS, only version 3.0.20 or later.
383 JH/06 Verification of the server certificate for a TLS connection is now tried
384 (but not required) by default. The verification status is now logged by
385 default, for both outbound TLS and client-certificate supplying inbound
388 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
391 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
392 Status Notification (bounce) messages are now MIME format per RFC 3464.
393 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
394 under the control of the dsn_advertise_hosts option, and routers may
395 have a dsn_lasthop option.
397 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
398 default, modifiable by a malware= option. The list separator for
399 the options can now be changed in the usual way. Bug 68.
401 JH/10 The smtp_receive_timeout main option is now expanded before use.
403 JH/11 The incoming_interface log option now also enables logging of the
404 local interface on delivery outgoing connections.
406 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
407 if the interface and destination host and port all match.
409 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
412 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
413 Patch from Andrew Lewis.
415 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
416 now supports optional time-restrictions, weighting, and priority
417 modifiers per server. Patch originally by <rommer@active.by>.
419 JH/16 The spamd_address main option now supports a mixed list of local
420 and remote servers. Remote servers can be IPv6 addresses, and
421 specify a port-range.
423 JH/17 Bug 68: The spamd_address main option now supports an optional
424 timeout value per server.
426 JH/18 Bug 1581: Router and transport options headers_add/remove can
427 now have the list separator specified.
429 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
432 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
435 JH/21 Support for the A6 type of dns record is withdrawn.
437 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
438 rather than the verbs used.
440 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
441 from 255 to 1024 chars.
443 JH/24 Verification callouts now attempt to use TLS by default.
445 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
446 are generic router options now. The defaults didn't change.
448 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
449 Original patch from Alexander Shikoff, worked over by JH.
451 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
452 files if ZCAT_COMMAND is not executable.
454 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
456 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
458 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
459 Normally benign, it bites when the pair was led to by a CNAME;
460 modern usage is to not canoicalize the domain to a CNAME target
461 (and we were inconsistent anyway for A-only vs AAAA+A).
463 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
465 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
466 when evaluating $sender_host_dnssec.
468 JH/31 Check the HELO verification lookup for DNSSEC, adding new
469 $sender_helo_dnssec variable.
471 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
473 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
475 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
477 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
478 documented as working, but never had. Support all but $spam_report.
480 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
481 added for tls authenticator.
483 HS/03 Add perl_taintmode main config option
489 TL/01 When running the test suite, the README says that variables such as
490 no_msglog_check are global and can be placed anywhere in a specific
491 test's script, however it was observed that placement needed to be near
492 the beginning for it to behave that way. Changed the runtest perl
493 script to read through the entire script once to detect and set these
494 variables, reset to the beginning of the script, and then run through
495 the script parsing/test process like normal.
497 TL/02 The BSD's have an arc4random API. One of the functions to induce
498 adding randomness was arc4random_stir(), but it has been removed in
499 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
500 function when detected.
502 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
503 cause callback expansion.
505 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
506 syntax errors in an expansion can be treated as a string instead of
507 logging or causing an error, due to the internal use of bool_lax
508 instead of bool when processing it.
510 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
511 server certificates when making smtp deliveries.
513 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
515 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
517 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
519 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
520 Merged patch from Sebastian Wiedenroth.
522 JH/05 Fix results-pipe from transport process. Several recipients, combined
523 with certificate use, exposed issues where response data items split
524 over buffer boundaries were not parsed properly. This eventually
525 resulted in duplicates being sent. This issue only became common enough
526 to notice due to the introduction of conection certificate information,
527 the item size being so much larger. Found and fixed by Wolfgang Breyha.
529 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
530 size buffer was used, resulting in syntax errors when an expansion
533 JH/07 Add support for directories of certificates when compiled with a GnuTLS
534 version 3.3.6 or later.
536 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
537 is EXPERIMENTAL_EVENT, the main-configuration and transport options
538 both become "event_action", the variables become $event_name, $event_data
539 and $event_defer_errno. There is a new variable $verify_mode, usable in
540 routers, transports and related events. The tls:cert event is now also
541 raised for inbound connections, if the main configuration event_action
544 TL/06 In test suite, disable OCSP for old versions of openssl which contained
545 early OCSP support, but no stapling (appears to be less than 1.0.0).
547 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
548 server certificate names available under the smtp transport option
549 "tls_verify_cert_hostname" now do not permit multi-component wildcard
552 JH/10 Time-related extraction expansions from certificates now use the main
553 option "timezone" setting for output formatting, and are consistent
554 between OpenSSL and GnuTLS compilations. Bug 1541.
556 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
557 encoded parameter in the incoming message. Bug 1558.
559 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
560 include certificate info, eximon was claiming there were spoolfile
563 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
565 JH/14 Log delivery-related information more consistently, using the sequence
566 "H=<name> [<ip>]" wherever possible.
568 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
569 are problematic for Debian distribution, omit them from the release
572 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
574 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
577 JH/17 Fix a null-indirection in certextract expansions when a nondefault
578 output list separator was used.
583 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
584 checkers that were complaining about end of non-void function with no
587 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
588 This was a regression intruduced in 4.83 by another bugfix.
590 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
592 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
593 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
599 TF/01 Correctly close the server side of TLS when forking for delivery.
601 When a message was received over SMTP with TLS, Exim failed to clear up
602 the incoming connection properly after forking off the child process to
603 deliver the message. In some situations the subsequent outgoing
604 delivery connection happened to have the same fd number as the incoming
605 connection previously had. Exim would try to use TLS and fail, logging
606 a "Bad file descriptor" error.
608 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
609 utilities have not been installed.
611 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
612 temporary space as the ACL may create new global variables.
614 TL/01 LDAP support uses per connection or global context settings, depending
615 upon the detected version of the libraries at build time.
617 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
618 to extract and use the src ip:port in logging and expansions as if it
619 were a direct connection from the outside internet. PPv2 support was
620 updated based on HAProxy spec change in May 2014.
622 JH/02 Add ${listextract {number}{list}{success}{fail}}.
624 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
625 Properly escape header and check for NULL return.
627 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
630 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
632 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
633 characters in header names, implemented as a verify condition.
634 Contributed by Michael Fischer v. Mollard.
636 TL/05 Rename SPF condition results err_perm and err_temp to standardized
637 results permerror and temperror. Previous values are deprecated but
638 still accepted. In a future release, err_perm and err_temp will be
639 completely removed, which will be a backward incompatibility if the
640 ACL tests for either of these two old results. Patch contributed by
641 user bes-internal on the mailing list.
643 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
645 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
646 selectors, in both main and reject logs.
648 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
651 JH/07 Add malware type "sock" for talking to simple daemon.
653 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
655 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
656 routers/transports under cutthrough routing.
658 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
659 numbers. Touch up "bool" conditional to keep the same definition.
661 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
663 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
665 JH/12 Expand items in router/transport headers_add or headers_remove lists
666 individually rather than the list as a whole. Bug 1452.
668 Required for reasonable handling of multiple headers_ options when
669 they may be empty; requires that headers_remove items with embedded
670 colons must have them doubled (or the list-separator changed).
672 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
673 view the policy declared in the DMARC record. Currently, $dmarc_status
674 is a combined value of both the record presence and the result of the
677 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
679 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
680 dnslookup router and the smtp transport (applying to the forward
683 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
684 of ldap servers used for a specific lookup. Patch provided by Heiko
687 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
688 New variable $lookup_dnssec_authenticated for observability.
690 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
691 Patch submitted by Lars Timman.
693 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
695 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
696 Requires trusted mode and valid format message id, aborts otherwise.
697 Patch contributed by Heiko Schlichting.
699 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
700 certextract with support for various fields. Bug 1358.
702 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
703 is requested by default, modifiable by smtp transport option
706 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
707 operate on certificate variables to give certificate fingerprints
708 Also new ${sha256:cert_variable}.
710 JH/23 The PRDR feature is moved from being Experimental into the mainline.
712 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
713 Christian Aistleitner.
715 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
717 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
718 file. Patch from Wolfgang Breyha.
720 JH/25 Expand the coverage of the delivery $host and $host_address to
721 client authenticators run in verify callout. Bug 1476.
723 JH/26 Port service names are now accepted for tls_on_connect_ports, to
724 align with daemon_smtp_ports. Bug 72.
726 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
727 support and error reporting did not work properly.
729 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
730 and is readable. Patch from Andrew Colin Kissa.
732 TL/14 Enhance documentation of ${run expansion and how it parses the
733 commandline after expansion, particularly in the case when an
734 unquoted variable expansion results in an empty value.
736 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
738 PP/02 Fix internal collision of T_APL on systems which support RFC3123
739 by renaming away from it. Addresses GH issue 15, reported by
742 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
744 TL/15 SECURITY: prevent double expansion in math comparison functions
745 (can expand unsanitized data). Not remotely exploitable.
752 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
754 PP/02 Make -n do something, by making it not do something.
755 When combined with -bP, the name of an option is not output.
757 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
760 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
761 $sender_host_name and config options to manage this, and basic check
764 PP/05 DSCP support for outbound connections and control modifier for inbound.
766 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
767 (Only plugin which currently uses this is kerberos4, which nobody should
768 be using, but we should make it available and other future plugins might
769 conceivably use it, even though it would break NAT; stuff *should* be
770 using channel bindings instead).
772 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
773 name; added for Sendmail compatibility; requires admin caller.
774 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
775 just ignore it); requires trusted caller.
776 Also parse but ignore: -Ac -Am -X<logfile>
779 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
781 TL/02 Add +smtp_confirmation as a default logging option.
783 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
784 Patch by Magnus Holmgren from 2007-02-20.
786 TL/04 Bugzilla 1281 - Spec typo.
787 Bugzilla 1283 - Spec typo.
788 Bugzilla 1290 - Spec grammar fixes.
790 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
792 TL/06 Add Experimental DMARC support using libopendmarc libraries.
794 TL/07 Fix an out of order global option causing a segfault. Reported to dev
795 mailing list by by Dmitry Isaikin.
797 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
799 JH/02 Support "G" suffix to numbers in ${if comparisons.
801 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
803 NM/01 Bugzilla 1197 - Spec typo
804 Bugzilla 1196 - Spec examples corrections
806 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
808 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
809 gnutls_enable_pkcs11, but renamed to more accurately indicate its
812 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
813 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
815 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
816 "acl {{name}{arg}...}", and optional args on acl condition
819 JH/05 Permit multiple router/transport headers_add/remove lines.
821 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
823 JH/07 Avoid using a waiting database for a single-message-only transport.
824 Performance patch from Paul Fisher. Bugzilla 1262.
826 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
829 JH/09 Add $headers_added variable, with content from use of ACL modifier
830 add_header (but not yet added to the message). Bugzilla 199.
832 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
833 Pulled from Bugzilla 817 by Wolfgang Breyha.
835 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
837 (nb: this is the same fix as in Exim 4.80.1)
839 JH/11 Add A= logging on delivery lines, and a client_set_id option on
842 JH/12 Add optional authenticated_sender logging to A= and a log_selector
845 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
847 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
848 advertise SMTP AUTH mechanism to us, instead of a generic
849 protocol violation error. Also, make Exim more robust to bad
850 data from the Dovecot auth socket.
852 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
854 When a queue runner is handling a message, Exim first routes the
855 recipient addresses, during which it prunes them based on the retry
856 hints database. After that it attempts to deliver the message to
857 any remaining recipients. It then updates the hints database using
860 So if a recipient address works intermittently, it can get repeatedly
861 deferred at routing time. The retry hints record remains fresh so the
862 address never reaches the final cutoff time.
864 This is a fairly common occurrence when a user is bumping up against
865 their storage quota. Exim had some logic in its local delivery code
866 to deal with this. However it did not apply to per-recipient defers
867 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
869 This change adds a proper retry rule check during routing so that the
870 final cutoff time is checked against the message's age. We only do
871 this check if there is an address retry record and there is not a
872 domain retry record; this implies that previous attempts to handle
873 the address had the retry_use_local_parts option turned on. We use
874 this as an approximation for the destination being like a local
875 delivery, as in LMTP.
877 I suspect this new check makes the old local delivery cutoff check
878 redundant, but I have not verified this so I left the code in place.
880 TF/02 Correct gecos expansion when From: is a prefix of the username.
882 Test 0254 submits a message to Exim with the header
886 When I ran the test suite under the user fanf2, Exim expanded
887 the header to contain my full name, whereas it should have added
888 a Resent-Sender: header. It erroneously treats any prefix of the
889 username as equal to the username.
891 This change corrects that bug.
893 GF/01 DCC debug and logging tidyup
894 Error conditions log to paniclog rather than rejectlog.
895 Debug lines prefixed by "DCC: " to remove any ambiguity.
897 TF/03 Avoid unnecessary rebuilds of lookup-related code.
899 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
900 Bug spotted by Jeremy Harris; was flawed since initial commit.
901 Would have resulted in OCSP responses post-SNI triggering an Exim
902 NULL dereference and crash.
904 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
906 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
907 Bug detection, analysis and fix by Samuel Thibault.
908 Bugzilla 1331, Debian bug #698092.
910 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
912 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
913 Server implementation by Todd Lyons, client by JH.
914 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
915 config variable "prdr_enable" controls whether the server
916 advertises the facility. If the client requests PRDR a new
917 acl_data_smtp_prdr ACL is called once for each recipient, after
918 the body content is received and before the acl_smtp_data ACL.
919 The client is controlled by bolth of: a hosts_try_prdr option
920 on the smtp transport, and the server advertisement.
921 Default client logging of deliveries and rejections involving
922 PRDR are flagged with the string "PRDR".
924 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
925 fclose(). Diagnosis by Todd Lyons.
927 PP/17 Update configure.default to handle IPv6 localhost better.
928 Patch by Alain Williams (plus minor tweaks).
931 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
932 This is now consistent with GnuTLS, and is now documented: the
933 previous undocumented portable approach to treating the option as
934 unset was to force an expansion failure. That still works, and
935 an empty string is now equivalent.
937 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
938 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
939 not performing validation itself.
941 PP/20 Added force_command boolean option to pipe transport.
942 Patch from Nick Koston, of cPanel Inc.
944 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
947 TF/04 Added udpsend ACL modifer and hexquote expansion operator
949 PP/21 Fix eximon continuous updating with timestamped log-files.
950 Broken in a format-string cleanup in 4.80, missed when I repaired the
951 other false fix of the same issue.
952 Report and fix from Heiko Schlichting.
955 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
956 Report from Prashanth Katuri.
958 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
959 It's SecureTransport, so affects any MacOS clients which use the
960 system-integrated TLS libraries, including email clients.
962 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
963 using a MIME ACL for non-SMTP local injection.
964 Report and assistance in diagnosis by Warren Baker.
966 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
968 JH/16 Fix comparisons for 64b. Bugzilla 1385.
970 TL/09 Add expansion variable $authenticated_fail_id to keep track of
971 last id that failed so it may be referenced in subsequent ACL's.
973 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
976 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
977 ldap library initialization, allowing self-signed CA's to be
978 used. Also properly sets require_cert option later in code by
979 using NULL (global ldap config) instead of ldap handle (per
980 session). Bug diagnosis and testing by alxgomz.
982 TL/12 Enhanced documentation in the ratelimit.pl script provided in
983 the src/util/ subdirectory.
985 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
986 renamed to Transport Post Delivery Action by Jeremy Harris, as
989 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
990 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
991 redis_servers = needs to be configured which will be used by the redis
992 lookup. Patch from Warren Baker, of The Packet Hub.
994 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
996 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
997 hostname or reverse DNS when processing a host list. Used suggestions
998 from multiple comments on this bug.
1000 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1002 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1003 Missed a few lines, added it to make the runtest require no keyboard
1006 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1007 contains upper case chars. Make router use caseful_local_part.
1009 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1010 support when GnuTLS has been built with p11-kit.
1016 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1018 This, or similar/improved, will also be change PP/11 of 4.82.
1024 PP/01 Handle short writes when writing local log-files.
1025 In practice, only affects FreeBSD (8 onwards).
1026 Bugzilla 1053, with thanks to Dmitry Isaikin.
1028 NM/01 Bugzilla 949 - Documentation tweak
1030 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1033 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1035 PP/02 Implemented gsasl authenticator.
1037 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1039 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1040 `pkg-config foo` for cflags/libs.
1042 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1043 with rest of GSASL and with heimdal_gssapi.
1045 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1046 `pkg-config foo` for cflags/libs for the TLS implementation.
1048 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1049 properties get this fed in as external SSF. A number of robustness
1050 and debugging improvements to the cyrus_sasl authenticator.
1052 PP/08 cyrus_sasl server now expands the server_realm option.
1054 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1055 Patch by Jeremy Harris.
1057 PP/10 Added dbmjz lookup type.
1059 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1061 PP/12 MAIL args handles TAB as well as SP, for better interop with
1062 non-compliant senders.
1063 Analysis and variant patch by Todd Lyons.
1065 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1066 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1067 Patch from Dirk Mueller <dmueller@suse.com>
1069 PP/13 tls_peerdn now print-escaped for spool files.
1070 Observed some $tls_peerdn in wild which contained \n, which resulted
1071 in spool file corruption.
1073 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1074 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1075 or write after TLS renegotiation, which otherwise led to messages
1078 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1079 as a tracking header (ie: a signed header comes before the signature).
1080 Patch from Wolfgang Breyha.
1082 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1083 comma-sep list; embedded commas doubled.
1085 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1087 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1089 Report and patch from Dmitry Banschikov.
1091 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
1092 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1093 failures. We appear to now support TLS1.1+ with Exim.
1095 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1096 lets Exim select keys and certificates based upon TLS SNI from client.
1097 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1098 before an outbound SMTP session. New log_selector, +tls_sni.
1100 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1101 NULL dereference. Report and patch from Alun Jones.
1103 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1104 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1105 Not seeing resolver debug output on NetBSD, but suspect this is a
1106 resolver implementation change.
1108 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1109 Left warnings. Added "eximon gdb" invocation mode.
1111 PP/21 Defaulting "accept_8bitmime" to true, not false.
1113 PP/22 Added -bw for inetd wait mode support.
1115 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1116 locate the relevant includes and libraries. Made this the default.
1118 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1119 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1121 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1122 This may cause build issues on older platforms.
1124 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1125 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1126 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1127 Added SNI support via GnuTLS too.
1128 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1130 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1132 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1133 Applied second patch from Janne, implementing suggestion to default
1134 multiple-strings-in-record handling to match SPF spec.
1136 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1138 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1139 read-only, out of scope).
1140 Patch from Wolfgang Breyha, report from Stuart Northfield.
1142 PP/29 Fix three issues highlighted by clang analyser static analysis.
1143 Only crash-plausible issue would require the Cambridge-specific
1144 iplookup router and a misconfiguration.
1145 Report from Marcin Mirosław.
1147 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1149 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1150 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1151 As part of this, removing so much warning spew let me fix some minor
1152 real issues in debug logging.
1154 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1155 assignment on my part. Fixed.
1157 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1158 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1159 Janne Snabb (who went above and beyond: thank you).
1161 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1162 string otherwise requires a connection and a bunch more work and it's
1163 relatively easy to get wrong. Should also expose TLS library linkage
1166 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1167 64-bit ${eval} (JH/03).
1169 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1170 GNU libc to support some of the 64-bit stuff, should not lead to
1171 conflicts. Defined before os.h is pulled in, so if a given platform
1172 needs to override this, it can.
1174 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1175 protection layer was required, which is not implemented.
1176 Bugzilla 1254, patch from Wolfgang Breyha.
1178 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1179 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1180 tls_dhparam take prime identifiers. Also unbreak combination of
1181 OpenSSL+DH_params+TLSSNI.
1183 PP/39 Disable SSLv2 by default in OpenSSL support.
1189 PP/01 Solaris build fix for Oracle's LDAP libraries.
1190 Bugzilla 1109, patch from Stephen Usher.
1192 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1194 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1197 TF/02 Fix a couple more cases where we did not log the error message
1198 when unlink() failed. See also change 4.74-TF/03.
1200 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1201 lock up or crash if it happened to be inside a call to libc when it
1202 got a SIGUSR1 from exiwhat.
1204 The SIGUSR1 handler appends the current process status to the process
1205 log which is later printed by exiwhat. It used to use the general
1206 purpose logging code to do this, but several functions it calls are
1207 not safe for signals.
1209 The new output code in the SIGUSR1 handler is specific to the process
1210 log, and simple enough that it's easy to inspect for signal safety.
1211 Removing some special cases also simplifies the general logging code.
1212 Removing the spurious timestamps from the process log simplifies
1215 TF/04 Improved ratelimit ACL condition.
1217 The /noupdate option has been deprecated in favour of /readonly which
1218 has clearer semantics. The /leaky, /strict, and /readonly update modes
1219 are mutually exclusive. The update mode is no longer included in the
1220 database key; it just determines when the database is updated. (This
1221 means that when you upgrde Exim will forget old rate measurements.)
1223 Exim now checks that the per_* options are used with an update mode that
1224 makes sense for the current ACL. For example, when Exim is processing a
1225 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1226 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1227 must specify per_mail/readonly. If you omit the update mode it defaults to
1228 /leaky where that makes sense (as before) or /readonly where required.
1230 The /noupdate option is now undocumented but still supported for
1231 backwards compatibility. It is equivalent to /readonly except that in
1232 ACLs where /readonly is required you may specify /leaky/noupdate or
1233 /strict/noupdate which are treated the same as /readonly.
1235 A useful new feature is the /count= option. This is a generalization
1236 of the per_byte option, so that you can measure the throughput of other
1237 aggregate values. For example, the per_byte option is now equivalent
1238 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1240 The per_rcpt option has been generalized using the /count= mechanism
1241 (though it's more complicated than the per_byte equivalence). When it is
1242 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1243 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1244 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1245 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1246 Note that using per_rcpt with a non-readonly update mode in more than
1247 one ACL will cause the recipients to be double-counted. (The per_mail
1248 and per_byte options don't have this problem.)
1250 The handling of very low rates has changed slightly. If the computed rate
1251 is less than the event's count (usually one) then this event is the first
1252 after a long gap. In this case the rate is set to the same as this event's
1253 count, so that the first message of a spam run is counted properly.
1255 The major new feature is a mechanism for counting the rate of unique
1256 events. The new per_addr option counts the number of different
1257 recipients that someone has sent messages to in the last time period. It
1258 behaves like per_rcpt if all the recipient addresses are different, but
1259 duplicate recipient addresses do not increase the measured rate. Like
1260 the /count= option this is a general mechanism, so the per_addr option
1261 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1262 example, measure the rate that a client uses different sender addresses
1263 with the options per_mail/unique=$sender_address. There are further
1264 details in the main documentation.
1266 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1268 TF/06 Removed a few PCRE remnants.
1270 TF/07 Automatically extract Exim's version number from tags in the git
1271 repository when doing development or release builds.
1273 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1274 Bugzilla 879. Patch from Paul Fisher.
1276 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1277 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1280 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1282 PP/05 Variable $av_failed, true if the AV scanner deferred.
1283 Bugzilla 1078. Patch from John Horne.
1285 PP/06 Stop make process more reliably on build failure.
1286 Bugzilla 1087. Patch from Heiko Schlittermann.
1288 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1289 Bugzilla 1089. Patch from Heiko Schlittermann.
1291 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1292 Bugzilla 1131. Patch from Holger Weiß.
1294 PP/09 Handle IPv6 addresses with SPF.
1295 Bugzilla 860. Patch from Wolfgang Breyha.
1297 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1299 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1302 PP/11 match_* no longer expand right-hand-side by default.
1303 New compile-time build option, EXPAND_LISTMATCH_RHS.
1304 New expansion conditions, "inlist", "inlisti".
1306 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1308 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1310 PP/14 fix log_write() format string regression from TF/03.
1311 Bugzilla 1152. Patch from Dmitry Isaikin.
1317 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1319 PP/02 Harmonised TLS library version reporting; only show if debugging.
1320 Layout now matches that introduced for other libraries in 4.74 PP/03.
1322 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1324 PP/04 New "dns_use_edns0" global option.
1326 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1329 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1330 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1332 TK/01 Updated PolarSSL code to 0.14.2.
1333 Bugzilla 1097. Patch from Andreas Metzler.
1335 PP/07 Catch divide-by-zero in ${eval:...}.
1336 Fixes bugzilla 1102.
1338 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1341 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1342 format-string attack -- SECURITY: remote arbitrary code execution.
1344 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1345 time unintentionally subject to list matching rules, letting the header
1346 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1347 arbitrary string expansion). This allowed for information disclosure.
1349 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1350 INT_MIN/-1 -- value coerced to INT_MAX.
1356 NM/01 Workround for PCRE version dependency in version reporting
1359 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1360 This fixes portability to compilers other than gcc, notably
1361 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1363 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1364 makefiles for portability to HP-UX and POSIX correctness.
1366 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1367 Also via indented variable definition in the Makefile.
1368 (Debugging by Oliver Heesakkers).
1370 PP/02 Restore caching of spamd results with expanded spamd_address.
1371 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1373 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1374 Improves build reliability. Fix from: Frank Elsner
1376 NM/02 Fix wide character breakage in the rfc2047 coding
1377 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1379 NM/03 Allow underscore in dnslist lookups
1380 Fixes bug 1026. Patch from Graeme Fowler
1382 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1383 Code patches from Adam Ciarcinski of NetBSD.
1385 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1388 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1389 is logged, to avoid truncation. Patch from John Horne.
1391 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1392 Patch from Jakob Hirsch.
1394 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1395 SQL string expansion failure details.
1396 Patch from Andrey Oktyabrski.
1398 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1399 Patch from Simon Arlott.
1401 PP/09 New lookups functionality failed to compile on old gcc which rejects
1402 extern declarations in function scope.
1403 Patch from Oliver Fleischmann
1405 PP/10 Use sig_atomic_t for flags set from signal handlers.
1406 Check getgroups() return and improve debugging.
1407 Fixed developed for diagnosis in bug 927 (which turned out to be
1410 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1411 Patch from Mark Zealey.
1413 PP/12 Bugzilla 1056: Improved spamd server selection.
1414 Patch from Mark Zealey.
1416 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1417 Based on patch from Heiko Schlittermann.
1419 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1420 Patch from Uwe Doering, sign-off by Michael Haardt.
1422 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1423 variable declaration deep within a block. Bug and patch from
1426 PP/15 lookups-Makefile IRIX compatibilty coercion.
1428 PP/16 Make DISABLE_DKIM build knob functional.
1430 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1431 Patch by Simon Arlott
1433 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1434 variable argument macros. Our copy now differs from upstream.
1440 TF/01 Failure to get a lock on a hints database can have serious
1441 consequences so log it to the panic log.
1443 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1444 controlled using the smtp_confirmation log selector.
1446 TF/03 Include the error message when we fail to unlink a spool file.
1448 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1449 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1450 for maintaining out-of-tree patches for some time.
1452 PP/01 Bugzilla 139: Documentation and portability issues.
1453 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1454 Handle per-OS dynamic-module compilation flags.
1456 PP/02 Let /dev/null have normal permissions.
1457 The 4.73 fixes were a little too stringent and complained about the
1458 permissions on /dev/null. Exempt it from some checks.
1459 Reported by Andreas M. Kirchwitz.
1461 PP/03 Report version information for many libraries, including
1462 Exim version information for dynamically loaded libraries. Created
1463 version.h, now support a version extension string for distributors
1464 who patch heavily. Dynamic module ABI change.
1466 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1467 privilege escalation vulnerability whereby the Exim run-time user
1468 can cause root to append content of the attacker's choosing to
1471 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1474 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1475 If dropping privileges for untrusted macros, we disabled normal logging
1476 on the basis that it would fail; for the Exim run-time user, this is not
1477 the case, and it resulted in successful deliveries going unlogged.
1478 Fixed. Reported by Andreas Metzler.
1484 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1485 only prepend for the Resent-* case. Fixes regression introduced in
1486 Exim 4.70 by NM/22 for Bugzilla 607.
1488 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1489 increasing numbers of administrators be bitten by this.
1491 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1493 PP/03 Bugzilla 994: added openssl_options main configuration option.
1495 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1497 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1499 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1501 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1502 without a peer certificate, leading to a segfault because of an
1503 assumption that peers always have certificates. Be a little more
1504 paranoid. Problem reported by Martin Tscholak.
1506 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1507 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1508 NB: ClamAV planning to remove STREAM in "middle of 2010".
1509 CL also introduces -bmalware, various -d+acl logging additions and
1510 more caution in buffer sizes.
1512 PP/09 Implemented reverse_ip expansion operator.
1514 PP/10 Bugzilla 937: provide a "debug" ACL control.
1516 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1518 PP/12 Bugzilla 973: Implement --version.
1520 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1522 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1524 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1526 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1527 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1528 ignore trailing whitespace.
1530 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1532 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1535 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1536 Notification from Dr Andrew Aitchison.
1538 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1539 ExtendedDetectionInfo response format.
1540 Notification from John Horne.
1542 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1545 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1546 XSL and documented dependency on system catalogs, with examples of how
1549 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1552 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1553 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1554 configuration file which is writeable by the Exim user or group.
1556 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1557 of configuration files to cover files specified with the -C option if
1558 they are going to be used with root privileges, not just the default
1561 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1562 option (effectively making it always true).
1564 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1565 files to be used while preserving root privileges.
1567 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1568 that rogue child processes cannot use them.
1570 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1571 run-time user, instead of root.
1573 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1574 Exim run-time user without dropping privileges.
1576 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1577 result string, instead of calling string_vformat() twice with the same
1580 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1581 for other users. Others should always drop root privileges if they use
1582 -C on the command line, even for a whitelisted configure file.
1584 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1586 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1592 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1593 $data_path, and $header_path variables; fixed documentation bugs and
1596 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1597 exipick to access non-standard spools, including the "frozen" queue
1600 NM/01 Bugzilla 965: Support mysql stored procedures.
1601 Patch from Alain Williams
1603 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1605 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1606 Patch from Andreas Metzler
1608 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1609 Patch from Kirill Miazine
1611 NM/05 Bugzilla 671: Added umask to procmail example.
1613 JJ/03 installed exipick 20100323.0, fixing doc bug
1615 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1616 directory. Notification and patch from Dan Rosenberg.
1618 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1620 TK/02 Improve log output when DKIM signing operation fails.
1622 MH/01 Treat the transport option dkim_domain as a colon separated
1623 list, not as a single string, and sign the message with each element,
1624 omitting multiple occurences of the same signer.
1626 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1627 Bugzilla 985, 986. Patch by Simon Arlott
1629 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1630 Patch by Simon Arlott
1632 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1633 MBX locking. Notification from Dan Rosenberg.
1639 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1641 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1643 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1645 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1647 NM/04 Bugzilla 915: Flag broken perl installation during build.
1653 TK/01 Added patch by Johannes Berg that expands the main option
1654 "spamd_address" if it starts with a dollar sign.
1656 TK/02 Write list of recipients to X-Envelope-Sender header when building
1657 the mbox-format spool file for content scanning (suggested by Jakob
1660 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1661 (http://www.dcc-servers.net/) support via dccifd. Activated by
1662 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1664 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1665 by Mark Daniel Reidel <mr@df.eu>.
1667 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1668 When building exim an external PCRE library is now needed -
1669 PCRE is a system library on the majority of modern systems.
1670 See entry on PCRE_LIBS in EDITME file.
1672 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1673 conversation. Added nologin parameter to request.
1674 Patch contributed by Kirill Miazine.
1676 TF/01 Do not log submission mode rewrites if they do not change the address.
1678 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1680 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1681 log files in place. Contributed by Roberto Lima.
1683 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1685 TF/03 Bugzilla 615: When checking the local_parts router precondition
1686 after a local_part_suffix or local_part_prefix option, Exim now
1687 does not use the address's named list lookup cache, since this
1688 contains cached lookups for the whole local part.
1690 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1691 Robert Millan. Documentation is in experimental-spec.txt.
1693 TF/04 Bugzilla 668: Fix parallel build (make -j).
1695 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1697 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1698 Patch provided by Jan Srzednicki.
1700 TF/05 Leading white space used to be stripped from $spam_report which
1701 wrecked the formatting. Now it is preserved.
1703 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1704 that they are available at delivery time.
1706 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1708 TF/08 TLS error reporting now respects the incoming_interface and
1709 incoming_port log selectors.
1711 TF/09 Produce a more useful error message if an SMTP transport's hosts
1712 setting expands to an empty string.
1714 NM/06 Bugzilla 744: EXPN did not work under TLS.
1715 Patch provided by Phil Pennock.
1717 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1718 Patch provided by Richard Godbee.
1720 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1721 acl_smtp_notquit, added index entry.
1723 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1724 Patch provided by Eugene Bujak.
1726 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1727 accept(). Patch provided by Maxim Dounin.
1729 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1730 Patch provided by Phil Pennock.
1732 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1734 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1735 Patch provided by Brad "anomie" Jorsch.
1737 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1738 Patch provided by Dean Brooks.
1740 TK/05 Add native DKIM support (does not depend on external libraries).
1742 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1743 Patch provided by Graeme Fowler.
1745 NM/16 Bugzilla 851: Documentation example syntax fix.
1747 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1749 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1752 NM/19 Bugzilla 745: TLS version reporting.
1753 Patch provided by Phil Pennock.
1755 NM/20 Bugzilla 167: bool: condition support.
1756 Patch provided by Phil Pennock.
1758 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1759 clients. Patch provided by Phil Pennock.
1761 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1762 Patch provided by Brad "anomie" Jorsch.
1764 NM/23 Bugzilla 687: Fix misparses in eximstats.
1765 Patch provided by Heiko Schlittermann.
1767 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1768 Patch provided by Heiko Schlittermann.
1770 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1771 plus update to original patch.
1773 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1775 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1776 Patch provided by David Brownlee.
1778 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1780 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1782 NM/30 Bugzilla 888: TLS documentation bugfixes.
1784 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1786 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1787 Unlike the original bugzilla I have changed all shell scripts in src tree.
1789 NM/33 Bugzilla 898: Transport filter timeout fix.
1790 Patch by Todd Rinaldo.
1792 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1793 Patch by Serge Demonchaux.
1795 NM/35 Bugzilla 39: Base64 decode bug fixes.
1796 Patch by Jakob Hirsch.
1798 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1800 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1802 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1804 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1810 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1811 ALT-N's libdkim that I have put here:
1812 http://duncanthrax.net/exim-experimental/
1814 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1815 were called 'true' and it seems that C99 defines that as a reserved
1816 keyword to be used with 'bool' variable types. That means you could
1817 not include C99-style headers which use bools without triggering
1818 build errors in sieve.c.
1820 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1821 as mailq or other aliases. Changed the --help handling significantly
1822 to do whats expected. exim_usage() emits usage/help information.
1824 SC/01 Added the -bylocaldomain option to eximstats.
1826 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1828 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1830 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1836 PH/01 Another patch from the Sieve maintainer.
1838 PH/02 When an IPv6 address is converted to a string for single-key lookup
1839 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1840 dots are used instead of colons so that keys in lsearch files need not
1841 contain colons. This was done some time before quoting was made available
1842 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1843 (notated using the quote facility) so as to distinguish them from IPv4
1844 keys. This meant that lookups for IP addresses in host lists did not work
1845 for iplsearch lookups.
1847 This has been fixed by arranging for IPv6 addresses to be expressed with
1848 colons if the lookup type is iplsearch. This is not incompatible, because
1849 previously such lookups could never work.
1851 The situation is now rather anomolous, since one *can* have colons in
1852 ordinary lsearch keys. However, making the change in all cases is
1853 incompatible and would probably break a number of configurations.
1855 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1858 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1859 conversion specification without a maximum field width, thereby enabling
1860 a rogue spamd server to cause a buffer overflow. While nobody in their
1861 right mind would setup Exim to query an untrusted spamd server, an
1862 attacker that gains access to a server running spamd could potentially
1863 exploit this vulnerability to run arbitrary code as the Exim user.
1865 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1866 $primary_hostname instead of what libspf2 thinks the hosts name is.
1868 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1869 a directory entry by the name of the lookup key. Previously, if a
1870 symlink pointed to a non-existing file or a file in a directory that
1871 Exim lacked permissions to read, a lookup for a key matching that
1872 symlink would fail. Now it is enough that a matching directory entry
1873 exists, symlink or not. (Bugzilla 503.)
1875 PH/03 The body_linecount and body_zerocount variables are now exported in the
1878 PH/04 Added the $dnslist_matched variable.
1880 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1881 This means they are set thereafter only if the connection becomes
1884 PH/06 Added the client_condition to authenticators so that some can be skipped
1885 by clients under certain conditions.
1887 PH/07 The error message for a badly-placed control=no_multiline_responses left
1888 "_responses" off the end of the name.
1890 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1892 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1893 (without spaces) instead of just copying the configuration text.
1895 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1897 PH/11 Added $max_received_linelength.
1899 PH/12 Added +ignore_defer and +include_defer to host lists.
1901 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1902 way in which PCRE > 7.0 is built.
1904 PH/14 Implemented queue_only_load_latch.
1906 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1907 MAIL command. The effect was to mangle the value on 64-bit systems.
1909 PH/16 Another patch from the Sieve maintainer.
1911 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1913 PH/18 If a system quota error occurred while trying to create the file for
1914 a maildir delivery, the message "Mailbox is full" was not appended to the
1915 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1916 only to a quota excession during the actual writing of the file.
1918 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1919 characters?) which causes problems in log lines. The DN values are now
1920 passed through string_printing() before being added to log lines.
1922 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1923 and InterBase are left for another time.)
1925 PH/21 Added message_body_newlines option.
1927 PH/22 Guard against possible overflow in moan_check_errorcopy().
1929 PH/23 POSIX allows open() to be a macro; guard against that.
1931 PH/24 If the recipient of an error message contained an @ in the local part
1932 (suitably quoted, of course), incorrect values were put in $domain and
1933 $local_part during the evaluation of errors_copy.
1939 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1940 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1943 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1944 issue a MAIL command.
1946 PH/02 In an ACL statement such as
1948 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1950 if a client was not listed at all, or was listed with a value other than
1951 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1952 the condition was not true (as it should be), so access was not denied.
1953 The bug was that the ! inversion was incorrectly passed on to the second
1954 item. This has been fixed.
1956 PH/03 Added additional dnslists conditions == and =& which are different from
1957 = and & when the dns lookup returns more than one IP address.
1959 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1960 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1962 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1963 FSYNC, which compiles an option called disable_fsync that allows for
1964 bypassing fsync(). The documentation is heavily laced with warnings.
1966 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1968 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1969 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1970 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1971 including adding "make clean"; (3) Added -fPIC when compiling the test
1972 dynamically loaded module, to get rid of a warning.
1974 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1975 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1976 = 0s. The bug is otherwise harmless.
1978 PH/07 There was a bug in the dovecot authenticator such that the value of
1979 $auth1 could be overwritten, and so not correctly preserved, after a
1980 successful authentication. This usually meant that the value preserved by
1981 the server_setid option was incorrect.
1983 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1985 PH/09 Installed PCRE release 7.0.
1987 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1988 run for batched SMTP input. It is now run at the start of every message
1989 in the batch. While fixing this I discovered that the process information
1990 (output by running exiwhat) was not always getting set for -bs and -bS
1991 input. This is fixed, and it now also says "batched" for BSMTP.
1993 PH/11 Added control=no_pipelining.
1995 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1996 patch, slightly modified), and move the expansion of helo_data till after
1997 the connection is made in the smtp transport (so it can use these
2000 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2002 PH/14 Added log_selector = +pid.
2004 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2006 PH/16 Add ${if forany and ${if forall.
2008 PH/17 Added dsn_from option to vary the From: line in DSNs.
2010 PH/18 Flush SMTP output before performing a callout, unless control =
2011 no_callout_flush is set.
2013 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2014 was true (the default) a successful delivery failed to delete the retry
2015 item, thus causing premature timeout of the address. The bug is now
2018 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2020 PH/21 Long custom messages for fakedefer and fakereject are now split up
2021 into multiline reponses in the same way that messages for "deny" and
2022 other ACL rejections are.
2024 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2025 with slight modification.
2027 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2028 draft, changing the syntax and factoring some duplicate code".
2030 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2031 for deliveries of the second and subsequent messages over the same SMTP
2034 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2035 ${reduce, with only minor "tidies".
2037 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2039 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2040 expansion side effects.
2042 PH/27 When a message times out after an over-quota error from an Exim-imposed
2043 quota, the bounce message says "mailbox is full". This message was not
2044 being given when it was a system quota that was exceeded. It now should
2047 MH/03 Made $recipients available in local_scan(). local_scan() already has
2048 better access to the recipient list through recipients_list[], but
2049 $recipients can be useful in postmaster-provided expansion strings.
2051 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2052 in the case of a MAIL command with additional options following the
2053 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2054 were accidentally chopped off.
2056 PH/29 SMTP synchronization checks are implemented when a command is read -
2057 there is a check that no more input is waiting when there shouldn't be
2058 any. However, for some commands, a delay in an ACL can mean that it is
2059 some time before the response is written. In this time, more input might
2060 arrive, invalidly. So now there are extra checks after an ACL has run for
2061 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2062 pipelining has not been advertised.
2064 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2066 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2067 returned in the SMTP response when smtp_return_error_details was false.
2068 This has been fixed.
2070 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2071 instead of the C I/O that was originally supplied, because problems were
2072 reported on Solaris.
2074 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2075 Exim which did not show up earlier: it was assuming that a call to
2076 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2077 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2078 was a macro that became an assignment, so it seemed to work. This has
2079 changed to a proper function call with a void return, hence the compile
2080 error. Exim's code has been fixed.
2082 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2085 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2087 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2089 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2090 because it turns out to be too big or there is a timeout) there is no
2091 "Completed" line in the log. When some messages of this type were
2092 selected by exigrep, they were listed as "not completed". Others were
2093 picked up by some special patterns. I have improved the selection
2094 criteria to be more general.
2096 PH/38 The host_find_failed option in the manualroute router can now be set
2097 to "ignore", to completely ignore a host whose IP address cannot be
2098 found. If all hosts are ignored, the behaviour is controlled by the new
2099 host_all_ignored option.
2101 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2102 homing or because of multiple MX records with /mx) generated more than
2103 one IP address, and the following item turned out to be the local host,
2104 all the secondary addresses of the first item were incorrectly removed
2105 from the list, along with the local host and any following hosts (which
2106 is what is supposed to happen).
2108 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2109 whoever called Exim into the -H file. In the case of the daemon it was
2110 behaving confusingly. When first started, it used values for whoever
2111 started the daemon, but after a SIGHUP it used the Exim user (because it
2112 calls itself on a restart). I have changed the code so that it now always
2115 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2116 message are rejected with the same error (e.g. no authentication or bad
2117 sender address), and a DATA command is nevertheless sent (as can happen
2118 with PIPELINING or a stupid MUA), the error message that was given to the
2119 RCPT commands is included in the rejection of the DATA command. This is
2120 intended to be helpful for MUAs that show only the final error to their
2123 PH/42 Another patch from the Sieve maintainer.
2125 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2126 Eximstats - Fixed some broken HTML links and added missing column headers
2128 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2129 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2131 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2133 PH/43 Yet another patch from the Sieve maintainer.
2135 PH/44 I found a way to check for a TCP/IP connection going away before sending
2136 the response to the final '.' that terminates a message, but only in the
2137 case where the client has not sent further data following the '.'
2138 (unfortunately, this is allowed). However, in many cases there won't be
2139 any further data because there won't be any more messages to send. A call
2140 to select() can be used: if it shows that the input is "ready", there is
2141 either input waiting, or the socket has been closed. An attempt to read
2142 the next input character can distinguish the two cases. Previously, Exim
2143 would have sent an OK response which the client would never have see.
2144 This could lead to message repetition. This fix should cure that, at
2145 least in a lot of common cases.
2147 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2148 advertised in response to EHLO.
2154 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2155 fixed by 4.65/MH/01 (is this a record?) are fixed:
2157 (i) An empty string was always treated as zero by the numeric comparison
2158 operators. This behaviour has been restored.
2160 (ii) It is documented that the numeric comparison operators always treat
2161 their arguments as decimal numbers. This was broken in that numbers
2162 starting with 0 were being interpreted as octal.
2164 While fixing these problems I realized that there was another issue that
2165 hadn't been noticed. Values of message_size_limit (both the global option
2166 and the transport option) were treated as octal if they started with 0.
2167 The documentation was vague. These values are now always treated as
2168 decimal, and I will make that clear in the documentation.
2174 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2175 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2178 MH/01 Don't check that the operands of numeric comparison operators are
2179 integers when their expansion is in "skipping" mode (fixes bug
2180 introduced by 4.64-PH/07).
2182 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2183 child addresses, Exim now panics and dies. Previously, because the count
2184 is held in a short int, deliveries were likely to be lost. As such a
2185 large number of recipients for a single message is ridiculous
2186 (performance will be very, very poor), I have chosen to impose a limit
2187 rather than extend the field.
2193 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2194 leftover -K file (the existence of which was triggered by #402).
2195 While we were at it, introduced process PID as part of the -K
2196 filename. This should rule out race conditions when creating
2199 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2200 processing considerably. Previous code took too long for large mails,
2201 triggering a timeout which in turn triggers #401.
2203 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2204 in the DK code in transports.c. sendfile() is not really portable,
2205 hence the _LINUX specificness.
2207 TF/01 In the add_headers option to the mail command in an Exim filter,
2208 there was a bug that Exim would claim a syntax error in any
2209 header after the first one which had an odd number of characters
2212 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2213 callout verification, Exim cached a "reject" for the entire domain. This
2214 is correct for most verifications, but it is not correct for a recipient
2215 verification with use_sender or use_postmaster set, because in that case
2216 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2217 case of MAIL FROM:<> rejection from other early rejections (e.g.
2218 rejection of HELO). When verifying a recipient using a non-null MAIL
2219 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2220 Whatever the result of the callout, the value of the domain cache is
2221 left unchanged (for any other kind of callout, getting as far as trying
2222 RCPT means that the domain itself is ok).
2224 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2227 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2228 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2229 session, the daemon ignores ECONNECT errors and logs others; it now
2230 ignores EPIPE as well.
2232 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2233 (quoted-printable decoding).
2235 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2236 later the small subsequent patch to fix an introduced bug.
2238 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2240 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2242 PH/08 An error is now given if message_size_limit is specified negative.
2244 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2245 to be given (somewhat) arbitrary names.
2247 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2250 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2251 miscellaneous code fixes
2253 PH/10 Added the log_reject_target ACL modifier to specify where to log
2256 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2257 hostname. This is wrong, because it relates to the incoming message (and
2258 probably the interface on which it is arriving) and not to the outgoing
2259 callout (which could be using a different interface). This has been
2260 changed to use the value of the helo_data option from the smtp transport
2261 instead - this is what is used when a message is actually being sent. If
2262 there is no remote transport (possible with a router that sets up host
2263 addresses), $smtp_active_hostname is used.
2265 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2266 tweaks were necessary in order to get it to work (see also 21 below):
2267 (a) The code assumed that strncpy() returns a negative number on buffer
2268 overflow, which isn't the case. Replaced with Exim's string_format()
2270 (b) There were several signed/unsigned issues. I just did the minimum
2271 hacking in of casts. There is scope for a larger refactoring.
2272 (c) The code used strcasecmp() which is not a standard C function.
2273 Replaced with Exim's strcmpic() function.
2274 (d) The code set only $1; it now sets $auth1 as well.
2275 (e) A simple test gave the error "authentication client didn't specify
2276 service in request". It would seem that Dovecot has changed its
2277 interface. Fortunately there's a specification; I followed it and
2278 changed what the client sends and it appears to be working now.
2280 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2283 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2284 address A is aliased to B and C, where B exists and C does not. Without
2285 -v the output is "A verified" because verification stops after a
2286 successful redirection if more than one address is generated. However,
2287 with -v the child addresses are also verified. Exim was outputting "A
2288 failed to verify" and then showing the successful verification for C,
2289 with its parentage. It now outputs "B failed to verify", showing B's
2290 parentage before showing the successful verification of C.
2292 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2293 look up a TXT record in a specific list after matching in a combined
2296 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2297 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2298 they consult the DNS. I had assumed they would set it the way they
2299 wanted; and indeed my experiments on Linux seem to show that in some
2300 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2301 To be on the safe side, however, I have now made the interface to
2302 host_find_byname() similar to host_find_bydns(), with an argument
2303 containing the DNS resolver options. The host_find_byname() function now
2304 sets these options at its start, just as host_find_bydns() does. The smtp
2305 transport options dns_qualify_single and dns_search_parents are passed to
2306 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2307 of host_find_byname() use the default settings of RES_DEFNAMES
2308 (qualify_single) but not RES_DNSRCH (search_parents).
2310 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2311 spool_read_header() do less string testing, by means of a preliminary
2312 switch on the second character of optional "-foo" lines. (This is
2313 overdue, caused by the large number of possibilities that now exist.
2314 Originally there were few.) While I was there, I also converted the
2315 str(n)cmp tests so they don't re-test the leading "-" and the first
2316 character, in the hope this might squeeze out yet more improvement.
2318 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2319 flag allowing group syntax was set by the header_syntax check but not
2320 turned off, possible causing trouble later; (2) The flag was not being
2321 set at all for the header_verify test, causing "group"-style headers to
2322 be rejected. I have now set it in this case, and also caused header_
2323 verify to ignore an empty address taken from a group. While doing this, I
2324 came across some other cases where the code for allowing group syntax
2325 while scanning a header line wasn't quite right (mostly, not resetting
2326 the flag correctly in the right place). These bugs could have caused
2327 trouble for malformed header lines. I hope it is now all correct.
2329 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2330 with the "reply" argument non-NULL. The code, however (which originally
2331 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2332 but it didn't always do it. This confused somebody who was copying the
2333 code for some other use. I have removed all the tests.
2335 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2336 feature that was used to support insecure browsers during the U.S. crypto
2337 embargo. It requires special client support, and Exim is probably the
2338 only MTA that supported it -- and would never use it because real RSA is
2339 always available. This code has been removed, because it had the bad
2340 effect of slowing Exim down by computing (never used) parameters for the
2341 RSA_EXPORT functionality.
2343 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2344 authenticator to fail if there's a tab character in the incoming data
2345 (there should never be unless someone is messing about, as it's supposed
2346 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2347 if the connection is using TLS or if the remote IP is the same as the
2348 local IP, and the "valid-client-cert option" if a client certificate has
2351 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2352 authenticators. This can be used for authorization after authentication
2353 succeeds. (In the case of plaintext, it servers for both authentication
2356 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2357 if any retry times were supplied.
2359 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2360 connection, where there is no client IP address to check. In this
2361 situation, the verify now always succeeds.
2363 PH/25 Applied John Jetmore's -Mset patch.
2365 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2367 PH/27 In a string expansion for a processed (not raw) header when multiple
2368 headers of the same name were present, leading whitespace was being
2369 removed from all of them, but trailing whitespace was being removed only
2370 from the last one. Now trailing whitespace is removed from each header
2371 before concatenation. Completely empty headers in a concatenation (as
2372 before) are ignored.
2374 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2375 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2377 PH/29 [Removed. This was a change that I later backed out, and forgot to
2378 correct the ChangeLog entry (that I had efficiently created) before
2379 committing the later change.]
2381 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2382 address errors (4xx response to RCPT) over the same connection as other
2383 messages routed to the same hosts. Such deliveries are always "forced",
2384 so retry times are not inspected. This resulted in far too many retries
2385 for the affected addresses. The effect occurred only when there were more
2386 hosts than the hosts_max_try setting in the smtp transport when it had
2387 the 4xx errors. Those hosts that it had tried were not added to the list
2388 of hosts for which the message was waiting, so if all were tried, there
2389 was no problem. Two fixes have been applied:
2391 (i) If there are any address or message errors in an SMTP delivery, none
2392 of the hosts (tried or untried) are now added to the list of hosts
2393 for which the message is waiting, so the message should not be a
2394 candidate for sending over the same connection that was used for a
2395 successful delivery of some other message. This seems entirely
2396 reasonable: after all the message is NOT "waiting for some host".
2397 This is so "obvious" that I'm not sure why it wasn't done
2398 previously. Hope I haven't missed anything, but it can't do any
2399 harm, as the worst effect is to miss an optimization.
2401 (ii) If, despite (i), such a delivery is accidentally attempted, the
2402 routing retry time is respected, so at least it doesn't keep
2403 hammering the server.
2405 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2406 in ${readsocket because some servers need this prod.
2408 PH/32 Added some extra debug output when updating a wait-xxx database.
2410 PH/33 The hint "could be header name not terminated by colon", which has been
2411 given for certain expansion errors for a long time, was not being given
2412 for the ${if def:h_colon_omitted{... case.
2414 PH/34 The spec says: "With one important exception, whenever a domain list is
2415 being scanned, $domain contains the subject domain." There was at least
2416 one case where this was not true.
2418 PH/35 The error "getsockname() failed: connection reset by peer" was being
2419 written to the panic log as well as the main log, but it isn't really
2420 panic-worthy as it just means the connection died rather early on. I have
2421 removed the panic log writing for the ECONNRESET error when getsockname()
2424 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2425 runs only) independently of the message's sender address. This meant
2426 that, if the 4xx error was in fact related to the sender, a different
2427 message to the same recipient with a different sender could confuse
2428 things. In particualar, this can happen when sending to a greylisting
2429 server, but other circumstances could also provoke similar problems.
2430 I have changed the default so that the retry time for these errors is now
2431 based a combination of the sender and recipient addresses. This change
2432 can be overridden by setting address_retry_include_sender=false in the
2435 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2436 remote server are returned as part of bounce messages. This was not
2437 happening for LMTP over a pipe (the lmtp transport), but now it is the
2438 same for both kinds of LMTP.
2440 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2441 in header lines that were in fact CNAMEs. This is no longer the case.
2443 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2444 and queue runs started by the daemon processed all messages. This has
2445 been fixed so that -R and -S can now usefully be given with -q<time>.
2447 PH/40 Import PCRE release 6.7 (fixes some bugs).
2449 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2451 PH/42 Give an error if -q is specified more than once.
2453 PH/43 Renamed the variables $interface_address and $interface_port as
2454 $received_ip_address and $received_port, to make it clear that these
2455 values apply to message reception, and not to the outgoing interface when
2456 a message is delivered. (The old names remain recognized, of course.)
2458 PH/44 There was no timeout on the connect() call when using a Unix domain
2459 socket in the ${readsocket expansion. There now is.
2461 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2462 be meaningful with "accept".
2464 SC/01 Eximstats V1.43
2465 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2467 SC/02 Eximstats V1.44
2468 Use a glob alias rather than an array ref in the generated
2469 parser. This improves both readability and performance.
2471 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2472 Collect SpamAssassin and rejection statistics.
2473 Don't display local sender or destination tables unless
2474 there is data to show.
2475 Added average volumes into the top table text output.
2477 SC/04 Eximstats V1.46
2478 Collect data on the number of addresses (recipients)
2479 as well as the number of messages.
2481 SC/05 Eximstats V1.47
2482 Added 'Message too big' to the list of mail rejection
2483 reasons (thanks to Marco Gaiarin).
2485 SC/06 Eximstats V1.48
2486 Mainlog lines which have GMT offsets and are too short to
2487 have a flag are now skipped.
2489 SC/07 Eximstats V1.49 (Alain Williams)
2490 Added the -emptyok flag.
2492 SC/08 Eximstats V1.50
2493 Fixes for obtaining the IP address from reject messages.
2495 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2496 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2497 whitesspace changes from 4.64-PH/27
2499 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2502 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2503 are found, allow negative numbers in numeric criteria)
2505 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2507 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2508 to match changes made in 4.64-PH/43
2510 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2512 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2513 "domain" is actually an IP address, and force a failure. This locks out
2514 those revolvers/nameservers that support "A-for-A" lookups, in
2515 contravention of the specifications.
2517 PH/48 When a host name was looked up from an IP address, and the subsequent
2518 forward lookup of the name timed out, the host name was left in
2519 $sender_host_name, contrary to the specification.
2521 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2522 restricted to single-key lookups, Exim was not diagnosing an error if
2523 * or *@ was used with a query-style lookup.
2525 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2527 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2528 long ago, but noone interested enough thought of it. Let's just say that
2529 the "1.1" means that there are some new functions that weren't there at
2530 some point in the past.
2532 PH/51 Error processing for expansion failure of helo_data from an smtp
2533 transport during callout processing was broken.
2535 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2536 tested/used via the -bh/-bhc/-bs options.
2538 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2539 bug, fixed in subsequent PCRE releases).
2541 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2542 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2544 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2550 SC/01 Use a glob alias rather than an array ref in eximstats generated
2551 parser. This improves both readability and performance.
2553 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2554 Don't display local sender or destination tables in eximstats unless
2555 there is data to show.
2556 Added average volumes into the eximstats top table text output.
2558 SC/03 Collect data on the number of addresses (recipients) as well
2559 as the number of messages in eximstats.
2561 TF/01 Correct an error in the documentation for the redirect router. Exim
2562 does (usually) call initgroups() when daemonizing.
2564 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2565 with consistent privilege compared to when running as a daemon.
2567 TF/03 Note in the spec that $authenticated_id is not set for local
2568 submissions from trusted users.
2570 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2571 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2573 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2574 by adding some example configuration directives to the default
2575 configuration file. A little bit of work is required to uncomment the
2576 directives and define how usernames and passwords are checked, but
2577 there is now a framework to start from.
2579 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2580 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2581 without this. I don't know how relevant this is to other LDAP libraries.
2583 PH/02 Add the verb name to the "unknown ACL verb" error.
2585 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2587 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2589 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2590 directory not expanded when it should be if an expanded home directory
2591 was set for the address (which is overridden by the transport).
2593 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2596 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2597 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2598 because it is too late at that time, and has no effect.
2600 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2601 security issue with \' (bugzilla #107). I could not use the
2602 PQescapeStringConn() function, because it needs a PGconn value as one of
2605 PH/08 When testing addresses using -bt, indicate those final addresses that
2606 are duplicates that would not cause an additional delivery. At least one
2607 person was confused, thinking that -bt output corresponded to deliveries.
2608 (Suppressing duplicates isn't a good idea as you lose the information
2609 about possibly different redirections that led to the duplicates.)
2611 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2612 systems where poll() doesn't work, in particular OS X.
2614 PH/10 Added more information to debugging output for retry time not reached.
2616 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2617 operations in malware.c.
2619 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2622 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2623 syslog_duplication set false, log lines that would normally be written
2624 both the the main log and to the reject log were not written to syslog at
2627 PH/14 In the default configuration, change the use of "message" in ACL warn
2628 statements to "add_header".
2630 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2631 not followed by a command (e.g. "seen endif").
2633 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2634 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2637 PH/17 Added extra conditions to the default value of delay_warning_condition
2641 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2642 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2643 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2646 The Auto-Submitted: and various List- headers are standardised, whereas I
2647 don't think Precedence: ever was.
2649 PH/18 Refactored debugging code in route_finduser() to show more information,
2650 in particular, the error code if getpwnam() issues one.
2652 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2653 This is apparently needed in addition to the PH/07 change above to avoid
2654 any possible encoding problems.
2656 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2657 but not after initializing Perl.
2659 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2660 output them only if debugging. By default they are written stderr,
2661 apparently, which is not desirable.
2663 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2666 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2669 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2671 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2672 authenticated or an ident call has been made. Suppress the default
2673 values for $authenticated_id and $authenticated_sender (but permit -oMai
2674 and -oMas) when testing with -bh.
2676 PH/24 Re-jigged the order of the tests in the default configuration so that the
2677 tests for valid domains and recipients precede the DNS black list and CSA
2678 tests, on the grounds that those ones are more expensive.
2680 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2681 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2682 This bug exists in every version of Exim that I still have, right back to
2685 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2686 However, an attempt to turn on case-sensitivity in a regex key by
2687 including (?-i) didn't work because the subject string was already
2688 lowercased, and the effects were non-intuitive. It turns out that a
2689 one-line patch can be used to allow (?-i) to work as expected.
2695 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2696 other effects) broke the use of negated acl sub-conditions.
2698 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2701 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2702 "Deny" causes Exim to reject the incoming connection with a 554 error.
2703 Unfortunately, if there is a major crisis, such as a disk failure,
2704 tcp-wrappers gives "deny", whereas what one would like would be some
2705 kind of temporary error. A kludge has been added to help with this.
2706 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2707 554 error is used if errno is still zero or contains ENOENT (which occurs
2708 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2711 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2713 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2714 errors. Otherwise a message that provokes a temporary error (when other
2715 messages do not) can cause a whole host to time out.
2717 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2718 addresses were routed directly to files or pipes from a redirect router.
2719 File deliveries just didn't batch; pipe deliveries might have suffered
2722 PH/06 A failure to get a lock for a hints database would erroneously always say
2723 "Failed to get write lock", even when it was really a read lock.
2725 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2726 of 0600. This has been changed to use the value of the lockfile_mode
2727 option (which defaults to 0600).
2729 PH/08 Applied small patch from the Sieve maintainer.
2731 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2732 folder from quota calculations, a direct delivery into this folder messed
2733 up the contents of the maildirsize file. This was because the regex was
2734 used only to exclude .Trash (or whatever) when the size of the mailbox
2735 was calculated. There was no check that a delivery was happening into an
2736 excluded directory. This bug has been fixed by ignoring all quota
2737 processing for deliveries into excluded directories.
2739 PH/10 Added the maildirfolder_create_regex option to appendfile.
2745 PH/01 The code for finding all the local interface addresses on a FreeBSD
2746 system running IPv6 was broken. This may well have applied to all BSD
2747 systems, as well as to others that have similar system calls. The broken
2748 code found IPv4 interfaces correctly, but gave incorrect values for the
2749 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2750 that it would not match correctly against @[] and not recognize the IPv6
2753 PH/02 The ipliteral router was not recognizing addresses of the form user@
2754 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2756 PH/03 Added disable_ipv6.
2758 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2759 decoded form, because it is most often used to construct To: headers
2760 lines in autoreplies, and the decoded form may well be syntactically
2761 invalid. However, $reply_address has leading white space removed, and all
2762 newlines turned into spaces so that the autoreply transport does not
2765 PH/05 If group was specified without a user on a router, and no group or user
2766 was specified on a transport, the group from the router was ignored.
2768 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2769 for visible compile-time settings that can be used to change these
2770 numbers, for those that want even more. Backwards compatibility with old
2771 spool files has been maintained. However, going back to a previous Exim
2772 release will lost any variables that are in spool files.
2774 PH/07 Two small changes when running in the test harness: increase delay when
2775 passing a TCP/IP connection to a new process, in case the original
2776 process has to generate a bounce, and remove special handling of
2777 127.0.0.2 (sic), which is no longer necessary.
2779 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2780 be the same on different OS.
2782 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2785 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2786 whether --show-vars was specified or not
2788 JJ/02 exipick: Added support for new ACL variable spool format introduced
2791 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2792 syntactically invalid From: or Reply-to: line, and a filter used this to
2793 generate an autoreply, and therefore failed to obtain an address for the
2794 autoreply, Exim could try to deliver to a non-existent relative file
2795 name, causing unrelated and misleading errors. What now happens is that
2796 it logs this as a hard delivery error, but does not attempt to create a
2799 PH/11 The exinext utility has a -C option for testing purposes, but although
2800 the given file was scanned by exinext itself; it wasn't being passed on
2801 when Exim was called.
2803 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2804 an end-of-file indication when reading a command response.
2806 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2807 compiled. In many other places in Exim, IPv6 addresses are always
2808 recognized, so I have changed this. It also means that IPv4 domain
2809 literals of the form [IPV4:n.n.n.n] are now always recognized.
2811 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2812 used if the router is not running as root, for example, when verifying at
2813 ACL time, or when using -bh. The debugging output from this situation was
2814 non-existent - all you got was a failure to exec. I have made two
2817 (a) Failures to set uid/gid, the current directory, or a process leader
2818 in a subprocess such as that created by queryprogram now generate
2819 suitable debugging ouput when -d is set.
2821 (b) The queryprogram router detects when it is not running as root,
2822 outputs suitable debugging information if -d is set, and then runs
2823 the subprocess without attempting to change uid/gid.
2825 PH/15 Minor change to Makefile for building test_host (undocumented testing
2828 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2829 additional section of a DNS packet that returns MX or SRV records.
2830 Instead, it always explicitly searches for A/AAAA records. This avoids
2831 major problems that occur when a DNS server includes only records of one
2832 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2833 fixed another bug: if SRV records were looked up and the corresponding
2834 address records were *not* found in the additional section, the port
2835 values from the SRV records were lost.
2837 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2838 using the correct key (the original address) when searching the retry
2839 rules in order to find which one to use for generating the retry hint.
2841 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2842 adding the default one. Similarly, if it contains a Reply-To: header, the
2843 errors_reply_to option, if set, is not used.
2845 PH/19 When calculating a retry time, Exim used to measure the "time since
2846 failure" by looking at the "first failed" field in the retry record. Now
2847 it does not use this if it is later than than the arrival time of the
2848 message. Instead it uses the arrival time. This makes for better
2849 behaviour in cases where some deliveries succeed, thus re-setting the
2850 "first failed" field. An example is a quota failure for a huge message
2851 when small messages continue to be delivered. Without this change, the
2852 "time since failure" will always be short, possible causing more frequent
2853 delivery attempts for the huge message than are intended.
2854 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2856 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2857 $1, $2, $3) because the numerical variables can be reset during some
2858 expansion items (e.g. "match"), thereby losing the authentication data.
2860 PH/21 Make -bV show the size of off_t variables so that the test suite can
2861 decide whether to run tests for quotas > 2G.
2863 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2864 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2865 filecount value is greater than 2G or if a quota value is greater than 2G
2866 on a system where the size of off_t is not greater than 4, a panic error
2869 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2870 never match. The debug and -bh output now contains an explicit error
2871 message indicating a malformed IPv4 address or mask.
2873 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2874 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2875 PH/23 above applies.
2877 PH/25 Do not write to syslog when running in the test harness. The only
2878 occasion when this arises is a failure to open the main or panic logs
2879 (for which there is an explicit test).
2881 PH/26 Added the /no_tell option to "control=freeze".
2883 PH/27 If a host name lookup failed very early in a connection, for example, if
2884 the IP address matched host_lookup and the reverse lookup yielded a name
2885 that did not have a forward lookup, an error message of the form "no IP
2886 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2887 could be logged. Now it outputs the IP address instead of "NULL".
2889 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2890 allows the sender and the authenticated sender to be set when
2891 submitting a message from within Exim. Since child_open_exim() is
2892 documented for local_scan(), the new function should be too.
2894 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2895 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2896 results in an empty string is now treated as unset.
2898 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2900 PH/31 Added sender_verify_fail logging option.
2902 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2903 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2904 tidied the source and removed it altogether.
2906 PH/33 When a queue run was abandoned because the load average was too high, a
2907 log line was always written; now it is written only if the queue_run log
2908 selector is set. In addition, the log line for abandonment now contains
2909 information about the queue run such as the pid. This is always present
2910 in "start" and "stop" lines but was omitted from the "abandon" line.
2912 PH/34 Omit spaces between a header name and the colon in the error message that
2913 is given when verify = headers_syntax fails (if there are lots of them,
2914 the message gets confusing).
2916 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2917 names, as there are now some PTR records that contain slashes. This check
2918 is only to protect against broken name servers that fall over on strange
2919 characters, so the fact that it applies to all lookups doesn't matter.
2921 PH/36 Now that the new test suite is complete, we can remove some of the
2922 special code in Exim that was needed for the old test suite. For example,
2923 sorting DNS records because real resolvers return them in an arbitrary
2924 order. The new test suite's fake resolver always returns records in the
2927 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2928 bounces) except when queue_only is set, to avoid logging races between
2929 the different processes.
2931 PH/38 Panic-die if .include specifies a non-absolute path.
2933 PH/39 A tweak to the "H" retry rule from its user.
2935 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2936 a label. They prevented compilation on older perls.
2938 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2939 a warning to be raised on newish perls.
2941 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2942 on queue. Changes to match documented behaviour of showing count of
2943 messages matching specified criteria.
2945 PH/40 Changed the default ident timeout from 30s to 5s.
2947 PH/41 Added support for the use of login_cap features, on those BSD systems
2948 that have them, for controlling the resources used by pipe deliveries.
2950 PH/42 The content-scanning code uses fopen() to create files in which to put
2951 message data. Previously it was not paying any attention to the mode of
2952 the files. Exim runs with umask(0) because the rest of the code creates
2953 files with open(), and sets the required mode explicitly. Thus, these
2954 files were ending up world-writeable. This was not a big issue, because,
2955 being within the spool directory, they were not world-accessible. I have
2956 created a function called modefopen, which takes an additional mode
2957 argument. It sets umask(777), creates the file, chmods it to the required
2958 mode, then resets the umask. All the relevant calls to fopen() in the
2959 content scanning code have been changed to use this function.
2961 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2962 to 24 hours. This avoids potential overflow problems when processing G
2963 and H retry rules. I suspect nobody ever tinkers with this value.
2965 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2967 PH/45 When the plaintext authenticator is running as a client, the server's
2968 challenges are checked to ensure they are valid base64 strings. By
2969 default, the authentication attempt is cancelled if an invalid string is
2970 received. Setting client_ignore_invalid_base64 true ignores these errors.
2971 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2972 they are received. Thus, the responses can be made to depend on the
2973 challenges. If an invalid string is ignored, an empty string is placed in
2976 PH/46 Messages that are created by the autoreply transport now contains a
2977 References: header, in accordance with RFCs 2822 and 3834.
2979 PH/47 Added authenticated_sender_force to the smtp transport.
2981 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2983 PH/49 Installed latest patch from the Sieve maintainer.
2985 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2986 was also set, the appendfile transport was unnecessarily scanning a
2987 directory of message files (e.g. for maildir delivery) to find the count
2988 of files (along with the size), even though it did not need this
2989 information. It now does the scan only if it needs to find either the
2990 size of the count of files.
2992 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2994 PH/52 Two bugs concerned with error handling when the smtp transport is
2997 (i) Exim was not creating retry information for temporary errors given
2998 for individual recipients after the DATA command when the smtp transport
2999 was used in LMTP mode. This meant that they could be retried too
3000 frequently, and not timed out correctly.
3002 (ii) Exim was setting the flag that allows error details to be returned
3003 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3004 recipients that were returned after the DATA command.
3006 PH/53 This is related to PH/52, but is more general: for any failing address,
3007 when detailed error information was permitted to be returned to the
3008 sender, but the error was temporary, then after the final timeout, only
3009 "retry timeout exceeded" was returned. Now it returns the full error as
3010 well as "retry timeout exceeded".
3012 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3013 do this, and (what is worse) MTAs that accept it.
3015 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3016 will now be deprecated.
3018 PH/56 New os.c-cygwin from the Cygwin maintainer.
3020 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3021 formats (previously only available in exim formats via -bpr, -bpru,
3022 and -bpra. Now also available in native and exiqgrep formats)
3024 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3025 with very large, slow to parse queues
3027 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3029 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3031 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3032 responses to authentication challenges, though it was showing the
3033 challenges; (ii) I've removed the CR characters from the debug output for
3036 PH/58 Allow for the insertion of a newline as well as a space when a string
3037 is turned into more than one encoded-word during RFC 2047 encoding. The
3038 Sieve code now uses this.
3040 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3041 data_4xx, lost_connection, tls_required.
3043 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3044 message was being sent as an SMTP response.
3046 PH/61 Add -l and -k options to exicyclog.
3048 PH/62 When verifying, if an address was redirected to one new address, so that
3049 verification continued, and the new address failed or deferred after
3050 having set something in $address_data, the value of $address_data was not
3051 passed back to the ACL. This was different to the case when no
3052 redirection occurred. The value is now passed back in both cases.
3054 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3055 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3056 use login_cap.h, so on its own it isn't the distinguishing feature. The
3057 new name refers directly to the setclassresources() function.
3059 PH/65 Added configuration files for NetBSD3.
3061 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3063 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3064 is preferred over IPv4.
3066 PH/68 The bounce_return_message and bounce_return_body options were not being
3067 honoured for bounces generated during the reception of non-SMTP messages.
3068 In particular, this applied to messages rejected by the ACL. This bug has
3069 been fixed. However, if bounce_return_message is true and bounce_return_
3070 body is false, the headers that are returned for a non-SMTP message
3071 include only those that have been read before the error was detected.
3072 (In the case of an ACL rejection, they have all been read.)
3074 PH/69 The HTML version of the specification is now built in a directory called
3075 spec_html instead of spec.html, because the latter looks like a path with
3076 a MIME-type, and this confuses some software.
3078 PH/70 Catch two compiler warnings in sieve.c.
3080 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3081 function verify_get_ident() calls ip_connect() to connect a socket, but
3082 if the "connect()" function timed out, ip_connect() used to close the
3083 socket. However, verify_get_ident() also closes the socket later, and in
3084 between Exim writes to the log, which may get opened at this point. When
3085 the socket was closed in ip_connect(), the log could get the same file
3086 descriptor number as the socket. This naturally causes chaos. The fix is
3087 not to close the socket in ip_connect(); the socket should be closed by
3088 the function that creates it. There was only one place in the code where
3089 this was missing, in the iplookup router, which I don't think anybody now
3090 uses, but I've fixed it anyway.
3092 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3093 well as to direct DNS lookups. Otherwise the handling of names in host
3094 lists is inconsistent and therefore confusing.
3100 PH/01 Two changes to the default runtime configuration:
3102 (1) Move the checks for relay_from_hosts and authenticated clients from
3103 after to before the (commented out) DNS black list checks.
3105 (2) Add control=submission to the relay_from_hosts and authenticated
3106 clients checks, on the grounds that messages accepted by these
3107 statements are most likely to be submissions.
3109 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3111 (1) Generate an error if the third argument for the ${prvs expansion is
3114 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3117 (3) Reset the variables that are obtained from the first argument of
3118 ${prvscheck and used in the second argument before leaving the code,
3119 because their memory is reclaimed, so using them afterwards may do
3122 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3123 one (it's much easier than Tom thought :-).
3125 (5) Because of (4), we can now allow for the use of $prvscheck_result
3126 inside the third argument.
3128 PH/03 For some reason, the default setting of PATH when running a command from
3129 a pipe transport was just "/usr/bin". I have changed it to
3132 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3133 anything to be listed in the output from -bV.
3135 PH/05 When a filter generated an autoreply, the entire To: header line was
3136 quoted in the delivery log line, like this:
3138 => >A.N.Other <ano@some.domain> <original@ddress> ...
3140 This has been changed so that it extracts the operative address. There
3141 may be more than one such address. If so, they are comma-separated, like
3144 => >ano@some.domain,ona@other.domain <original@ddress> ...
3146 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3147 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3148 not being looked up in the rDNS to get a host name, Exim was showing the
3149 IP address twice in Received: lines, even though the IP addresses were
3150 identical. For example:
3152 Received: from [1.2.3.4] (helo=[1.2.3.4])
3154 However, if the real host name was known, it was omitting the HELO data
3155 if it matched the actual IP address. This has been tidied up so that it
3156 doesn't show the same IP address twice.
3158 PH/07 When both +timestamp and +memory debugging was on, the value given by
3159 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3160 called by the debug printing, thereby overwriting the timestamp buffer.
3161 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3163 PH/08 When the original message was included in an autoreply transport, it
3164 always said "this is a copy of the message, including all the headers",
3165 even if body_only or headers_only was set. It now gives an appropriate
3168 PH/09 Applied a patch from the Sieve maintainer which:
3170 o fixes some comments
3171 o adds the (disabled) notify extension core
3172 o adds some debug output for the result of if/elsif tests
3173 o points to the current vacation draft in the documentation
3174 and documents the missing references header update
3178 o fixes a bug in processing the envelope test (when testing
3179 multiple envelope elements, the last element determinted the
3182 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3183 Electronic Mail") by including:
3185 Auto-submitted: auto-generated
3187 in the messages that it generates (bounce messages and others, such as
3188 warnings). In the case of bounce messages for non-SMTP mesages, there was
3189 also a typo: it was using "Auto_submitted" (underscore instead of
3190 hyphen). Since every message generated by Exim is necessarily in response
3191 to another message, thes have all been changed to:
3193 Auto-Submitted: auto-replied
3195 in accordance with these statements in the RFC:
3197 The auto-replied keyword:
3199 - SHOULD be used on messages sent in direct response to another
3200 message by an automatic process,
3202 - MUST NOT be used on manually-generated messages,
3204 - MAY be used on Delivery Status Notifications (DSNs) and Message
3205 Disposition Notifications (MDNs),
3207 - MUST NOT be used on messages generated by automatic or periodic
3208 processes, except for messages which are automatic responses to
3211 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3212 to the default Received: header definition.
3214 PH/12 Added log selector acl_warn_skipped (default on).
3216 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3217 variables because (a) they are in the wrong storage pool and (b) even if
3218 they were copied, it wouldn't work properly because of the caching.
3220 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3221 checking when decoding. Apparently there are clients that generate
3222 overlong encoded strings. Why am I not surprised?
3224 PH/15 If the first argument of "${if match_address" was not empty, but did not
3225 contain an "@" character, Exim crashed. Now it writes a panic log message
3226 and treats the condition as false.
3228 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3230 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3231 extension "envelope-auth". The code is finished and in agreement with
3232 other implementations, but there is no documentation so far and in fact,
3233 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3234 not changing the active code.
3236 Print executed "if" and "elsif" statements when debugging is used. This
3237 helps a great deal to understand what a filter does.
3239 Document more things not specified clearly in RFC3028. I had all this
3240 sorted out, when out of a sudden new issues came to my mind. Oops."
3242 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3245 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3246 canonical form (as documented). However, after a host name lookup from
3247 the IP address, check_host() was doing a simple string comparison with
3248 addresses acquired from the DNS when checking that the found name did
3249 have the original IP as one of its addresses. Since any found IPv6
3250 addresses are likely to be in abbreviated form, the comparison could
3251 fail. Luckily, there already exists a function for doing the comparison
3252 by converting both addresses to binary, so now that is used instead of
3253 the text comparison.
3255 PH/20 There was another similar case to PH/19, when a complete host name was
3256 given in a host list; looking up its IP address could give an abbreviated
3257 form, whereas the current host's name might or might not be abbreviated.
3258 The same fix has been applied.
3264 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3265 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3268 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3269 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3271 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3273 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3274 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3275 Also, for "Auto-Submitted", treat anything other than "no" as
3276 non-personal, in accordance with RFC 3834. (Previously it treated
3277 anything starting "auto-" as non-personal.)
3279 TF/01 The control=submission/name=... option had a problem with syntax
3280 errors if the name included a slash character. The /name= option
3281 now slurps the rest of the string, so it can include any characters
3282 but it must come last in the list of options (after /sender_retain
3285 PH/05 Some modifications to the interface to the fake nameserver for the new
3293 TK/01 Added the "success_on_redirect" address verification option. See
3294 NewStuff for rationale and an example.
3296 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3298 PH/02 Patch to exigrep to allow it to work on syslog lines.
3300 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3301 fread() to skip over the body file's header line, because in Cygwin the
3302 header line is locked and is inaccessible.
3304 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3305 co-exist for some time) to make it clear that it is the Exim ID that is
3306 referenced, not the Message-ID: header line.
3308 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3309 string_format() function, because snprintf() does not exist on all
3312 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3313 ${stat: expansion item.
3315 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3316 protocol synchronization error", to keep the pedants happy.
3318 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3319 well as for IRIX systems, when gcc is being used. See the host.c source
3322 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3324 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3327 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3328 transport and to the smtp transport in LMTP mode.
3330 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3332 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3333 run a filter in a subprocess. This could lead to confusion in subsequent
3334 lookups in the parent process. There should also be a search_tidyup() at
3335 the end of the subprocess.
3337 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3338 only if the host matched helo_try_verify_hosts, which caused the
3339 verification to occur when the EHLO/HELO command was issued. The ACL just
3340 tested the remembered result. Now, if a previous verification attempt has
3341 not happened, "verify = helo" does it there and then.
3343 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3345 TK/03 Fix log output including CR from clamd.
3347 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3348 exist provoked a memory error which could cause a segfault.
3350 PH/15 Installed PCRE 6.2
3352 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3354 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3355 of the problem. Specifically, suggested +O2 rather than +O1 for the
3358 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3360 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3361 record, the port was not correctly logged when the outgoing_port log
3362 selector was set (it logged the transort's default port).
3364 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3365 fallback_hosts, and "hosts" in the smtp transport.
3367 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3368 host errors such as "Connection refused".
3370 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3371 authentication with radiusclient 0.4.9:
3373 - Error returned from rc_read_config was caught wrongly
3374 - Username/password not passed on to radius server due to wrong length.
3376 The presumption is that some radiusclient API changes for 4.51/PH/17
3377 were not taken care of correctly. The code is still untested by me (my
3378 Linux distribution still has 0.3.2 of radiusclient), but it was
3379 contributed by a Radius user.
3381 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3382 expanding the "port" option of the smtp transport.
3384 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3385 while reading a MIME header. Thanks to Tom Hughes for a patch.
3387 PH/24 Include config.h inside local_scan.h so that configuration settings are
3390 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3391 that in an ACL for RCPT (for example), you can examine exactly what was
3394 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3395 commands, but it was not correctly comparing the address with the actual
3396 client host address. Thus, it would show the EHLO address in Received:
3397 header lines when this was not necessary.
3399 PH/27 Added the % operator to ${eval:}.
3401 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3402 it should be ignoring failures (because with -C, for example, it has lost
3403 privilege). It wasn't ignoring creation failures other than "already
3406 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3409 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3410 because an input line was too long, either on its own, or by virtue of
3411 too many continuations, the temporary file was not being removed, and the
3412 return code was incorrect.
3414 PH/31 Missing "BOOL" in function definition in filtertest.c.
3416 PH/32 Applied Sieve patches from the maintainer.
3418 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3420 PH/33 Added "verify = not_blind".
3422 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3423 Local/Makefile (with some defaults set). These are used in built scripts
3424 such as exicyclog, but they have never been used in the exim_install
3425 script (though there are many overriding facilities there). I have
3426 arranged that the exim_install script now takes note of these two
3429 PH/35 Installed configuration files for Dragonfly.
3431 PH/36 When a locally submitted message by a trusted user did not contain a
3432 From: header, and the sender address was obtained from -f or from an SMTP
3433 MAIL command, and the trusted user did not use -F to supply a sender
3434 name, $originator_name was incorrectly used when constructing a From:
3435 header. Furthermore, $originator_name was used for submission mode
3436 messages from external hosts without From: headers in a similar way,
3437 which is clearly wrong.
3439 PH/37 Added control=suppress_local_fixups.
3441 PH/38 When log_selector = +received_sender was set, and the addition of the
3442 sender made the log line's construction buffer exactly full, or one byte
3443 less than full, an overflow happened when the terminating "\n" was
3446 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3447 when the result of a list match is failure because a DNS lookup failed.
3449 PH/40 RM_COMMAND is now used in the building process.
3451 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3452 the "build-* directories that it finds.
3454 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3455 address in a domain literal was a prefix of an interface address.
3457 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3458 when verifying a sender address, unless rewrite_headers is false.
3460 PH/44 Wrote a long comment about why errors_to addresses are verified as
3461 recipients, not senders.
3463 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3464 the ratelimit ACL was added.
3466 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3468 PH/46 Added extra information about PostgreSQL errors to the error string.
3470 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3471 avoiding the need to install special zones in a real server. This is
3472 backwards compatible; if it can't find the fake resolver, it drops back.
3473 Thus, both old and new test suites can be run.
3475 TF/02 Added util/ratelimit.pl
3477 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3478 clock is set back in time.
3480 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3481 Candler <B.Candler@pobox.com>.
3483 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3484 OK for addresses that are the result of redirections.
3486 PH/48 A number of further additions for the benefit of the new test suite,
3487 including a fake gethostbyname() that interfaces to the fake DNS resolver
3490 TF/06 The fix for widen_domains has also been applied to qualify_single and
3491 search_parents which are the other dnslookup options that can cause
3494 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3497 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3499 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3500 service" if it encountered an MX record with an empty target hostname.
3501 The message is now "an MX or SRV record indicated no SMTP service".
3503 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3504 if the DNS of the sending site is misconfigured. This is quite a
3505 common situation. This change restores the behaviour of treating a
3506 helo verification defer as a failure.
3508 PH/51 If self=fail was set on a router, the bounce message did not include the
3509 actual error message.
3515 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3517 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3518 command itself ended in error, the underlying message about the transport
3519 filter timeout was being overwritten with the pipe command error. Now the
3520 underlying error message should be appended to the second error message.
3522 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3524 PH/02 Reduce the amount of output that "make" produces by default. Full output
3525 can still be requested.
3527 PH/03 The warning log line about a condition test deferring for a "warn" verb
3528 was being output only once per connection, rather than after each
3529 occurrence (because it was using the same function as for successful
3530 "warn" verbs). This seems wrong, so I have changed it.
3532 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3533 it should not have, which might have caused a crash in the right
3534 circumstances, but probably never did.
3536 PH/04 Installed a modified version of Tony Finch's patch to make submission
3537 mode fix the return path as well as the Sender: header line, and to
3538 add a /name= option so that you can make the user's friendly name appear
3541 TF/03 Added the control = fakedefer ACL modifier.
3543 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3544 Mark Lowes for thorough testing.
3546 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3548 TK/03 Merged latest SRS patch from Miles Wilton.
3550 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3551 with the definition in sysexits.h (which is #included earlier).
3552 Fortunately, Exim does not actually use EX_OK. The code used to try to
3553 preserve the sysexits.h value, by assumimg that macro definitions were
3554 scanned for macro replacements. I have been disabused of this notion,
3555 so now the code just undefines EX_OK before #including unistd.h.
3557 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3558 in the smtp transport. When a block could not be written in a single
3559 write() function, the timeout was being re-applied to each part-write.
3560 This seems wrong - if the receiver was accepting one byte at a time it
3561 would take for ever. The timeout is now adjusted when this happens. It
3562 doesn't have to be particularly precise.
3564 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3565 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3567 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3568 without a domain if the check to <postmaster@domain> fails.
3570 SC/01 Eximstats: added -xls and the ability to specify output files
3571 (patch written by Frank Heydlauf).
3573 SC/02 Eximstats: use FileHandles for outputing results.
3575 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3577 SC/04 Eximstats: fixed display of large numbers with -nvr option
3579 SC/05 Eximstats: fixed merging of reports with empty tables.
3581 SC/06 Eximstats: added the -include_original_destination flag
3583 SC/07 Eximstats: removed tabs and trailing whitespace.
3585 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3587 TK/06 MBOX spool code: Add real "From " MBOX separator line
3588 so the .eml file is really in mbox format (even though
3589 most programs do not really care). Patch from Alex Miller.
3591 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3592 The latter is generated from $received_to and is only set if the
3593 message has one envelope recipient. SA can use these headers,
3594 obviously out-of-the-box. Patch from Alex Miller.
3596 PH/08 The ${def test on a variable was returning false if the variable's
3597 value was "0", contrary to what the specification has always said!
3598 The result should be true unless the variable is empty.
3600 PH/09 The syntax error of a character other than { following "${if
3601 def:variable_name" (after optional whitespace) was not being diagnosed.
3602 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3603 accidental colon was present, for example, could give incorrect results.
3605 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3606 result is used (not including appendfile, where other changes are about
3609 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3610 This involved changing a lot of size variables from int to off_t. It
3611 should work with maildirs and everything.
3613 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3614 spamd dying while we are connected to it.
3616 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3619 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3620 function for BDB 4.3.
3622 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3624 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3625 the sources for PCRE have been re-arranged and more files are now
3628 PH/15 The code I had for printing potentially long long variables in PH/11
3629 above was not the best (it lost precision). The length of off_t variables
3630 is now inspected at build time, and an appropriate printing format (%ld
3631 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3632 to be "long long int" or "long int". This is needed for the internal
3633 formatting function string_vformat().
3635 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3636 the configuration file to be ":syslog", then the script "guesses" where
3637 the logs files are, rather than using the compiled in default. In our
3638 case the guess is not the same as the compiled default, so the script
3639 suddenly stopped working when I started to use syslog. The patch checks
3640 to see if log_file_path is "". If so, it attempts to read it from exim
3641 with no configuration file to get the compiled in version, before it
3642 falls back to the previous guessing code."
3644 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3645 implementing BATV in an Exim configuration. See NewStuff for the gory
3648 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3649 Makefile that are specific to HP-UX.
3651 PH/18 If the "use_postmaster" option was set for a recipient callout together
3652 with the "random" option, the postmaster address was used as the MAIL
3653 FROM address for the random test, but not for the subsequent recipient
3654 test. It is now used for both.
3656 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3657 patch removes a few documentation additions to RFC 3028, because the
3658 latest draft now contains them. It adds the new en;ascii-case comparator
3659 and a new error check for 8bit text in MIME parts. Comparator and
3660 require names are now matched exactly. I enabled the subaddress
3661 extension, but it is not well tested yet (read: it works for me)."
3663 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3664 rework some of the code of TK/09 above to avoid the hardwired use of
3665 "%lld" and "long long". Replaced the call to snprintf() with a call to
3668 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3669 records point to non-existent hosts", "retry timeout exceeded", and
3670 "retry time not reached for any host after a long failure period".
3672 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3673 experimental DomainKeys support:
3675 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3676 (2) On an error such as an illegally used "control", the wrong name for
3677 the control was given.
3679 These problems did NOT occur unless DomainKeys support was compiled.
3681 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3683 PH/24 Added ${if match_ip condition.
3685 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3686 databases so that it will be absolutely obvious if a crash occurs in the
3687 DB library. This is a regular occurrence (often caused by mis-matched
3690 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3691 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3692 picked up on a user's system that detects such things. There doesn't seem
3693 to be a gcc warning option for this - only an attribute that has to be
3694 put on the function's prototype. It seems that in Fedora Core 4 they have
3695 set this on a number of new functions. No doubt there will be more in due
3698 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3699 specify a transport. However, if an address that was verified by such a
3700 router was the subject of a callout, Exim crashed because it tried to
3701 read the rcpt_include_affixes from the non-existent transport. Now it
3702 just assumes that the setting of that option is false. This bug was
3703 introduced by 4.51/PH/31.
3705 PH/28 Changed -d+all to exclude +memory, because that information is very
3706 rarely of interest, but it makes the output a lot bigger. People tend to
3707 do -d+all out of habit.
3709 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3710 code in os-type was giving problems when libc.so lives in lib64, like on
3713 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3714 aren't the modern standard, and it seems that some systems' include files
3715 don't always have them. Exim was already checking for some of the newer
3716 ones like T_AAAA, and defining it itself. I've added checks for all the
3717 record types that Exim uses.
3719 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3720 not automatically generating a new one, as it is supposed to. This
3721 prevented TLS from working. If the file did exist, but contained invalid
3722 data, a new version was generated, as expected. It was only the case of a
3723 non-existent file that was broken.
3725 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3726 with a change in libdomainkeys > 0.64.
3728 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3729 from DNS. If the selector record carries the flag, it now has
3730 precedence over the domain-wide flag.
3732 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3734 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3735 the use of a port name that isn't defined in /etc/services) occurred, the
3736 message was deferred as in a normal delivery, and thus remained on the
3737 spool, instead of being failed because of the mua_wrapper setting. This
3738 is now fixed, and I tidied up some of the mua_wrapper messages at the
3741 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3742 the messages in a hash of arrays rather than using individual hashes.
3743 This is a bit cleaner and results in dramatic memory savings, albeit
3744 at a slight CPU cost.
3746 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3747 as requested by Marc Sherman.
3749 SC/10 Eximstats: added histograms for user specified patterns as requested
3752 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3754 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3755 fopen() in the content-scanning modules that did not already have it.
3761 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3762 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3764 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3766 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3768 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3769 to test Sieve filters that use "vacation".
3771 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3772 that changes the way the GnuTLS parameters are stored in the cache file.
3773 The new format can be generated externally. For backward compatibility,
3774 if the data in the cache doesn't make sense, Exim assumes it has read an
3775 old-format file, and it generates new data and writes a new file. This
3776 means that you can't go back to an older release without removing the
3779 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3780 work if there are any delivery delays because "one_time" forces the
3781 parent to be marked "delivered", so its unseen clone is never tried
3782 again. For this reason, Exim now forbids the simultaneous setting of
3785 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3786 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3787 entry if you want to know the details. The fix, however, neglected to
3788 consider the case when local delivery batching is involved. The test for
3789 "previously delivered" was not happening when checking to see if an
3790 address could be batched with a previous (undelivered) one; under
3791 certain circumstances this could lead to multiple deliveries to the same
3794 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3795 in its include files, and this causes problems building Exim.
3797 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3798 header_syntax) but Exim was just ignoring anything given after a slash.
3799 In particular, this caused confusion with an attempt to use "verify =
3800 reverse_host_lookup/defer_ok". An error is now given when options are
3801 supplied for verify items that do not have them. (Maybe reverse_host_
3802 lookup should have a defer_ok option, but that's a different point.)
3804 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3805 defined by RFC 821) to 2048, because there were problems with some AUTH
3806 commands, and RFC 1869 says the size should be increased for extended
3807 SMTP commands that take arguments.
3809 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3812 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3813 "unknown" error; now it says that the functionality isn't in the binary.
3815 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3816 an address' error message when a string expansion fails (syntax or
3817 whatever). Otherwise the password may appear in the log. Following change
3818 PH/42 below, there is no longer a chance of it appearing in a bounce
3821 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3823 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3824 of its addresses was ever tried. (Bugzilla bug #2.)
3826 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3827 the result incorrectly in the debug output. (It correctly added a newline
3828 to what was transported.)
3830 TF/01 Added $received_time.
3832 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3833 commented out examples of how to interface to a virus scanner and to
3834 SpamAssassin. Also added commented examples of av_scanner and
3835 spamd_address settings.
3837 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3838 and controls are allowed in which ACLs. There were a couple of minor
3839 errors. Some of the entries in the conditions table (which is a table of
3840 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3841 negation of where the condition IS allowed.
3843 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3845 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3846 header file does not have a version number, so I've had to invent a new
3847 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3848 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3849 radiusclient), but it was contributed by a Radius user.
3851 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3852 files or directories, for OpenSSL.
3854 PH/19 When an Exim process that is running as root has to create an Exim log
3855 file, it does so in a subprocess that runs as exim:exim so as to get the
3856 ownership right at creation (otherwise, other Exim processes might see
3857 the file with the wrong ownership). There was no test for failure of this
3858 fork() call, which would lead to the process getting stuck as it waited
3859 for a non-existent subprocess. Forks do occasionally fail when resources
3860 run out. I reviewed all the other calls to fork(); they all seem to check
3863 PH/20 When checking for unexpected SMTP input at connect time (before writing
3864 the banner), Exim was not dealing correctly with a non-positive return
3865 from the read() function. If the client had disconnected by this time,
3866 the result was a log entry for a synchronization error with an empty
3867 string after "input=" when read() returned zero. If read() returned -1
3868 (an event I could not check), uninitialized data bytes were printed.
3869 There were reports of junk text (parts of files, etc) appearing after
3872 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3874 PH/22 Added support for macro redefinition, and (re)definition in between
3875 driver and ACL definitions.
3877 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3878 forgetting to use the resulting value; it was using the unexpanded value.
3880 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3881 hadn't been configured. The fix is from Juergen Kreileder, who
3882 understands it better than I do:
3884 "Here's what I see happening with three configured cyrus_sasl
3885 authenticators configured (plain, login, cram-md5):
3887 On startup auth_cyrus_sasl_init() gets called for each of these.
3888 This means three calls to sasl_listmech() without a specified mech_list.
3889 => SASL tests which mechs of all available mechs actually work
3890 => three warnings about OTP not working
3891 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3893 With the patch, sasl_listmech() also gets called three times. But now
3894 SASL's mech_list option is set to the server_mech specified in the the
3895 authenticator. Or in other words, the answer from sasl_listmech()
3896 gets limited to just the mech you're testing for (which is different
3898 => the return list contains just 'plain' or 'login', 'cram-md5' or
3899 nothing depending on the value of ob->server_mech.
3901 I've just tested the patch: Authentication still works fine,
3902 unavailable mechs specified in the exim configuration are still
3903 caught, and the auth.log warnings about OTP are gone."
3905 PH/25 When debugging is enabled, the contents of the command line are added
3906 to the debugging output, even when log_selector=+arguments is not
3909 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3910 answer is "GNU", and only if the return is "GNU/something" is the answer
3913 PH/27 $acl_verify_message is now set immediately after the failure of a
3914 verification in an ACL, and so is available in subsequent modifiers. In
3915 particular, the message can be preserved by coding like this:
3917 warn !verify = sender
3918 set acl_m0 = $acl_verify_message
3920 Previously, $acl_verify_message was set only while expanding "message"
3921 and "log_message" when a very denied access.
3923 PH/28 Modified OS/os.c-Linux with
3925 -#ifndef OS_LOAD_AVERAGE
3926 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3928 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3929 nomenclature these days.)
3931 PH/29 Installed patch from the Sieve maintainer that adds the options
3932 sieve_useraddress and sieve_subaddress to the redirect router.
3934 PH/30 In these circumstances:
3935 . Two addresses routed to the same list of hosts;
3936 . First host does not offer TLS;
3937 . First host accepts first address;
3938 . First host gives temporary error to second address;
3939 . Second host offers TLS and a TLS session is established;
3940 . Second host accepts second address.
3941 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3942 and peerdn, if requested) that were in fact used only for the second
3945 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3946 attention to any local part prefix or suffix that was matched by the
3947 router that accepted the address. It now behaves in the same way as it
3948 does for delivery: the affixes are removed from the local part unless
3949 rcpt_include_affixes is set on the transport.
3951 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3952 timeout during the DATA phase of an incoming message.
3954 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3955 applied a patch sanctioned by the Sieve maintainer.
3957 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3958 the uid or gid is negative. A case of a negative gid caused this to be
3959 noticed. The fix allows for either to be negative.
3961 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3962 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3965 PH/36 The change PH/12 above was broken. Fixed it.
3967 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3968 the grounds that routing the same address twice would always produce the
3969 same answer. This might have been true once, but it is certainly no
3970 longer true now. Routing a child address may depend on the previous
3971 routing that produced that child. Some complicated redirection strategies
3972 went wrong when messages had multiple recipients, and made Exim's
3973 behaviour dependent on the order in which the addresses were given.
3975 I have moved the duplicate checking until after the routing is complete.
3976 Exim scans the addresses that are assigned to local and remote
3977 transports, and removes any duplicates. This means that more work will be
3978 done, as duplicates will always all be routed, but duplicates are
3979 presumably rare, so I don't expect this is of any significance.
3981 For deliveries to pipes, files, and autoreplies, the duplicate checking
3982 still happens during the routing process, since they are not going to be
3985 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3986 It corrects a timeout issue with spamd. This is Ian's comment: "The
3987 background is that sometimes spamd either never reads data from a
3988 connection it has accepted, or it never writes response data. The exiscan
3989 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3990 blindly assumes that writes won't block so it may never time out."
3992 PH/39 Allow G after quota size as well as K and M.
3994 PH/40 The value set for $authenticated_id in an authenticator may not contain
3995 binary zeroes or newlines because the value is written to log lines and
3996 to spool files. There was no check on this. Now the value is run through
3997 the string_printing() function so that such characters are converted to
3998 printable escape sequences.
4000 PH/41 $message_linecount is a new variable that contains the total number of
4001 lines in the message. Compare $body_linecount, which is the count for the
4004 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4005 bounce and delay warning messages, except in certain special cases, which
4008 (a) An SMTP error message from a remote host;
4009 (b) A message specified in a :fail: redirection;
4010 (c) A message specified in a "fail" command in a system filter;
4011 (d) A message specified in a FAIL return from the queryprogram router;
4012 (e) A message specified by the cannot_route_message router option.
4014 In these cases only, Exim does include the error details in bounce and
4015 warning messages. There are also a few cases where bland messages such
4016 as "unrouteable address" or "local delivery error" are given.
4018 PH/43 $value is now also set for the "else" part of a ${run expansion.
4020 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4021 being worked on, but at least Exim now implements the latest version to
4024 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4025 process to complete was treated as a delivery failure, a timeout while
4026 writing the message to the pipe was logged, but erroneously treated as a
4027 successful delivery. Such timeouts include transport filter timeouts. For
4028 consistency with the overall process timeout, these timeouts are now
4029 treated as errors, giving rise to delivery failures by default. However,
4030 there is now a new Boolean option for the pipe transport called
4031 timeout_defer, which, if set TRUE, converts the failures into defers for
4032 both kinds of timeout. A transport filter timeout is now identified in
4035 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4036 systems where "make" and "gmake" are different, calling "gmake" at top
4037 level broke things. I've arranged for the value of $(MAKE) to be passed
4038 from the Makefile to this script so that it can call the same version of
4042 A note about Exim versions 4.44 and 4.50
4043 ----------------------------------------
4045 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4046 changes of various kinds. As a consequence, a big documentation update was
4047 needed. This delayed the release for rather longer than seemed good, especially
4048 in the light of a couple of (minor) security issues. Therefore, the changes
4049 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4050 release. So 4.44 and 4.50 are in effect two different branches that both start
4053 I have left the 4.50 change log unchanged; it contains all the changes since
4054 4.43. The change log for 4.44 is below; many of its items are identical to
4055 those for 4.50. This seems to be the most sensible way to preserve the
4056 historical information.
4062 1. Minor wording change to the doc/README.SIEVE file.
4064 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4065 computation of the current number of files was incorrect.
4067 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4068 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4071 4. Give more explanation in the error message when the command for a transport
4072 filter fails to execute.
4074 5. There are several places where Exim runs a non-Exim command in a
4075 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4076 was being done only for the command run by the queryprogram router. It is
4077 now done for all such subprocesses. The other cases are: ${run, transport
4078 filters, and the commands run by the lmtp and pipe transports.
4080 6. Added CONFIGURE_GROUP build-time option.
4082 7. Some older OS have a limit of 256 on the maximum number of file
4083 descriptors. Exim was using setrlimit() to set 1000 as a large value
4084 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4085 systems. I've change it so that if it can't get 1000, it tries for 256.
4087 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4088 was an oversight, and furthermore, ever since the addition of extra
4089 controls (e.g. 4.43/32), the checks on when to allow different forms of
4090 "control" were broken. There should now be diagnostics for all cases when a
4091 control that does not make sense is encountered.
4093 9. Added the /retain_sender option to "control=submission".
4095 10. $recipients is now available in the predata ACL (oversight).
4097 11. Tidy the search cache before the fork to do a delivery from a message
4098 received from the command line. Otherwise the child will trigger a lookup
4099 failure and thereby defer the delivery if it tries to use (for example) a
4100 cached ldap connection that the parent has called unbind on.
4102 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4103 of $address_data from the recipient verification was clobbered by the
4104 sender verification.
4106 13. The value of address_data from a sender verification is now available in
4107 $sender_address_data in subsequent conditions in the ACL statement.
4109 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4111 15. Added a new option "connect=<time>" to callout options, to set a different
4114 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4115 was its contents. (It was OK if the option was not defined at all.)
4117 17. A "Completed" log line is now written for messages that are removed from
4118 the spool by the -Mrm option.
4120 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4121 information about exactly what failed.
4123 19. Added -dd to debug only the daemon process.
4125 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4126 handles timeouts, both on the server side and network timeouts. Renamed the
4127 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4129 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4130 It is now set to "smtps".
4132 22. $host_address is now set to the target address during the checking of
4133 ignore_target_hosts.
4135 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4136 being passed; this would have caused $sender_host_name to have been used if
4137 matching the list had actually called for a host name (not very likely,
4138 since this list is usually IP addresses). A host name is now passed as
4141 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4142 code that specifies a non-restarting handler (typically sigaction() in
4143 modern systems) in an attempt to fix a rare and obscure crash bug.
4145 25. Narrowed the window for a race in the daemon that could cause it to ignore
4146 SIGCHLD signals. This is not a major problem, because they are used only to
4147 wake it up if nothing else does.
4149 26. A malformed maildirsize file could cause Exim to calculate negative values
4150 for the mailbox size or file count. Odd effects could occur as a result.
4151 The maildirsize information is now recalculated if the size or filecount
4154 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4155 support for a long time. Removed HAVE_SYS_VFS_H.
4157 28. Installed the latest version of exipick from John Jetmore.
4159 29. In an address list, if the pattern was not a regular expression, an empty
4160 subject address (from a bounce message) matched only if the pattern was an
4161 empty string. Non-empty patterns were not even tested. This was the wrong
4162 because it is perfectly reasonable to use an empty address as part of a
4163 database query. An empty address is now tested by patterns that are
4164 lookups. However, all the other forms of pattern expect the subject to
4165 contain a local part and a domain, and therefore, for them, an empty
4166 address still always fails if the pattern is not itself empty.
4168 30. Exim went into a mad DNS loop when attempting to do a callout where the
4169 host was specified on an smtp transport, and looking it up yielded more
4170 than one IP address.
4172 31. Re-factored the code for checking spool and log partition space into a
4173 function that finds that data and another that does the check. The former
4174 is then used to implement four new variables: $spool_space, $log_space,
4175 $spool_inodes, and $log_inodes.
4177 32. The RFC2047 encoding function was originally intended for short strings
4178 such as real names; it was not keeping to the 75-character limit for
4179 encoded words that the RFC imposes. It now respects the limit, and
4180 generates multiple encoded words if necessary. To be on the safe side, I
4181 have increased the buffer size for the ${rfc2047: expansion operator from
4184 33. It is now permitted to omit both strings after an "if" condition; if the
4185 condition is true, the result is "true". As before, when the second string
4186 is omitted, a false condition yields an empty string. This makes it less
4187 cumbersome to write custom ACL and router conditions.
4189 34. Failure to deliver a bounce message always caused it to be frozen, even if
4190 there was an errors_to setting on the router. The errors_to setting is now
4193 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4194 canonical form (fully expanded) before being placed in
4195 $sender_host_address.
4197 36. The table in the code that translates DNS record types into text (T_A to
4198 "A" for instance) was missing entries for NS and CNAME. It is just possible
4199 that this could have caused confusion if both these types were looked up
4200 for the same domain, because the text type is used as part of Exim's
4201 per-process caching. But the chance of anyone hitting this buglet seems
4204 37. The dnsdb lookup has been extended in a number of ways.
4206 (1) There is a new type, "zns", which walks up the domain tree until it
4207 finds some nameserver records. It should be used with care.
4209 (2) There is a new type, "mxh", which is like "mx" except that it returns
4210 just the host names, not the priorities.
4212 (3) It is now possible to give a list of domains (or IP addresses) to be
4213 looked up. The behaviour when one of the lookups defers can be
4214 controlled by a keyword.
4216 (4) It is now possible to specify the separator character for use when
4217 multiple records are returned.
4219 38. The dnslists ACL condition has been extended: it is now possible to supply
4220 a list of IP addresses and/or domains to be looked up in a particular DNS
4223 39. Added log_selector=+queue_time_overall.
4225 40. When running the queue in the test harness, wait just a tad after forking a
4226 delivery process, to get repeatability of debugging output.
4228 41. Include certificate and key file names in error message when GnuTLS fails
4229 to set them up, because the GnuTLS error message doesn't include the name
4230 of the failing file when there is a problem reading it.
4232 42. Allow both -bf and -bF in the same test run.
4234 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4236 44. The "Exiscan patch" is now merged into the mainline Exim source.
4238 45. Sometimes the final signoff response after QUIT could fail to get
4239 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4240 < 0 before doing a fflush(). This bug looks as though it goes back to the
4241 introduction of TLS in release 3.20, but "sometimes" must have been rare
4242 because the tests only now provoked it.
4244 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4245 (this can affect the format of dates).
4247 47. exim_tidydb, when checking for the continued existence of a message for
4248 which it has found a message-specific retry record, was not finding
4249 messages that were in split spool directories. Consequently, it was
4250 deleting retry records that should have stayed in existence.
4252 48. Steve fixed some bugs in eximstats.
4254 49. The SPA authentication driver was not abandoning authentication and moving
4255 on to the next authenticator when an expansion was forced to fail,
4256 contradicting the general specification for all authenticators. Instead it
4257 was generating a temporary error. It now behaves as specified.
4259 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4260 (the order specifies the preference for clients). The order is now AES256,
4261 AES128, 3DES, ARCFOUR128.
4263 51. Small patch to Sieve code - explicitly set From: when generating an
4266 52. Exim crashed if a remote delivery caused a very long error message to be
4267 recorded - for instance if somebody sent an entire SpamAssassin report back
4268 as a large number of 550 error lines. This bug was coincidentally fixed by
4269 increasing the size of one of Exim's internal buffers (big_buffer) that
4270 happened as part of the Exiscan merge. However, to be on the safe side, I
4271 have made the code more robust (and fixed the comments that describe what
4274 53. Now that there can be additional text after "Completed" in log lines (if
4275 the queue_time_overall log selector is set), a one-byte patch to exigrep
4276 was needed to allow it to recognize "Completed" as not the last thing in
4279 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4280 patch that reportedly fixes this has been added. I am not expert enough to
4281 create a test for it. This is what the patch creator wrote:
4283 "I found a little strange behaviour of ldap code when working with
4284 Windows 2003 AD Domain, where users was placed in more than one
4285 Organization Units. When I tried to give exim partial DN, the exit code
4286 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4287 But simultaneously result of request was absolutely normal ldap result,
4288 so I produce this patch..."
4290 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4291 REFERENCE, so I have modified the code to exclude the patch when that macro
4294 55. Some experimental protocols are using DNS PTR records for new purposes. The
4295 keys for these records are domain names, not reversed IP addresses. The
4296 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4297 leaves it alone. Component reversal etc. now happens only for IP addresses.
4300 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4302 57. Double the size of the debug message buffer (to 2048) so that more of very
4303 long debug lines gets shown.
4305 58. The exicyclog utility now does better if the number of log files to keep
4306 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4308 59. Two changes related to the smtp_active_hostname option:
4310 (1) $smtp_active_hostname is now available as a variable.
4311 (2) The default for smtp_banner uses $smtp_active_hostname instead
4312 of $primary_hostname.
4314 60. The host_aton() function is supposed to be passed a string that is known
4315 to be a valid IP address. However, in the case of IPv6 addresses, it was
4316 not checking this. This is a hostage to fortune. Exim now panics and dies
4317 if the condition is not met. A case was found where this could be provoked
4318 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4319 components; fortuitously, this particular loophole had already been fixed
4320 by change 4.50/55 above.
4322 If there are any other similar loopholes, the new check in host_aton()
4323 itself should stop them being exploited. The report I received stated that
4324 data on the command line could provoke the exploit when Exim was running as
4325 exim, but did not say which command line option was involved. All I could
4326 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4327 running as the user.
4330 61. There was a buffer overflow vulnerability in the SPA authentication code
4331 (which came originally from the Samba project). I have added a test to the
4332 spa_base64_to_bits() function which I hope fixes it.
4335 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4336 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4338 63. The daemon start-up calls getloadavg() while still root for those OS that
4339 need the first call to be done as root, but it missed one case: when
4340 deliver_queue_load_max is set with deliver_drop_privilege. This is
4341 necessary for the benefit of the queue runner, because there is no re-exec
4342 when deliver_drop_privilege is set.
4344 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4345 This has been fixed.
4347 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4348 was in use, was not putting the data itself into the right store pool;
4349 consequently, it could be overwritten for a subsequent message in the same
4350 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4353 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4355 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4356 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4357 the function were treating the return as a boolean value, which happened to
4358 work because 0=false and not-0=true, but is not correct code.
4360 68. The host_aton() function was not handling scoped IPv6 addresses (those
4361 with, for example, "%eth0" on the end) correctly.
4363 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4364 negated items (that is, ~something) in unsigned ints. Some compilers
4365 apparently mutter when there is no cast.
4367 70. If an address verification called from an ACL failed, and did not produce a
4368 user-specific message (i.e. there was only a "system" message), nothing was
4369 put in $acl_verify_message. In this situation, it now puts the system
4372 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4373 session; change 4.31/43 added the unwanted input to the log line - except
4374 that it did not do this in the start of session case. It now does.
4376 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4377 This is wrong and can cause the other end to generate a synchronization
4378 error if it is another Exim or anything else that does the synchronization
4379 check. A QUIT command is no longer sent after a timeout.
4381 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4382 during host lookups.
4384 74. The defer_ok option of callout verification was not working if it was used
4385 when verifying addresses in header lines, that is, for this case:
4387 verify = header_sender/callout=defer_ok
4389 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4390 those file descriptors could be used for SMTP connections. If anything
4391 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4392 could be sent to the SMTP client, causing chaos. The daemon now opens
4393 stdin, stdout, and stderr to /dev/null when it puts itself into the
4396 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4397 log by default. The user can override this with suitable Perl magic.
4399 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4400 the log message when discard triggers, was not working for the DATA ACL or
4401 for the non-SMTP ACL.
4403 78. Error message wording change in sieve.c.
4405 79. If smtp_accept_max_per_host was set, the number of connections could be
4406 restricted to fewer than expected, because the daemon was trying to set up
4407 a new connection before checking whether the processes handling previous
4408 connections had finished. The check for completed processes is now done
4409 earlier. On busy systems, this bug wouldn't be noticed because something
4410 else would have woken the daemon, and it would have reaped the completed
4413 80. If a message was submitted locally by a user whose login name contained one
4414 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4415 It caused a spool format error. I have fixed the spool reading code. A
4416 related problem was that the "from" clause in the Received: line became
4417 illegal because of the space(s). It is now covered by ${quote_local_part.
4419 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4422 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4424 83. Patch from Sieve maintainer to fix unterminated string problem in
4425 "vacation" handling.
4427 84. Some minor changes to the Linux configuration files to help with other
4428 OS variants using glibc.
4430 85. One more patch for Sieve to update vacation handling to latest spec.
4433 ----------------------------------------------------
4434 See the note above about the 4.44 and 4.50 releases.
4435 ----------------------------------------------------
4441 1. Change 4.43/35 introduced a bug that caused file counts to be
4442 incorrectly computed when quota_filecount was set in an appendfile
4445 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4446 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4449 3. Give more explanation in the error message when the command for a transport
4450 filter fails to execute.
4452 4. There are several places where Exim runs a non-Exim command in a
4453 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4454 was being done only for the command run by the queryprogram router. It is
4455 now done for all such subprocesses. The other cases are: ${run, transport
4456 filters, and the commands run by the lmtp and pipe transports.
4458 5. Some older OS have a limit of 256 on the maximum number of file
4459 descriptors. Exim was using setrlimit() to set 1000 as a large value
4460 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4461 systems. I've change it so that if it can't get 1000, it tries for 256.
4463 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4464 was an oversight, and furthermore, ever since the addition of extra
4465 controls (e.g. 4.43/32), the checks on when to allow different forms of
4466 "control" were broken. There should now be diagnostics for all cases when a
4467 control that does not make sense is encountered.
4469 7. $recipients is now available in the predata ACL (oversight).
4471 8. Tidy the search cache before the fork to do a delivery from a message
4472 received from the command line. Otherwise the child will trigger a lookup
4473 failure and thereby defer the delivery if it tries to use (for example) a
4474 cached ldap connection that the parent has called unbind on.
4476 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4477 of $address_data from the recipient verification was clobbered by the
4478 sender verification.
4480 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4481 was its contents. (It was OK if the option was not defined at all.)
4483 11. A "Completed" log line is now written for messages that are removed from
4484 the spool by the -Mrm option.
4486 12. $host_address is now set to the target address during the checking of
4487 ignore_target_hosts.
4489 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4490 being passed; this would have caused $sender_host_name to have been used if
4491 matching the list had actually called for a host name (not very likely,
4492 since this list is usually IP addresses). A host name is now passed as
4495 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4496 code that specifies a non-restarting handler (typically sigaction() in
4497 modern systems) in an attempt to fix a rare and obscure crash bug.
4499 15. Narrowed the window for a race in the daemon that could cause it to ignore
4500 SIGCHLD signals. This is not a major problem, because they are used only to
4501 wake it up if nothing else does.
4503 16. A malformed maildirsize file could cause Exim to calculate negative values
4504 for the mailbox size or file count. Odd effects could occur as a result.
4505 The maildirsize information is now recalculated if the size or filecount
4508 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4509 support for a long time. Removed HAVE_SYS_VFS_H.
4511 18. Updated exipick to current release from John Jetmore.
4513 19. Allow an empty sender to be matched against a lookup in an address list.
4514 Previously the only cases considered were a regular expression, or an
4517 20. Exim went into a mad DNS lookup loop when doing a callout where the
4518 host was specified on the transport, if the DNS lookup yielded more than
4521 21. The RFC2047 encoding function was originally intended for short strings
4522 such as real names; it was not keeping to the 75-character limit for
4523 encoded words that the RFC imposes. It now respects the limit, and
4524 generates multiple encoded words if necessary. To be on the safe side, I
4525 have increased the buffer size for the ${rfc2047: expansion operator from
4528 22. Failure to deliver a bounce message always caused it to be frozen, even if
4529 there was an errors_to setting on the router. The errors_to setting is now
4532 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4533 canonical form (fully expanded) before being placed in
4534 $sender_host_address.
4536 24. Updated eximstats to version 1.33
4538 25. Include certificate and key file names in error message when GnuTLS fails
4539 to set them up, because the GnuTLS error message doesn't include the name
4540 of the failing file when there is a problem reading it.
4542 26. Expand error message when OpenSSL has problems setting up cert/key files.
4545 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4546 (this can affect the format of dates).
4548 28. exim_tidydb, when checking for the continued existence of a message for
4549 which it has found a message-specific retry record, was not finding
4550 messages that were in split spool directories. Consequently, it was
4551 deleting retry records that should have stayed in existence.
4553 29. eximstats updated to version 1.35
4554 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4555 1.35 - bugfix such that pie charts by volume are generated correctly
4557 30. The SPA authentication driver was not abandoning authentication and moving
4558 on to the next authenticator when an expansion was forced to fail,
4559 contradicting the general specification for all authenticators. Instead it
4560 was generating a temporary error. It now behaves as specified.
4562 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4563 (the order specifies the preference for clients). The order is now AES256,
4564 AES128, 3DES, ARCFOUR128.
4566 31. Small patch to Sieve code - explicitly set From: when generating an
4569 32. Exim crashed if a remote delivery caused a very long error message to be
4570 recorded - for instance if somebody sent an entire SpamAssassin report back
4571 as a large number of 550 error lines. This bug was coincidentally fixed by
4572 increasing the size of one of Exim's internal buffers (big_buffer) that
4573 happened as part of the Exiscan merge. However, to be on the safe side, I
4574 have made the code more robust (and fixed the comments that describe what
4577 33. Some experimental protocols are using DNS PTR records for new purposes. The
4578 keys for these records are domain names, not reversed IP addresses. The
4579 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4580 leaves it alone. Component reversal etc. now happens only for IP addresses.
4583 34. The host_aton() function is supposed to be passed a string that is known
4584 to be a valid IP address. However, in the case of IPv6 addresses, it was
4585 not checking this. This is a hostage to fortune. Exim now panics and dies
4586 if the condition is not met. A case was found where this could be provoked
4587 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4588 components; fortuitously, this particular loophole had already been fixed
4589 by change 4.50/55 or 4.44/33 above.
4591 If there are any other similar loopholes, the new check in host_aton()
4592 itself should stop them being exploited. The report I received stated that
4593 data on the command line could provoke the exploit when Exim was running as
4594 exim, but did not say which command line option was involved. All I could
4595 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4596 running as the user.
4599 35. There was a buffer overflow vulnerability in the SPA authentication code
4600 (which came originally from the Samba project). I have added a test to the
4601 spa_base64_to_bits() function which I hope fixes it.
4604 36. The daemon start-up calls getloadavg() while still root for those OS that
4605 need the first call to be done as root, but it missed one case: when
4606 deliver_queue_load_max is set with deliver_drop_privilege. This is
4607 necessary for the benefit of the queue runner, because there is no re-exec
4608 when deliver_drop_privilege is set.
4610 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4611 was in use, was not putting the data itself into the right store pool;
4612 consequently, it could be overwritten for a subsequent message in the same
4613 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4616 38. Sometimes the final signoff response after QUIT could fail to get
4617 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4618 < 0 before doing a fflush(). This bug looks as though it goes back to the
4619 introduction of TLS in release 3.20, but "sometimes" must have been rare
4620 because the tests only now provoked it.
4626 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4627 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4628 Now it returns an integer. A number of calls were still expecting a T/F
4629 return. Fortuitously, in all cases, the tests worked in OK situations,
4630 which is the norm. However, things would have gone wrong on any write
4631 failures on the smtp file descriptor. This function is used when sending
4632 messages over SMTP and also when doing verify callouts.
4634 2. When Exim is called to do synchronous delivery of a locally submitted
4635 message (the -odf or -odi options), it no longer closes stderr before doing
4638 3. Implemented the mua_wrapper option.
4640 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4642 5. Implemented the functions header_remove(), header_testname(),
4643 header_add_at_position(), and receive_remove_recipient(), and exported them
4646 6. If an ACL "warn" statement specified the addition of headers, Exim already
4647 inserted X-ACL-Warn: at the start if there was no header name. However, it
4648 was not making this test for the second and subsequent header lines if
4649 there were newlines in the string. This meant that an invalid header could
4650 be inserted if Exim was badly configured.
4652 7. Allow an ACL "warn" statement to add header lines at the start or after all
4653 the Received: headers, as well as at the end.
4655 8. Added the rcpt_4xx retry error code.
4657 9. Added postmaster_mailfrom=xxx to callout verification option.
4659 10. Added mailfrom=xxxx to the callout verification option, for verify=
4662 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4663 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4665 12. Inserted some casts to stop certain compilers warning when using pointer
4666 differences as field lengths or precisions in printf-type calls (mostly
4667 affecting debugging statements).
4669 13. Added optional readline() support for -be (dynamically loaded).
4671 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4672 same clock tick as a message's arrival, so that its received time was the
4673 same as the "first fail" time on the retry record, and that message
4674 remained on the queue past the ultimate address timeout, every queue runner
4675 would try a delivery (because it was past the ultimate address timeout) but
4676 after another failure, the ultimate address timeout, which should have then
4677 bounced the address, did not kick in. This was a "< instead of <=" error;
4678 in most cases the first failure would have been in the next clock tick
4679 after the received time, and all would be well.
4681 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4682 being recognized when the domain list was tested by the match_domain
4683 condition in an expansion string.
4685 16. Added the ${str2b64: operator.
4687 17. Exim was always calling setrlimit() to set a large limit for the number of
4688 processes, without checking whether the existing limit was already
4689 adequate. (It did check for the limit on file descriptors.) Furthermore,
4690 errors from getrlimit() and setrlimit() were being ignored. Now they are
4691 logged to the main and panic logs, but Exim does carry on, to try to do its
4692 job under whatever limits there are.
4694 18. Imported PCRE 5.0.
4696 19. Trivial typo in log message " temporarily refused connection" (the leading
4699 20. If the log selector return_path_on_delivery was set and an address was
4700 redirected to /dev/null, the delivery process crashed because it assumed
4701 that a return path would always be set for a "successful" delivery. In this
4702 case, the whole delivery is bypassed as an optimization, and therefore no
4705 21. Internal re-arrangement: the function for sending a challenge and reading
4706 a response while authentication was assuming a zero-terminated challenge
4707 string. It's now changed to take a pointer and a length, to allow for
4708 binary data in such strings.
4710 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4712 23. Exim was not respecting finduser_retries when seeking the login of the
4713 uid under which it was called; it was always trying 10 times. (The default
4714 setting of finduser_retries is zero.) Also, it was sleeping after the final
4715 failure, which is pointless.
4717 24. Implemented tls_on_connect_ports.
4719 25. Implemented acl_smtp_predata.
4721 26. If the domain in control=submission is set empty, Exim assumes that the
4722 authenticated id is a complete email address when it generates From: or
4723 Sender: header lines.
4725 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4726 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4727 chown and chgrp in /bin and hostname in /usr/bin.
4729 28. Exim was keeping the "process log" file open after each use, just as it
4730 does for the main log. This opens the possibility of it remaining open for
4731 long periods when the USR1 signal hits a daemon. Occasional processlog
4732 errors were reported, that could have been caused by this. Anyway, it seems
4733 much more sensible not to leave this file open at all, so that is what now
4736 29. The long-running daemon process does not normally write to the log once it
4737 has entered its main loop, and it closes the log before doing so. This is
4738 so that log files can straightforwardly be renamed and moved. However,
4739 there are a couple of unusual error situations where the daemon does write
4740 log entries, and I had neglected to close the log afterwards.
4742 30. The text of an SMTP error response that was received during a remote
4743 delivery was being truncated at 512 bytes. This is too short for some of
4744 the long messages that one sometimes sees. I've increased the limit to
4747 31. It is now possible to make retry rules that apply only when a message has a
4748 specific sender, in particular, an empty sender.
4750 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4751 it possible to be selective about when SMTP synchronization is enforced.
4753 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4755 32. Implemented hosts_connection_nolog.
4757 33. Added an ACL for QUIT.
4759 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4762 35. Added mailbox_size and mailbox_filecount to appendfile.
4764 36. Added control = no_multiline_responses to ACLs.
4766 37. There was a bug in the logic of the code that waits for the clock to tick
4767 in the case where the clock went backwards by a substantial amount such
4768 that the microsecond fraction of "now" was more than the microsecond
4769 fraction of "then" (but the whole seconds number was less).
4771 38. Added support for the libradius Radius client library this is found on
4772 FreeBSD (previously only the radiusclient library was supported).
4778 1. When certain lookups returned multiple values in the form name=value, the
4779 quoting of the values was not always being done properly. Specifically:
4780 (a) If the value started with a double quote, but contained no whitespace,
4782 (b) If the value contained whitespace other than a space character (i.e.
4783 tabs or newlines or carriage returns) it was not quoted.
4784 This fix has been applied to the mysql and pgsql lookups by writing a
4785 separate quoting function and calling it from the lookup code. The fix
4786 should probably also be applied to nisplus, ibase and oracle lookups, but
4787 since I cannot test any of those, I have not disturbed their existing code.
4789 2. A hit in the callout cache for a specific address caused a log line with no
4790 reason for rejecting RCPT. Now it says "Previous (cached) callout
4791 verification failure".
4793 3. There was an off-by-one bug in the queryprogram router. An over-long
4794 return line was truncated at 256 instead of 255 characters, thereby
4795 overflowing its buffer with the terminating zero. As well as fixing this, I
4796 have increased the buffer size to 1024 (and made a note to document this).
4798 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4799 when Exim is waiting for an SMTP response from a remote server, Exim
4800 restarts its select() call on the socket, thereby resetting its timeout.
4801 This is not a problem when such interrupts are rare. Somebody set up a cron
4802 job to run exiwhat every 2 minutes, which is less than the normal select()
4803 timeout (5 or 10 minutes). This meant that the select() timeout never
4804 kicked in because it was always reset. I have fixed this by comparing the
4805 time when an interrupt arrives with the time at the start of the first call
4806 to select(). If more time than the timeout has elapsed, the interrupt is
4807 treated as a timeout.
4809 5. Some internal re-factoring in preparation for the addition of Sieve
4810 extensions (by MH). In particular, the "personal" test is moved to a
4811 separate function, and given an option for scanning Cc: and Bcc: (which is
4812 not set for Exim filters).
4814 6. When Exim created an email address using the login of the caller as the
4815 local part (e.g. when creating a From: or Sender: header line), it was not
4816 quoting the local part when it contained special characters such as @.
4818 7. Installed new OpenBSD configuration files.
4820 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4821 try to make them clearer.
4823 9. Callout options, other than the timeout value, were being ignored when
4824 verifying sender addresses in header lines. For example, when using
4826 verify = header_sender/callout=no_cache
4828 the cache was (incorrectly) being used.
4830 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4831 only the Cygwin environment.
4833 11. When return_path_on_delivery was set as a log selector, if different remote
4834 addresses in the same message used different return paths and parallel
4835 remote delivery occurred, the wrong values would sometimes be logged.
4836 (Whenever a remote delivery process finished, the return path value from
4837 the most recently started remote delivery process was logged.)
4839 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4840 lines when AUTH and/or TLS are in use. This is the "received protocol"
4841 field. Exim used to use "asmtp" for authenticated SMTP, without any
4842 indication (in the protocol name) for TLS use. Now it follows the RFC and
4843 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4844 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4845 names appear in log lines as well as in Received: header lines.
4847 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4848 extensions, and comparison tests, and to fix some bugs.
4850 14. Changes to the "personal" filter test:
4852 (1) The test was buggy in that it was just doing the equivalent of
4853 "contains" tests on header lines. For example, if a user's address was
4854 anne@some.where, the "personal" test would incorrectly be true for
4856 To: susanne@some.where
4858 This test is now done by extracting each address from the header in turn,
4859 and checking the entire address. Other tests that are part of "personal"
4860 are now done using regular expressions (for example, to check local parts
4861 of addresses in From: header lines).
4863 (2) The list of non-personal local parts in From: addresses has been
4864 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4865 taken from the Sieve specification recommendations.
4867 (3) If the message contains any header line starting with "List-" it is
4868 treated as non-personal.
4870 (4) The test for "circular" in the Subject: header line has been removed
4871 because it now seems ill-conceived.
4873 15. Minor typos in src/EDITME comments corrected.
4875 16. Installed latest exipick from John Jetmore.
4877 17. If headers_add on a router specified a text string that was too long for
4878 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4879 of string_sprintf() is now avoided.
4881 18. $message_body_size was not set (it was always zero) when running the DATA
4882 ACL and the local_scan() function.
4884 19. For the "mail" command in an Exim filter, no default was being set for
4885 the once_repeat time, causing a random time value to be used if "once" was
4886 specified. (If the value happened to be <= 0, no repeat happened.) The
4887 default is now 0s, meaning "never repeat". The "vacation" command was OK
4888 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4889 (I found it when inspecting the code).
4891 20. There is now an overall timeout for performing a callout verification. It
4892 defaults to 4 times the callout timeout, which applies to individual SMTP
4893 commands during the callout. The overall timeout applies when there is more
4894 than one host that can be tried. The timeout is checked before trying the
4895 next host. This prevents very long delays if there are a large number of
4896 hosts and all are timing out (e.g. when the network connections are timing
4897 out). The value of the overall timeout can be changed by specifying an
4898 additional sub-option for "callout", called "maxwait". For example:
4900 verify = sender/callout=5s,maxwait=20s
4902 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4903 to the end before writing, but this should make it even safer).
4905 22. Exim was forgetting that it had advertised PIPELINING for the second and
4906 subsequent messages on an SMTP connection. It was also not resetting its
4907 memory on STARTTLS and an internal HELO.
4909 23. When Exim logs an SMTP synchronization error within a session, it now
4910 records whether PIPELINING has been advertised or not.
4912 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4913 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4914 rather than long int.
4916 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4918 26. Added the never_mail option to autoreply.
4924 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4925 crash if the getsockname() call failed; this can happen if a connection is
4926 closed very soon after it is established. The problem was simply in the
4927 order in which certain operations were done, causing Exim to try to write
4928 to the SMTP stream before it had set up the file descriptor. The bug has
4929 been fixed by making things happen in the correct order.
4935 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4936 before the connection was closed, thus losing the rejection response.
4938 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4939 some early Solaris releases, but causes trouble in current releases where
4940 socklen_t is defined.
4942 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4945 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4948 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4949 that releasing the top of it at the end releases what was used for sub-
4950 expansions (unless the block got too big). However, discard this block if
4951 the first thing is a variable or header, so that we can use its block when
4952 it is dynamic (useful for very large $message_headers, for example).
4954 6. Lookups now cache *every* query, not just the most recent. A new, separate
4955 store pool is used for this. It can be recovered when all lookup caches are
4956 flushed. Lookups now release memory at the end of their result strings.
4957 This has involved some general refactoring of the lookup sources.
4959 7. Some code has been added to the store_xxx() functions to reduce the amount
4960 of flapping under certain conditions.
4962 8. log_incoming_interface used to affect only the <= reception log lines. Now
4963 it causes the local interface and port to be added to several more SMTP log
4964 lines, for example "SMTP connection from", and rejection lines.
4966 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4968 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4970 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4971 could be overwritten at the end of the current message (or the start of a
4972 new message if it was set in a HELO ACL). The value is now preserved for
4973 the duration of the SMTP connection.
4975 12. If a transport had a headers_rewrite setting, and a matching header line
4976 contained an unqualified address, that address was qualified, even if it
4977 did not match any rewriting rules. The underlying bug was that the values
4978 of the flags that permit the existence of unqualified sender and recipient
4979 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4980 non-local messages, and by -bnq for local messages) were not being
4981 preserved with the message after it was received.
4983 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4984 "next input=" as part of the text comprising the host identity instead of
4985 the correct text. The code was using the same buffer for two different
4986 strings. However, depending on which order the printing function evaluated
4987 its arguments, the bug did not always show up. Under Linux, for example, my
4988 test suite worked just fine.
4990 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4991 doesn't work with some older versions of Perl. It has been changed to "my",
4992 which in any case is probably the better facility to use.
4994 15. A really picky compiler found some instances of statements for creating
4995 error messages that either had too many or two few arguments for the format
4998 16. The size of the buffer for calls to the DNS resolver has been increased
4999 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5000 for addresses that have a lot of PTR records. This alleviates a problem; it
5001 does not fully solve it.
5003 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5004 buffer now truncates the list and logs the incident, which is the same
5005 action as happens when Exim is looking up a host name and its aliases.
5006 Previously in this situation something unpredictable would happen;
5007 sometimes it was "internal error: store_reset failed".
5009 18. If a server dropped the connection unexpectedly when an Exim client was
5010 using GnuTLS and trying to read a response, the client delivery process
5011 crashed while trying to generate an error log message.
5013 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5014 string, for example:
5016 warn message = H1: something\nH2: something
5018 the text was added as a single header line from Exim's point of view
5019 though it ended up OK in the delivered message. However, searching for the
5020 second and subsequent header lines using $h_h2: did not work. This has been
5021 fixed. Similarly, if a system filter added multiple headers in this way,
5022 the routers could not see them.
5024 20. Expanded the error message when iplsearch is called with an invalid key to
5025 suggest using net-iplsearch in a host list.
5027 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5028 ACLs are no longer actually imposed (and a message to that effect is
5031 22. If a "gecos" field in a passwd entry contained escaped characters, in
5032 particular, if it contained a \" sequence, Exim got it wrong when building
5033 a From: or a Sender: header from that name. A second bug also caused
5034 incorrect handling when an unquoted " was present following a character
5035 that needed quoting.
5037 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5038 was not being matched caselessly.
5040 24. Arranged for all hyphens in the exim.8 source to be escaped with
5043 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5044 properly. Recipient callout cache records were still being keyed to include
5045 the sender, even when use_sender was set false. This led to far more
5046 callouts that were necessary. The sender is no longer included in the key
5047 when use_sender is false.
5049 26. Added "control = submission" modifier to ACLs.
5051 27. Added the ${base62d: operator to decode base 62 numbers.
5053 28. dnsdb lookups can now access SRV records.
5055 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5056 the configuration file.
5058 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5059 (-v) mode. This makes the output for a verbose queue run more intelligible.
5061 31. Added a use_postmaster feature to recipient callouts.
5063 32. Added the $body_zerocount variable, containing the number of binary zero
5064 bytes in the message body.
5066 33. The time of last modification of the "new" subdirectory is now used as the
5067 "mailbox time last read" when there is a quota error for a maildir
5070 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5072 35. Added +ignore_unknown as a special item in host lists.
5074 36. Code for decoding IPv6 addresses in host lists is now included, even if
5075 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5076 address was recognized as an IP address, but was then not correctly decoded
5077 into binary, causing unexpected and incorrect effects when compared with
5084 1. Very minor rewording of debugging text in manualroute to say "list of
5085 hosts" instead of "hostlist".
5087 2. If verify=header_syntax was set, and a header line with an unqualified
5088 address (no domain) and a large number of spaces between the end of the
5089 name and the colon was received, the reception process suffered a buffer
5090 overflow, and (when I tested it) crashed. This was caused by some obsolete
5091 code that should have been removed. The fix is to remove it!
5093 3. When running in the test harness, delay a bit after writing a bounce
5094 message to get a bit more predictability in the log output.
5096 4. Added a call to search_tidyup() just before forking a reception process. In
5097 theory, someone could use a lookup in the expansion of smtp_accept_max_
5098 per_host which, without the tidyup, could leave open a database connection.
5100 5. Added the variables $recipient_data and $sender_data which get set from a
5101 lookup success in an ACL "recipients" or "senders" condition, or a router
5102 "senders" option, similar to $domain_data and $local_part_data.
5104 6. Moved the writing of debug_print from before to after the "senders" test
5107 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5108 problems for message scanning, either using a data ACL, or using
5109 local_scan() because the Received: header was not generated till after they
5110 were called (in order to set the time as the time of reception completion).
5111 I have revised the way this works. The header is now generated after the
5112 body is received, but before the ACL or local_scan() are called. After they
5113 are run, the timestamp in the header is updated.
5119 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5120 before starting a queue runner without re-exec. This happened only when
5121 deliver_drop_privilege was set or when the Exim user was set to root. The
5122 effect of the bug was that timeouts during subsequent deliveries caused
5123 crashes instead of being properly handled. The handler is now left at its
5124 default (and expected) setting.
5126 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5127 message, again when deliver_drop_privilege is set or Exim is run as root.
5128 The bug described in (1) was not present in this case, but the tidying up
5129 of the other signals was missing. I have made the two cases consistent.
5131 3. The ignore_target_hosts setting on a manualroute router was being ignored
5132 for hosts that were looked up using the /MX notation.
5134 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5137 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5138 operated on the sender address. After changing the $sender_address to <>
5139 for the sender address verify, Exim was re-instated it as the original
5140 (before rewriting) address, but remembering that it had rewritten it, so it
5141 wasn't rewriting it again. This bug also had the effect of breaking the
5142 sender address verification caching when the sender address was rewritten.
5144 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5145 This has been changed so that if the ip literal address matches
5146 ignore_target_hosts, the router declines.
5148 7. Added expansion conditions match_domain, match_address, and match_local_
5149 part (NOT match_host).
5151 8. The placeholder for the Received: header didn't have a length field set.
5153 9. Added code to Exim itself and to exim_lock to test for a specific race
5154 condition that could lead to file corruption when using MBX delivery. The
5155 issue is with the lockfile that is created in /tmp. If this file is removed
5156 after a process has opened it but before that process has acquired a lock,
5157 there is the potential for a second process to recreate the file and also
5158 acquire a lock. This could lead to two Exim processes writing to the file
5159 at the same time. The added code performs the same test as UW imapd; it
5160 checks after acquiring the lock that its file descriptor still refers to
5161 the same named file.
5163 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5164 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5167 11. Added the smtp_active_hostname option. If used, this will typically be made
5168 to depend on the incoming interface address. Because $interface_address is
5169 not set up until the daemon has forked a reception process, error responses
5170 that can happen earlier (such as "too many connections") no longer contain
5173 12. If an expansion in a condition on a "warn" statement fails because a lookup
5174 defers, the "warn" statement is abandoned, and the next ACL statement is
5175 processed. Previously this caused the whole ACL to be aborted.
5177 13. Added the iplsearch lookup type.
5179 14. Added ident_timeout as a log selector.
5181 15. Added tls_certificate_verified as a log selector.
5183 16. Added a global option tls_require_ciphers (compare the smtp transport
5184 option of the same name). This controls incoming TLS connections.
5186 17. I finally figured out how to make tls_require_ciphers do a similar thing
5187 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5188 before starting the TLS session.
5190 18. Tabs are now shown as \t in -bP output.
5192 19. If the log selector return_path_on_delivery was set, Exim crashed when
5193 bouncing a message because it had too many Received: header lines.
5195 20. If two routers both had headers_remove settings, and the first one included
5196 a superfluous trailing colon, the final name in the first list and the
5197 first name in the second list were incorrectly joined into one item (with a
5198 colon in the middle).
5204 1. Added -C and -D options to the exinext utility, mainly to make it easier
5205 to include in the automated testing, but these could be helpful when
5206 multiple configurations are in use.
5208 2. The exinext utility was not formatting the output nicely when there was
5209 an alternate port involved in the retry record key, nor when there was a
5210 message id as well (for retries that were specific to a specific message
5211 and a specific host). It was also confused by IPv6 addresses, because of
5212 the additional colons they contain. I have fixed the IPv4 problem, and
5213 patched it up to do a reasonable job for IPv6.
5215 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5216 delivery, the log line now contains "pipelined" if PIPELINING was used.
5218 4. An SMTP transport process used to panic and die if the bind() call to set
5219 an explicit outgoing interface failed. This has been changed; it is now
5220 treated in the same way as a connect() failure.
5222 5. A reference to $sender_host_name in the part of a conditional expansion
5223 that was being skipped was still causing a DNS lookup. This no longer
5226 6. The def: expansion condition was not recognizing references to header lines
5227 that used bh_ and bheader_.
5229 7. Added the _cache feature to named lists.
5231 8. The code for checking quota_filecount in the appendfile transport was
5232 allowing one more file than it should have been.
5234 9. For compatibility with Sendmail, the command line option
5242 and sets the incoming protocol and host name (for trusted callers). The
5243 host name and its colon can be omitted when only the protocol is to be set.
5244 Note the Exim already has two private options, -pd and -ps, that refer to
5245 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5246 "s", but I don't think that's a major issue.
5248 10. A number of refactoring changes to the code, none of which should affect
5251 (a) The number of logging options was getting close to filling up the
5252 32-bit word that was used as a bit map. I have split them into two classes:
5253 those that are passed in the argument to log_write(), and those that are
5254 only ever tested independently outside of that function. These are now in
5255 separate 32-bit words, so there is plenty of room for expansion again.
5256 There is no change in the user interface or the logging behaviour.
5258 (b) When building, for example, log lines, the code previously used a
5259 macro that called string_cat() twice, in order to add two strings. This is
5260 not really sufficiently general. Furthermore, there was one instance where
5261 it was actually wrong because one of the argument was used twice, and in
5262 one call a function was used. (As it happened, calling the function twice
5263 did not affect the overall behaviour.) The macro has been replaced by a
5264 function that can join an arbitrary number of extra strings onto a growing
5267 (c) The code for expansion conditions now uses a table and a binary chop
5268 instead of a serial search (which was left over from when there were very
5269 few conditions). Also, it now recognizes conditions like "pam" even when
5270 the relevant support is not compiled in: a suitably worded error message is
5271 given if an attempt is made to use such a condition.
5273 11. Added ${time_interval:xxxxx}.
5275 12. A bug was causing one of the ddress fields not to be passed back correctly
5276 from remote delivery subprocesses. The field in question was not being
5277 subsequently used, so this caused to problems in practice.
5279 13. Added new log selectors queue_time and deliver_time.
5281 14. Might have fixed a bug in maildirsizefile handling that threw up
5282 "unexpected character" debug warnings, and recalculated the data
5283 unnecessarily. In any case, I expanded the warning message to give more
5286 15. Added the message "Restricted characters in address" to the statements in
5287 the default ACL that block characters like @ and % in local parts.
5289 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5290 Three changes have been made:
5292 (a) There was a serious bug; a negative response to MAIL caused the whole
5293 recipient domain to be cached as invalid, thereby blocking all messages
5294 to all local parts at the same domain, from all senders. This bug has
5295 been fixed. The domain is no longer cached after a negative response to
5296 MAIL if the sender used is not empty.
5298 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5301 (c) A new callout option, "use_sender" has been added for people who want
5302 the modified behaviour.
5308 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5311 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5312 indeed breaks things for older releases.
5314 3. Added additional logging to the case where there is a problem reading data
5315 from a filter that is running in a subprocess using a pipe, in order to
5316 try to track down a specific problem.
5318 4. Testing facility fudge: when running in the test harness and attempting
5319 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5320 getting "No route to host". Convert this to a timeout.
5322 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5325 6. Some OS don't have socklen_t but use size_t instead. This affects the
5326 fifth argument of getsockopt() amongst other things. This is now
5327 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5328 can be set for individual OS. I have set it for SunOS5, OSF1, and
5329 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5330 some earlier ones do not.
5332 7. Change 4.30/15 was not doing the test caselessly.
5334 8. The standard form for an IPv6 address literal was being rejected by address
5335 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5336 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5337 this, as well as the form without the "IPv6" on the front (but only when
5338 address literals are enabled, of course).
5340 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5342 10. Exim crashed if a message with an empty sender address specified by -f
5343 encountered a router with an errors_to setting. This could be provoked only
5344 by a command such as
5348 where an empty string was supplied; "<>" did not hit this bug.
5350 11. Installed PCRE release 4.5.
5352 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5353 remained set. It is now erased.
5355 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5356 times from message ids (which are base 36 rather than the normal 62).
5358 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5359 were being counted as actual protocol errors, and logged if the log
5360 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5361 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5362 and DATA following a set of rejected RCPTs do not count as protocol errors.
5363 In other words, Exim assumes they were pipelined, though this may not
5364 actually be the case. Of course, in all cases the client gets an
5365 appropriate error code.
5367 15. If a lookup fails in an ACL condition, a message about the failure may
5368 be available; it is used if testing the ACL cannot continue, because most
5369 such messages specify what the cause of the deferral is. However, some
5370 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5371 that caused an old message to be retained and used if a later statement
5372 caused a defer, replacing the real cause of the deferral.
5374 16. If an IP address had so many PTR records that the DNS lookup buffer
5375 was not large enough to hold them, Exim could crash while trying to process
5376 the truncated data. It now detects and logs this case.
5378 17. Further to 4.21/58, another change has been made: if (and only if) the
5379 first line of a message (the first header line) ends with CRLF, a bare LF
5380 in a subsequent header line has a space inserted after it, so as not to
5381 terminate the header.
5383 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5384 unnecessarily, used atoi() instead of strtol(), and didn't check the
5385 termination when getting file sizes from file names by regex.
5387 19. Completely re-implemented the support for maildirsize files, in the light
5388 of a number of problems with the previous contributed implementation
5389 (4.30/29). In particular:
5391 . If the quota is zero, the maildirsize file is maintained, but no quota is
5394 . If the maildir directory does not exist, it is created before any attempt
5395 to write a maildirsize file.
5397 . The quota value in the file is just a cache; if the quota is changed in
5398 the transport, the new value overrides.
5400 . A regular expression is available for excluding directories from the
5403 20. The autoreply transport checks the characters in options that define the
5404 message's headers; it allows continued headers, but it was checking with
5405 isspace() after an embedded newline instead of explicitly looking for a
5408 21. If all the "regular" hosts to which an address was routed had passed their
5409 expiry times, and had not reached their retry times, the address was
5410 bounced, even if fallback hosts were defined. Now Exim should go on to try
5413 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5414 equivalent code in the SMTP transport. Some hosts send humungous responses
5415 to HELO/EHLO, more than 1024 it seems.
5417 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5418 gives compiler warnings in some environments. I've now done it "properly",
5421 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5422 (because of problems with the built-in one) was declared to return uschar *
5423 instead of char *, causing compiler failure.
5425 25. Fixed a file descriptor leak when processing alias/forward files.
5427 26. Fixed a minor format string issue in dbfn.c.
5429 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5431 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5432 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5435 29. When a sender address is verified, it is cached, to save repeating the test
5436 when there is more than one recipient in a message. However, when the
5437 verification involves a callout, it is possible for different callout
5438 options to be set for different recipients. It is too complicated to keep
5439 track of this in the cache, so now Exim always runs a verification when a
5440 callout is required, relying on the callout cache for the optimization.
5441 The overhead is duplication of the address routing, but this should not be
5444 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5445 to be verified with callout=postmaster, and the main callout worked but the
5446 postmaster check failed, the verification correctly failed. However, if a
5447 subsequent RCPT command asked for sender verification *without* the
5448 postmaster check, incorrect caching caused this verification also to fail,
5451 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5452 it was not caching the DNS options (qualify_single, search_parents) that
5453 were used when the lookup failed. A subsequent lookup with different
5454 options therefore always gave the same answer, though there were cases
5455 where it should not have. (Example: a "domains = !$mx_any" option on a
5456 dnslookup router: the "domains" option is always processed without any
5457 widening, but the router might have qualify_single set.) Now Exim uses the
5458 cached value only when the same options are set.
5460 32. Added John Jetmore's "exipick" utility to the distribution.
5462 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5463 than a timeout (e.g. a certificate is required, and is not provided), an
5464 Exim server now closes the connection immediately. Previously it waited for
5465 the client to close - but if the client is SSL, it seems that they each
5466 wait for each other, leading to a delay before one of them times out.
5468 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5469 maintained 0.8.x compatibility because I don't think many are using it, and
5470 it is clearly obsolete.
5472 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5475 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5476 client certificate was expired. A simple patch fixes this, though I don't
5477 understand the full logic of why the verify callback is called multiple
5480 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5481 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5482 which causes problems with some clients (such as the Certicom SSL Plus
5483 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5484 disables the coutermeasure allowing Eudora to connect."
5486 38. Exim was not checking that a write() to a log file succeeded. This could
5487 lead to Bad Things if a log got too big, in particular if it hit a file
5488 size limit. Exim now panics and dies if it cannot write to a log file, just
5489 as it does if it cannot open a log file.
5491 39. Modified OS/Makefile-Linux so that it now contains
5493 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5495 The two -D definitions ensure that Exim is compiled with large file
5496 support, which makes it possible to handle log files that are bigger than
5499 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5500 instance) a domain was checked against a named list that involved a lookup,
5501 causing $domain_data to be set, then another domain was checked against the
5502 same list, then the first domain was re-checked, the value of $domain_data
5503 after the final check could be wrong. In particular, if the second check
5504 failed, it could be set empty. This bug probably also applied to
5507 41. The strip_trailing_dot option was not being applied to the address given
5508 with the -f command-line option.
5510 42. The code for reading a message's header from the spool was incrementing
5511 $received_count, but never initializing it. This meant that the value was
5512 incorrect (doubled) while delivering a message in the same process in which
5513 it was received. In the most common configuration of Exim, this never
5514 happens - a fresh exec is done - but it can happen when
5515 deliver_drop_privilege is set.
5517 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5518 it now includes up to 150 characters of the unexpected data in the log
5521 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5522 and building data strings. The size of both of these buffers was 10 000
5523 bytes - far larger than anybody would *ever* want, thought I. Needless to
5524 say, somebody hit the limit. I have increased the maximum line length to
5525 20 000 and the maximum data length of concatenated lines to 100 000. I have
5526 also fixed two bugs, because there was no checking on these buffers. Tsk,
5527 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5528 buffer is too small.
5530 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5531 lsearch lookups. Now it does.
5533 46. When parsing a route_list item in a manualroute router, a fixed-length
5534 buffer was used for the list of hosts. I made this 1024 bytes long,
5535 thinking that nobody would ever have a list of hosts that long. Wrong.
5536 Somebody had a whole pile of complicated expansion conditions, and the
5537 string was silently truncated, leading to an expansion error. It turns out
5538 that it is easier to change to an unlimited length (owing to other changes
5539 that have happened since this code was originally written) than to build
5540 structure for giving a limitation error. The length of the item that
5541 expands into the list of hosts is now unlimited.
5543 47. The lsearch lookup could not handle data where the length of text line was
5544 more than 4095 characters. Such lines were truncated, leading to shortened
5545 data being returned. It should now handle lines of any length.
5547 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5548 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5551 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5552 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5553 no longer happens in comment lines. A list of replacements is now placed
5554 at the head of all of the source files, except those whose only change is
5555 to replace PERL_COMMAND in the very first #! line.
5557 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5558 messages on the queue, with a bottom-up merge sort, using code contributed
5559 by Michael Haardt. This should make operations like -bp somewhat faster on
5560 large queues. It won't affect queue runners, except when queue_run_in_order
5563 51. Installed eximstats 1.31 in the distribution.
5565 52. Added support for SRV lookups to the dnslookup router.
5567 53. If an ACL referred to $message_body or $message_body_end, the value was not
5568 reset for any messages that followed in the same SMTP session.
5570 54. The store-handling optimization for building very long strings was not
5571 differentiating between the different store pools. I don't think this
5572 actually made any difference in practice, but I've tidied it.
5574 55. While running the routers to verify a sender address, $sender_address
5575 was still set to the sender address. This is wrong, because when routing to
5576 send a bounce to the sender, it would be empty. Therefore, I have changed
5577 it so that, while verifying a sender address, $sender_address is set to <>.
5578 (There is no change to what happens when verifying a recipient address.)
5580 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5581 target A or AAAA records (if not already returned) without resetting the
5582 qualify_single or search_parents options of the DNS resolver. These are
5583 inappropriate in this case because the targets of MX and SRV records must
5584 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5585 target that, when qualified, matched something in the local domain. These
5586 two options are now turned off when doing these lookups.
5588 57. It seems that at least some releases of Reiserfs (which does not have the
5589 concept of a fixed number of inodes) returns zero and not -1 for the
5590 number of available inodes. This interacted badly with check_spool_inodes,
5591 which assumed that -1 was the "no such thing" setting. What I have done is
5592 to check that the total number of inodes is greater than zero before doing
5593 the test of how many are available.
5595 58. When a "warn" ACL statement has a log_message modifier, the message is
5596 remembered, and not repeated. This is to avoid a lot of repetition when a
5597 message has many recipients that cause the same warning to be written.
5598 Howewer, Exim was preserving the list of already written lines for an
5599 entire SMTP session, which doesn't seem right. The memory is now reset if a
5600 new message is started.
5602 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5603 debugging output unless log_rewrite was also set.
5605 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5606 of (int)(handle) when we know that handle contains (void *)(-1).
5608 61. The Exim daemon panic-logs an error return when it closes the incoming
5609 connection. However "connection reset by peer" seems to be common, and
5610 isn't really an error worthy of noting specially, so that particular error
5613 62. When Exim is trying to find all the local interfaces, it used to panic and
5614 die if the ioctl to get the interface flags failed. However, it seems that
5615 on at least one OS (Solaris 9) it is possible to have an interface that is
5616 included in the list of interfaces, but for which you get a failure error
5617 for this call. This happens when the interface is not "plumbed" into a
5618 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5619 failure of the "get flags" call assumes that the interface is down.
5621 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5622 makes life easier for people who are doing arithmetic on fields extracted
5623 from dates, where you often get leading zeros that should not be
5624 interpreted as octal.
5626 64. Added qualify_domain to the redirect router, to override the global
5629 65. If a pathologically long header line contained very many addresses (the
5630 report of this problem mentioned 10 000) and each of them was rewritten,
5631 Exim could use up a very large amount of memory. (It kept on making new
5632 copies of the header line as it rewrote, and never released the old ones.)
5633 At the expense of a bit more processing, the header rewriting function has
5634 been changed so that it no longer eats memory in this way.
5636 66. The generation of the Received: header has been moved from the time that a
5637 message starts to be received, to the time that it finishes. The timestamp
5638 in the Received: header should now be very close to that of the <= log
5639 line. There are two side-effects of this change:
5641 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5642 logged header lines no longer include the local Received: line, because
5643 it has not yet been created. The same applies to a copy of the message
5644 that is returned to a non-SMTP sender when a message is rejected.
5646 (b) When a filter file is tested using -bf, no additional Received: header
5647 is added to the test message. After some thought, I decided that this
5650 This change does not affect the value of $received_for. It is still set
5651 after address rewriting, but before local_scan() is called.
5653 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5655 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5656 gave an unhelpful panic error message, and a defer error. I have managed to
5657 change this behaviour so that it now rejects any supplied certificate,
5658 which seems right, as the list of acceptable certificates is empty.
5660 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5661 gave an unhelpful defer error. I have not managed to make this reject any
5662 supplied certificates, but the error message it gives is "no certificate
5663 supplied", which is not helpful.
5665 70. exigrep's output now also includes lines that are not associated with any
5666 message, but which match the given pattern. Implemented by a patch from
5667 Martin Sluka, which also tidied up the Perl a bit.
5669 71. Recipient callout verification, like sender verification, was using <> in
5670 the MAIL FROM command. This isn't really the right thing, since the actual
5671 sender may affect whether the remote host accepts the recipient or not. I
5672 have changed it to use the actual sender in the callout; this means that
5673 the cache record is now keyed on a recipient/sender pair, not just the
5674 recipient address. There doesn't seem to be a real danger of callout loops,
5675 since a callout by the remote host to check the sender would use <>.
5676 [SEE ABOVE: changed after hitting problems.]
5678 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5679 temporary errors. However, in the case of such a code being given after
5680 the end of a data transmission (i.e. after ".") Exim was failing to write
5681 a retry record for the message. (Yes, there was some broken host that was
5682 actually sending 8xx at this point.)
5684 73. An unknown lookup type in a host list could cause Exim to panic-die when
5685 the list was checked. (An example that provoked this was putting <; in the
5686 middle of a list instead of at the start.) If this happened during a DATA
5687 ACL check, a -D file could be left lying around. This kind of configuration
5688 error no longer causes Exim to die; instead it causes a defer errror. The
5689 incident is still logged to the main and panic logs.
5691 74. Buglet left over from Exim 3 conversion. The message "too many messages
5692 in one connection" was written to the rejectlog but not the mainlog, except
5693 when address rewriting (yes!) was being logged.
5695 75. Added write_rejectlog option.
5697 76. When a system filter was run not as root (that is, when system_filter_user
5698 was set), the values of the $n variables were not being returned to the
5699 main process; thus, they were not subsequently available in the $sn
5702 77. Added +return_path_on_delivery log selector.
5704 78. A connection timeout was being treated differently from recipients deferred
5705 when testing hosts_max_try with a message that was older than the host's
5706 retry timeout. (The host should not be counted, thus allowing all hosts to
5707 be tried at least once before bouncing.) This may have been the cause of an
5708 occasionally reported bug whereby a message would remain on the queue
5709 longer than the retry timeout, but would be bounced if a delivery was
5710 forced. I say "may" because I never totally pinned down the problem;
5711 setting up timeout/retry tests is difficult. See also the next item.
5713 79. The ultimate address timeout was not being applied to errors that involved
5714 a combination of host plus message (for example, a timeout on a MAIL
5715 command). When an address resolved to a number of possible hosts, and they
5716 were not all tried for each delivery (e.g. because of hosts_max_try), a
5717 message could remain on the queue longer than the retry timeout.
5719 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5722 81. Fixed an obscure SMTP outgoing bug which required at least the following
5723 conditions: (a) there was another message waiting for the same server;
5724 (b) the server returned 5xx to all RCPT commands in the first message so
5725 that the message was not completed; (c) the server dropped the connection
5726 or gave a negative response to the RSET that Exim sends to abort the
5727 transaction. The observed case was a dropped connection after DATA that had
5728 been sent in pipelining mode. That is, the server had advertised PIPELINING
5729 but was not implementing it correctly. The effect of the bug was incorrect
5730 behaviour, such as trying another host, and this could lead to a crash.
5736 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5737 and daemon.c were passed as pointers to ints; they should have been
5738 pointers to socklen_t variables (which are typically unsigned ints).
5740 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5743 3. Fixed a really odd bug that affected only the testing scheme; patching a
5744 certain fixed string in the binary changed the value of another string that
5745 happened to be identical to the end of the original first string.
5747 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5748 name", it returns that address as the IP address. On some operating
5749 systems (e.g. Solaris), it also passes back the IP address string as the
5750 "host name". However, on others (e.g. Linux), it passes back an empty
5751 string. Exim wasn't checking for this, and was changing the host name to an
5752 empty string, assuming it had been canonicized.
5754 5. Although rare, it is permitted to have more than one PTR record for a given
5755 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5756 all the names associated with an address, because they do in Solaris.
5757 However, it seems that they do not in Linux for data that comes from the
5758 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5759 I found this out when I moved to a new Linux workstation and tried to run
5760 the Exim test suite.
5762 To get round this problem I have changed the code so that it now does its
5763 own call to the DNS to look up PTR records when searching for a host name.
5764 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5765 addresses that are only in /etc/hosts are still found.
5767 This behaviour is, however, controlled by an option called host_lookup_
5768 order, which defaults to "bydns:byaddr". If people want to use the other
5769 order, or indeed, just use one or the other means of lookup, they can
5770 specify it in this variable.
5772 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5773 some operating systems, this comes back from gethostbyaddr() as an empty
5774 string, and this is what Exim used to test for. However, it seems that in
5775 other systems, "." is yielded. Exim now tests for this case too.
5777 7. The values of check_spool_space and check_log_space are now held internally
5778 as a number of kilobytes instead of an absolute number of bytes. If a
5779 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5780 kilobyte. This means that much larger values can be stored.
5782 8. Exim monitor: an attempt to get the action menu when not actually pointing
5783 at a message produces an empty menu entitled "No message selected". This
5784 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5785 no entries in it ("Shell widget menu has zero width and/or height"). So I
5786 have added a single, blank menu entry in this case.
5788 9. Added ${quote_local_part.
5790 10. MIME decoding is now applied to the contents of Subject: header lines when
5793 11. Now that a reference to $sender_host_address automatically causes a reverse
5794 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5795 host lookup before query-style lookups in lists that might use this
5796 variable. This has therefore been abolished, and the "net-" prefix is no
5797 longer necessary for query-style lookups.
5799 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5800 have been a typo for LFLAGS, so it has been changed.
5802 13. The install script calls Exim with "-C /dev/null" in order to find the
5803 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5804 to be output. Howeve, since Exim outputs its version number before the
5805 error, it didn't break the script. It just looked ugly. I fixed this by
5806 always allowing "-C /dev/null" if the caller is root.
5808 14. Ignore overlarge ACL variable number when reading spool file - insurance
5809 against a later release with more variables having written the file.
5811 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5812 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5813 this, as well as the form without the "IPv6" on the front.
5815 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5816 OS/Makefile-Darwin file.
5818 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5819 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5821 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5822 variable. However, some OS use ints for inodes. I've added cast to long int
5823 to get rid of the compiler warning.
5825 19. I had forgotten to lock out "/../" in configuration file names when
5826 ALT_CONFIG_PREFIX was set.
5828 20. Routers used for verification do not need to specify transports. However,
5829 if such a router generated a host list, and callout was configured, Exim
5830 crashed, because it could not find a port number from the (non-existent)
5831 transport. It now assumes port 25 in this circumstance.
5833 21. Added the -t option to exigrep.
5835 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5836 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5837 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5838 them is set, LOOKUP_LSEARCH is forced.
5840 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5842 24. Added sender and host information to the "rejected by local_scan()" log
5843 line; previously there was no indication of these.
5845 25. Added .include_if_exists.
5847 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5848 Linux. It turns out that not all file systems support this. Apparently some
5849 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5850 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5851 not supported on the file descriptor, is now ignored when Exim is trying to
5852 sync a directory. This applies only to Linux.
5854 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5856 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5857 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5860 29. Added support for maildirsize files from supplied patch (modified a bit).
5862 30. The use of :fail: followed by an empty string could lead Exim to respond to
5863 sender verification failures with (e.g.):
5865 550 Verification failed for <xxx>
5866 550 Sender verify failed
5868 where the first response line was missing the '-' that indicates it is not
5869 the final line of the response.
5871 31. The loop for finding the name of the user that called Exim had a hardwired
5872 limit of 10; it now uses the value of finduser_retries, which is used for
5873 all other user lookups.
5875 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5878 33. Exim was neglecting to zero errno before one call of strtol() when
5879 expanding a string and expecting an integer value. On some systems this
5880 resulted in spurious "integer overflow" errors. Also, it was casting the
5881 result into an int without checking.
5883 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5884 did not work. The code looks as if it has *never* worked, though it appears
5885 to have been documented since at least releast 1.62. I have made it work.
5887 35. The "timeout_DNS" error in retry rules, also documented since at least
5888 1.62, also never worked. As it isn't clear exactly what this means, and
5889 clearly it isn't a major issue, I have abolished the feature by treating it
5890 as "timeout", and writing a warning to the main and panic logs.
5892 36. The display of retry rules for -brt wasn't always showing the error code
5895 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5896 timeout_connect_A, timeout_connect_MX.
5898 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5899 to the empty sender.
5901 39. The daemon was not analysing the content of -oX till after it had closed
5902 stderr and disconnected from the controlling terminal. This meant that any
5903 syntax errors were only noted on the panic log, and the return code from
5904 the command was 0. By re-arranging the code a little, I've made the
5905 decoding happen first, so such errors now appear on stderr, and the return
5906 code is 1. However, the actual setting up of the sockets still happens in
5907 the disconnected process, so errors there are still only recorded on the
5910 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5911 connections (as happens on some IP stacks) was logged at start up time as
5912 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5913 it from "IPv6 and IPv4", which means that two separate sockets are being
5916 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5917 says whether AF_INET or AF_INET6 was passed as an argument.
5919 42. Exiwhat output was messed up when time zones were included in log
5922 43. Exiwhat now gives more information about the daemon's listening ports,
5923 and whether -tls-on-connect was used.
5925 44. The "port" option of the smtp transport is now expanded.
5927 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5928 silently ignored. Now an error message is written to the main and panic
5931 46. There's a new ACL modifier called "logwrite" which writes to a log file
5932 as soon as it is encountered.
5934 47. Added $local_user_uid and $local_user_gid at routing time.
5936 48. Exim crashed when trying to verify a sender address that was being
5939 49. Exim was recognizing only a space character after ".include". It now also
5940 recognizes a tab character.
5942 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5943 extracting the relevant information from the specification. The man page no
5944 longer contains scrambled data for the -d option, and I've added a section
5945 at the front about calling Exim under different names.
5947 51. Added "extra_headers" argument to the "mail" command in filter files.
5949 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5952 53. Installed eximstats 1.29.
5954 54. Added transport_filter_timeout as a generic transport option.
5956 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5957 Cc: header lines. This was required by RFC 822, but it not required by RFC
5960 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5961 incoming messages that did not have them. Now it does so only if the
5962 message originates locally, that is, if there is no associated remote host
5963 address. When Resent- header lines are present, this applies to the Resent-
5964 lines rather than the non-Resent- lines.
5966 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5967 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5969 58. Messages for configuration errors now include the name of the main
5970 configuration file - useful now that there may be more than one file in a
5971 list (.included file names were always shown).
5973 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5974 for those rare installations that do not start the daemon as root or run it
5975 setuid root. I've cut out the call to initgroups() if the daemon is not
5978 60. The Exim user and group can now be bound into the binary as text strings
5979 that are looked up at the start of Exim's processing.
5981 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5983 62. Added $mailstore_basename variable.
5985 63. Installed patch to sieve.c from Michael Haardt.
5987 64. When Exim failed to open the panic log after failing to open the main log,
5988 the original message it was trying to log was written to stderr and debug
5989 output, but if they were not available (the usual case in production), it
5990 was lost. Now it is written to syslog before the two lines that record the
5991 failures to open the logs.
5993 65. Users' Exim filters run in subprocesses under the user's uid. It is
5994 possible for a "deliver" command or an alias in a "personal" command to
5995 provoke an address rewrite. If logging of address rewriting is configured,
5996 this fails because the process is not running as root or exim. There may be
5997 a better way of dealing with this, but for the moment (because 4.30 needs
5998 to be released), I have disabled address rewrite logging when running a
5999 filter in a non-root, non-exim process.
6005 1. The buildconfig auxiliary program wasn't quoting the value set for
6006 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6007 not defined. This bug was masked in 4.22 by the effect that was fixed in
6010 2. Some messages that were rejected after a message id was allocated were
6011 shown as "incomplete" by exigrep. It no longer does this for messages that
6012 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6014 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6015 have allow_domain_literals set, the ID did not get logged in the <= line.
6016 Domain literals are now always recognized in Message-ID: header lines.
6018 4. The first argument for a ${extract expansion item is the key name or field
6019 number. Leading and trailing spaces in this item were not being ignored,
6020 causing some misleading effects.
6022 5. When deliver_drop_privilege was set, single queue runner processes started
6023 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6024 same command in the process it spins off) were not dropping privilege.
6026 6. When the daemon running as "exim" started a queue runner, it always
6027 re-executed Exim in the spun-off process. This is a waste of effort when
6028 deliver_drop_privilege is set. The new process now just calls the
6029 queue-runner function directly.
6035 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6038 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6039 ignored. Though the use of -oP was forcing the writing of a pid file, it
6040 was always written to the default place.
6042 3. If the message "no IP address found for host xxxx" is generated during
6043 incoming verification, it is now followed by identification of the incoming
6044 connection (so you can more easily find what provoked it).
6046 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6048 5. Added some features to "harden" Exim a bit more against certain attacks:
6050 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6051 be put in Local/Makefile. This is like the never_users runtime option,
6052 but it cannot be overridden. The default setting is "root".
6054 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6055 prefix string with which any file named in a -C command line option
6058 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6059 is retained for -C and -D only if the caller of Exim is root. Without
6060 it, the exim user may also use -C and -D and retain privilege.
6062 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6063 command line option is disabled.
6065 6. Macro names set by the -D option must start with an upper case letter, just
6066 like macro names defined in the configuration file.
6068 7. Added "dereference=" facility to LDAP.
6070 8. Two instances of the typo "uknown" in the source files are fixed.
6072 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6073 the Configure-Makefile script screwed up while processing it.
6075 10. Incorporated PCRE 4.4.
6077 11. The SMTP synchronization check was not operating right at the start of an
6078 SMTP session. For example, it could not catch a HELO sent before the client
6079 waited for the greeting. There is now a check for outstanding input at the
6080 point when the greeting is written. Because of the duplex, asynchronous
6081 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6082 way, but not yet received, when the check is performed.
6084 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6085 on TCP/IP sockets, because this apparently causes some broken clients to
6088 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6089 unchanged) from the Cygwin maintainer.
6091 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6092 was supported instead of testing for mbx. Effectively a typo.
6094 15. The spa authenticator server code was not checking that the input it
6095 received was valid base64.
6097 16. The debug output line for the "set" modifier in ACLs was not showing the
6098 name of the variable that was being set.
6100 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6102 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6103 lookup on its own. Something else was needed to trigger the lookup. For
6104 example, a match in host_lookup or the need for a host name in a host list.
6105 Now, if $sender_host_name is referenced and the host name has not yet been
6106 looked up, a lookup is performed. If the lookup fails, the variable remains
6107 empty, and $host_lookup_failed is set to "1".
6109 19. Added "eqi" as a case-independent comparison operator.
6111 20. The saslauthd authentication condition could segfault if neither service
6112 nor realm was specified.
6114 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6115 error message that was logged was misleading, and incoming SMTP
6116 connections were dropped. The message is now more accurate, and temporary
6117 errors are given to SMTP connections.
6119 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6120 (except RSET) with a 421 error, until QUIT is received. However, it was
6121 failing to send a response to QUIT.
6123 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6124 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6125 ACL is now run after the helo_try_verify_hosts code.
6127 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6128 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6129 case-independent, so other case variants are also recognized). Apparently
6130 some systems use these upper case variants.
6132 25. If more than two messages were waiting for the same host, and a transport
6133 filter was specified for the transport, Exim sent two messages over the
6134 same TCP/IP connection, and then failed with "socket operation on non-
6135 socket" when it tried to send the third.
6137 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6139 27. The extern definition of crypt16() in expand.c was not being excluded when
6140 the OS had its own crypt16() function.
6142 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6143 as a preferred synonym for return_size_limit, both as an option and as an
6146 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6148 30. Changed the default configuration ACL to relax the local part checking rule
6149 for addresses that are not in any local domains. For these addresses,
6150 slashes and pipe symbols are allowed within local parts, but the sequence
6151 /../ is explicitly forbidden.
6153 31. SPA server authentication was not clearing the challenge buffer before
6156 32. log_message in a "warn" ACL statement was writing to the reject log as
6157 well as to the main log, which contradicts the documentation and doesn't
6158 seem right (because no rejection is happening). So I have stopped it.
6160 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6161 However, I am unable to do any testing of this.
6163 34. Fixed an infelicity in the appendfile transport. When checking directories
6164 for a mailbox, to see if any needed to be created, it was accidentally
6165 using path names with one or more superfluous leading slashes; tracing
6166 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6168 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6169 contents are added to the log line that is written for every discarded
6170 recipient. (Previously a log_message setting was ignored.)
6172 36. The ${quote: operator now quotes the string if it is empty.
6174 37. The install script runs exim in order to find its version number. If for
6175 some reason other than non-existence or emptiness, which it checks, it
6176 could not run './exim', it was installing it with an empty version number,
6177 i.e. as "exim-". This error state is now caught, and the installation is
6180 38. An argument was missing from the function that creates an error message
6181 when Exim fails to connect to the socket for saslauthd authentication.
6182 This could cause Exim to crash, or give a corrupted message.
6184 39. Added isip, isip4, and isip6 to ${if conditions.
6186 40. The ACL variables $acl_xx are now saved with the message, and can be
6187 accessed later in routers, transports, and filters.
6189 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6190 strings in the file are not string-expanded.
6192 42. If a MAIL command specified a SIZE value that was too large to fit into an
6193 int variable, the check against message_size_limit failed. Such values are
6194 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6195 day this will have to be increased, but I don't think I want to be around
6196 when emails are that large.
6203 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6204 iconv() is not standard in FreeBSD.
6206 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6207 IPv6 enabled. The observed symptom was a segmentation fault on return from
6208 the function os_common_find_running_interfaces() in src/os.c.
6210 3. In the check_special_case() function in daemon.c I had used "errno" as an
6211 argument name, which causes warnings on some systems. This was basically a
6212 typo, since it was named "eno" in the comments!
6214 4. The code that waits for the clock to tick (at a resolution of some fraction
6215 of a second) so as to ensure message-id uniqueness was always waiting for
6216 at least one whole tick, when it could have waited for less. [This is
6217 almost certainly not relevant at current processor speeds, where it is
6218 unlikely to ever wait at all. But we try to future-proof.]
6220 5. The function that sleeps for a time interval that includes fractions of a
6221 second contained a race. It did not block SIGALRM between setting the
6222 timer, and suspending (a couple of lines later). If the interval was short
6223 and the sigsuspend() was delayed until after it had expired, the suspension
6224 never ended. On busy systems this could lead to processes getting stuck for
6227 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6228 process, before it forks any delivery processes. The open lookup caching
6229 mechanism meant that the open file or database connection was passed into
6230 the delivery process. The problem was that delivery processes always tidy
6231 up cached lookup data. This could cause a problem for the next delivery
6232 process started by the queue runner, because the external queue runner
6233 process does not know about the closure. So the next delivery process
6234 still has data in the lookup cache. In the case of a file lookup, there was
6235 no problem because closing a file descriptor in a subprocess doesn't affect
6236 the parent. However, if the lookup was caching a connection to a database,
6237 the connection was closed, and the second delivery process was likely to
6238 see errors such as "PGSQL: query failed: server closed the connection
6239 unexpectedly". The problem has been fixed by closing all cached lookups
6240 in a queue runner before running a delivery process.
6242 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6243 seem to have the "const" qualifier which it has on other OS. I've
6246 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6247 *of the same type* (client or server) with the same public name that an
6248 error should be diagnosed.
6250 9. When Exim looked up a host name for an IP address, but failed to find the
6251 original IP address when looking up the host name (a safety check), it
6252 output the message "<ip address> does not match any IP for NULL", which was
6253 confusing, to say the least. The bug was that the host name should have
6254 appeared instead of "NULL".
6256 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6257 user that is built into the binary, and the -C or -D options is used, root
6258 privilege is dropped before the configuration file is read. In addition,
6259 logging is switched to stderr instead of the normal log files. If the
6260 configuration then re-defines the Exim user, the unprivileged environment
6261 is probably not what is expected, so Exim logs a panic warning message (but
6264 However, if deliver_drop_privilege is set, the unprivileged state may well
6265 be exactly what is intended, so the warning has been cut out in that case,
6266 and Exim is allowed to try to write to its normal log files.
6272 1. smtp_return_error_details was not giving details for temporary sender
6273 or receiver verification errors.
6275 2. Diagnose a configuration error if two authenticators have the same public
6278 3. Exim used not to create the message log file for a message until the first
6279 delivery attempt. This could be confusing when incoming messages were held
6280 for policy or load reasons. The message log file is now created at the time
6281 the message is received, and an initial "Received" line is written to it.
6283 4. The automatically generated man page for command line options had a minor
6284 bug that caused no ill effects; however, a more serious problem was that
6285 the procedure for building the man page automatically didn't always
6286 operate. Consequently, release 4.20 contains an out-of-date version. This
6287 shouldn't happen again.
6289 5. When building Exim with embedded Perl support, the script that builds the
6290 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6291 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6293 6. The freeze_tell option was not being used for messages that were frozen on
6294 arrival, either by an ACL or by local_scan().
6296 7. Added the smtp_incomplete_transaction log selector.
6298 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6299 was accepting AUTH without a new EHLO.
6301 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6302 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6305 10. Logging of TCP/IP connections (when configured) now happens in the main
6306 daemon process instead of the child process, so that the TCP/IP connection
6307 count is more accurate (but it can never be perfect).
6309 11. The use of "drop" in a nested ACL was not being handled correctly in the
6310 outer ACL. Now, if condition failure induced by the nested "drop" causes
6311 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6312 or "require"), the connection is dropped.
6314 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6315 that yield "discard" can now be used with an "accept" or a "discard" verb,
6316 but an error is generated for any others (because I can't see a useful way
6317 to define what should happen).
6319 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6320 are now processed in the same way as lines in the Exim configuration file.
6321 In particular, continuation lines are supported.
6323 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6325 15. Added -ti meaning -t -i.
6327 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6328 domains, and warn by logging if others are found.
6330 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6331 returned from GIFCONF when Exim is trying to find the list of interfaces on
6332 a host. The code in os.c has been modified to copy each ifreq to an aligned
6333 structure in all cases.
6335 Also, in some cases, the returned ifreq's were being copied to a 'struct
6336 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6337 means the last couple of bytes of an IPv6 address could be chopped if the
6338 ifreq contained only a normal sockaddr (14 bytes storage).
6340 18. Named domain lists were not supported in the hosts_treat_as_local option.
6341 An entry such as +xxxx was not recognized, and was treated as a literal
6344 19. Ensure that header lines added by a DATA ACL are included in the reject log
6345 if the ACL subsequently rejects the message.
6347 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6348 MD5 (which is deprecated).
6350 21. When testing a filter file using -bf, Exim was writing a message when it
6351 took the sender from a "From " line in the message, but it was not doing so
6352 when it took $return_path from a Return-Path: header line. It now does.
6354 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6355 with a valid header line field name (a series of printing characters
6356 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6358 23. Changed "disc" in the source to "disk" to conform to the documentation and
6359 the book and for uniformity.
6361 24. Ignore Sendmail's -Ooption=value command line item.
6363 25. When execve() failed while trying to run a command in a pipe transport,
6364 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6365 could be confused with a return value of 69 from the command itself. This
6366 has been changed to 127, the value the shell returns if it is asked to run
6367 a non-existent command. The wording for the related log line suggests a
6368 non-existent command as the problem.
6370 26. If received_header_text expands to an empty string, do not add a Received:
6371 header line to the message. (Well, it adds a token one on the spool, but
6372 marks it "old" so that it doesn't get used or transmitted.)
6374 27. Installed eximstats 1.28 (addition of -nt option).
6376 28. There was no check for failure on the call to getsockname() in the daemon
6377 code. This can fail if there is a shortage of resources on the system, with
6378 ENOMEM, for example. A temporary error is now given on failure.
6380 29. Contrary to the C standard, it seems that in some environments, the
6381 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6382 program. Exim now does this explicitly; it affects the formatting of
6383 timestamps using strftime().
6385 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6386 complaints. I've now initialized all the variables, to avoid this.
6388 32. Header lines added by a system filter were not being "seen" during
6389 transport-time rewrites.
6391 33. The info_callback() function passed to OpenSSL is set up with type void
6392 (*)(SSL *, int, int), as described somewhere. However, when calling the
6393 function (actually a macro) that sets it up, the type void(*)() is
6394 expected. I've put in a cast to prevent warnings from picky compilers.
6396 34. If a DNS black list lookup found a CNAME record, but there were no A
6397 records associated with the domain it pointed at, Exim crashed.
6399 35. If a DNS black list lookup returned more than one A record, Exim ignored
6400 all but the first. It now scans all returned addresses if a particular IP
6401 value is being sought. In this situation, the contents of the
6402 $dnslist_value variable are a list of all the addresses, separated by a
6405 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6406 to accept a host name and all its aliases if the forward lookup for any of
6407 them yielded the IP address of the incoming connection. Now it accepts only
6408 those names whose forward lookup yields the correct IP address. Any other
6409 names are discarded. This closes a loophole whereby a rogue DNS
6410 administrator could create reverse DNS records to break through a
6411 wildcarded host restriction in an ACL.
6413 37. If a user filter or a system filter that ran in a subprocess used any of
6414 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6415 the wrong values were passed to the pipe command ($thisaddress had the
6416 value of $0, $0 had the value of $1, etc). This bug was introduced by
6417 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6419 38. Improved the line breaking for long SMTP error messages from ACLs.
6420 Previously, if there was no break point between 40 and 75 characters, Exim
6421 left the rest of the message alone. Two changes have been made: (a) I've
6422 reduced the minimum length to 35 characters; (b) if it can't find a break
6423 point between 35 and 75 characters, it looks ahead and uses the first one
6424 that it finds. This may give the occasional overlong line, but at least the
6425 remaining text gets split now.
6427 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6428 file descriptors might be low, and that setting 1000 would always raise it.
6429 It turns out that in some environments, the limit is already over 1000 and
6430 that lowering it causes trouble. So now Exim takes care not to decrease it.
6432 40. When delivering a message, the value of $return_path is set to $sender_
6433 address at the start of routing (routers may change the value). By an
6434 oversight, this default was not being set up when an address was tested by
6435 -bt or -bv, which affected the outcome if any router or filter referred to
6438 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6439 log only when "message" or "log_message" are set. However, if one of the
6440 conditions was an address verification, or a call to a nested ACL, the
6441 messages generated by the underlying test were being passed through. This
6442 no longer happens. The underlying message is available in $acl_verify_
6443 message for both "message" and "log_message" expansions, so it can be
6444 passed through if needed.
6446 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6447 new expansion $bh_ to give the encoded byte string without charset
6448 translation. Translation happens only if iconv() is available; HAVE_ICONV
6449 indicates this at build time. HEADERS_CHARSET gives the charset to
6450 translate to; headers_charset can change it in the configuration, and
6451 "headers charset" can change it in an individual filter file.
6453 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6454 that creates RFC 2047 encoded "words" labels them as that charset instead
6455 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6456 expansion operator; (ii) when Exim creates a From: line for a local
6457 message; (iii) when a header line is rewritten to include a "phrase" part.
6459 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6460 buggy, causing it to skip the first lines of messages whose message ID
6461 ended in 'D'. This would not have bitten before Exim release 4.14, because
6462 message IDs were unlikely to end in 'D' before then. The effect was to have
6463 incorrect size information for certain domains.
6465 45. #include "config.h" was missing at the start of the crypt16.c module. This
6466 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6469 46. If there was a timeout during a "random" callout check, Exim treated it as
6470 a failure of the random address, and carried on sending RSET and the real
6471 address. If the delay was just some slowness somewhere, the response to the
6472 original RCPT would be taken as a response to RSET and so on, causing
6473 mayhem of various kinds.
6475 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6476 when I implemented it. It didn't allow for the fact that some option values
6477 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6478 the right test for positive values.
6480 48. Domain names in DNS records are case-independent. Exim always looks them up
6481 in lower case. Some resolvers return domain names in exactly the case they
6482 appear in the zone file, that is, they may contain uppercase letters. Not
6483 all resolvers do this - some return always lower case. Exim was treating a
6484 change of case by a resolver as a change of domain, similar to a widening
6485 of a domain abbreviation. This triggered its re-routing code and so it was
6486 trying to route what was effectively the same domain again. This normally
6487 caused routing to fail (because the router wouldn't handle the domain
6488 twice). Now Exim checks for this case specially, and just changes the
6489 casing of the domain that it ultimately uses when it transmits the message
6492 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6495 50. If a filter generated a file delivery with a non-absolute name (possible if
6496 no home directory exists for the router), the forbid_file option was not
6499 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6500 the existing equality matching.
6502 52. Exim was using ints instead of ino_t variables in some places where it was
6503 dealing with inode numbers.
6505 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6506 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6507 called TMPDIR, and if it finds it is different, it changes its value.
6509 54. The smtp_printf() function is now made available to local_scan() so
6510 additional output lines can be written before returning. There is also an
6511 smtp_fflush() function to enable the detection of a dropped connection.
6512 The variables smtp_input and smtp_batched_input are exported to
6515 55. Changed the default runtime configuration: the message "Unknown user"
6516 has been removed from the ACL, and instead placed on the localuser router,
6517 using the cannot_route_message feature. This means that any verification
6518 failures that generate their own messages won't get overridden. Similarly,
6519 the "Unrouteable address" message that was in the ACL for unverifiable
6520 relay addresses has also been removed.
6522 56. Added hosts_avoid_esmtp to the smtp transport.
6524 57. The exicyclog script was not checking for the esoteric option
6525 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6526 will work only if exicyclog is run under the appropriate euid.
6528 58. Following a discussion on the list, the rules by which Exim recognises line
6529 endings on incoming messages have been changed. The -dropcr and drop_cr
6530 options are now no-ops, retained only for backwards compatibility. The
6531 following line terminators are recognized: LF CRLF CR. However, special
6532 processing applies to CR:
6534 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6535 nor a local message in the state where . is a terminator.
6537 (ii) If a bare CR is encountered in a header line, an extra space is added
6538 after the line terminator so as not to end the header. The reasoning
6539 behind this is that bare CRs in header lines are most likely either
6540 to be mistakes, or people trying to play silly games.
6542 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6543 was being incorrectly given as 18 bytes larger than it should have been.
6544 This is a VOB (very old bug).
6546 60. This may never have affected anything current, but just in case it has:
6547 When the local host is found other than at the start of a list of hosts,
6548 the local host, those with the same MX, and any that follow, are discarded.
6549 When the list in question was part of a longer list of hosts, the following
6550 hosts (not currently being processed) were also being discarded. This no
6551 longer happens. I'm not sure if this situation could ever has previously
6554 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6557 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6558 header. This is something that is recommended in a new Internet Draft, and
6559 is something that is documented as being done by Sendmail. There are two
6560 possible values. For messages generated by the autoreply transport, Exim
6563 Auto-Submitted: auto-replied
6565 whereas for all other generated messages (e.g. bounces) it adds
6567 Auto-Submitted: auto-generated
6569 63. The "personal" condition in filters now includes a test for the
6570 Auto-Submitted: header. If it contains the string "auto-" the message it
6571 not considered personal.
6573 64. Added rcpt_include_affixes as a generic transport option.
6575 65. Added queue_only_override (default true).
6577 66. Added the syslog_duplication option.
6579 67. If what should have been the first header line of a message consisted of
6580 a space followed by a colon, Exim was mis-interpreting it as a header line.
6581 It isn't of course - it is syntactically invalid and should therefore be
6582 treated as the start of the message body. The misbehaviour could have
6583 caused a number of strange effects, including loss of data in subsequent
6584 header lines, and spool format errors.
6586 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6587 client host had authenticated. This control can now be exercised by an ACL
6588 for more flexibility.
6590 69. By default, callouts do not happen when testing with -bh. There is now a
6591 variant, -bhc, which does actually run the callout code, including
6592 consulting and updating the callout cache.
6594 70. Added support for saslauthd authentication, courtesy of Alexander
6597 71. If statvfs() failed on the spool or log directories while checking their
6598 size for availability, Exim confusingly gave the error "space shortage".
6599 Furthermore, in debugging mode it crashed with a floating point exception.
6600 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6601 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6602 serious problem, Exim now writes to the main and panic logs when this
6603 happens, with details of the failure. It then refuses to accept the
6604 incoming message, giving the message "spool directory problem" or "log
6605 directory problem" with a 421 code for SMTP messages.
6607 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6608 0, 1, and 2 exist, because some OS complain for execs without them (see
6609 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6610 descriptors. However, the code omitted to check that the open succeeded,
6611 causing mysterious errors if for some reason the permissions on /dev/null
6612 got screwed. Now Exim writes a message to the main and panic logs, and
6613 bombs out if it can't open /dev/null.
6615 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6616 interact so that it is all more flexible. It is supposed to remain
6617 backwards compatible. Also added extra_local_interfaces.
6619 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6620 to bomb out with an assertion failure - to the client this appears as a
6621 connection drop. This problem occurs in the part of the code that was taken
6622 from the Samba project. Fortunately, the assertion is in a very simple
6623 function, so I have fixed this by reproducing the function inline in the
6624 one place where it is called, and arranging for authentication to fail
6625 instead of killing the process with assert().
6627 75. The SPA client code was not working when the server requested OEM rather
6628 than Unicode encoding.
6630 76. Added code to make require_files with a specific uid setting more usable in
6631 the case where statting the file as root fails - usually a non-root-mounted
6632 NFS file system. When this happens and the failure is EACCES, Exim now
6633 forks a subprocess and does the per-uid checking as the relevant uid.
6635 77. Added process_log_path.
6637 78. If log_file_path was not explicitly set, a setting of check_log_space or
6638 check_log_inodes was ignored.
6640 79. If a space check for the spool or log partitions fails, the incident is now
6641 logged. Of course, in the latter case the data may get lost...
6643 80. Added the %p formatting code to string_format() so that it can be used to
6644 print addresses in debug_print(). Adjusted all the address printing in the
6645 debugging in store.c to use %p rather than %d.
6647 81. There was a concern that a line of code in smtp_in.c could overflow a
6648 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6649 initially expressed, the concern was not well-founded, because trailing
6650 spaces are removed early. However, if the trailing spaces were followed by
6651 a NULL, they did not get removed, so the overflow was possible. Two fixes
6654 (a) I re-wrote the offending code in a cleaner fashion.
6655 (b) If an incoming SMTP command contains a NULL character, it is rejected
6658 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6659 runs initgroups(), so that if the Exim user is in any additional groups,
6660 they will be used during message reception.
6666 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob