1 Change log file for Exim from version 4.21
2 -------------------------------------------
6 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
7 supports it and a size is available (ie. the sending peer gave us one).
9 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
10 years of being deprecated). The replacements are the ACLs
11 acl_smtp_mime and acl_not_smtp_mime.
13 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
14 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
15 or even an in-clear connection were permitted. Now, if the host lookup
16 was dnssec and dane was requested then the host is only used if the
17 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
18 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
19 if one fails this test.
20 This means that a poorly-configured remote DNS will make it incommunicado;
21 but it protects against a DNS-interception attack on it.
23 JH/04 Bug 1810: make continued-use of an open smtp transport connection
24 non-noisy when a race steals the message being considered.
26 JH/05 If main configuration option tls_certificate is unset, generate a
27 selfsigned certificate for inbound TLS connections.
29 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
30 in rewrites and routers.
32 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
33 and logged a warning sing 4.83; now they are a configuration file error.
35 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
36 (lacking @domain). Apply the same qualification processing as RCPT.
38 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
40 JH/10 Support ${sha256:} applied to a string (as well as the previous
43 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
44 a cutthrough deliver is pending, as we always want to make a connection.
45 This also avoids re-routing the message when later placing the cutthrough
46 connection after a verify cache hit.
47 Do not update it with the verify result either.
49 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
50 when routing results in more than one destination address.
52 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
53 signing (which inhibits the cutthrough capability). Previously only
54 the presence of an option was tested; now an expansion evaluating as
55 empty is permissible (obviously it should depend only on data available
56 when the cutthrough connection is made).
58 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
59 the relevant preceding SMTP command did not note the pipelining mode.
61 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
62 Previously they were not counted.
64 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
65 as one having no matching records. Previously we deferred the message
66 that needed the lookup.
68 JH/17 Fakereject: previously logged as a norml message arrival "<="; now
69 distinguished as "(=".
71 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
72 for missing MX records. Previously it only worked for missing A records.
74 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
76 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
77 after the data-go-ahead and data-ack. Patch from Jason Betts.
79 JH/21 Bug 1846: Send DMARC forensic reports for reject and quaratine results,
80 even for a "none" policy. Patch from Tony Meyer.
82 JH/22 Fix continued use of a connection for further deliveries. If a port was
83 specified by a router, it must also match for the delivery to be
86 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
87 When one of the recipients of a message was unsuitable for the connection
88 (has no matching addresses), we lost track of needing to mark it
89 deferred. As a result mail would be lost.
91 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
93 JH/25 Decoding ACL controls is now done using a binary search; the sourcecode
94 takes up less space and should be simpler to maintain. Merge the ACL
95 condition decode tables also, with similar effect.
97 JH/26 Fix problem with one_time used on a redirect router which returned the
98 parent address unchanged. A retry would see the parent address marked as
99 delivered, so not attempt the (identical) child. As a result mail would
102 JH/27 Fix a possible security hole, wherein a process operating with the Exim
103 UID can gain a root shell. Credit to http://www.halfdog.net/ for
104 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
107 JH/28 Enable {spool,log} filesystem space and inode checks as default.
108 Main config options check_{log,spool}_{inodes,space} are now
109 100 inodes, 10MB unless set otherwise in the configuration.
111 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
112 Previously it only applied to the main-section connection policy
115 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
117 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
118 by me. Added RFC7919 DH primes as an alternative.
120 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
121 are not in the system include path.
123 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
124 GnuTLS, when a sennsion startup failed (eg because the client disconnected)
125 Exim did stdio operations after fclose. This was exposed by a recent
126 change which nulled out the file handle after the fclose.
132 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
133 and 3.4.4 - once the server is enabled to respond to an OCSP request
134 it does even when not requested, resulting in a stapling non-aware
135 client dropping the TLS connection.
137 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
138 support variable-length bit vectors. No functional change.
140 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
141 The I= interface field on outgoing lines is now after the H= remote
142 host field, same as incoming lines. There is a separate
143 outgoing_interface log selector which allows you to disable the
146 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
147 If not running log_selector +smtp_connection the mainlog would be held
148 open indefinitely after a "too many connections" event, including to a
149 deleted file after a log rotate. Leave the per net connection logging
150 leaving it open for efficiency as that will be quickly detected by the
151 check on the next write.
153 HS/01 Bug 1671: Fix post transport crash.
154 Processing the wait-<transport> messages could crash the delivery
155 process if the message IDs didn't exist for some reason. When
156 using 'split_spool_directory=yes' the construction of the spool
157 file name failed already, exposing the same netto behaviour.
159 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
160 mime_regex ACL conditions.
162 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
163 to DSN fail messages (bounces): remote IP, remote greeting, remote response
164 to HELO, local diagnostic string.
166 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
167 log line to debug. Even when configured with a tls authenticator many
168 client connections are expected to not authenticate in this way, so
169 an authenticate fail is not an error.
171 HS/02 Add the Exim version string to the process info. This way exiwhat
172 gives some more detail about the running daemon.
174 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
175 matter for fast-change records such as DNSBLs.
177 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
178 retry record, even if constant. There may be multiple transports with
179 different interface settings and the retry behaviour needs to be kept
182 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
184 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
186 JH/10 Bug 840: fix log_defer_output option of pipe transport
188 JH/11 Bug 830: use same host for all RCPTS of a message, even under
189 hosts_randomize. This matters a lot when combined with mua_wrapper.
191 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
192 ${quote_pgsql:<string>} operator.
194 JH/13 Bug 1708: avoid misaligned access in cached lookup.
196 JH/14 Change header file name for freeradius-client. Relevant if compiling
197 with Radius support; from the Gentoo tree and checked under Fedora.
199 JH/15 Bug 1712: Introduce $prdr_requested flag variable
201 JH/16 Bug 1714: Permit an empty string as expansion result for transport
202 option transport_filter, meaning no filtering.
204 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
206 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
207 defaults to "*" (all hosts). The variable is now available when not built
208 with TLS, default unset, mainly to enable keeping the testuite sane.
209 If a server certificate is not supplied (via tls_certificate) an error is
210 logged, and clients will find TLS connections fail on startup. Presumably
211 they will retry in-clear.
212 Packagers of Exim are strongly encouraged to create a server certificate
213 at installation time.
215 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
216 with the $config_file variable.
218 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
219 in transport context, after the attempt, and per-recipient. The latter type
220 is per host attempted. The event data is the error message, and the errno
221 information encodes the lookup type (A vs. MX) used for the (first) host,
222 and the trailing two digits of the smtp 4xx reponse.
224 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
225 to write to mainlog (or rejectlog, paniclog) in the window between file
226 creation and permissions/ownership being changed. Particularly affects
227 installations where exicyclog is run as root, rather than exim user;
228 result is that the running daemon panics and dies.
230 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
232 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
233 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
234 "pri" and "weight". Note that the previous implicit priority given by the
235 list order is no longer honoured.
237 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
240 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
241 by defining SUPPORT_SOCKS.
243 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
244 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
245 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
246 variables are renamed to proxy_{local,external}_{address,port}.
248 JH/27 Move Internationalisation support from Experimental to mainline, enabled
249 for a build by defining SUPPORT_I18N
251 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
252 of the query string, and make ${quote_redis:} do that quoting.
254 JH/29 Move Events support from Experimental to mainline, enabled by default
255 and removable for a build by defining DISABLE_EVENT.
257 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
259 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
260 cached by the daemon.
262 JH/32 Move Redis support from Experimental to mainline, enabled for a build
263 by defining LOOKUP_REDIS. The libhiredis library is required.
265 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
266 keys are given for lookup.
268 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
269 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
270 only supported when built with TLS support. The PolarSSL SHA routines
271 are still used when the TLS library is too old for convenient support.
273 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
274 openssl_options), for security. OpenSSL forces this from version 1.1.0
275 server-side so match that on older versions.
277 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
278 allocation for $value could be released as the expansion processing
279 concluded, but leaving the global pointer active for it.
281 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
282 and to use the domains and local_parts ACL conditions.
284 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
285 incorrectly not doubled on cutthrough transmission, hence seen as a
286 body-termination at the receiving system - resulting in truncated mails.
287 Commonly the sender saw a TCP-level error, and retransmitted the message
288 via the normal store-and-forward channel. This could result in duplicates
289 received - but deduplicating mailstores were liable to retain only the
290 initial truncated version.
292 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
294 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
296 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
297 we're in there, support oversigning also; bug 1309.
299 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
301 HS/04 Add support for keep_environment and add_environment options.
303 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
304 either intentional arithmetic overflow during PRNG, or testing config-
307 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
308 delivery resulted in actual delivery. Cancel cutthrough before DATA
311 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
312 rejects a recipient: pass the reject to the originator.
314 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
315 Many were false-positives and ignorable, but it's worth fixing the
318 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
319 for the new environment-manipulation done at startup. Move the routines
320 from being local to tls.c to being global via the os.c file.
322 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
323 an extract embedded as result-arg for a map, the first arg for extract
324 is unavailable so we cannot tell if this is a numbered or keyed
325 extraction. Accept either.
331 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
334 JH/02 The smtp transport option "multi_domain" is now expanded.
336 JH/03 The smtp transport now requests PRDR by default, if the server offers
339 JH/04 Certificate name checking on server certificates, when exim is a client,
340 is now done by default. The transport option tls_verify_cert_hostnames
341 can be used to disable this per-host. The build option
342 EXPERIMENTAL_CERTNAMES is withdrawn.
344 JH/05 The value of the tls_verify_certificates smtp transport and main options
345 default to the word "system" to access the system default CA bundle.
346 For GnuTLS, only version 3.0.20 or later.
348 JH/06 Verification of the server certificate for a TLS connection is now tried
349 (but not required) by default. The verification status is now logged by
350 default, for both outbound TLS and client-certificate supplying inbound
353 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
356 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
357 Status Notification (bounce) messages are now MIME format per RFC 3464.
358 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
359 under the control of the dsn_advertise_hosts option, and routers may
360 have a dsn_lasthop option.
362 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
363 default, modifiable by a malware= option. The list separator for
364 the options can now be changed in the usual way. Bug 68.
366 JH/10 The smtp_receive_timeout main option is now expanded before use.
368 JH/11 The incoming_interface log option now also enables logging of the
369 local interface on delivery outgoing connections.
371 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
372 if the interface and destination host and port all match.
374 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
377 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
378 Patch from Andrew Lewis.
380 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
381 now supports optional time-restrictions, weighting, and priority
382 modifiers per server. Patch originally by <rommer@active.by>.
384 JH/16 The spamd_address main option now supports a mixed list of local
385 and remote servers. Remote servers can be IPv6 addresses, and
386 specify a port-range.
388 JH/17 Bug 68: The spamd_address main option now supports an optional
389 timeout value per server.
391 JH/18 Bug 1581: Router and transport options headers_add/remove can
392 now have the list separator specified.
394 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
397 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
400 JH/21 Support for the A6 type of dns record is withdrawn.
402 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
403 rather than the verbs used.
405 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
406 from 255 to 1024 chars.
408 JH/24 Verification callouts now attempt to use TLS by default.
410 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
411 are generic router options now. The defaults didn't change.
413 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
414 Original patch from Alexander Shikoff, worked over by JH.
416 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
417 files if ZCAT_COMMAND is not executable.
419 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
421 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
423 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
424 Normally benign, it bites when the pair was led to by a CNAME;
425 modern usage is to not canoicalize the domain to a CNAME target
426 (and we were inconsistent anyway for A-only vs AAAA+A).
428 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
430 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
431 when evaluating $sender_host_dnssec.
433 JH/31 Check the HELO verification lookup for DNSSEC, adding new
434 $sender_helo_dnssec variable.
436 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
438 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
440 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
442 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
443 documented as working, but never had. Support all but $spam_report.
445 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
446 added for tls authenticator.
448 HS/03 Add perl_taintmode main config option
453 TL/01 When running the test suite, the README says that variables such as
454 no_msglog_check are global and can be placed anywhere in a specific
455 test's script, however it was observed that placement needed to be near
456 the beginning for it to behave that way. Changed the runtest perl
457 script to read through the entire script once to detect and set these
458 variables, reset to the beginning of the script, and then run through
459 the script parsing/test process like normal.
461 TL/02 The BSD's have an arc4random API. One of the functions to induce
462 adding randomness was arc4random_stir(), but it has been removed in
463 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
464 function when detected.
466 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
467 cause callback expansion.
469 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
470 syntax errors in an expansion can be treated as a string instead of
471 logging or causing an error, due to the internal use of bool_lax
472 instead of bool when processing it.
474 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
475 server certificates when making smtp deliveries.
477 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
479 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
481 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
483 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
484 Merged patch from Sebastian Wiedenroth.
486 JH/05 Fix results-pipe from transport process. Several recipients, combined
487 with certificate use, exposed issues where response data items split
488 over buffer boundaries were not parsed properly. This eventually
489 resulted in duplicates being sent. This issue only became common enough
490 to notice due to the introduction of conection certificate information,
491 the item size being so much larger. Found and fixed by Wolfgang Breyha.
493 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
494 size buffer was used, resulting in syntax errors when an expansion
497 JH/07 Add support for directories of certificates when compiled with a GnuTLS
498 version 3.3.6 or later.
500 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
501 is EXPERIMENTAL_EVENT, the main-configuration and transport options
502 both become "event_action", the variables become $event_name, $event_data
503 and $event_defer_errno. There is a new variable $verify_mode, usable in
504 routers, transports and related events. The tls:cert event is now also
505 raised for inbound connections, if the main configuration event_action
508 TL/06 In test suite, disable OCSP for old versions of openssl which contained
509 early OCSP support, but no stapling (appears to be less than 1.0.0).
511 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
512 server certificate names available under the smtp transport option
513 "tls_verify_cert_hostname" now do not permit multi-component wildcard
516 JH/10 Time-related extraction expansions from certificates now use the main
517 option "timezone" setting for output formatting, and are consistent
518 between OpenSSL and GnuTLS compilations. Bug 1541.
520 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
521 encoded parameter in the incoming message. Bug 1558.
523 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
524 include certificate info, eximon was claiming there were spoolfile
527 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
529 JH/14 Log delivery-related information more consistently, using the sequence
530 "H=<name> [<ip>]" wherever possible.
532 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
533 are problematic for Debian distribution, omit them from the release
536 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
538 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
541 JH/17 Fix a null-indirection in certextract expansions when a nondefault
542 output list separator was used.
547 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
548 checkers that were complaining about end of non-void function with no
551 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
552 This was a regression intruduced in 4.83 by another bugfix.
554 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
556 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
557 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
563 TF/01 Correctly close the server side of TLS when forking for delivery.
565 When a message was received over SMTP with TLS, Exim failed to clear up
566 the incoming connection properly after forking off the child process to
567 deliver the message. In some situations the subsequent outgoing
568 delivery connection happened to have the same fd number as the incoming
569 connection previously had. Exim would try to use TLS and fail, logging
570 a "Bad file descriptor" error.
572 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
573 utilities have not been installed.
575 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
576 temporary space as the ACL may create new global variables.
578 TL/01 LDAP support uses per connection or global context settings, depending
579 upon the detected version of the libraries at build time.
581 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
582 to extract and use the src ip:port in logging and expansions as if it
583 were a direct connection from the outside internet. PPv2 support was
584 updated based on HAProxy spec change in May 2014.
586 JH/02 Add ${listextract {number}{list}{success}{fail}}.
588 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
589 Properly escape header and check for NULL return.
591 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
594 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
596 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
597 characters in header names, implemented as a verify condition.
598 Contributed by Michael Fischer v. Mollard.
600 TL/05 Rename SPF condition results err_perm and err_temp to standardized
601 results permerror and temperror. Previous values are deprecated but
602 still accepted. In a future release, err_perm and err_temp will be
603 completely removed, which will be a backward incompatibility if the
604 ACL tests for either of these two old results. Patch contributed by
605 user bes-internal on the mailing list.
607 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
609 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
610 selectors, in both main and reject logs.
612 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
615 JH/07 Add malware type "sock" for talking to simple daemon.
617 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
619 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
620 routers/transports under cutthrough routing.
622 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
623 numbers. Touch up "bool" conditional to keep the same definition.
625 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
627 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
629 JH/12 Expand items in router/transport headers_add or headers_remove lists
630 individually rather than the list as a whole. Bug 1452.
632 Required for reasonable handling of multiple headers_ options when
633 they may be empty; requires that headers_remove items with embedded
634 colons must have them doubled (or the list-separator changed).
636 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
637 view the policy declared in the DMARC record. Currently, $dmarc_status
638 is a combined value of both the record presence and the result of the
641 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
643 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
644 dnslookup router and the smtp transport (applying to the forward
647 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
648 of ldap servers used for a specific lookup. Patch provided by Heiko
651 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
652 New variable $lookup_dnssec_authenticated for observability.
654 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
655 Patch submitted by Lars Timman.
657 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
659 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
660 Requires trusted mode and valid format message id, aborts otherwise.
661 Patch contributed by Heiko Schlichting.
663 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
664 certextract with support for various fields. Bug 1358.
666 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
667 is requested by default, modifiable by smtp transport option
670 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
671 operate on certificate variables to give certificate fingerprints
672 Also new ${sha256:cert_variable}.
674 JH/23 The PRDR feature is moved from being Experimental into the mainline.
676 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
677 Christian Aistleitner.
679 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
681 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
682 file. Patch from Wolfgang Breyha.
684 JH/25 Expand the coverage of the delivery $host and $host_address to
685 client authenticators run in verify callout. Bug 1476.
687 JH/26 Port service names are now accepted for tls_on_connect_ports, to
688 align with daemon_smtp_ports. Bug 72.
690 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
691 support and error reporting did not work properly.
693 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
694 and is readable. Patch from Andrew Colin Kissa.
696 TL/14 Enhance documentation of ${run expansion and how it parses the
697 commandline after expansion, particularly in the case when an
698 unquoted variable expansion results in an empty value.
700 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
702 PP/02 Fix internal collision of T_APL on systems which support RFC3123
703 by renaming away from it. Addresses GH issue 15, reported by
706 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
708 TL/15 SECURITY: prevent double expansion in math comparison functions
709 (can expand unsanitized data). Not remotely exploitable.
716 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
718 PP/02 Make -n do something, by making it not do something.
719 When combined with -bP, the name of an option is not output.
721 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
724 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
725 $sender_host_name and config options to manage this, and basic check
728 PP/05 DSCP support for outbound connections and control modifier for inbound.
730 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
731 (Only plugin which currently uses this is kerberos4, which nobody should
732 be using, but we should make it available and other future plugins might
733 conceivably use it, even though it would break NAT; stuff *should* be
734 using channel bindings instead).
736 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
737 name; added for Sendmail compatibility; requires admin caller.
738 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
739 just ignore it); requires trusted caller.
740 Also parse but ignore: -Ac -Am -X<logfile>
743 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
745 TL/02 Add +smtp_confirmation as a default logging option.
747 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
748 Patch by Magnus Holmgren from 2007-02-20.
750 TL/04 Bugzilla 1281 - Spec typo.
751 Bugzilla 1283 - Spec typo.
752 Bugzilla 1290 - Spec grammar fixes.
754 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
756 TL/06 Add Experimental DMARC support using libopendmarc libraries.
758 TL/07 Fix an out of order global option causing a segfault. Reported to dev
759 mailing list by by Dmitry Isaikin.
761 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
763 JH/02 Support "G" suffix to numbers in ${if comparisons.
765 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
767 NM/01 Bugzilla 1197 - Spec typo
768 Bugzilla 1196 - Spec examples corrections
770 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
772 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
773 gnutls_enable_pkcs11, but renamed to more accurately indicate its
776 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
777 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
779 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
780 "acl {{name}{arg}...}", and optional args on acl condition
783 JH/05 Permit multiple router/transport headers_add/remove lines.
785 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
787 JH/07 Avoid using a waiting database for a single-message-only transport.
788 Performance patch from Paul Fisher. Bugzilla 1262.
790 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
793 JH/09 Add $headers_added variable, with content from use of ACL modifier
794 add_header (but not yet added to the message). Bugzilla 199.
796 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
797 Pulled from Bugzilla 817 by Wolfgang Breyha.
799 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
801 (nb: this is the same fix as in Exim 4.80.1)
803 JH/11 Add A= logging on delivery lines, and a client_set_id option on
806 JH/12 Add optional authenticated_sender logging to A= and a log_selector
809 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
811 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
812 advertise SMTP AUTH mechanism to us, instead of a generic
813 protocol violation error. Also, make Exim more robust to bad
814 data from the Dovecot auth socket.
816 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
818 When a queue runner is handling a message, Exim first routes the
819 recipient addresses, during which it prunes them based on the retry
820 hints database. After that it attempts to deliver the message to
821 any remaining recipients. It then updates the hints database using
824 So if a recipient address works intermittently, it can get repeatedly
825 deferred at routing time. The retry hints record remains fresh so the
826 address never reaches the final cutoff time.
828 This is a fairly common occurrence when a user is bumping up against
829 their storage quota. Exim had some logic in its local delivery code
830 to deal with this. However it did not apply to per-recipient defers
831 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
833 This change adds a proper retry rule check during routing so that the
834 final cutoff time is checked against the message's age. We only do
835 this check if there is an address retry record and there is not a
836 domain retry record; this implies that previous attempts to handle
837 the address had the retry_use_local_parts option turned on. We use
838 this as an approximation for the destination being like a local
839 delivery, as in LMTP.
841 I suspect this new check makes the old local delivery cutoff check
842 redundant, but I have not verified this so I left the code in place.
844 TF/02 Correct gecos expansion when From: is a prefix of the username.
846 Test 0254 submits a message to Exim with the header
850 When I ran the test suite under the user fanf2, Exim expanded
851 the header to contain my full name, whereas it should have added
852 a Resent-Sender: header. It erroneously treats any prefix of the
853 username as equal to the username.
855 This change corrects that bug.
857 GF/01 DCC debug and logging tidyup
858 Error conditions log to paniclog rather than rejectlog.
859 Debug lines prefixed by "DCC: " to remove any ambiguity.
861 TF/03 Avoid unnecessary rebuilds of lookup-related code.
863 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
864 Bug spotted by Jeremy Harris; was flawed since initial commit.
865 Would have resulted in OCSP responses post-SNI triggering an Exim
866 NULL dereference and crash.
868 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
870 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
871 Bug detection, analysis and fix by Samuel Thibault.
872 Bugzilla 1331, Debian bug #698092.
874 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
876 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
877 Server implementation by Todd Lyons, client by JH.
878 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
879 config variable "prdr_enable" controls whether the server
880 advertises the facility. If the client requests PRDR a new
881 acl_data_smtp_prdr ACL is called once for each recipient, after
882 the body content is received and before the acl_smtp_data ACL.
883 The client is controlled by bolth of: a hosts_try_prdr option
884 on the smtp transport, and the server advertisement.
885 Default client logging of deliveries and rejections involving
886 PRDR are flagged with the string "PRDR".
888 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
889 fclose(). Diagnosis by Todd Lyons.
891 PP/17 Update configure.default to handle IPv6 localhost better.
892 Patch by Alain Williams (plus minor tweaks).
895 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
896 This is now consistent with GnuTLS, and is now documented: the
897 previous undocumented portable approach to treating the option as
898 unset was to force an expansion failure. That still works, and
899 an empty string is now equivalent.
901 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
902 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
903 not performing validation itself.
905 PP/20 Added force_command boolean option to pipe transport.
906 Patch from Nick Koston, of cPanel Inc.
908 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
911 TF/04 Added udpsend ACL modifer and hexquote expansion operator
913 PP/21 Fix eximon continuous updating with timestamped log-files.
914 Broken in a format-string cleanup in 4.80, missed when I repaired the
915 other false fix of the same issue.
916 Report and fix from Heiko Schlichting.
919 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
920 Report from Prashanth Katuri.
922 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
923 It's SecureTransport, so affects any MacOS clients which use the
924 system-integrated TLS libraries, including email clients.
926 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
927 using a MIME ACL for non-SMTP local injection.
928 Report and assistance in diagnosis by Warren Baker.
930 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
932 JH/16 Fix comparisons for 64b. Bugzilla 1385.
934 TL/09 Add expansion variable $authenticated_fail_id to keep track of
935 last id that failed so it may be referenced in subsequent ACL's.
937 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
940 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
941 ldap library initialization, allowing self-signed CA's to be
942 used. Also properly sets require_cert option later in code by
943 using NULL (global ldap config) instead of ldap handle (per
944 session). Bug diagnosis and testing by alxgomz.
946 TL/12 Enhanced documentation in the ratelimit.pl script provided in
947 the src/util/ subdirectory.
949 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
950 renamed to Transport Post Delivery Action by Jeremy Harris, as
953 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
954 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
955 redis_servers = needs to be configured which will be used by the redis
956 lookup. Patch from Warren Baker, of The Packet Hub.
958 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
960 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
961 hostname or reverse DNS when processing a host list. Used suggestions
962 from multiple comments on this bug.
964 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
966 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
967 Missed a few lines, added it to make the runtest require no keyboard
970 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
971 contains upper case chars. Make router use caseful_local_part.
973 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
974 support when GnuTLS has been built with p11-kit.
980 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
982 This, or similar/improved, will also be change PP/11 of 4.82.
988 PP/01 Handle short writes when writing local log-files.
989 In practice, only affects FreeBSD (8 onwards).
990 Bugzilla 1053, with thanks to Dmitry Isaikin.
992 NM/01 Bugzilla 949 - Documentation tweak
994 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
997 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
999 PP/02 Implemented gsasl authenticator.
1001 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1003 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1004 `pkg-config foo` for cflags/libs.
1006 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1007 with rest of GSASL and with heimdal_gssapi.
1009 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1010 `pkg-config foo` for cflags/libs for the TLS implementation.
1012 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1013 properties get this fed in as external SSF. A number of robustness
1014 and debugging improvements to the cyrus_sasl authenticator.
1016 PP/08 cyrus_sasl server now expands the server_realm option.
1018 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1019 Patch by Jeremy Harris.
1021 PP/10 Added dbmjz lookup type.
1023 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1025 PP/12 MAIL args handles TAB as well as SP, for better interop with
1026 non-compliant senders.
1027 Analysis and variant patch by Todd Lyons.
1029 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1030 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1031 Patch from Dirk Mueller <dmueller@suse.com>
1033 PP/13 tls_peerdn now print-escaped for spool files.
1034 Observed some $tls_peerdn in wild which contained \n, which resulted
1035 in spool file corruption.
1037 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1038 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1039 or write after TLS renegotiation, which otherwise led to messages
1042 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1043 as a tracking header (ie: a signed header comes before the signature).
1044 Patch from Wolfgang Breyha.
1046 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1047 comma-sep list; embedded commas doubled.
1049 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1051 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1053 Report and patch from Dmitry Banschikov.
1055 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
1056 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1057 failures. We appear to now support TLS1.1+ with Exim.
1059 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1060 lets Exim select keys and certificates based upon TLS SNI from client.
1061 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1062 before an outbound SMTP session. New log_selector, +tls_sni.
1064 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1065 NULL dereference. Report and patch from Alun Jones.
1067 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1068 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1069 Not seeing resolver debug output on NetBSD, but suspect this is a
1070 resolver implementation change.
1072 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1073 Left warnings. Added "eximon gdb" invocation mode.
1075 PP/21 Defaulting "accept_8bitmime" to true, not false.
1077 PP/22 Added -bw for inetd wait mode support.
1079 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1080 locate the relevant includes and libraries. Made this the default.
1082 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1083 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1085 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1086 This may cause build issues on older platforms.
1088 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1089 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1090 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1091 Added SNI support via GnuTLS too.
1092 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1094 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1096 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1097 Applied second patch from Janne, implementing suggestion to default
1098 multiple-strings-in-record handling to match SPF spec.
1100 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1102 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1103 read-only, out of scope).
1104 Patch from Wolfgang Breyha, report from Stuart Northfield.
1106 PP/29 Fix three issues highlighted by clang analyser static analysis.
1107 Only crash-plausible issue would require the Cambridge-specific
1108 iplookup router and a misconfiguration.
1109 Report from Marcin Mirosław.
1111 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1113 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1114 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1115 As part of this, removing so much warning spew let me fix some minor
1116 real issues in debug logging.
1118 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1119 assignment on my part. Fixed.
1121 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1122 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1123 Janne Snabb (who went above and beyond: thank you).
1125 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1126 string otherwise requires a connection and a bunch more work and it's
1127 relatively easy to get wrong. Should also expose TLS library linkage
1130 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1131 64-bit ${eval} (JH/03).
1133 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1134 GNU libc to support some of the 64-bit stuff, should not lead to
1135 conflicts. Defined before os.h is pulled in, so if a given platform
1136 needs to override this, it can.
1138 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1139 protection layer was required, which is not implemented.
1140 Bugzilla 1254, patch from Wolfgang Breyha.
1142 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1143 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1144 tls_dhparam take prime identifiers. Also unbreak combination of
1145 OpenSSL+DH_params+TLSSNI.
1147 PP/39 Disable SSLv2 by default in OpenSSL support.
1153 PP/01 Solaris build fix for Oracle's LDAP libraries.
1154 Bugzilla 1109, patch from Stephen Usher.
1156 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1158 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1161 TF/02 Fix a couple more cases where we did not log the error message
1162 when unlink() failed. See also change 4.74-TF/03.
1164 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1165 lock up or crash if it happened to be inside a call to libc when it
1166 got a SIGUSR1 from exiwhat.
1168 The SIGUSR1 handler appends the current process status to the process
1169 log which is later printed by exiwhat. It used to use the general
1170 purpose logging code to do this, but several functions it calls are
1171 not safe for signals.
1173 The new output code in the SIGUSR1 handler is specific to the process
1174 log, and simple enough that it's easy to inspect for signal safety.
1175 Removing some special cases also simplifies the general logging code.
1176 Removing the spurious timestamps from the process log simplifies
1179 TF/04 Improved ratelimit ACL condition.
1181 The /noupdate option has been deprecated in favour of /readonly which
1182 has clearer semantics. The /leaky, /strict, and /readonly update modes
1183 are mutually exclusive. The update mode is no longer included in the
1184 database key; it just determines when the database is updated. (This
1185 means that when you upgrde Exim will forget old rate measurements.)
1187 Exim now checks that the per_* options are used with an update mode that
1188 makes sense for the current ACL. For example, when Exim is processing a
1189 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1190 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1191 must specify per_mail/readonly. If you omit the update mode it defaults to
1192 /leaky where that makes sense (as before) or /readonly where required.
1194 The /noupdate option is now undocumented but still supported for
1195 backwards compatibility. It is equivalent to /readonly except that in
1196 ACLs where /readonly is required you may specify /leaky/noupdate or
1197 /strict/noupdate which are treated the same as /readonly.
1199 A useful new feature is the /count= option. This is a generalization
1200 of the per_byte option, so that you can measure the throughput of other
1201 aggregate values. For example, the per_byte option is now equivalent
1202 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1204 The per_rcpt option has been generalized using the /count= mechanism
1205 (though it's more complicated than the per_byte equivalence). When it is
1206 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1207 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1208 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1209 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1210 Note that using per_rcpt with a non-readonly update mode in more than
1211 one ACL will cause the recipients to be double-counted. (The per_mail
1212 and per_byte options don't have this problem.)
1214 The handling of very low rates has changed slightly. If the computed rate
1215 is less than the event's count (usually one) then this event is the first
1216 after a long gap. In this case the rate is set to the same as this event's
1217 count, so that the first message of a spam run is counted properly.
1219 The major new feature is a mechanism for counting the rate of unique
1220 events. The new per_addr option counts the number of different
1221 recipients that someone has sent messages to in the last time period. It
1222 behaves like per_rcpt if all the recipient addresses are different, but
1223 duplicate recipient addresses do not increase the measured rate. Like
1224 the /count= option this is a general mechanism, so the per_addr option
1225 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1226 example, measure the rate that a client uses different sender addresses
1227 with the options per_mail/unique=$sender_address. There are further
1228 details in the main documentation.
1230 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1232 TF/06 Removed a few PCRE remnants.
1234 TF/07 Automatically extract Exim's version number from tags in the git
1235 repository when doing development or release builds.
1237 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1238 Bugzilla 879. Patch from Paul Fisher.
1240 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1241 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1244 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1246 PP/05 Variable $av_failed, true if the AV scanner deferred.
1247 Bugzilla 1078. Patch from John Horne.
1249 PP/06 Stop make process more reliably on build failure.
1250 Bugzilla 1087. Patch from Heiko Schlittermann.
1252 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1253 Bugzilla 1089. Patch from Heiko Schlittermann.
1255 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1256 Bugzilla 1131. Patch from Holger Weiß.
1258 PP/09 Handle IPv6 addresses with SPF.
1259 Bugzilla 860. Patch from Wolfgang Breyha.
1261 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1263 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1266 PP/11 match_* no longer expand right-hand-side by default.
1267 New compile-time build option, EXPAND_LISTMATCH_RHS.
1268 New expansion conditions, "inlist", "inlisti".
1270 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1272 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1274 PP/14 fix log_write() format string regression from TF/03.
1275 Bugzilla 1152. Patch from Dmitry Isaikin.
1281 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1283 PP/02 Harmonised TLS library version reporting; only show if debugging.
1284 Layout now matches that introduced for other libraries in 4.74 PP/03.
1286 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1288 PP/04 New "dns_use_edns0" global option.
1290 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1293 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1294 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1296 TK/01 Updated PolarSSL code to 0.14.2.
1297 Bugzilla 1097. Patch from Andreas Metzler.
1299 PP/07 Catch divide-by-zero in ${eval:...}.
1300 Fixes bugzilla 1102.
1302 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1305 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1306 format-string attack -- SECURITY: remote arbitrary code execution.
1308 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1309 time unintentionally subject to list matching rules, letting the header
1310 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1311 arbitrary string expansion). This allowed for information disclosure.
1313 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1314 INT_MIN/-1 -- value coerced to INT_MAX.
1320 NM/01 Workround for PCRE version dependency in version reporting
1323 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1324 This fixes portability to compilers other than gcc, notably
1325 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1327 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1328 makefiles for portability to HP-UX and POSIX correctness.
1330 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1331 Also via indented variable definition in the Makefile.
1332 (Debugging by Oliver Heesakkers).
1334 PP/02 Restore caching of spamd results with expanded spamd_address.
1335 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1337 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1338 Improves build reliability. Fix from: Frank Elsner
1340 NM/02 Fix wide character breakage in the rfc2047 coding
1341 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1343 NM/03 Allow underscore in dnslist lookups
1344 Fixes bug 1026. Patch from Graeme Fowler
1346 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1347 Code patches from Adam Ciarcinski of NetBSD.
1349 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1352 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1353 is logged, to avoid truncation. Patch from John Horne.
1355 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1356 Patch from Jakob Hirsch.
1358 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1359 SQL string expansion failure details.
1360 Patch from Andrey Oktyabrski.
1362 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1363 Patch from Simon Arlott.
1365 PP/09 New lookups functionality failed to compile on old gcc which rejects
1366 extern declarations in function scope.
1367 Patch from Oliver Fleischmann
1369 PP/10 Use sig_atomic_t for flags set from signal handlers.
1370 Check getgroups() return and improve debugging.
1371 Fixed developed for diagnosis in bug 927 (which turned out to be
1374 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1375 Patch from Mark Zealey.
1377 PP/12 Bugzilla 1056: Improved spamd server selection.
1378 Patch from Mark Zealey.
1380 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1381 Based on patch from Heiko Schlittermann.
1383 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1384 Patch from Uwe Doering, sign-off by Michael Haardt.
1386 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1387 variable declaration deep within a block. Bug and patch from
1390 PP/15 lookups-Makefile IRIX compatibilty coercion.
1392 PP/16 Make DISABLE_DKIM build knob functional.
1394 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1395 Patch by Simon Arlott
1397 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1398 variable argument macros. Our copy now differs from upstream.
1404 TF/01 Failure to get a lock on a hints database can have serious
1405 consequences so log it to the panic log.
1407 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1408 controlled using the smtp_confirmation log selector.
1410 TF/03 Include the error message when we fail to unlink a spool file.
1412 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1413 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1414 for maintaining out-of-tree patches for some time.
1416 PP/01 Bugzilla 139: Documentation and portability issues.
1417 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1418 Handle per-OS dynamic-module compilation flags.
1420 PP/02 Let /dev/null have normal permissions.
1421 The 4.73 fixes were a little too stringent and complained about the
1422 permissions on /dev/null. Exempt it from some checks.
1423 Reported by Andreas M. Kirchwitz.
1425 PP/03 Report version information for many libraries, including
1426 Exim version information for dynamically loaded libraries. Created
1427 version.h, now support a version extension string for distributors
1428 who patch heavily. Dynamic module ABI change.
1430 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1431 privilege escalation vulnerability whereby the Exim run-time user
1432 can cause root to append content of the attacker's choosing to
1435 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1438 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1439 If dropping privileges for untrusted macros, we disabled normal logging
1440 on the basis that it would fail; for the Exim run-time user, this is not
1441 the case, and it resulted in successful deliveries going unlogged.
1442 Fixed. Reported by Andreas Metzler.
1448 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1449 only prepend for the Resent-* case. Fixes regression introduced in
1450 Exim 4.70 by NM/22 for Bugzilla 607.
1452 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1453 increasing numbers of administrators be bitten by this.
1455 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1457 PP/03 Bugzilla 994: added openssl_options main configuration option.
1459 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1461 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1463 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1465 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1466 without a peer certificate, leading to a segfault because of an
1467 assumption that peers always have certificates. Be a little more
1468 paranoid. Problem reported by Martin Tscholak.
1470 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1471 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1472 NB: ClamAV planning to remove STREAM in "middle of 2010".
1473 CL also introduces -bmalware, various -d+acl logging additions and
1474 more caution in buffer sizes.
1476 PP/09 Implemented reverse_ip expansion operator.
1478 PP/10 Bugzilla 937: provide a "debug" ACL control.
1480 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1482 PP/12 Bugzilla 973: Implement --version.
1484 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1486 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1488 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1490 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1491 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1492 ignore trailing whitespace.
1494 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1496 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1499 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1500 Notification from Dr Andrew Aitchison.
1502 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1503 ExtendedDetectionInfo response format.
1504 Notification from John Horne.
1506 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1509 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1510 XSL and documented dependency on system catalogs, with examples of how
1513 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1516 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1517 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1518 configuration file which is writeable by the Exim user or group.
1520 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1521 of configuration files to cover files specified with the -C option if
1522 they are going to be used with root privileges, not just the default
1525 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1526 option (effectively making it always true).
1528 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1529 files to be used while preserving root privileges.
1531 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1532 that rogue child processes cannot use them.
1534 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1535 run-time user, instead of root.
1537 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1538 Exim run-time user without dropping privileges.
1540 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1541 result string, instead of calling string_vformat() twice with the same
1544 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1545 for other users. Others should always drop root privileges if they use
1546 -C on the command line, even for a whitelisted configure file.
1548 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1550 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1556 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1557 $data_path, and $header_path variables; fixed documentation bugs and
1560 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1561 exipick to access non-standard spools, including the "frozen" queue
1564 NM/01 Bugzilla 965: Support mysql stored procedures.
1565 Patch from Alain Williams
1567 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1569 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1570 Patch from Andreas Metzler
1572 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1573 Patch from Kirill Miazine
1575 NM/05 Bugzilla 671: Added umask to procmail example.
1577 JJ/03 installed exipick 20100323.0, fixing doc bug
1579 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1580 directory. Notification and patch from Dan Rosenberg.
1582 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1584 TK/02 Improve log output when DKIM signing operation fails.
1586 MH/01 Treat the transport option dkim_domain as a colon separated
1587 list, not as a single string, and sign the message with each element,
1588 omitting multiple occurences of the same signer.
1590 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1591 Bugzilla 985, 986. Patch by Simon Arlott
1593 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1594 Patch by Simon Arlott
1596 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1597 MBX locking. Notification from Dan Rosenberg.
1603 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1605 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1607 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1609 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1611 NM/04 Bugzilla 915: Flag broken perl installation during build.
1617 TK/01 Added patch by Johannes Berg that expands the main option
1618 "spamd_address" if it starts with a dollar sign.
1620 TK/02 Write list of recipients to X-Envelope-Sender header when building
1621 the mbox-format spool file for content scanning (suggested by Jakob
1624 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1625 (http://www.dcc-servers.net/) support via dccifd. Activated by
1626 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1628 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1629 by Mark Daniel Reidel <mr@df.eu>.
1631 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1632 When building exim an external PCRE library is now needed -
1633 PCRE is a system library on the majority of modern systems.
1634 See entry on PCRE_LIBS in EDITME file.
1636 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1637 conversation. Added nologin parameter to request.
1638 Patch contributed by Kirill Miazine.
1640 TF/01 Do not log submission mode rewrites if they do not change the address.
1642 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1644 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1645 log files in place. Contributed by Roberto Lima.
1647 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1649 TF/03 Bugzilla 615: When checking the local_parts router precondition
1650 after a local_part_suffix or local_part_prefix option, Exim now
1651 does not use the address's named list lookup cache, since this
1652 contains cached lookups for the whole local part.
1654 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1655 Robert Millan. Documentation is in experimental-spec.txt.
1657 TF/04 Bugzilla 668: Fix parallel build (make -j).
1659 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1661 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1662 Patch provided by Jan Srzednicki.
1664 TF/05 Leading white space used to be stripped from $spam_report which
1665 wrecked the formatting. Now it is preserved.
1667 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1668 that they are available at delivery time.
1670 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1672 TF/08 TLS error reporting now respects the incoming_interface and
1673 incoming_port log selectors.
1675 TF/09 Produce a more useful error message if an SMTP transport's hosts
1676 setting expands to an empty string.
1678 NM/06 Bugzilla 744: EXPN did not work under TLS.
1679 Patch provided by Phil Pennock.
1681 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1682 Patch provided by Richard Godbee.
1684 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1685 acl_smtp_notquit, added index entry.
1687 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1688 Patch provided by Eugene Bujak.
1690 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1691 accept(). Patch provided by Maxim Dounin.
1693 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1694 Patch provided by Phil Pennock.
1696 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1698 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1699 Patch provided by Brad "anomie" Jorsch.
1701 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1702 Patch provided by Dean Brooks.
1704 TK/05 Add native DKIM support (does not depend on external libraries).
1706 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1707 Patch provided by Graeme Fowler.
1709 NM/16 Bugzilla 851: Documentation example syntax fix.
1711 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1713 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1716 NM/19 Bugzilla 745: TLS version reporting.
1717 Patch provided by Phil Pennock.
1719 NM/20 Bugzilla 167: bool: condition support.
1720 Patch provided by Phil Pennock.
1722 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1723 clients. Patch provided by Phil Pennock.
1725 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1726 Patch provided by Brad "anomie" Jorsch.
1728 NM/23 Bugzilla 687: Fix misparses in eximstats.
1729 Patch provided by Heiko Schlittermann.
1731 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1732 Patch provided by Heiko Schlittermann.
1734 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1735 plus update to original patch.
1737 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1739 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1740 Patch provided by David Brownlee.
1742 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1744 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1746 NM/30 Bugzilla 888: TLS documentation bugfixes.
1748 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1750 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1751 Unlike the original bugzilla I have changed all shell scripts in src tree.
1753 NM/33 Bugzilla 898: Transport filter timeout fix.
1754 Patch by Todd Rinaldo.
1756 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1757 Patch by Serge Demonchaux.
1759 NM/35 Bugzilla 39: Base64 decode bug fixes.
1760 Patch by Jakob Hirsch.
1762 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1764 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1766 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1768 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1774 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1775 ALT-N's libdkim that I have put here:
1776 http://duncanthrax.net/exim-experimental/
1778 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1779 were called 'true' and it seems that C99 defines that as a reserved
1780 keyword to be used with 'bool' variable types. That means you could
1781 not include C99-style headers which use bools without triggering
1782 build errors in sieve.c.
1784 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1785 as mailq or other aliases. Changed the --help handling significantly
1786 to do whats expected. exim_usage() emits usage/help information.
1788 SC/01 Added the -bylocaldomain option to eximstats.
1790 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1792 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1794 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1800 PH/01 Another patch from the Sieve maintainer.
1802 PH/02 When an IPv6 address is converted to a string for single-key lookup
1803 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1804 dots are used instead of colons so that keys in lsearch files need not
1805 contain colons. This was done some time before quoting was made available
1806 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1807 (notated using the quote facility) so as to distinguish them from IPv4
1808 keys. This meant that lookups for IP addresses in host lists did not work
1809 for iplsearch lookups.
1811 This has been fixed by arranging for IPv6 addresses to be expressed with
1812 colons if the lookup type is iplsearch. This is not incompatible, because
1813 previously such lookups could never work.
1815 The situation is now rather anomolous, since one *can* have colons in
1816 ordinary lsearch keys. However, making the change in all cases is
1817 incompatible and would probably break a number of configurations.
1819 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1822 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1823 conversion specification without a maximum field width, thereby enabling
1824 a rogue spamd server to cause a buffer overflow. While nobody in their
1825 right mind would setup Exim to query an untrusted spamd server, an
1826 attacker that gains access to a server running spamd could potentially
1827 exploit this vulnerability to run arbitrary code as the Exim user.
1829 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1830 $primary_hostname instead of what libspf2 thinks the hosts name is.
1832 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1833 a directory entry by the name of the lookup key. Previously, if a
1834 symlink pointed to a non-existing file or a file in a directory that
1835 Exim lacked permissions to read, a lookup for a key matching that
1836 symlink would fail. Now it is enough that a matching directory entry
1837 exists, symlink or not. (Bugzilla 503.)
1839 PH/03 The body_linecount and body_zerocount variables are now exported in the
1842 PH/04 Added the $dnslist_matched variable.
1844 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1845 This means they are set thereafter only if the connection becomes
1848 PH/06 Added the client_condition to authenticators so that some can be skipped
1849 by clients under certain conditions.
1851 PH/07 The error message for a badly-placed control=no_multiline_responses left
1852 "_responses" off the end of the name.
1854 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1856 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1857 (without spaces) instead of just copying the configuration text.
1859 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1861 PH/11 Added $max_received_linelength.
1863 PH/12 Added +ignore_defer and +include_defer to host lists.
1865 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1866 way in which PCRE > 7.0 is built.
1868 PH/14 Implemented queue_only_load_latch.
1870 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1871 MAIL command. The effect was to mangle the value on 64-bit systems.
1873 PH/16 Another patch from the Sieve maintainer.
1875 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1877 PH/18 If a system quota error occurred while trying to create the file for
1878 a maildir delivery, the message "Mailbox is full" was not appended to the
1879 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1880 only to a quota excession during the actual writing of the file.
1882 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1883 characters?) which causes problems in log lines. The DN values are now
1884 passed through string_printing() before being added to log lines.
1886 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1887 and InterBase are left for another time.)
1889 PH/21 Added message_body_newlines option.
1891 PH/22 Guard against possible overflow in moan_check_errorcopy().
1893 PH/23 POSIX allows open() to be a macro; guard against that.
1895 PH/24 If the recipient of an error message contained an @ in the local part
1896 (suitably quoted, of course), incorrect values were put in $domain and
1897 $local_part during the evaluation of errors_copy.
1903 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1904 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1907 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1908 issue a MAIL command.
1910 PH/02 In an ACL statement such as
1912 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1914 if a client was not listed at all, or was listed with a value other than
1915 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1916 the condition was not true (as it should be), so access was not denied.
1917 The bug was that the ! inversion was incorrectly passed on to the second
1918 item. This has been fixed.
1920 PH/03 Added additional dnslists conditions == and =& which are different from
1921 = and & when the dns lookup returns more than one IP address.
1923 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1924 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1926 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1927 FSYNC, which compiles an option called disable_fsync that allows for
1928 bypassing fsync(). The documentation is heavily laced with warnings.
1930 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1932 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1933 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1934 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1935 including adding "make clean"; (3) Added -fPIC when compiling the test
1936 dynamically loaded module, to get rid of a warning.
1938 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1939 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1940 = 0s. The bug is otherwise harmless.
1942 PH/07 There was a bug in the dovecot authenticator such that the value of
1943 $auth1 could be overwritten, and so not correctly preserved, after a
1944 successful authentication. This usually meant that the value preserved by
1945 the server_setid option was incorrect.
1947 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1949 PH/09 Installed PCRE release 7.0.
1951 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1952 run for batched SMTP input. It is now run at the start of every message
1953 in the batch. While fixing this I discovered that the process information
1954 (output by running exiwhat) was not always getting set for -bs and -bS
1955 input. This is fixed, and it now also says "batched" for BSMTP.
1957 PH/11 Added control=no_pipelining.
1959 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1960 patch, slightly modified), and move the expansion of helo_data till after
1961 the connection is made in the smtp transport (so it can use these
1964 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1966 PH/14 Added log_selector = +pid.
1968 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1970 PH/16 Add ${if forany and ${if forall.
1972 PH/17 Added dsn_from option to vary the From: line in DSNs.
1974 PH/18 Flush SMTP output before performing a callout, unless control =
1975 no_callout_flush is set.
1977 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1978 was true (the default) a successful delivery failed to delete the retry
1979 item, thus causing premature timeout of the address. The bug is now
1982 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1984 PH/21 Long custom messages for fakedefer and fakereject are now split up
1985 into multiline reponses in the same way that messages for "deny" and
1986 other ACL rejections are.
1988 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1989 with slight modification.
1991 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1992 draft, changing the syntax and factoring some duplicate code".
1994 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1995 for deliveries of the second and subsequent messages over the same SMTP
1998 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1999 ${reduce, with only minor "tidies".
2001 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2003 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2004 expansion side effects.
2006 PH/27 When a message times out after an over-quota error from an Exim-imposed
2007 quota, the bounce message says "mailbox is full". This message was not
2008 being given when it was a system quota that was exceeded. It now should
2011 MH/03 Made $recipients available in local_scan(). local_scan() already has
2012 better access to the recipient list through recipients_list[], but
2013 $recipients can be useful in postmaster-provided expansion strings.
2015 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2016 in the case of a MAIL command with additional options following the
2017 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2018 were accidentally chopped off.
2020 PH/29 SMTP synchronization checks are implemented when a command is read -
2021 there is a check that no more input is waiting when there shouldn't be
2022 any. However, for some commands, a delay in an ACL can mean that it is
2023 some time before the response is written. In this time, more input might
2024 arrive, invalidly. So now there are extra checks after an ACL has run for
2025 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2026 pipelining has not been advertised.
2028 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2030 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2031 returned in the SMTP response when smtp_return_error_details was false.
2032 This has been fixed.
2034 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2035 instead of the C I/O that was originally supplied, because problems were
2036 reported on Solaris.
2038 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2039 Exim which did not show up earlier: it was assuming that a call to
2040 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2041 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2042 was a macro that became an assignment, so it seemed to work. This has
2043 changed to a proper function call with a void return, hence the compile
2044 error. Exim's code has been fixed.
2046 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2049 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2051 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2053 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2054 because it turns out to be too big or there is a timeout) there is no
2055 "Completed" line in the log. When some messages of this type were
2056 selected by exigrep, they were listed as "not completed". Others were
2057 picked up by some special patterns. I have improved the selection
2058 criteria to be more general.
2060 PH/38 The host_find_failed option in the manualroute router can now be set
2061 to "ignore", to completely ignore a host whose IP address cannot be
2062 found. If all hosts are ignored, the behaviour is controlled by the new
2063 host_all_ignored option.
2065 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2066 homing or because of multiple MX records with /mx) generated more than
2067 one IP address, and the following item turned out to be the local host,
2068 all the secondary addresses of the first item were incorrectly removed
2069 from the list, along with the local host and any following hosts (which
2070 is what is supposed to happen).
2072 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2073 whoever called Exim into the -H file. In the case of the daemon it was
2074 behaving confusingly. When first started, it used values for whoever
2075 started the daemon, but after a SIGHUP it used the Exim user (because it
2076 calls itself on a restart). I have changed the code so that it now always
2079 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2080 message are rejected with the same error (e.g. no authentication or bad
2081 sender address), and a DATA command is nevertheless sent (as can happen
2082 with PIPELINING or a stupid MUA), the error message that was given to the
2083 RCPT commands is included in the rejection of the DATA command. This is
2084 intended to be helpful for MUAs that show only the final error to their
2087 PH/42 Another patch from the Sieve maintainer.
2089 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2090 Eximstats - Fixed some broken HTML links and added missing column headers
2092 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2093 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2095 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2097 PH/43 Yet another patch from the Sieve maintainer.
2099 PH/44 I found a way to check for a TCP/IP connection going away before sending
2100 the response to the final '.' that terminates a message, but only in the
2101 case where the client has not sent further data following the '.'
2102 (unfortunately, this is allowed). However, in many cases there won't be
2103 any further data because there won't be any more messages to send. A call
2104 to select() can be used: if it shows that the input is "ready", there is
2105 either input waiting, or the socket has been closed. An attempt to read
2106 the next input character can distinguish the two cases. Previously, Exim
2107 would have sent an OK response which the client would never have see.
2108 This could lead to message repetition. This fix should cure that, at
2109 least in a lot of common cases.
2111 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2112 advertised in response to EHLO.
2118 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2119 fixed by 4.65/MH/01 (is this a record?) are fixed:
2121 (i) An empty string was always treated as zero by the numeric comparison
2122 operators. This behaviour has been restored.
2124 (ii) It is documented that the numeric comparison operators always treat
2125 their arguments as decimal numbers. This was broken in that numbers
2126 starting with 0 were being interpreted as octal.
2128 While fixing these problems I realized that there was another issue that
2129 hadn't been noticed. Values of message_size_limit (both the global option
2130 and the transport option) were treated as octal if they started with 0.
2131 The documentation was vague. These values are now always treated as
2132 decimal, and I will make that clear in the documentation.
2138 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2139 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2142 MH/01 Don't check that the operands of numeric comparison operators are
2143 integers when their expansion is in "skipping" mode (fixes bug
2144 introduced by 4.64-PH/07).
2146 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2147 child addresses, Exim now panics and dies. Previously, because the count
2148 is held in a short int, deliveries were likely to be lost. As such a
2149 large number of recipients for a single message is ridiculous
2150 (performance will be very, very poor), I have chosen to impose a limit
2151 rather than extend the field.
2157 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2158 leftover -K file (the existence of which was triggered by #402).
2159 While we were at it, introduced process PID as part of the -K
2160 filename. This should rule out race conditions when creating
2163 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2164 processing considerably. Previous code took too long for large mails,
2165 triggering a timeout which in turn triggers #401.
2167 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2168 in the DK code in transports.c. sendfile() is not really portable,
2169 hence the _LINUX specificness.
2171 TF/01 In the add_headers option to the mail command in an Exim filter,
2172 there was a bug that Exim would claim a syntax error in any
2173 header after the first one which had an odd number of characters
2176 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2177 callout verification, Exim cached a "reject" for the entire domain. This
2178 is correct for most verifications, but it is not correct for a recipient
2179 verification with use_sender or use_postmaster set, because in that case
2180 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2181 case of MAIL FROM:<> rejection from other early rejections (e.g.
2182 rejection of HELO). When verifying a recipient using a non-null MAIL
2183 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2184 Whatever the result of the callout, the value of the domain cache is
2185 left unchanged (for any other kind of callout, getting as far as trying
2186 RCPT means that the domain itself is ok).
2188 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2191 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2192 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2193 session, the daemon ignores ECONNECT errors and logs others; it now
2194 ignores EPIPE as well.
2196 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2197 (quoted-printable decoding).
2199 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2200 later the small subsequent patch to fix an introduced bug.
2202 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2204 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2206 PH/08 An error is now given if message_size_limit is specified negative.
2208 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2209 to be given (somewhat) arbitrary names.
2211 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2214 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2215 miscellaneous code fixes
2217 PH/10 Added the log_reject_target ACL modifier to specify where to log
2220 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2221 hostname. This is wrong, because it relates to the incoming message (and
2222 probably the interface on which it is arriving) and not to the outgoing
2223 callout (which could be using a different interface). This has been
2224 changed to use the value of the helo_data option from the smtp transport
2225 instead - this is what is used when a message is actually being sent. If
2226 there is no remote transport (possible with a router that sets up host
2227 addresses), $smtp_active_hostname is used.
2229 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2230 tweaks were necessary in order to get it to work (see also 21 below):
2231 (a) The code assumed that strncpy() returns a negative number on buffer
2232 overflow, which isn't the case. Replaced with Exim's string_format()
2234 (b) There were several signed/unsigned issues. I just did the minimum
2235 hacking in of casts. There is scope for a larger refactoring.
2236 (c) The code used strcasecmp() which is not a standard C function.
2237 Replaced with Exim's strcmpic() function.
2238 (d) The code set only $1; it now sets $auth1 as well.
2239 (e) A simple test gave the error "authentication client didn't specify
2240 service in request". It would seem that Dovecot has changed its
2241 interface. Fortunately there's a specification; I followed it and
2242 changed what the client sends and it appears to be working now.
2244 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2247 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2248 address A is aliased to B and C, where B exists and C does not. Without
2249 -v the output is "A verified" because verification stops after a
2250 successful redirection if more than one address is generated. However,
2251 with -v the child addresses are also verified. Exim was outputting "A
2252 failed to verify" and then showing the successful verification for C,
2253 with its parentage. It now outputs "B failed to verify", showing B's
2254 parentage before showing the successful verification of C.
2256 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2257 look up a TXT record in a specific list after matching in a combined
2260 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2261 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2262 they consult the DNS. I had assumed they would set it the way they
2263 wanted; and indeed my experiments on Linux seem to show that in some
2264 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2265 To be on the safe side, however, I have now made the interface to
2266 host_find_byname() similar to host_find_bydns(), with an argument
2267 containing the DNS resolver options. The host_find_byname() function now
2268 sets these options at its start, just as host_find_bydns() does. The smtp
2269 transport options dns_qualify_single and dns_search_parents are passed to
2270 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2271 of host_find_byname() use the default settings of RES_DEFNAMES
2272 (qualify_single) but not RES_DNSRCH (search_parents).
2274 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2275 spool_read_header() do less string testing, by means of a preliminary
2276 switch on the second character of optional "-foo" lines. (This is
2277 overdue, caused by the large number of possibilities that now exist.
2278 Originally there were few.) While I was there, I also converted the
2279 str(n)cmp tests so they don't re-test the leading "-" and the first
2280 character, in the hope this might squeeze out yet more improvement.
2282 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2283 flag allowing group syntax was set by the header_syntax check but not
2284 turned off, possible causing trouble later; (2) The flag was not being
2285 set at all for the header_verify test, causing "group"-style headers to
2286 be rejected. I have now set it in this case, and also caused header_
2287 verify to ignore an empty address taken from a group. While doing this, I
2288 came across some other cases where the code for allowing group syntax
2289 while scanning a header line wasn't quite right (mostly, not resetting
2290 the flag correctly in the right place). These bugs could have caused
2291 trouble for malformed header lines. I hope it is now all correct.
2293 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2294 with the "reply" argument non-NULL. The code, however (which originally
2295 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2296 but it didn't always do it. This confused somebody who was copying the
2297 code for some other use. I have removed all the tests.
2299 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2300 feature that was used to support insecure browsers during the U.S. crypto
2301 embargo. It requires special client support, and Exim is probably the
2302 only MTA that supported it -- and would never use it because real RSA is
2303 always available. This code has been removed, because it had the bad
2304 effect of slowing Exim down by computing (never used) parameters for the
2305 RSA_EXPORT functionality.
2307 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2308 authenticator to fail if there's a tab character in the incoming data
2309 (there should never be unless someone is messing about, as it's supposed
2310 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2311 if the connection is using TLS or if the remote IP is the same as the
2312 local IP, and the "valid-client-cert option" if a client certificate has
2315 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2316 authenticators. This can be used for authorization after authentication
2317 succeeds. (In the case of plaintext, it servers for both authentication
2320 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2321 if any retry times were supplied.
2323 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2324 connection, where there is no client IP address to check. In this
2325 situation, the verify now always succeeds.
2327 PH/25 Applied John Jetmore's -Mset patch.
2329 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2331 PH/27 In a string expansion for a processed (not raw) header when multiple
2332 headers of the same name were present, leading whitespace was being
2333 removed from all of them, but trailing whitespace was being removed only
2334 from the last one. Now trailing whitespace is removed from each header
2335 before concatenation. Completely empty headers in a concatenation (as
2336 before) are ignored.
2338 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2339 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2341 PH/29 [Removed. This was a change that I later backed out, and forgot to
2342 correct the ChangeLog entry (that I had efficiently created) before
2343 committing the later change.]
2345 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2346 address errors (4xx response to RCPT) over the same connection as other
2347 messages routed to the same hosts. Such deliveries are always "forced",
2348 so retry times are not inspected. This resulted in far too many retries
2349 for the affected addresses. The effect occurred only when there were more
2350 hosts than the hosts_max_try setting in the smtp transport when it had
2351 the 4xx errors. Those hosts that it had tried were not added to the list
2352 of hosts for which the message was waiting, so if all were tried, there
2353 was no problem. Two fixes have been applied:
2355 (i) If there are any address or message errors in an SMTP delivery, none
2356 of the hosts (tried or untried) are now added to the list of hosts
2357 for which the message is waiting, so the message should not be a
2358 candidate for sending over the same connection that was used for a
2359 successful delivery of some other message. This seems entirely
2360 reasonable: after all the message is NOT "waiting for some host".
2361 This is so "obvious" that I'm not sure why it wasn't done
2362 previously. Hope I haven't missed anything, but it can't do any
2363 harm, as the worst effect is to miss an optimization.
2365 (ii) If, despite (i), such a delivery is accidentally attempted, the
2366 routing retry time is respected, so at least it doesn't keep
2367 hammering the server.
2369 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2370 in ${readsocket because some servers need this prod.
2372 PH/32 Added some extra debug output when updating a wait-xxx database.
2374 PH/33 The hint "could be header name not terminated by colon", which has been
2375 given for certain expansion errors for a long time, was not being given
2376 for the ${if def:h_colon_omitted{... case.
2378 PH/34 The spec says: "With one important exception, whenever a domain list is
2379 being scanned, $domain contains the subject domain." There was at least
2380 one case where this was not true.
2382 PH/35 The error "getsockname() failed: connection reset by peer" was being
2383 written to the panic log as well as the main log, but it isn't really
2384 panic-worthy as it just means the connection died rather early on. I have
2385 removed the panic log writing for the ECONNRESET error when getsockname()
2388 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2389 runs only) independently of the message's sender address. This meant
2390 that, if the 4xx error was in fact related to the sender, a different
2391 message to the same recipient with a different sender could confuse
2392 things. In particualar, this can happen when sending to a greylisting
2393 server, but other circumstances could also provoke similar problems.
2394 I have changed the default so that the retry time for these errors is now
2395 based a combination of the sender and recipient addresses. This change
2396 can be overridden by setting address_retry_include_sender=false in the
2399 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2400 remote server are returned as part of bounce messages. This was not
2401 happening for LMTP over a pipe (the lmtp transport), but now it is the
2402 same for both kinds of LMTP.
2404 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2405 in header lines that were in fact CNAMEs. This is no longer the case.
2407 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2408 and queue runs started by the daemon processed all messages. This has
2409 been fixed so that -R and -S can now usefully be given with -q<time>.
2411 PH/40 Import PCRE release 6.7 (fixes some bugs).
2413 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2415 PH/42 Give an error if -q is specified more than once.
2417 PH/43 Renamed the variables $interface_address and $interface_port as
2418 $received_ip_address and $received_port, to make it clear that these
2419 values apply to message reception, and not to the outgoing interface when
2420 a message is delivered. (The old names remain recognized, of course.)
2422 PH/44 There was no timeout on the connect() call when using a Unix domain
2423 socket in the ${readsocket expansion. There now is.
2425 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2426 be meaningful with "accept".
2428 SC/01 Eximstats V1.43
2429 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2431 SC/02 Eximstats V1.44
2432 Use a glob alias rather than an array ref in the generated
2433 parser. This improves both readability and performance.
2435 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2436 Collect SpamAssassin and rejection statistics.
2437 Don't display local sender or destination tables unless
2438 there is data to show.
2439 Added average volumes into the top table text output.
2441 SC/04 Eximstats V1.46
2442 Collect data on the number of addresses (recipients)
2443 as well as the number of messages.
2445 SC/05 Eximstats V1.47
2446 Added 'Message too big' to the list of mail rejection
2447 reasons (thanks to Marco Gaiarin).
2449 SC/06 Eximstats V1.48
2450 Mainlog lines which have GMT offsets and are too short to
2451 have a flag are now skipped.
2453 SC/07 Eximstats V1.49 (Alain Williams)
2454 Added the -emptyok flag.
2456 SC/08 Eximstats V1.50
2457 Fixes for obtaining the IP address from reject messages.
2459 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2460 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2461 whitesspace changes from 4.64-PH/27
2463 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2466 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2467 are found, allow negative numbers in numeric criteria)
2469 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2471 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2472 to match changes made in 4.64-PH/43
2474 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2476 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2477 "domain" is actually an IP address, and force a failure. This locks out
2478 those revolvers/nameservers that support "A-for-A" lookups, in
2479 contravention of the specifications.
2481 PH/48 When a host name was looked up from an IP address, and the subsequent
2482 forward lookup of the name timed out, the host name was left in
2483 $sender_host_name, contrary to the specification.
2485 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2486 restricted to single-key lookups, Exim was not diagnosing an error if
2487 * or *@ was used with a query-style lookup.
2489 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2491 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2492 long ago, but noone interested enough thought of it. Let's just say that
2493 the "1.1" means that there are some new functions that weren't there at
2494 some point in the past.
2496 PH/51 Error processing for expansion failure of helo_data from an smtp
2497 transport during callout processing was broken.
2499 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2500 tested/used via the -bh/-bhc/-bs options.
2502 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2503 bug, fixed in subsequent PCRE releases).
2505 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2506 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2508 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2514 SC/01 Use a glob alias rather than an array ref in eximstats generated
2515 parser. This improves both readability and performance.
2517 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2518 Don't display local sender or destination tables in eximstats unless
2519 there is data to show.
2520 Added average volumes into the eximstats top table text output.
2522 SC/03 Collect data on the number of addresses (recipients) as well
2523 as the number of messages in eximstats.
2525 TF/01 Correct an error in the documentation for the redirect router. Exim
2526 does (usually) call initgroups() when daemonizing.
2528 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2529 with consistent privilege compared to when running as a daemon.
2531 TF/03 Note in the spec that $authenticated_id is not set for local
2532 submissions from trusted users.
2534 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2535 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2537 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2538 by adding some example configuration directives to the default
2539 configuration file. A little bit of work is required to uncomment the
2540 directives and define how usernames and passwords are checked, but
2541 there is now a framework to start from.
2543 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2544 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2545 without this. I don't know how relevant this is to other LDAP libraries.
2547 PH/02 Add the verb name to the "unknown ACL verb" error.
2549 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2551 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2553 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2554 directory not expanded when it should be if an expanded home directory
2555 was set for the address (which is overridden by the transport).
2557 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2560 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2561 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2562 because it is too late at that time, and has no effect.
2564 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2565 security issue with \' (bugzilla #107). I could not use the
2566 PQescapeStringConn() function, because it needs a PGconn value as one of
2569 PH/08 When testing addresses using -bt, indicate those final addresses that
2570 are duplicates that would not cause an additional delivery. At least one
2571 person was confused, thinking that -bt output corresponded to deliveries.
2572 (Suppressing duplicates isn't a good idea as you lose the information
2573 about possibly different redirections that led to the duplicates.)
2575 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2576 systems where poll() doesn't work, in particular OS X.
2578 PH/10 Added more information to debugging output for retry time not reached.
2580 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2581 operations in malware.c.
2583 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2586 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2587 syslog_duplication set false, log lines that would normally be written
2588 both the the main log and to the reject log were not written to syslog at
2591 PH/14 In the default configuration, change the use of "message" in ACL warn
2592 statements to "add_header".
2594 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2595 not followed by a command (e.g. "seen endif").
2597 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2598 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2601 PH/17 Added extra conditions to the default value of delay_warning_condition
2605 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2606 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2607 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2610 The Auto-Submitted: and various List- headers are standardised, whereas I
2611 don't think Precedence: ever was.
2613 PH/18 Refactored debugging code in route_finduser() to show more information,
2614 in particular, the error code if getpwnam() issues one.
2616 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2617 This is apparently needed in addition to the PH/07 change above to avoid
2618 any possible encoding problems.
2620 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2621 but not after initializing Perl.
2623 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2624 output them only if debugging. By default they are written stderr,
2625 apparently, which is not desirable.
2627 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2630 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2633 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2635 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2636 authenticated or an ident call has been made. Suppress the default
2637 values for $authenticated_id and $authenticated_sender (but permit -oMai
2638 and -oMas) when testing with -bh.
2640 PH/24 Re-jigged the order of the tests in the default configuration so that the
2641 tests for valid domains and recipients precede the DNS black list and CSA
2642 tests, on the grounds that those ones are more expensive.
2644 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2645 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2646 This bug exists in every version of Exim that I still have, right back to
2649 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2650 However, an attempt to turn on case-sensitivity in a regex key by
2651 including (?-i) didn't work because the subject string was already
2652 lowercased, and the effects were non-intuitive. It turns out that a
2653 one-line patch can be used to allow (?-i) to work as expected.
2659 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2660 other effects) broke the use of negated acl sub-conditions.
2662 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2665 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2666 "Deny" causes Exim to reject the incoming connection with a 554 error.
2667 Unfortunately, if there is a major crisis, such as a disk failure,
2668 tcp-wrappers gives "deny", whereas what one would like would be some
2669 kind of temporary error. A kludge has been added to help with this.
2670 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2671 554 error is used if errno is still zero or contains ENOENT (which occurs
2672 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2675 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2677 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2678 errors. Otherwise a message that provokes a temporary error (when other
2679 messages do not) can cause a whole host to time out.
2681 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2682 addresses were routed directly to files or pipes from a redirect router.
2683 File deliveries just didn't batch; pipe deliveries might have suffered
2686 PH/06 A failure to get a lock for a hints database would erroneously always say
2687 "Failed to get write lock", even when it was really a read lock.
2689 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2690 of 0600. This has been changed to use the value of the lockfile_mode
2691 option (which defaults to 0600).
2693 PH/08 Applied small patch from the Sieve maintainer.
2695 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2696 folder from quota calculations, a direct delivery into this folder messed
2697 up the contents of the maildirsize file. This was because the regex was
2698 used only to exclude .Trash (or whatever) when the size of the mailbox
2699 was calculated. There was no check that a delivery was happening into an
2700 excluded directory. This bug has been fixed by ignoring all quota
2701 processing for deliveries into excluded directories.
2703 PH/10 Added the maildirfolder_create_regex option to appendfile.
2709 PH/01 The code for finding all the local interface addresses on a FreeBSD
2710 system running IPv6 was broken. This may well have applied to all BSD
2711 systems, as well as to others that have similar system calls. The broken
2712 code found IPv4 interfaces correctly, but gave incorrect values for the
2713 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2714 that it would not match correctly against @[] and not recognize the IPv6
2717 PH/02 The ipliteral router was not recognizing addresses of the form user@
2718 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2720 PH/03 Added disable_ipv6.
2722 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2723 decoded form, because it is most often used to construct To: headers
2724 lines in autoreplies, and the decoded form may well be syntactically
2725 invalid. However, $reply_address has leading white space removed, and all
2726 newlines turned into spaces so that the autoreply transport does not
2729 PH/05 If group was specified without a user on a router, and no group or user
2730 was specified on a transport, the group from the router was ignored.
2732 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2733 for visible compile-time settings that can be used to change these
2734 numbers, for those that want even more. Backwards compatibility with old
2735 spool files has been maintained. However, going back to a previous Exim
2736 release will lost any variables that are in spool files.
2738 PH/07 Two small changes when running in the test harness: increase delay when
2739 passing a TCP/IP connection to a new process, in case the original
2740 process has to generate a bounce, and remove special handling of
2741 127.0.0.2 (sic), which is no longer necessary.
2743 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2744 be the same on different OS.
2746 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2749 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2750 whether --show-vars was specified or not
2752 JJ/02 exipick: Added support for new ACL variable spool format introduced
2755 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2756 syntactically invalid From: or Reply-to: line, and a filter used this to
2757 generate an autoreply, and therefore failed to obtain an address for the
2758 autoreply, Exim could try to deliver to a non-existent relative file
2759 name, causing unrelated and misleading errors. What now happens is that
2760 it logs this as a hard delivery error, but does not attempt to create a
2763 PH/11 The exinext utility has a -C option for testing purposes, but although
2764 the given file was scanned by exinext itself; it wasn't being passed on
2765 when Exim was called.
2767 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2768 an end-of-file indication when reading a command response.
2770 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2771 compiled. In many other places in Exim, IPv6 addresses are always
2772 recognized, so I have changed this. It also means that IPv4 domain
2773 literals of the form [IPV4:n.n.n.n] are now always recognized.
2775 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2776 used if the router is not running as root, for example, when verifying at
2777 ACL time, or when using -bh. The debugging output from this situation was
2778 non-existent - all you got was a failure to exec. I have made two
2781 (a) Failures to set uid/gid, the current directory, or a process leader
2782 in a subprocess such as that created by queryprogram now generate
2783 suitable debugging ouput when -d is set.
2785 (b) The queryprogram router detects when it is not running as root,
2786 outputs suitable debugging information if -d is set, and then runs
2787 the subprocess without attempting to change uid/gid.
2789 PH/15 Minor change to Makefile for building test_host (undocumented testing
2792 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2793 additional section of a DNS packet that returns MX or SRV records.
2794 Instead, it always explicitly searches for A/AAAA records. This avoids
2795 major problems that occur when a DNS server includes only records of one
2796 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2797 fixed another bug: if SRV records were looked up and the corresponding
2798 address records were *not* found in the additional section, the port
2799 values from the SRV records were lost.
2801 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2802 using the correct key (the original address) when searching the retry
2803 rules in order to find which one to use for generating the retry hint.
2805 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2806 adding the default one. Similarly, if it contains a Reply-To: header, the
2807 errors_reply_to option, if set, is not used.
2809 PH/19 When calculating a retry time, Exim used to measure the "time since
2810 failure" by looking at the "first failed" field in the retry record. Now
2811 it does not use this if it is later than than the arrival time of the
2812 message. Instead it uses the arrival time. This makes for better
2813 behaviour in cases where some deliveries succeed, thus re-setting the
2814 "first failed" field. An example is a quota failure for a huge message
2815 when small messages continue to be delivered. Without this change, the
2816 "time since failure" will always be short, possible causing more frequent
2817 delivery attempts for the huge message than are intended.
2818 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2820 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2821 $1, $2, $3) because the numerical variables can be reset during some
2822 expansion items (e.g. "match"), thereby losing the authentication data.
2824 PH/21 Make -bV show the size of off_t variables so that the test suite can
2825 decide whether to run tests for quotas > 2G.
2827 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2828 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2829 filecount value is greater than 2G or if a quota value is greater than 2G
2830 on a system where the size of off_t is not greater than 4, a panic error
2833 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2834 never match. The debug and -bh output now contains an explicit error
2835 message indicating a malformed IPv4 address or mask.
2837 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2838 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2839 PH/23 above applies.
2841 PH/25 Do not write to syslog when running in the test harness. The only
2842 occasion when this arises is a failure to open the main or panic logs
2843 (for which there is an explicit test).
2845 PH/26 Added the /no_tell option to "control=freeze".
2847 PH/27 If a host name lookup failed very early in a connection, for example, if
2848 the IP address matched host_lookup and the reverse lookup yielded a name
2849 that did not have a forward lookup, an error message of the form "no IP
2850 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2851 could be logged. Now it outputs the IP address instead of "NULL".
2853 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2854 allows the sender and the authenticated sender to be set when
2855 submitting a message from within Exim. Since child_open_exim() is
2856 documented for local_scan(), the new function should be too.
2858 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2859 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2860 results in an empty string is now treated as unset.
2862 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2864 PH/31 Added sender_verify_fail logging option.
2866 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2867 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2868 tidied the source and removed it altogether.
2870 PH/33 When a queue run was abandoned because the load average was too high, a
2871 log line was always written; now it is written only if the queue_run log
2872 selector is set. In addition, the log line for abandonment now contains
2873 information about the queue run such as the pid. This is always present
2874 in "start" and "stop" lines but was omitted from the "abandon" line.
2876 PH/34 Omit spaces between a header name and the colon in the error message that
2877 is given when verify = headers_syntax fails (if there are lots of them,
2878 the message gets confusing).
2880 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2881 names, as there are now some PTR records that contain slashes. This check
2882 is only to protect against broken name servers that fall over on strange
2883 characters, so the fact that it applies to all lookups doesn't matter.
2885 PH/36 Now that the new test suite is complete, we can remove some of the
2886 special code in Exim that was needed for the old test suite. For example,
2887 sorting DNS records because real resolvers return them in an arbitrary
2888 order. The new test suite's fake resolver always returns records in the
2891 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2892 bounces) except when queue_only is set, to avoid logging races between
2893 the different processes.
2895 PH/38 Panic-die if .include specifies a non-absolute path.
2897 PH/39 A tweak to the "H" retry rule from its user.
2899 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2900 a label. They prevented compilation on older perls.
2902 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2903 a warning to be raised on newish perls.
2905 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2906 on queue. Changes to match documented behaviour of showing count of
2907 messages matching specified criteria.
2909 PH/40 Changed the default ident timeout from 30s to 5s.
2911 PH/41 Added support for the use of login_cap features, on those BSD systems
2912 that have them, for controlling the resources used by pipe deliveries.
2914 PH/42 The content-scanning code uses fopen() to create files in which to put
2915 message data. Previously it was not paying any attention to the mode of
2916 the files. Exim runs with umask(0) because the rest of the code creates
2917 files with open(), and sets the required mode explicitly. Thus, these
2918 files were ending up world-writeable. This was not a big issue, because,
2919 being within the spool directory, they were not world-accessible. I have
2920 created a function called modefopen, which takes an additional mode
2921 argument. It sets umask(777), creates the file, chmods it to the required
2922 mode, then resets the umask. All the relevant calls to fopen() in the
2923 content scanning code have been changed to use this function.
2925 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2926 to 24 hours. This avoids potential overflow problems when processing G
2927 and H retry rules. I suspect nobody ever tinkers with this value.
2929 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2931 PH/45 When the plaintext authenticator is running as a client, the server's
2932 challenges are checked to ensure they are valid base64 strings. By
2933 default, the authentication attempt is cancelled if an invalid string is
2934 received. Setting client_ignore_invalid_base64 true ignores these errors.
2935 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2936 they are received. Thus, the responses can be made to depend on the
2937 challenges. If an invalid string is ignored, an empty string is placed in
2940 PH/46 Messages that are created by the autoreply transport now contains a
2941 References: header, in accordance with RFCs 2822 and 3834.
2943 PH/47 Added authenticated_sender_force to the smtp transport.
2945 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2947 PH/49 Installed latest patch from the Sieve maintainer.
2949 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2950 was also set, the appendfile transport was unnecessarily scanning a
2951 directory of message files (e.g. for maildir delivery) to find the count
2952 of files (along with the size), even though it did not need this
2953 information. It now does the scan only if it needs to find either the
2954 size of the count of files.
2956 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2958 PH/52 Two bugs concerned with error handling when the smtp transport is
2961 (i) Exim was not creating retry information for temporary errors given
2962 for individual recipients after the DATA command when the smtp transport
2963 was used in LMTP mode. This meant that they could be retried too
2964 frequently, and not timed out correctly.
2966 (ii) Exim was setting the flag that allows error details to be returned
2967 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2968 recipients that were returned after the DATA command.
2970 PH/53 This is related to PH/52, but is more general: for any failing address,
2971 when detailed error information was permitted to be returned to the
2972 sender, but the error was temporary, then after the final timeout, only
2973 "retry timeout exceeded" was returned. Now it returns the full error as
2974 well as "retry timeout exceeded".
2976 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2977 do this, and (what is worse) MTAs that accept it.
2979 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2980 will now be deprecated.
2982 PH/56 New os.c-cygwin from the Cygwin maintainer.
2984 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2985 formats (previously only available in exim formats via -bpr, -bpru,
2986 and -bpra. Now also available in native and exiqgrep formats)
2988 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2989 with very large, slow to parse queues
2991 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2993 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2995 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2996 responses to authentication challenges, though it was showing the
2997 challenges; (ii) I've removed the CR characters from the debug output for
3000 PH/58 Allow for the insertion of a newline as well as a space when a string
3001 is turned into more than one encoded-word during RFC 2047 encoding. The
3002 Sieve code now uses this.
3004 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3005 data_4xx, lost_connection, tls_required.
3007 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3008 message was being sent as an SMTP response.
3010 PH/61 Add -l and -k options to exicyclog.
3012 PH/62 When verifying, if an address was redirected to one new address, so that
3013 verification continued, and the new address failed or deferred after
3014 having set something in $address_data, the value of $address_data was not
3015 passed back to the ACL. This was different to the case when no
3016 redirection occurred. The value is now passed back in both cases.
3018 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3019 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3020 use login_cap.h, so on its own it isn't the distinguishing feature. The
3021 new name refers directly to the setclassresources() function.
3023 PH/65 Added configuration files for NetBSD3.
3025 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3027 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3028 is preferred over IPv4.
3030 PH/68 The bounce_return_message and bounce_return_body options were not being
3031 honoured for bounces generated during the reception of non-SMTP messages.
3032 In particular, this applied to messages rejected by the ACL. This bug has
3033 been fixed. However, if bounce_return_message is true and bounce_return_
3034 body is false, the headers that are returned for a non-SMTP message
3035 include only those that have been read before the error was detected.
3036 (In the case of an ACL rejection, they have all been read.)
3038 PH/69 The HTML version of the specification is now built in a directory called
3039 spec_html instead of spec.html, because the latter looks like a path with
3040 a MIME-type, and this confuses some software.
3042 PH/70 Catch two compiler warnings in sieve.c.
3044 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3045 function verify_get_ident() calls ip_connect() to connect a socket, but
3046 if the "connect()" function timed out, ip_connect() used to close the
3047 socket. However, verify_get_ident() also closes the socket later, and in
3048 between Exim writes to the log, which may get opened at this point. When
3049 the socket was closed in ip_connect(), the log could get the same file
3050 descriptor number as the socket. This naturally causes chaos. The fix is
3051 not to close the socket in ip_connect(); the socket should be closed by
3052 the function that creates it. There was only one place in the code where
3053 this was missing, in the iplookup router, which I don't think anybody now
3054 uses, but I've fixed it anyway.
3056 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3057 well as to direct DNS lookups. Otherwise the handling of names in host
3058 lists is inconsistent and therefore confusing.
3064 PH/01 Two changes to the default runtime configuration:
3066 (1) Move the checks for relay_from_hosts and authenticated clients from
3067 after to before the (commented out) DNS black list checks.
3069 (2) Add control=submission to the relay_from_hosts and authenticated
3070 clients checks, on the grounds that messages accepted by these
3071 statements are most likely to be submissions.
3073 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3075 (1) Generate an error if the third argument for the ${prvs expansion is
3078 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3081 (3) Reset the variables that are obtained from the first argument of
3082 ${prvscheck and used in the second argument before leaving the code,
3083 because their memory is reclaimed, so using them afterwards may do
3086 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3087 one (it's much easier than Tom thought :-).
3089 (5) Because of (4), we can now allow for the use of $prvscheck_result
3090 inside the third argument.
3092 PH/03 For some reason, the default setting of PATH when running a command from
3093 a pipe transport was just "/usr/bin". I have changed it to
3096 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3097 anything to be listed in the output from -bV.
3099 PH/05 When a filter generated an autoreply, the entire To: header line was
3100 quoted in the delivery log line, like this:
3102 => >A.N.Other <ano@some.domain> <original@ddress> ...
3104 This has been changed so that it extracts the operative address. There
3105 may be more than one such address. If so, they are comma-separated, like
3108 => >ano@some.domain,ona@other.domain <original@ddress> ...
3110 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3111 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3112 not being looked up in the rDNS to get a host name, Exim was showing the
3113 IP address twice in Received: lines, even though the IP addresses were
3114 identical. For example:
3116 Received: from [1.2.3.4] (helo=[1.2.3.4])
3118 However, if the real host name was known, it was omitting the HELO data
3119 if it matched the actual IP address. This has been tidied up so that it
3120 doesn't show the same IP address twice.
3122 PH/07 When both +timestamp and +memory debugging was on, the value given by
3123 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3124 called by the debug printing, thereby overwriting the timestamp buffer.
3125 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3127 PH/08 When the original message was included in an autoreply transport, it
3128 always said "this is a copy of the message, including all the headers",
3129 even if body_only or headers_only was set. It now gives an appropriate
3132 PH/09 Applied a patch from the Sieve maintainer which:
3134 o fixes some comments
3135 o adds the (disabled) notify extension core
3136 o adds some debug output for the result of if/elsif tests
3137 o points to the current vacation draft in the documentation
3138 and documents the missing references header update
3142 o fixes a bug in processing the envelope test (when testing
3143 multiple envelope elements, the last element determinted the
3146 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3147 Electronic Mail") by including:
3149 Auto-submitted: auto-generated
3151 in the messages that it generates (bounce messages and others, such as
3152 warnings). In the case of bounce messages for non-SMTP mesages, there was
3153 also a typo: it was using "Auto_submitted" (underscore instead of
3154 hyphen). Since every message generated by Exim is necessarily in response
3155 to another message, thes have all been changed to:
3157 Auto-Submitted: auto-replied
3159 in accordance with these statements in the RFC:
3161 The auto-replied keyword:
3163 - SHOULD be used on messages sent in direct response to another
3164 message by an automatic process,
3166 - MUST NOT be used on manually-generated messages,
3168 - MAY be used on Delivery Status Notifications (DSNs) and Message
3169 Disposition Notifications (MDNs),
3171 - MUST NOT be used on messages generated by automatic or periodic
3172 processes, except for messages which are automatic responses to
3175 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3176 to the default Received: header definition.
3178 PH/12 Added log selector acl_warn_skipped (default on).
3180 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3181 variables because (a) they are in the wrong storage pool and (b) even if
3182 they were copied, it wouldn't work properly because of the caching.
3184 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3185 checking when decoding. Apparently there are clients that generate
3186 overlong encoded strings. Why am I not surprised?
3188 PH/15 If the first argument of "${if match_address" was not empty, but did not
3189 contain an "@" character, Exim crashed. Now it writes a panic log message
3190 and treats the condition as false.
3192 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3194 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3195 extension "envelope-auth". The code is finished and in agreement with
3196 other implementations, but there is no documentation so far and in fact,
3197 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3198 not changing the active code.
3200 Print executed "if" and "elsif" statements when debugging is used. This
3201 helps a great deal to understand what a filter does.
3203 Document more things not specified clearly in RFC3028. I had all this
3204 sorted out, when out of a sudden new issues came to my mind. Oops."
3206 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3209 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3210 canonical form (as documented). However, after a host name lookup from
3211 the IP address, check_host() was doing a simple string comparison with
3212 addresses acquired from the DNS when checking that the found name did
3213 have the original IP as one of its addresses. Since any found IPv6
3214 addresses are likely to be in abbreviated form, the comparison could
3215 fail. Luckily, there already exists a function for doing the comparison
3216 by converting both addresses to binary, so now that is used instead of
3217 the text comparison.
3219 PH/20 There was another similar case to PH/19, when a complete host name was
3220 given in a host list; looking up its IP address could give an abbreviated
3221 form, whereas the current host's name might or might not be abbreviated.
3222 The same fix has been applied.
3228 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3229 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3232 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3233 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3235 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3237 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3238 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3239 Also, for "Auto-Submitted", treat anything other than "no" as
3240 non-personal, in accordance with RFC 3834. (Previously it treated
3241 anything starting "auto-" as non-personal.)
3243 TF/01 The control=submission/name=... option had a problem with syntax
3244 errors if the name included a slash character. The /name= option
3245 now slurps the rest of the string, so it can include any characters
3246 but it must come last in the list of options (after /sender_retain
3249 PH/05 Some modifications to the interface to the fake nameserver for the new
3257 TK/01 Added the "success_on_redirect" address verification option. See
3258 NewStuff for rationale and an example.
3260 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3262 PH/02 Patch to exigrep to allow it to work on syslog lines.
3264 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3265 fread() to skip over the body file's header line, because in Cygwin the
3266 header line is locked and is inaccessible.
3268 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3269 co-exist for some time) to make it clear that it is the Exim ID that is
3270 referenced, not the Message-ID: header line.
3272 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3273 string_format() function, because snprintf() does not exist on all
3276 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3277 ${stat: expansion item.
3279 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3280 protocol synchronization error", to keep the pedants happy.
3282 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3283 well as for IRIX systems, when gcc is being used. See the host.c source
3286 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3288 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3291 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3292 transport and to the smtp transport in LMTP mode.
3294 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3296 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3297 run a filter in a subprocess. This could lead to confusion in subsequent
3298 lookups in the parent process. There should also be a search_tidyup() at
3299 the end of the subprocess.
3301 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3302 only if the host matched helo_try_verify_hosts, which caused the
3303 verification to occur when the EHLO/HELO command was issued. The ACL just
3304 tested the remembered result. Now, if a previous verification attempt has
3305 not happened, "verify = helo" does it there and then.
3307 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3309 TK/03 Fix log output including CR from clamd.
3311 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3312 exist provoked a memory error which could cause a segfault.
3314 PH/15 Installed PCRE 6.2
3316 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3318 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3319 of the problem. Specifically, suggested +O2 rather than +O1 for the
3322 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3324 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3325 record, the port was not correctly logged when the outgoing_port log
3326 selector was set (it logged the transort's default port).
3328 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3329 fallback_hosts, and "hosts" in the smtp transport.
3331 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3332 host errors such as "Connection refused".
3334 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3335 authentication with radiusclient 0.4.9:
3337 - Error returned from rc_read_config was caught wrongly
3338 - Username/password not passed on to radius server due to wrong length.
3340 The presumption is that some radiusclient API changes for 4.51/PH/17
3341 were not taken care of correctly. The code is still untested by me (my
3342 Linux distribution still has 0.3.2 of radiusclient), but it was
3343 contributed by a Radius user.
3345 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3346 expanding the "port" option of the smtp transport.
3348 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3349 while reading a MIME header. Thanks to Tom Hughes for a patch.
3351 PH/24 Include config.h inside local_scan.h so that configuration settings are
3354 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3355 that in an ACL for RCPT (for example), you can examine exactly what was
3358 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3359 commands, but it was not correctly comparing the address with the actual
3360 client host address. Thus, it would show the EHLO address in Received:
3361 header lines when this was not necessary.
3363 PH/27 Added the % operator to ${eval:}.
3365 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3366 it should be ignoring failures (because with -C, for example, it has lost
3367 privilege). It wasn't ignoring creation failures other than "already
3370 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3373 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3374 because an input line was too long, either on its own, or by virtue of
3375 too many continuations, the temporary file was not being removed, and the
3376 return code was incorrect.
3378 PH/31 Missing "BOOL" in function definition in filtertest.c.
3380 PH/32 Applied Sieve patches from the maintainer.
3382 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3384 PH/33 Added "verify = not_blind".
3386 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3387 Local/Makefile (with some defaults set). These are used in built scripts
3388 such as exicyclog, but they have never been used in the exim_install
3389 script (though there are many overriding facilities there). I have
3390 arranged that the exim_install script now takes note of these two
3393 PH/35 Installed configuration files for Dragonfly.
3395 PH/36 When a locally submitted message by a trusted user did not contain a
3396 From: header, and the sender address was obtained from -f or from an SMTP
3397 MAIL command, and the trusted user did not use -F to supply a sender
3398 name, $originator_name was incorrectly used when constructing a From:
3399 header. Furthermore, $originator_name was used for submission mode
3400 messages from external hosts without From: headers in a similar way,
3401 which is clearly wrong.
3403 PH/37 Added control=suppress_local_fixups.
3405 PH/38 When log_selector = +received_sender was set, and the addition of the
3406 sender made the log line's construction buffer exactly full, or one byte
3407 less than full, an overflow happened when the terminating "\n" was
3410 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3411 when the result of a list match is failure because a DNS lookup failed.
3413 PH/40 RM_COMMAND is now used in the building process.
3415 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3416 the "build-* directories that it finds.
3418 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3419 address in a domain literal was a prefix of an interface address.
3421 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3422 when verifying a sender address, unless rewrite_headers is false.
3424 PH/44 Wrote a long comment about why errors_to addresses are verified as
3425 recipients, not senders.
3427 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3428 the ratelimit ACL was added.
3430 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3432 PH/46 Added extra information about PostgreSQL errors to the error string.
3434 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3435 avoiding the need to install special zones in a real server. This is
3436 backwards compatible; if it can't find the fake resolver, it drops back.
3437 Thus, both old and new test suites can be run.
3439 TF/02 Added util/ratelimit.pl
3441 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3442 clock is set back in time.
3444 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3445 Candler <B.Candler@pobox.com>.
3447 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3448 OK for addresses that are the result of redirections.
3450 PH/48 A number of further additions for the benefit of the new test suite,
3451 including a fake gethostbyname() that interfaces to the fake DNS resolver
3454 TF/06 The fix for widen_domains has also been applied to qualify_single and
3455 search_parents which are the other dnslookup options that can cause
3458 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3461 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3463 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3464 service" if it encountered an MX record with an empty target hostname.
3465 The message is now "an MX or SRV record indicated no SMTP service".
3467 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3468 if the DNS of the sending site is misconfigured. This is quite a
3469 common situation. This change restores the behaviour of treating a
3470 helo verification defer as a failure.
3472 PH/51 If self=fail was set on a router, the bounce message did not include the
3473 actual error message.
3479 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3481 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3482 command itself ended in error, the underlying message about the transport
3483 filter timeout was being overwritten with the pipe command error. Now the
3484 underlying error message should be appended to the second error message.
3486 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3488 PH/02 Reduce the amount of output that "make" produces by default. Full output
3489 can still be requested.
3491 PH/03 The warning log line about a condition test deferring for a "warn" verb
3492 was being output only once per connection, rather than after each
3493 occurrence (because it was using the same function as for successful
3494 "warn" verbs). This seems wrong, so I have changed it.
3496 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3497 it should not have, which might have caused a crash in the right
3498 circumstances, but probably never did.
3500 PH/04 Installed a modified version of Tony Finch's patch to make submission
3501 mode fix the return path as well as the Sender: header line, and to
3502 add a /name= option so that you can make the user's friendly name appear
3505 TF/03 Added the control = fakedefer ACL modifier.
3507 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3508 Mark Lowes for thorough testing.
3510 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3512 TK/03 Merged latest SRS patch from Miles Wilton.
3514 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3515 with the definition in sysexits.h (which is #included earlier).
3516 Fortunately, Exim does not actually use EX_OK. The code used to try to
3517 preserve the sysexits.h value, by assumimg that macro definitions were
3518 scanned for macro replacements. I have been disabused of this notion,
3519 so now the code just undefines EX_OK before #including unistd.h.
3521 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3522 in the smtp transport. When a block could not be written in a single
3523 write() function, the timeout was being re-applied to each part-write.
3524 This seems wrong - if the receiver was accepting one byte at a time it
3525 would take for ever. The timeout is now adjusted when this happens. It
3526 doesn't have to be particularly precise.
3528 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3529 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3531 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3532 without a domain if the check to <postmaster@domain> fails.
3534 SC/01 Eximstats: added -xls and the ability to specify output files
3535 (patch written by Frank Heydlauf).
3537 SC/02 Eximstats: use FileHandles for outputing results.
3539 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3541 SC/04 Eximstats: fixed display of large numbers with -nvr option
3543 SC/05 Eximstats: fixed merging of reports with empty tables.
3545 SC/06 Eximstats: added the -include_original_destination flag
3547 SC/07 Eximstats: removed tabs and trailing whitespace.
3549 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3551 TK/06 MBOX spool code: Add real "From " MBOX separator line
3552 so the .eml file is really in mbox format (even though
3553 most programs do not really care). Patch from Alex Miller.
3555 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3556 The latter is generated from $received_to and is only set if the
3557 message has one envelope recipient. SA can use these headers,
3558 obviously out-of-the-box. Patch from Alex Miller.
3560 PH/08 The ${def test on a variable was returning false if the variable's
3561 value was "0", contrary to what the specification has always said!
3562 The result should be true unless the variable is empty.
3564 PH/09 The syntax error of a character other than { following "${if
3565 def:variable_name" (after optional whitespace) was not being diagnosed.
3566 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3567 accidental colon was present, for example, could give incorrect results.
3569 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3570 result is used (not including appendfile, where other changes are about
3573 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3574 This involved changing a lot of size variables from int to off_t. It
3575 should work with maildirs and everything.
3577 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3578 spamd dying while we are connected to it.
3580 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3583 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3584 function for BDB 4.3.
3586 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3588 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3589 the sources for PCRE have been re-arranged and more files are now
3592 PH/15 The code I had for printing potentially long long variables in PH/11
3593 above was not the best (it lost precision). The length of off_t variables
3594 is now inspected at build time, and an appropriate printing format (%ld
3595 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3596 to be "long long int" or "long int". This is needed for the internal
3597 formatting function string_vformat().
3599 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3600 the configuration file to be ":syslog", then the script "guesses" where
3601 the logs files are, rather than using the compiled in default. In our
3602 case the guess is not the same as the compiled default, so the script
3603 suddenly stopped working when I started to use syslog. The patch checks
3604 to see if log_file_path is "". If so, it attempts to read it from exim
3605 with no configuration file to get the compiled in version, before it
3606 falls back to the previous guessing code."
3608 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3609 implementing BATV in an Exim configuration. See NewStuff for the gory
3612 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3613 Makefile that are specific to HP-UX.
3615 PH/18 If the "use_postmaster" option was set for a recipient callout together
3616 with the "random" option, the postmaster address was used as the MAIL
3617 FROM address for the random test, but not for the subsequent recipient
3618 test. It is now used for both.
3620 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3621 patch removes a few documentation additions to RFC 3028, because the
3622 latest draft now contains them. It adds the new en;ascii-case comparator
3623 and a new error check for 8bit text in MIME parts. Comparator and
3624 require names are now matched exactly. I enabled the subaddress
3625 extension, but it is not well tested yet (read: it works for me)."
3627 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3628 rework some of the code of TK/09 above to avoid the hardwired use of
3629 "%lld" and "long long". Replaced the call to snprintf() with a call to
3632 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3633 records point to non-existent hosts", "retry timeout exceeded", and
3634 "retry time not reached for any host after a long failure period".
3636 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3637 experimental DomainKeys support:
3639 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3640 (2) On an error such as an illegally used "control", the wrong name for
3641 the control was given.
3643 These problems did NOT occur unless DomainKeys support was compiled.
3645 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3647 PH/24 Added ${if match_ip condition.
3649 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3650 databases so that it will be absolutely obvious if a crash occurs in the
3651 DB library. This is a regular occurrence (often caused by mis-matched
3654 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3655 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3656 picked up on a user's system that detects such things. There doesn't seem
3657 to be a gcc warning option for this - only an attribute that has to be
3658 put on the function's prototype. It seems that in Fedora Core 4 they have
3659 set this on a number of new functions. No doubt there will be more in due
3662 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3663 specify a transport. However, if an address that was verified by such a
3664 router was the subject of a callout, Exim crashed because it tried to
3665 read the rcpt_include_affixes from the non-existent transport. Now it
3666 just assumes that the setting of that option is false. This bug was
3667 introduced by 4.51/PH/31.
3669 PH/28 Changed -d+all to exclude +memory, because that information is very
3670 rarely of interest, but it makes the output a lot bigger. People tend to
3671 do -d+all out of habit.
3673 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3674 code in os-type was giving problems when libc.so lives in lib64, like on
3677 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3678 aren't the modern standard, and it seems that some systems' include files
3679 don't always have them. Exim was already checking for some of the newer
3680 ones like T_AAAA, and defining it itself. I've added checks for all the
3681 record types that Exim uses.
3683 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3684 not automatically generating a new one, as it is supposed to. This
3685 prevented TLS from working. If the file did exist, but contained invalid
3686 data, a new version was generated, as expected. It was only the case of a
3687 non-existent file that was broken.
3689 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3690 with a change in libdomainkeys > 0.64.
3692 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3693 from DNS. If the selector record carries the flag, it now has
3694 precedence over the domain-wide flag.
3696 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3698 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3699 the use of a port name that isn't defined in /etc/services) occurred, the
3700 message was deferred as in a normal delivery, and thus remained on the
3701 spool, instead of being failed because of the mua_wrapper setting. This
3702 is now fixed, and I tidied up some of the mua_wrapper messages at the
3705 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3706 the messages in a hash of arrays rather than using individual hashes.
3707 This is a bit cleaner and results in dramatic memory savings, albeit
3708 at a slight CPU cost.
3710 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3711 as requested by Marc Sherman.
3713 SC/10 Eximstats: added histograms for user specified patterns as requested
3716 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3718 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3719 fopen() in the content-scanning modules that did not already have it.
3725 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3726 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3728 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3730 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3732 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3733 to test Sieve filters that use "vacation".
3735 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3736 that changes the way the GnuTLS parameters are stored in the cache file.
3737 The new format can be generated externally. For backward compatibility,
3738 if the data in the cache doesn't make sense, Exim assumes it has read an
3739 old-format file, and it generates new data and writes a new file. This
3740 means that you can't go back to an older release without removing the
3743 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3744 work if there are any delivery delays because "one_time" forces the
3745 parent to be marked "delivered", so its unseen clone is never tried
3746 again. For this reason, Exim now forbids the simultaneous setting of
3749 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3750 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3751 entry if you want to know the details. The fix, however, neglected to
3752 consider the case when local delivery batching is involved. The test for
3753 "previously delivered" was not happening when checking to see if an
3754 address could be batched with a previous (undelivered) one; under
3755 certain circumstances this could lead to multiple deliveries to the same
3758 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3759 in its include files, and this causes problems building Exim.
3761 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3762 header_syntax) but Exim was just ignoring anything given after a slash.
3763 In particular, this caused confusion with an attempt to use "verify =
3764 reverse_host_lookup/defer_ok". An error is now given when options are
3765 supplied for verify items that do not have them. (Maybe reverse_host_
3766 lookup should have a defer_ok option, but that's a different point.)
3768 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3769 defined by RFC 821) to 2048, because there were problems with some AUTH
3770 commands, and RFC 1869 says the size should be increased for extended
3771 SMTP commands that take arguments.
3773 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3776 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3777 "unknown" error; now it says that the functionality isn't in the binary.
3779 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3780 an address' error message when a string expansion fails (syntax or
3781 whatever). Otherwise the password may appear in the log. Following change
3782 PH/42 below, there is no longer a chance of it appearing in a bounce
3785 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3787 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3788 of its addresses was ever tried. (Bugzilla bug #2.)
3790 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3791 the result incorrectly in the debug output. (It correctly added a newline
3792 to what was transported.)
3794 TF/01 Added $received_time.
3796 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3797 commented out examples of how to interface to a virus scanner and to
3798 SpamAssassin. Also added commented examples of av_scanner and
3799 spamd_address settings.
3801 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3802 and controls are allowed in which ACLs. There were a couple of minor
3803 errors. Some of the entries in the conditions table (which is a table of
3804 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3805 negation of where the condition IS allowed.
3807 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3809 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3810 header file does not have a version number, so I've had to invent a new
3811 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3812 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3813 radiusclient), but it was contributed by a Radius user.
3815 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3816 files or directories, for OpenSSL.
3818 PH/19 When an Exim process that is running as root has to create an Exim log
3819 file, it does so in a subprocess that runs as exim:exim so as to get the
3820 ownership right at creation (otherwise, other Exim processes might see
3821 the file with the wrong ownership). There was no test for failure of this
3822 fork() call, which would lead to the process getting stuck as it waited
3823 for a non-existent subprocess. Forks do occasionally fail when resources
3824 run out. I reviewed all the other calls to fork(); they all seem to check
3827 PH/20 When checking for unexpected SMTP input at connect time (before writing
3828 the banner), Exim was not dealing correctly with a non-positive return
3829 from the read() function. If the client had disconnected by this time,
3830 the result was a log entry for a synchronization error with an empty
3831 string after "input=" when read() returned zero. If read() returned -1
3832 (an event I could not check), uninitialized data bytes were printed.
3833 There were reports of junk text (parts of files, etc) appearing after
3836 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3838 PH/22 Added support for macro redefinition, and (re)definition in between
3839 driver and ACL definitions.
3841 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3842 forgetting to use the resulting value; it was using the unexpanded value.
3844 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3845 hadn't been configured. The fix is from Juergen Kreileder, who
3846 understands it better than I do:
3848 "Here's what I see happening with three configured cyrus_sasl
3849 authenticators configured (plain, login, cram-md5):
3851 On startup auth_cyrus_sasl_init() gets called for each of these.
3852 This means three calls to sasl_listmech() without a specified mech_list.
3853 => SASL tests which mechs of all available mechs actually work
3854 => three warnings about OTP not working
3855 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3857 With the patch, sasl_listmech() also gets called three times. But now
3858 SASL's mech_list option is set to the server_mech specified in the the
3859 authenticator. Or in other words, the answer from sasl_listmech()
3860 gets limited to just the mech you're testing for (which is different
3862 => the return list contains just 'plain' or 'login', 'cram-md5' or
3863 nothing depending on the value of ob->server_mech.
3865 I've just tested the patch: Authentication still works fine,
3866 unavailable mechs specified in the exim configuration are still
3867 caught, and the auth.log warnings about OTP are gone."
3869 PH/25 When debugging is enabled, the contents of the command line are added
3870 to the debugging output, even when log_selector=+arguments is not
3873 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3874 answer is "GNU", and only if the return is "GNU/something" is the answer
3877 PH/27 $acl_verify_message is now set immediately after the failure of a
3878 verification in an ACL, and so is available in subsequent modifiers. In
3879 particular, the message can be preserved by coding like this:
3881 warn !verify = sender
3882 set acl_m0 = $acl_verify_message
3884 Previously, $acl_verify_message was set only while expanding "message"
3885 and "log_message" when a very denied access.
3887 PH/28 Modified OS/os.c-Linux with
3889 -#ifndef OS_LOAD_AVERAGE
3890 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3892 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3893 nomenclature these days.)
3895 PH/29 Installed patch from the Sieve maintainer that adds the options
3896 sieve_useraddress and sieve_subaddress to the redirect router.
3898 PH/30 In these circumstances:
3899 . Two addresses routed to the same list of hosts;
3900 . First host does not offer TLS;
3901 . First host accepts first address;
3902 . First host gives temporary error to second address;
3903 . Second host offers TLS and a TLS session is established;
3904 . Second host accepts second address.
3905 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3906 and peerdn, if requested) that were in fact used only for the second
3909 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3910 attention to any local part prefix or suffix that was matched by the
3911 router that accepted the address. It now behaves in the same way as it
3912 does for delivery: the affixes are removed from the local part unless
3913 rcpt_include_affixes is set on the transport.
3915 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3916 timeout during the DATA phase of an incoming message.
3918 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3919 applied a patch sanctioned by the Sieve maintainer.
3921 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3922 the uid or gid is negative. A case of a negative gid caused this to be
3923 noticed. The fix allows for either to be negative.
3925 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3926 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3929 PH/36 The change PH/12 above was broken. Fixed it.
3931 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3932 the grounds that routing the same address twice would always produce the
3933 same answer. This might have been true once, but it is certainly no
3934 longer true now. Routing a child address may depend on the previous
3935 routing that produced that child. Some complicated redirection strategies
3936 went wrong when messages had multiple recipients, and made Exim's
3937 behaviour dependent on the order in which the addresses were given.
3939 I have moved the duplicate checking until after the routing is complete.
3940 Exim scans the addresses that are assigned to local and remote
3941 transports, and removes any duplicates. This means that more work will be
3942 done, as duplicates will always all be routed, but duplicates are
3943 presumably rare, so I don't expect this is of any significance.
3945 For deliveries to pipes, files, and autoreplies, the duplicate checking
3946 still happens during the routing process, since they are not going to be
3949 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3950 It corrects a timeout issue with spamd. This is Ian's comment: "The
3951 background is that sometimes spamd either never reads data from a
3952 connection it has accepted, or it never writes response data. The exiscan
3953 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3954 blindly assumes that writes won't block so it may never time out."
3956 PH/39 Allow G after quota size as well as K and M.
3958 PH/40 The value set for $authenticated_id in an authenticator may not contain
3959 binary zeroes or newlines because the value is written to log lines and
3960 to spool files. There was no check on this. Now the value is run through
3961 the string_printing() function so that such characters are converted to
3962 printable escape sequences.
3964 PH/41 $message_linecount is a new variable that contains the total number of
3965 lines in the message. Compare $body_linecount, which is the count for the
3968 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3969 bounce and delay warning messages, except in certain special cases, which
3972 (a) An SMTP error message from a remote host;
3973 (b) A message specified in a :fail: redirection;
3974 (c) A message specified in a "fail" command in a system filter;
3975 (d) A message specified in a FAIL return from the queryprogram router;
3976 (e) A message specified by the cannot_route_message router option.
3978 In these cases only, Exim does include the error details in bounce and
3979 warning messages. There are also a few cases where bland messages such
3980 as "unrouteable address" or "local delivery error" are given.
3982 PH/43 $value is now also set for the "else" part of a ${run expansion.
3984 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3985 being worked on, but at least Exim now implements the latest version to
3988 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3989 process to complete was treated as a delivery failure, a timeout while
3990 writing the message to the pipe was logged, but erroneously treated as a
3991 successful delivery. Such timeouts include transport filter timeouts. For
3992 consistency with the overall process timeout, these timeouts are now
3993 treated as errors, giving rise to delivery failures by default. However,
3994 there is now a new Boolean option for the pipe transport called
3995 timeout_defer, which, if set TRUE, converts the failures into defers for
3996 both kinds of timeout. A transport filter timeout is now identified in
3999 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4000 systems where "make" and "gmake" are different, calling "gmake" at top
4001 level broke things. I've arranged for the value of $(MAKE) to be passed
4002 from the Makefile to this script so that it can call the same version of
4006 A note about Exim versions 4.44 and 4.50
4007 ----------------------------------------
4009 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4010 changes of various kinds. As a consequence, a big documentation update was
4011 needed. This delayed the release for rather longer than seemed good, especially
4012 in the light of a couple of (minor) security issues. Therefore, the changes
4013 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4014 release. So 4.44 and 4.50 are in effect two different branches that both start
4017 I have left the 4.50 change log unchanged; it contains all the changes since
4018 4.43. The change log for 4.44 is below; many of its items are identical to
4019 those for 4.50. This seems to be the most sensible way to preserve the
4020 historical information.
4026 1. Minor wording change to the doc/README.SIEVE file.
4028 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4029 computation of the current number of files was incorrect.
4031 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4032 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4035 4. Give more explanation in the error message when the command for a transport
4036 filter fails to execute.
4038 5. There are several places where Exim runs a non-Exim command in a
4039 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4040 was being done only for the command run by the queryprogram router. It is
4041 now done for all such subprocesses. The other cases are: ${run, transport
4042 filters, and the commands run by the lmtp and pipe transports.
4044 6. Added CONFIGURE_GROUP build-time option.
4046 7. Some older OS have a limit of 256 on the maximum number of file
4047 descriptors. Exim was using setrlimit() to set 1000 as a large value
4048 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4049 systems. I've change it so that if it can't get 1000, it tries for 256.
4051 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4052 was an oversight, and furthermore, ever since the addition of extra
4053 controls (e.g. 4.43/32), the checks on when to allow different forms of
4054 "control" were broken. There should now be diagnostics for all cases when a
4055 control that does not make sense is encountered.
4057 9. Added the /retain_sender option to "control=submission".
4059 10. $recipients is now available in the predata ACL (oversight).
4061 11. Tidy the search cache before the fork to do a delivery from a message
4062 received from the command line. Otherwise the child will trigger a lookup
4063 failure and thereby defer the delivery if it tries to use (for example) a
4064 cached ldap connection that the parent has called unbind on.
4066 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4067 of $address_data from the recipient verification was clobbered by the
4068 sender verification.
4070 13. The value of address_data from a sender verification is now available in
4071 $sender_address_data in subsequent conditions in the ACL statement.
4073 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4075 15. Added a new option "connect=<time>" to callout options, to set a different
4078 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4079 was its contents. (It was OK if the option was not defined at all.)
4081 17. A "Completed" log line is now written for messages that are removed from
4082 the spool by the -Mrm option.
4084 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4085 information about exactly what failed.
4087 19. Added -dd to debug only the daemon process.
4089 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4090 handles timeouts, both on the server side and network timeouts. Renamed the
4091 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4093 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4094 It is now set to "smtps".
4096 22. $host_address is now set to the target address during the checking of
4097 ignore_target_hosts.
4099 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4100 being passed; this would have caused $sender_host_name to have been used if
4101 matching the list had actually called for a host name (not very likely,
4102 since this list is usually IP addresses). A host name is now passed as
4105 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4106 code that specifies a non-restarting handler (typically sigaction() in
4107 modern systems) in an attempt to fix a rare and obscure crash bug.
4109 25. Narrowed the window for a race in the daemon that could cause it to ignore
4110 SIGCHLD signals. This is not a major problem, because they are used only to
4111 wake it up if nothing else does.
4113 26. A malformed maildirsize file could cause Exim to calculate negative values
4114 for the mailbox size or file count. Odd effects could occur as a result.
4115 The maildirsize information is now recalculated if the size or filecount
4118 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4119 support for a long time. Removed HAVE_SYS_VFS_H.
4121 28. Installed the latest version of exipick from John Jetmore.
4123 29. In an address list, if the pattern was not a regular expression, an empty
4124 subject address (from a bounce message) matched only if the pattern was an
4125 empty string. Non-empty patterns were not even tested. This was the wrong
4126 because it is perfectly reasonable to use an empty address as part of a
4127 database query. An empty address is now tested by patterns that are
4128 lookups. However, all the other forms of pattern expect the subject to
4129 contain a local part and a domain, and therefore, for them, an empty
4130 address still always fails if the pattern is not itself empty.
4132 30. Exim went into a mad DNS loop when attempting to do a callout where the
4133 host was specified on an smtp transport, and looking it up yielded more
4134 than one IP address.
4136 31. Re-factored the code for checking spool and log partition space into a
4137 function that finds that data and another that does the check. The former
4138 is then used to implement four new variables: $spool_space, $log_space,
4139 $spool_inodes, and $log_inodes.
4141 32. The RFC2047 encoding function was originally intended for short strings
4142 such as real names; it was not keeping to the 75-character limit for
4143 encoded words that the RFC imposes. It now respects the limit, and
4144 generates multiple encoded words if necessary. To be on the safe side, I
4145 have increased the buffer size for the ${rfc2047: expansion operator from
4148 33. It is now permitted to omit both strings after an "if" condition; if the
4149 condition is true, the result is "true". As before, when the second string
4150 is omitted, a false condition yields an empty string. This makes it less
4151 cumbersome to write custom ACL and router conditions.
4153 34. Failure to deliver a bounce message always caused it to be frozen, even if
4154 there was an errors_to setting on the router. The errors_to setting is now
4157 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4158 canonical form (fully expanded) before being placed in
4159 $sender_host_address.
4161 36. The table in the code that translates DNS record types into text (T_A to
4162 "A" for instance) was missing entries for NS and CNAME. It is just possible
4163 that this could have caused confusion if both these types were looked up
4164 for the same domain, because the text type is used as part of Exim's
4165 per-process caching. But the chance of anyone hitting this buglet seems
4168 37. The dnsdb lookup has been extended in a number of ways.
4170 (1) There is a new type, "zns", which walks up the domain tree until it
4171 finds some nameserver records. It should be used with care.
4173 (2) There is a new type, "mxh", which is like "mx" except that it returns
4174 just the host names, not the priorities.
4176 (3) It is now possible to give a list of domains (or IP addresses) to be
4177 looked up. The behaviour when one of the lookups defers can be
4178 controlled by a keyword.
4180 (4) It is now possible to specify the separator character for use when
4181 multiple records are returned.
4183 38. The dnslists ACL condition has been extended: it is now possible to supply
4184 a list of IP addresses and/or domains to be looked up in a particular DNS
4187 39. Added log_selector=+queue_time_overall.
4189 40. When running the queue in the test harness, wait just a tad after forking a
4190 delivery process, to get repeatability of debugging output.
4192 41. Include certificate and key file names in error message when GnuTLS fails
4193 to set them up, because the GnuTLS error message doesn't include the name
4194 of the failing file when there is a problem reading it.
4196 42. Allow both -bf and -bF in the same test run.
4198 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4200 44. The "Exiscan patch" is now merged into the mainline Exim source.
4202 45. Sometimes the final signoff response after QUIT could fail to get
4203 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4204 < 0 before doing a fflush(). This bug looks as though it goes back to the
4205 introduction of TLS in release 3.20, but "sometimes" must have been rare
4206 because the tests only now provoked it.
4208 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4209 (this can affect the format of dates).
4211 47. exim_tidydb, when checking for the continued existence of a message for
4212 which it has found a message-specific retry record, was not finding
4213 messages that were in split spool directories. Consequently, it was
4214 deleting retry records that should have stayed in existence.
4216 48. Steve fixed some bugs in eximstats.
4218 49. The SPA authentication driver was not abandoning authentication and moving
4219 on to the next authenticator when an expansion was forced to fail,
4220 contradicting the general specification for all authenticators. Instead it
4221 was generating a temporary error. It now behaves as specified.
4223 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4224 (the order specifies the preference for clients). The order is now AES256,
4225 AES128, 3DES, ARCFOUR128.
4227 51. Small patch to Sieve code - explicitly set From: when generating an
4230 52. Exim crashed if a remote delivery caused a very long error message to be
4231 recorded - for instance if somebody sent an entire SpamAssassin report back
4232 as a large number of 550 error lines. This bug was coincidentally fixed by
4233 increasing the size of one of Exim's internal buffers (big_buffer) that
4234 happened as part of the Exiscan merge. However, to be on the safe side, I
4235 have made the code more robust (and fixed the comments that describe what
4238 53. Now that there can be additional text after "Completed" in log lines (if
4239 the queue_time_overall log selector is set), a one-byte patch to exigrep
4240 was needed to allow it to recognize "Completed" as not the last thing in
4243 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4244 patch that reportedly fixes this has been added. I am not expert enough to
4245 create a test for it. This is what the patch creator wrote:
4247 "I found a little strange behaviour of ldap code when working with
4248 Windows 2003 AD Domain, where users was placed in more than one
4249 Organization Units. When I tried to give exim partial DN, the exit code
4250 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4251 But simultaneously result of request was absolutely normal ldap result,
4252 so I produce this patch..."
4254 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4255 REFERENCE, so I have modified the code to exclude the patch when that macro
4258 55. Some experimental protocols are using DNS PTR records for new purposes. The
4259 keys for these records are domain names, not reversed IP addresses. The
4260 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4261 leaves it alone. Component reversal etc. now happens only for IP addresses.
4264 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4266 57. Double the size of the debug message buffer (to 2048) so that more of very
4267 long debug lines gets shown.
4269 58. The exicyclog utility now does better if the number of log files to keep
4270 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4272 59. Two changes related to the smtp_active_hostname option:
4274 (1) $smtp_active_hostname is now available as a variable.
4275 (2) The default for smtp_banner uses $smtp_active_hostname instead
4276 of $primary_hostname.
4278 60. The host_aton() function is supposed to be passed a string that is known
4279 to be a valid IP address. However, in the case of IPv6 addresses, it was
4280 not checking this. This is a hostage to fortune. Exim now panics and dies
4281 if the condition is not met. A case was found where this could be provoked
4282 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4283 components; fortuitously, this particular loophole had already been fixed
4284 by change 4.50/55 above.
4286 If there are any other similar loopholes, the new check in host_aton()
4287 itself should stop them being exploited. The report I received stated that
4288 data on the command line could provoke the exploit when Exim was running as
4289 exim, but did not say which command line option was involved. All I could
4290 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4291 running as the user.
4294 61. There was a buffer overflow vulnerability in the SPA authentication code
4295 (which came originally from the Samba project). I have added a test to the
4296 spa_base64_to_bits() function which I hope fixes it.
4299 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4300 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4302 63. The daemon start-up calls getloadavg() while still root for those OS that
4303 need the first call to be done as root, but it missed one case: when
4304 deliver_queue_load_max is set with deliver_drop_privilege. This is
4305 necessary for the benefit of the queue runner, because there is no re-exec
4306 when deliver_drop_privilege is set.
4308 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4309 This has been fixed.
4311 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4312 was in use, was not putting the data itself into the right store pool;
4313 consequently, it could be overwritten for a subsequent message in the same
4314 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4317 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4319 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4320 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4321 the function were treating the return as a boolean value, which happened to
4322 work because 0=false and not-0=true, but is not correct code.
4324 68. The host_aton() function was not handling scoped IPv6 addresses (those
4325 with, for example, "%eth0" on the end) correctly.
4327 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4328 negated items (that is, ~something) in unsigned ints. Some compilers
4329 apparently mutter when there is no cast.
4331 70. If an address verification called from an ACL failed, and did not produce a
4332 user-specific message (i.e. there was only a "system" message), nothing was
4333 put in $acl_verify_message. In this situation, it now puts the system
4336 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4337 session; change 4.31/43 added the unwanted input to the log line - except
4338 that it did not do this in the start of session case. It now does.
4340 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4341 This is wrong and can cause the other end to generate a synchronization
4342 error if it is another Exim or anything else that does the synchronization
4343 check. A QUIT command is no longer sent after a timeout.
4345 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4346 during host lookups.
4348 74. The defer_ok option of callout verification was not working if it was used
4349 when verifying addresses in header lines, that is, for this case:
4351 verify = header_sender/callout=defer_ok
4353 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4354 those file descriptors could be used for SMTP connections. If anything
4355 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4356 could be sent to the SMTP client, causing chaos. The daemon now opens
4357 stdin, stdout, and stderr to /dev/null when it puts itself into the
4360 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4361 log by default. The user can override this with suitable Perl magic.
4363 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4364 the log message when discard triggers, was not working for the DATA ACL or
4365 for the non-SMTP ACL.
4367 78. Error message wording change in sieve.c.
4369 79. If smtp_accept_max_per_host was set, the number of connections could be
4370 restricted to fewer than expected, because the daemon was trying to set up
4371 a new connection before checking whether the processes handling previous
4372 connections had finished. The check for completed processes is now done
4373 earlier. On busy systems, this bug wouldn't be noticed because something
4374 else would have woken the daemon, and it would have reaped the completed
4377 80. If a message was submitted locally by a user whose login name contained one
4378 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4379 It caused a spool format error. I have fixed the spool reading code. A
4380 related problem was that the "from" clause in the Received: line became
4381 illegal because of the space(s). It is now covered by ${quote_local_part.
4383 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4386 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4388 83. Patch from Sieve maintainer to fix unterminated string problem in
4389 "vacation" handling.
4391 84. Some minor changes to the Linux configuration files to help with other
4392 OS variants using glibc.
4394 85. One more patch for Sieve to update vacation handling to latest spec.
4397 ----------------------------------------------------
4398 See the note above about the 4.44 and 4.50 releases.
4399 ----------------------------------------------------
4405 1. Change 4.43/35 introduced a bug that caused file counts to be
4406 incorrectly computed when quota_filecount was set in an appendfile
4409 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4410 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4413 3. Give more explanation in the error message when the command for a transport
4414 filter fails to execute.
4416 4. There are several places where Exim runs a non-Exim command in a
4417 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4418 was being done only for the command run by the queryprogram router. It is
4419 now done for all such subprocesses. The other cases are: ${run, transport
4420 filters, and the commands run by the lmtp and pipe transports.
4422 5. Some older OS have a limit of 256 on the maximum number of file
4423 descriptors. Exim was using setrlimit() to set 1000 as a large value
4424 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4425 systems. I've change it so that if it can't get 1000, it tries for 256.
4427 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4428 was an oversight, and furthermore, ever since the addition of extra
4429 controls (e.g. 4.43/32), the checks on when to allow different forms of
4430 "control" were broken. There should now be diagnostics for all cases when a
4431 control that does not make sense is encountered.
4433 7. $recipients is now available in the predata ACL (oversight).
4435 8. Tidy the search cache before the fork to do a delivery from a message
4436 received from the command line. Otherwise the child will trigger a lookup
4437 failure and thereby defer the delivery if it tries to use (for example) a
4438 cached ldap connection that the parent has called unbind on.
4440 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4441 of $address_data from the recipient verification was clobbered by the
4442 sender verification.
4444 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4445 was its contents. (It was OK if the option was not defined at all.)
4447 11. A "Completed" log line is now written for messages that are removed from
4448 the spool by the -Mrm option.
4450 12. $host_address is now set to the target address during the checking of
4451 ignore_target_hosts.
4453 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4454 being passed; this would have caused $sender_host_name to have been used if
4455 matching the list had actually called for a host name (not very likely,
4456 since this list is usually IP addresses). A host name is now passed as
4459 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4460 code that specifies a non-restarting handler (typically sigaction() in
4461 modern systems) in an attempt to fix a rare and obscure crash bug.
4463 15. Narrowed the window for a race in the daemon that could cause it to ignore
4464 SIGCHLD signals. This is not a major problem, because they are used only to
4465 wake it up if nothing else does.
4467 16. A malformed maildirsize file could cause Exim to calculate negative values
4468 for the mailbox size or file count. Odd effects could occur as a result.
4469 The maildirsize information is now recalculated if the size or filecount
4472 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4473 support for a long time. Removed HAVE_SYS_VFS_H.
4475 18. Updated exipick to current release from John Jetmore.
4477 19. Allow an empty sender to be matched against a lookup in an address list.
4478 Previously the only cases considered were a regular expression, or an
4481 20. Exim went into a mad DNS lookup loop when doing a callout where the
4482 host was specified on the transport, if the DNS lookup yielded more than
4485 21. The RFC2047 encoding function was originally intended for short strings
4486 such as real names; it was not keeping to the 75-character limit for
4487 encoded words that the RFC imposes. It now respects the limit, and
4488 generates multiple encoded words if necessary. To be on the safe side, I
4489 have increased the buffer size for the ${rfc2047: expansion operator from
4492 22. Failure to deliver a bounce message always caused it to be frozen, even if
4493 there was an errors_to setting on the router. The errors_to setting is now
4496 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4497 canonical form (fully expanded) before being placed in
4498 $sender_host_address.
4500 24. Updated eximstats to version 1.33
4502 25. Include certificate and key file names in error message when GnuTLS fails
4503 to set them up, because the GnuTLS error message doesn't include the name
4504 of the failing file when there is a problem reading it.
4506 26. Expand error message when OpenSSL has problems setting up cert/key files.
4509 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4510 (this can affect the format of dates).
4512 28. exim_tidydb, when checking for the continued existence of a message for
4513 which it has found a message-specific retry record, was not finding
4514 messages that were in split spool directories. Consequently, it was
4515 deleting retry records that should have stayed in existence.
4517 29. eximstats updated to version 1.35
4518 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4519 1.35 - bugfix such that pie charts by volume are generated correctly
4521 30. The SPA authentication driver was not abandoning authentication and moving
4522 on to the next authenticator when an expansion was forced to fail,
4523 contradicting the general specification for all authenticators. Instead it
4524 was generating a temporary error. It now behaves as specified.
4526 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4527 (the order specifies the preference for clients). The order is now AES256,
4528 AES128, 3DES, ARCFOUR128.
4530 31. Small patch to Sieve code - explicitly set From: when generating an
4533 32. Exim crashed if a remote delivery caused a very long error message to be
4534 recorded - for instance if somebody sent an entire SpamAssassin report back
4535 as a large number of 550 error lines. This bug was coincidentally fixed by
4536 increasing the size of one of Exim's internal buffers (big_buffer) that
4537 happened as part of the Exiscan merge. However, to be on the safe side, I
4538 have made the code more robust (and fixed the comments that describe what
4541 33. Some experimental protocols are using DNS PTR records for new purposes. The
4542 keys for these records are domain names, not reversed IP addresses. The
4543 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4544 leaves it alone. Component reversal etc. now happens only for IP addresses.
4547 34. The host_aton() function is supposed to be passed a string that is known
4548 to be a valid IP address. However, in the case of IPv6 addresses, it was
4549 not checking this. This is a hostage to fortune. Exim now panics and dies
4550 if the condition is not met. A case was found where this could be provoked
4551 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4552 components; fortuitously, this particular loophole had already been fixed
4553 by change 4.50/55 or 4.44/33 above.
4555 If there are any other similar loopholes, the new check in host_aton()
4556 itself should stop them being exploited. The report I received stated that
4557 data on the command line could provoke the exploit when Exim was running as
4558 exim, but did not say which command line option was involved. All I could
4559 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4560 running as the user.
4563 35. There was a buffer overflow vulnerability in the SPA authentication code
4564 (which came originally from the Samba project). I have added a test to the
4565 spa_base64_to_bits() function which I hope fixes it.
4568 36. The daemon start-up calls getloadavg() while still root for those OS that
4569 need the first call to be done as root, but it missed one case: when
4570 deliver_queue_load_max is set with deliver_drop_privilege. This is
4571 necessary for the benefit of the queue runner, because there is no re-exec
4572 when deliver_drop_privilege is set.
4574 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4575 was in use, was not putting the data itself into the right store pool;
4576 consequently, it could be overwritten for a subsequent message in the same
4577 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4580 38. Sometimes the final signoff response after QUIT could fail to get
4581 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4582 < 0 before doing a fflush(). This bug looks as though it goes back to the
4583 introduction of TLS in release 3.20, but "sometimes" must have been rare
4584 because the tests only now provoked it.
4590 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4591 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4592 Now it returns an integer. A number of calls were still expecting a T/F
4593 return. Fortuitously, in all cases, the tests worked in OK situations,
4594 which is the norm. However, things would have gone wrong on any write
4595 failures on the smtp file descriptor. This function is used when sending
4596 messages over SMTP and also when doing verify callouts.
4598 2. When Exim is called to do synchronous delivery of a locally submitted
4599 message (the -odf or -odi options), it no longer closes stderr before doing
4602 3. Implemented the mua_wrapper option.
4604 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4606 5. Implemented the functions header_remove(), header_testname(),
4607 header_add_at_position(), and receive_remove_recipient(), and exported them
4610 6. If an ACL "warn" statement specified the addition of headers, Exim already
4611 inserted X-ACL-Warn: at the start if there was no header name. However, it
4612 was not making this test for the second and subsequent header lines if
4613 there were newlines in the string. This meant that an invalid header could
4614 be inserted if Exim was badly configured.
4616 7. Allow an ACL "warn" statement to add header lines at the start or after all
4617 the Received: headers, as well as at the end.
4619 8. Added the rcpt_4xx retry error code.
4621 9. Added postmaster_mailfrom=xxx to callout verification option.
4623 10. Added mailfrom=xxxx to the callout verification option, for verify=
4626 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4627 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4629 12. Inserted some casts to stop certain compilers warning when using pointer
4630 differences as field lengths or precisions in printf-type calls (mostly
4631 affecting debugging statements).
4633 13. Added optional readline() support for -be (dynamically loaded).
4635 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4636 same clock tick as a message's arrival, so that its received time was the
4637 same as the "first fail" time on the retry record, and that message
4638 remained on the queue past the ultimate address timeout, every queue runner
4639 would try a delivery (because it was past the ultimate address timeout) but
4640 after another failure, the ultimate address timeout, which should have then
4641 bounced the address, did not kick in. This was a "< instead of <=" error;
4642 in most cases the first failure would have been in the next clock tick
4643 after the received time, and all would be well.
4645 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4646 being recognized when the domain list was tested by the match_domain
4647 condition in an expansion string.
4649 16. Added the ${str2b64: operator.
4651 17. Exim was always calling setrlimit() to set a large limit for the number of
4652 processes, without checking whether the existing limit was already
4653 adequate. (It did check for the limit on file descriptors.) Furthermore,
4654 errors from getrlimit() and setrlimit() were being ignored. Now they are
4655 logged to the main and panic logs, but Exim does carry on, to try to do its
4656 job under whatever limits there are.
4658 18. Imported PCRE 5.0.
4660 19. Trivial typo in log message " temporarily refused connection" (the leading
4663 20. If the log selector return_path_on_delivery was set and an address was
4664 redirected to /dev/null, the delivery process crashed because it assumed
4665 that a return path would always be set for a "successful" delivery. In this
4666 case, the whole delivery is bypassed as an optimization, and therefore no
4669 21. Internal re-arrangement: the function for sending a challenge and reading
4670 a response while authentication was assuming a zero-terminated challenge
4671 string. It's now changed to take a pointer and a length, to allow for
4672 binary data in such strings.
4674 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4676 23. Exim was not respecting finduser_retries when seeking the login of the
4677 uid under which it was called; it was always trying 10 times. (The default
4678 setting of finduser_retries is zero.) Also, it was sleeping after the final
4679 failure, which is pointless.
4681 24. Implemented tls_on_connect_ports.
4683 25. Implemented acl_smtp_predata.
4685 26. If the domain in control=submission is set empty, Exim assumes that the
4686 authenticated id is a complete email address when it generates From: or
4687 Sender: header lines.
4689 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4690 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4691 chown and chgrp in /bin and hostname in /usr/bin.
4693 28. Exim was keeping the "process log" file open after each use, just as it
4694 does for the main log. This opens the possibility of it remaining open for
4695 long periods when the USR1 signal hits a daemon. Occasional processlog
4696 errors were reported, that could have been caused by this. Anyway, it seems
4697 much more sensible not to leave this file open at all, so that is what now
4700 29. The long-running daemon process does not normally write to the log once it
4701 has entered its main loop, and it closes the log before doing so. This is
4702 so that log files can straightforwardly be renamed and moved. However,
4703 there are a couple of unusual error situations where the daemon does write
4704 log entries, and I had neglected to close the log afterwards.
4706 30. The text of an SMTP error response that was received during a remote
4707 delivery was being truncated at 512 bytes. This is too short for some of
4708 the long messages that one sometimes sees. I've increased the limit to
4711 31. It is now possible to make retry rules that apply only when a message has a
4712 specific sender, in particular, an empty sender.
4714 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4715 it possible to be selective about when SMTP synchronization is enforced.
4717 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4719 32. Implemented hosts_connection_nolog.
4721 33. Added an ACL for QUIT.
4723 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4726 35. Added mailbox_size and mailbox_filecount to appendfile.
4728 36. Added control = no_multiline_responses to ACLs.
4730 37. There was a bug in the logic of the code that waits for the clock to tick
4731 in the case where the clock went backwards by a substantial amount such
4732 that the microsecond fraction of "now" was more than the microsecond
4733 fraction of "then" (but the whole seconds number was less).
4735 38. Added support for the libradius Radius client library this is found on
4736 FreeBSD (previously only the radiusclient library was supported).
4742 1. When certain lookups returned multiple values in the form name=value, the
4743 quoting of the values was not always being done properly. Specifically:
4744 (a) If the value started with a double quote, but contained no whitespace,
4746 (b) If the value contained whitespace other than a space character (i.e.
4747 tabs or newlines or carriage returns) it was not quoted.
4748 This fix has been applied to the mysql and pgsql lookups by writing a
4749 separate quoting function and calling it from the lookup code. The fix
4750 should probably also be applied to nisplus, ibase and oracle lookups, but
4751 since I cannot test any of those, I have not disturbed their existing code.
4753 2. A hit in the callout cache for a specific address caused a log line with no
4754 reason for rejecting RCPT. Now it says "Previous (cached) callout
4755 verification failure".
4757 3. There was an off-by-one bug in the queryprogram router. An over-long
4758 return line was truncated at 256 instead of 255 characters, thereby
4759 overflowing its buffer with the terminating zero. As well as fixing this, I
4760 have increased the buffer size to 1024 (and made a note to document this).
4762 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4763 when Exim is waiting for an SMTP response from a remote server, Exim
4764 restarts its select() call on the socket, thereby resetting its timeout.
4765 This is not a problem when such interrupts are rare. Somebody set up a cron
4766 job to run exiwhat every 2 minutes, which is less than the normal select()
4767 timeout (5 or 10 minutes). This meant that the select() timeout never
4768 kicked in because it was always reset. I have fixed this by comparing the
4769 time when an interrupt arrives with the time at the start of the first call
4770 to select(). If more time than the timeout has elapsed, the interrupt is
4771 treated as a timeout.
4773 5. Some internal re-factoring in preparation for the addition of Sieve
4774 extensions (by MH). In particular, the "personal" test is moved to a
4775 separate function, and given an option for scanning Cc: and Bcc: (which is
4776 not set for Exim filters).
4778 6. When Exim created an email address using the login of the caller as the
4779 local part (e.g. when creating a From: or Sender: header line), it was not
4780 quoting the local part when it contained special characters such as @.
4782 7. Installed new OpenBSD configuration files.
4784 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4785 try to make them clearer.
4787 9. Callout options, other than the timeout value, were being ignored when
4788 verifying sender addresses in header lines. For example, when using
4790 verify = header_sender/callout=no_cache
4792 the cache was (incorrectly) being used.
4794 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4795 only the Cygwin environment.
4797 11. When return_path_on_delivery was set as a log selector, if different remote
4798 addresses in the same message used different return paths and parallel
4799 remote delivery occurred, the wrong values would sometimes be logged.
4800 (Whenever a remote delivery process finished, the return path value from
4801 the most recently started remote delivery process was logged.)
4803 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4804 lines when AUTH and/or TLS are in use. This is the "received protocol"
4805 field. Exim used to use "asmtp" for authenticated SMTP, without any
4806 indication (in the protocol name) for TLS use. Now it follows the RFC and
4807 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4808 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4809 names appear in log lines as well as in Received: header lines.
4811 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4812 extensions, and comparison tests, and to fix some bugs.
4814 14. Changes to the "personal" filter test:
4816 (1) The test was buggy in that it was just doing the equivalent of
4817 "contains" tests on header lines. For example, if a user's address was
4818 anne@some.where, the "personal" test would incorrectly be true for
4820 To: susanne@some.where
4822 This test is now done by extracting each address from the header in turn,
4823 and checking the entire address. Other tests that are part of "personal"
4824 are now done using regular expressions (for example, to check local parts
4825 of addresses in From: header lines).
4827 (2) The list of non-personal local parts in From: addresses has been
4828 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4829 taken from the Sieve specification recommendations.
4831 (3) If the message contains any header line starting with "List-" it is
4832 treated as non-personal.
4834 (4) The test for "circular" in the Subject: header line has been removed
4835 because it now seems ill-conceived.
4837 15. Minor typos in src/EDITME comments corrected.
4839 16. Installed latest exipick from John Jetmore.
4841 17. If headers_add on a router specified a text string that was too long for
4842 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4843 of string_sprintf() is now avoided.
4845 18. $message_body_size was not set (it was always zero) when running the DATA
4846 ACL and the local_scan() function.
4848 19. For the "mail" command in an Exim filter, no default was being set for
4849 the once_repeat time, causing a random time value to be used if "once" was
4850 specified. (If the value happened to be <= 0, no repeat happened.) The
4851 default is now 0s, meaning "never repeat". The "vacation" command was OK
4852 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4853 (I found it when inspecting the code).
4855 20. There is now an overall timeout for performing a callout verification. It
4856 defaults to 4 times the callout timeout, which applies to individual SMTP
4857 commands during the callout. The overall timeout applies when there is more
4858 than one host that can be tried. The timeout is checked before trying the
4859 next host. This prevents very long delays if there are a large number of
4860 hosts and all are timing out (e.g. when the network connections are timing
4861 out). The value of the overall timeout can be changed by specifying an
4862 additional sub-option for "callout", called "maxwait". For example:
4864 verify = sender/callout=5s,maxwait=20s
4866 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4867 to the end before writing, but this should make it even safer).
4869 22. Exim was forgetting that it had advertised PIPELINING for the second and
4870 subsequent messages on an SMTP connection. It was also not resetting its
4871 memory on STARTTLS and an internal HELO.
4873 23. When Exim logs an SMTP synchronization error within a session, it now
4874 records whether PIPELINING has been advertised or not.
4876 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4877 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4878 rather than long int.
4880 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4882 26. Added the never_mail option to autoreply.
4888 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4889 crash if the getsockname() call failed; this can happen if a connection is
4890 closed very soon after it is established. The problem was simply in the
4891 order in which certain operations were done, causing Exim to try to write
4892 to the SMTP stream before it had set up the file descriptor. The bug has
4893 been fixed by making things happen in the correct order.
4899 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4900 before the connection was closed, thus losing the rejection response.
4902 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4903 some early Solaris releases, but causes trouble in current releases where
4904 socklen_t is defined.
4906 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4909 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4912 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4913 that releasing the top of it at the end releases what was used for sub-
4914 expansions (unless the block got too big). However, discard this block if
4915 the first thing is a variable or header, so that we can use its block when
4916 it is dynamic (useful for very large $message_headers, for example).
4918 6. Lookups now cache *every* query, not just the most recent. A new, separate
4919 store pool is used for this. It can be recovered when all lookup caches are
4920 flushed. Lookups now release memory at the end of their result strings.
4921 This has involved some general refactoring of the lookup sources.
4923 7. Some code has been added to the store_xxx() functions to reduce the amount
4924 of flapping under certain conditions.
4926 8. log_incoming_interface used to affect only the <= reception log lines. Now
4927 it causes the local interface and port to be added to several more SMTP log
4928 lines, for example "SMTP connection from", and rejection lines.
4930 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4932 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4934 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4935 could be overwritten at the end of the current message (or the start of a
4936 new message if it was set in a HELO ACL). The value is now preserved for
4937 the duration of the SMTP connection.
4939 12. If a transport had a headers_rewrite setting, and a matching header line
4940 contained an unqualified address, that address was qualified, even if it
4941 did not match any rewriting rules. The underlying bug was that the values
4942 of the flags that permit the existence of unqualified sender and recipient
4943 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4944 non-local messages, and by -bnq for local messages) were not being
4945 preserved with the message after it was received.
4947 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4948 "next input=" as part of the text comprising the host identity instead of
4949 the correct text. The code was using the same buffer for two different
4950 strings. However, depending on which order the printing function evaluated
4951 its arguments, the bug did not always show up. Under Linux, for example, my
4952 test suite worked just fine.
4954 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4955 doesn't work with some older versions of Perl. It has been changed to "my",
4956 which in any case is probably the better facility to use.
4958 15. A really picky compiler found some instances of statements for creating
4959 error messages that either had too many or two few arguments for the format
4962 16. The size of the buffer for calls to the DNS resolver has been increased
4963 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4964 for addresses that have a lot of PTR records. This alleviates a problem; it
4965 does not fully solve it.
4967 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4968 buffer now truncates the list and logs the incident, which is the same
4969 action as happens when Exim is looking up a host name and its aliases.
4970 Previously in this situation something unpredictable would happen;
4971 sometimes it was "internal error: store_reset failed".
4973 18. If a server dropped the connection unexpectedly when an Exim client was
4974 using GnuTLS and trying to read a response, the client delivery process
4975 crashed while trying to generate an error log message.
4977 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4978 string, for example:
4980 warn message = H1: something\nH2: something
4982 the text was added as a single header line from Exim's point of view
4983 though it ended up OK in the delivered message. However, searching for the
4984 second and subsequent header lines using $h_h2: did not work. This has been
4985 fixed. Similarly, if a system filter added multiple headers in this way,
4986 the routers could not see them.
4988 20. Expanded the error message when iplsearch is called with an invalid key to
4989 suggest using net-iplsearch in a host list.
4991 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4992 ACLs are no longer actually imposed (and a message to that effect is
4995 22. If a "gecos" field in a passwd entry contained escaped characters, in
4996 particular, if it contained a \" sequence, Exim got it wrong when building
4997 a From: or a Sender: header from that name. A second bug also caused
4998 incorrect handling when an unquoted " was present following a character
4999 that needed quoting.
5001 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5002 was not being matched caselessly.
5004 24. Arranged for all hyphens in the exim.8 source to be escaped with
5007 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5008 properly. Recipient callout cache records were still being keyed to include
5009 the sender, even when use_sender was set false. This led to far more
5010 callouts that were necessary. The sender is no longer included in the key
5011 when use_sender is false.
5013 26. Added "control = submission" modifier to ACLs.
5015 27. Added the ${base62d: operator to decode base 62 numbers.
5017 28. dnsdb lookups can now access SRV records.
5019 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5020 the configuration file.
5022 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5023 (-v) mode. This makes the output for a verbose queue run more intelligible.
5025 31. Added a use_postmaster feature to recipient callouts.
5027 32. Added the $body_zerocount variable, containing the number of binary zero
5028 bytes in the message body.
5030 33. The time of last modification of the "new" subdirectory is now used as the
5031 "mailbox time last read" when there is a quota error for a maildir
5034 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5036 35. Added +ignore_unknown as a special item in host lists.
5038 36. Code for decoding IPv6 addresses in host lists is now included, even if
5039 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5040 address was recognized as an IP address, but was then not correctly decoded
5041 into binary, causing unexpected and incorrect effects when compared with
5048 1. Very minor rewording of debugging text in manualroute to say "list of
5049 hosts" instead of "hostlist".
5051 2. If verify=header_syntax was set, and a header line with an unqualified
5052 address (no domain) and a large number of spaces between the end of the
5053 name and the colon was received, the reception process suffered a buffer
5054 overflow, and (when I tested it) crashed. This was caused by some obsolete
5055 code that should have been removed. The fix is to remove it!
5057 3. When running in the test harness, delay a bit after writing a bounce
5058 message to get a bit more predictability in the log output.
5060 4. Added a call to search_tidyup() just before forking a reception process. In
5061 theory, someone could use a lookup in the expansion of smtp_accept_max_
5062 per_host which, without the tidyup, could leave open a database connection.
5064 5. Added the variables $recipient_data and $sender_data which get set from a
5065 lookup success in an ACL "recipients" or "senders" condition, or a router
5066 "senders" option, similar to $domain_data and $local_part_data.
5068 6. Moved the writing of debug_print from before to after the "senders" test
5071 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5072 problems for message scanning, either using a data ACL, or using
5073 local_scan() because the Received: header was not generated till after they
5074 were called (in order to set the time as the time of reception completion).
5075 I have revised the way this works. The header is now generated after the
5076 body is received, but before the ACL or local_scan() are called. After they
5077 are run, the timestamp in the header is updated.
5083 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5084 before starting a queue runner without re-exec. This happened only when
5085 deliver_drop_privilege was set or when the Exim user was set to root. The
5086 effect of the bug was that timeouts during subsequent deliveries caused
5087 crashes instead of being properly handled. The handler is now left at its
5088 default (and expected) setting.
5090 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5091 message, again when deliver_drop_privilege is set or Exim is run as root.
5092 The bug described in (1) was not present in this case, but the tidying up
5093 of the other signals was missing. I have made the two cases consistent.
5095 3. The ignore_target_hosts setting on a manualroute router was being ignored
5096 for hosts that were looked up using the /MX notation.
5098 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5101 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5102 operated on the sender address. After changing the $sender_address to <>
5103 for the sender address verify, Exim was re-instated it as the original
5104 (before rewriting) address, but remembering that it had rewritten it, so it
5105 wasn't rewriting it again. This bug also had the effect of breaking the
5106 sender address verification caching when the sender address was rewritten.
5108 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5109 This has been changed so that if the ip literal address matches
5110 ignore_target_hosts, the router declines.
5112 7. Added expansion conditions match_domain, match_address, and match_local_
5113 part (NOT match_host).
5115 8. The placeholder for the Received: header didn't have a length field set.
5117 9. Added code to Exim itself and to exim_lock to test for a specific race
5118 condition that could lead to file corruption when using MBX delivery. The
5119 issue is with the lockfile that is created in /tmp. If this file is removed
5120 after a process has opened it but before that process has acquired a lock,
5121 there is the potential for a second process to recreate the file and also
5122 acquire a lock. This could lead to two Exim processes writing to the file
5123 at the same time. The added code performs the same test as UW imapd; it
5124 checks after acquiring the lock that its file descriptor still refers to
5125 the same named file.
5127 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5128 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5131 11. Added the smtp_active_hostname option. If used, this will typically be made
5132 to depend on the incoming interface address. Because $interface_address is
5133 not set up until the daemon has forked a reception process, error responses
5134 that can happen earlier (such as "too many connections") no longer contain
5137 12. If an expansion in a condition on a "warn" statement fails because a lookup
5138 defers, the "warn" statement is abandoned, and the next ACL statement is
5139 processed. Previously this caused the whole ACL to be aborted.
5141 13. Added the iplsearch lookup type.
5143 14. Added ident_timeout as a log selector.
5145 15. Added tls_certificate_verified as a log selector.
5147 16. Added a global option tls_require_ciphers (compare the smtp transport
5148 option of the same name). This controls incoming TLS connections.
5150 17. I finally figured out how to make tls_require_ciphers do a similar thing
5151 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5152 before starting the TLS session.
5154 18. Tabs are now shown as \t in -bP output.
5156 19. If the log selector return_path_on_delivery was set, Exim crashed when
5157 bouncing a message because it had too many Received: header lines.
5159 20. If two routers both had headers_remove settings, and the first one included
5160 a superfluous trailing colon, the final name in the first list and the
5161 first name in the second list were incorrectly joined into one item (with a
5162 colon in the middle).
5168 1. Added -C and -D options to the exinext utility, mainly to make it easier
5169 to include in the automated testing, but these could be helpful when
5170 multiple configurations are in use.
5172 2. The exinext utility was not formatting the output nicely when there was
5173 an alternate port involved in the retry record key, nor when there was a
5174 message id as well (for retries that were specific to a specific message
5175 and a specific host). It was also confused by IPv6 addresses, because of
5176 the additional colons they contain. I have fixed the IPv4 problem, and
5177 patched it up to do a reasonable job for IPv6.
5179 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5180 delivery, the log line now contains "pipelined" if PIPELINING was used.
5182 4. An SMTP transport process used to panic and die if the bind() call to set
5183 an explicit outgoing interface failed. This has been changed; it is now
5184 treated in the same way as a connect() failure.
5186 5. A reference to $sender_host_name in the part of a conditional expansion
5187 that was being skipped was still causing a DNS lookup. This no longer
5190 6. The def: expansion condition was not recognizing references to header lines
5191 that used bh_ and bheader_.
5193 7. Added the _cache feature to named lists.
5195 8. The code for checking quota_filecount in the appendfile transport was
5196 allowing one more file than it should have been.
5198 9. For compatibility with Sendmail, the command line option
5206 and sets the incoming protocol and host name (for trusted callers). The
5207 host name and its colon can be omitted when only the protocol is to be set.
5208 Note the Exim already has two private options, -pd and -ps, that refer to
5209 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5210 "s", but I don't think that's a major issue.
5212 10. A number of refactoring changes to the code, none of which should affect
5215 (a) The number of logging options was getting close to filling up the
5216 32-bit word that was used as a bit map. I have split them into two classes:
5217 those that are passed in the argument to log_write(), and those that are
5218 only ever tested independently outside of that function. These are now in
5219 separate 32-bit words, so there is plenty of room for expansion again.
5220 There is no change in the user interface or the logging behaviour.
5222 (b) When building, for example, log lines, the code previously used a
5223 macro that called string_cat() twice, in order to add two strings. This is
5224 not really sufficiently general. Furthermore, there was one instance where
5225 it was actually wrong because one of the argument was used twice, and in
5226 one call a function was used. (As it happened, calling the function twice
5227 did not affect the overall behaviour.) The macro has been replaced by a
5228 function that can join an arbitrary number of extra strings onto a growing
5231 (c) The code for expansion conditions now uses a table and a binary chop
5232 instead of a serial search (which was left over from when there were very
5233 few conditions). Also, it now recognizes conditions like "pam" even when
5234 the relevant support is not compiled in: a suitably worded error message is
5235 given if an attempt is made to use such a condition.
5237 11. Added ${time_interval:xxxxx}.
5239 12. A bug was causing one of the ddress fields not to be passed back correctly
5240 from remote delivery subprocesses. The field in question was not being
5241 subsequently used, so this caused to problems in practice.
5243 13. Added new log selectors queue_time and deliver_time.
5245 14. Might have fixed a bug in maildirsizefile handling that threw up
5246 "unexpected character" debug warnings, and recalculated the data
5247 unnecessarily. In any case, I expanded the warning message to give more
5250 15. Added the message "Restricted characters in address" to the statements in
5251 the default ACL that block characters like @ and % in local parts.
5253 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5254 Three changes have been made:
5256 (a) There was a serious bug; a negative response to MAIL caused the whole
5257 recipient domain to be cached as invalid, thereby blocking all messages
5258 to all local parts at the same domain, from all senders. This bug has
5259 been fixed. The domain is no longer cached after a negative response to
5260 MAIL if the sender used is not empty.
5262 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5265 (c) A new callout option, "use_sender" has been added for people who want
5266 the modified behaviour.
5272 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5275 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5276 indeed breaks things for older releases.
5278 3. Added additional logging to the case where there is a problem reading data
5279 from a filter that is running in a subprocess using a pipe, in order to
5280 try to track down a specific problem.
5282 4. Testing facility fudge: when running in the test harness and attempting
5283 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5284 getting "No route to host". Convert this to a timeout.
5286 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5289 6. Some OS don't have socklen_t but use size_t instead. This affects the
5290 fifth argument of getsockopt() amongst other things. This is now
5291 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5292 can be set for individual OS. I have set it for SunOS5, OSF1, and
5293 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5294 some earlier ones do not.
5296 7. Change 4.30/15 was not doing the test caselessly.
5298 8. The standard form for an IPv6 address literal was being rejected by address
5299 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5300 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5301 this, as well as the form without the "IPv6" on the front (but only when
5302 address literals are enabled, of course).
5304 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5306 10. Exim crashed if a message with an empty sender address specified by -f
5307 encountered a router with an errors_to setting. This could be provoked only
5308 by a command such as
5312 where an empty string was supplied; "<>" did not hit this bug.
5314 11. Installed PCRE release 4.5.
5316 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5317 remained set. It is now erased.
5319 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5320 times from message ids (which are base 36 rather than the normal 62).
5322 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5323 were being counted as actual protocol errors, and logged if the log
5324 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5325 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5326 and DATA following a set of rejected RCPTs do not count as protocol errors.
5327 In other words, Exim assumes they were pipelined, though this may not
5328 actually be the case. Of course, in all cases the client gets an
5329 appropriate error code.
5331 15. If a lookup fails in an ACL condition, a message about the failure may
5332 be available; it is used if testing the ACL cannot continue, because most
5333 such messages specify what the cause of the deferral is. However, some
5334 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5335 that caused an old message to be retained and used if a later statement
5336 caused a defer, replacing the real cause of the deferral.
5338 16. If an IP address had so many PTR records that the DNS lookup buffer
5339 was not large enough to hold them, Exim could crash while trying to process
5340 the truncated data. It now detects and logs this case.
5342 17. Further to 4.21/58, another change has been made: if (and only if) the
5343 first line of a message (the first header line) ends with CRLF, a bare LF
5344 in a subsequent header line has a space inserted after it, so as not to
5345 terminate the header.
5347 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5348 unnecessarily, used atoi() instead of strtol(), and didn't check the
5349 termination when getting file sizes from file names by regex.
5351 19. Completely re-implemented the support for maildirsize files, in the light
5352 of a number of problems with the previous contributed implementation
5353 (4.30/29). In particular:
5355 . If the quota is zero, the maildirsize file is maintained, but no quota is
5358 . If the maildir directory does not exist, it is created before any attempt
5359 to write a maildirsize file.
5361 . The quota value in the file is just a cache; if the quota is changed in
5362 the transport, the new value overrides.
5364 . A regular expression is available for excluding directories from the
5367 20. The autoreply transport checks the characters in options that define the
5368 message's headers; it allows continued headers, but it was checking with
5369 isspace() after an embedded newline instead of explicitly looking for a
5372 21. If all the "regular" hosts to which an address was routed had passed their
5373 expiry times, and had not reached their retry times, the address was
5374 bounced, even if fallback hosts were defined. Now Exim should go on to try
5377 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5378 equivalent code in the SMTP transport. Some hosts send humungous responses
5379 to HELO/EHLO, more than 1024 it seems.
5381 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5382 gives compiler warnings in some environments. I've now done it "properly",
5385 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5386 (because of problems with the built-in one) was declared to return uschar *
5387 instead of char *, causing compiler failure.
5389 25. Fixed a file descriptor leak when processing alias/forward files.
5391 26. Fixed a minor format string issue in dbfn.c.
5393 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5395 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5396 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5399 29. When a sender address is verified, it is cached, to save repeating the test
5400 when there is more than one recipient in a message. However, when the
5401 verification involves a callout, it is possible for different callout
5402 options to be set for different recipients. It is too complicated to keep
5403 track of this in the cache, so now Exim always runs a verification when a
5404 callout is required, relying on the callout cache for the optimization.
5405 The overhead is duplication of the address routing, but this should not be
5408 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5409 to be verified with callout=postmaster, and the main callout worked but the
5410 postmaster check failed, the verification correctly failed. However, if a
5411 subsequent RCPT command asked for sender verification *without* the
5412 postmaster check, incorrect caching caused this verification also to fail,
5415 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5416 it was not caching the DNS options (qualify_single, search_parents) that
5417 were used when the lookup failed. A subsequent lookup with different
5418 options therefore always gave the same answer, though there were cases
5419 where it should not have. (Example: a "domains = !$mx_any" option on a
5420 dnslookup router: the "domains" option is always processed without any
5421 widening, but the router might have qualify_single set.) Now Exim uses the
5422 cached value only when the same options are set.
5424 32. Added John Jetmore's "exipick" utility to the distribution.
5426 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5427 than a timeout (e.g. a certificate is required, and is not provided), an
5428 Exim server now closes the connection immediately. Previously it waited for
5429 the client to close - but if the client is SSL, it seems that they each
5430 wait for each other, leading to a delay before one of them times out.
5432 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5433 maintained 0.8.x compatibility because I don't think many are using it, and
5434 it is clearly obsolete.
5436 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5439 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5440 client certificate was expired. A simple patch fixes this, though I don't
5441 understand the full logic of why the verify callback is called multiple
5444 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5445 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5446 which causes problems with some clients (such as the Certicom SSL Plus
5447 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5448 disables the coutermeasure allowing Eudora to connect."
5450 38. Exim was not checking that a write() to a log file succeeded. This could
5451 lead to Bad Things if a log got too big, in particular if it hit a file
5452 size limit. Exim now panics and dies if it cannot write to a log file, just
5453 as it does if it cannot open a log file.
5455 39. Modified OS/Makefile-Linux so that it now contains
5457 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5459 The two -D definitions ensure that Exim is compiled with large file
5460 support, which makes it possible to handle log files that are bigger than
5463 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5464 instance) a domain was checked against a named list that involved a lookup,
5465 causing $domain_data to be set, then another domain was checked against the
5466 same list, then the first domain was re-checked, the value of $domain_data
5467 after the final check could be wrong. In particular, if the second check
5468 failed, it could be set empty. This bug probably also applied to
5471 41. The strip_trailing_dot option was not being applied to the address given
5472 with the -f command-line option.
5474 42. The code for reading a message's header from the spool was incrementing
5475 $received_count, but never initializing it. This meant that the value was
5476 incorrect (doubled) while delivering a message in the same process in which
5477 it was received. In the most common configuration of Exim, this never
5478 happens - a fresh exec is done - but it can happen when
5479 deliver_drop_privilege is set.
5481 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5482 it now includes up to 150 characters of the unexpected data in the log
5485 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5486 and building data strings. The size of both of these buffers was 10 000
5487 bytes - far larger than anybody would *ever* want, thought I. Needless to
5488 say, somebody hit the limit. I have increased the maximum line length to
5489 20 000 and the maximum data length of concatenated lines to 100 000. I have
5490 also fixed two bugs, because there was no checking on these buffers. Tsk,
5491 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5492 buffer is too small.
5494 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5495 lsearch lookups. Now it does.
5497 46. When parsing a route_list item in a manualroute router, a fixed-length
5498 buffer was used for the list of hosts. I made this 1024 bytes long,
5499 thinking that nobody would ever have a list of hosts that long. Wrong.
5500 Somebody had a whole pile of complicated expansion conditions, and the
5501 string was silently truncated, leading to an expansion error. It turns out
5502 that it is easier to change to an unlimited length (owing to other changes
5503 that have happened since this code was originally written) than to build
5504 structure for giving a limitation error. The length of the item that
5505 expands into the list of hosts is now unlimited.
5507 47. The lsearch lookup could not handle data where the length of text line was
5508 more than 4095 characters. Such lines were truncated, leading to shortened
5509 data being returned. It should now handle lines of any length.
5511 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5512 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5515 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5516 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5517 no longer happens in comment lines. A list of replacements is now placed
5518 at the head of all of the source files, except those whose only change is
5519 to replace PERL_COMMAND in the very first #! line.
5521 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5522 messages on the queue, with a bottom-up merge sort, using code contributed
5523 by Michael Haardt. This should make operations like -bp somewhat faster on
5524 large queues. It won't affect queue runners, except when queue_run_in_order
5527 51. Installed eximstats 1.31 in the distribution.
5529 52. Added support for SRV lookups to the dnslookup router.
5531 53. If an ACL referred to $message_body or $message_body_end, the value was not
5532 reset for any messages that followed in the same SMTP session.
5534 54. The store-handling optimization for building very long strings was not
5535 differentiating between the different store pools. I don't think this
5536 actually made any difference in practice, but I've tidied it.
5538 55. While running the routers to verify a sender address, $sender_address
5539 was still set to the sender address. This is wrong, because when routing to
5540 send a bounce to the sender, it would be empty. Therefore, I have changed
5541 it so that, while verifying a sender address, $sender_address is set to <>.
5542 (There is no change to what happens when verifying a recipient address.)
5544 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5545 target A or AAAA records (if not already returned) without resetting the
5546 qualify_single or search_parents options of the DNS resolver. These are
5547 inappropriate in this case because the targets of MX and SRV records must
5548 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5549 target that, when qualified, matched something in the local domain. These
5550 two options are now turned off when doing these lookups.
5552 57. It seems that at least some releases of Reiserfs (which does not have the
5553 concept of a fixed number of inodes) returns zero and not -1 for the
5554 number of available inodes. This interacted badly with check_spool_inodes,
5555 which assumed that -1 was the "no such thing" setting. What I have done is
5556 to check that the total number of inodes is greater than zero before doing
5557 the test of how many are available.
5559 58. When a "warn" ACL statement has a log_message modifier, the message is
5560 remembered, and not repeated. This is to avoid a lot of repetition when a
5561 message has many recipients that cause the same warning to be written.
5562 Howewer, Exim was preserving the list of already written lines for an
5563 entire SMTP session, which doesn't seem right. The memory is now reset if a
5564 new message is started.
5566 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5567 debugging output unless log_rewrite was also set.
5569 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5570 of (int)(handle) when we know that handle contains (void *)(-1).
5572 61. The Exim daemon panic-logs an error return when it closes the incoming
5573 connection. However "connection reset by peer" seems to be common, and
5574 isn't really an error worthy of noting specially, so that particular error
5577 62. When Exim is trying to find all the local interfaces, it used to panic and
5578 die if the ioctl to get the interface flags failed. However, it seems that
5579 on at least one OS (Solaris 9) it is possible to have an interface that is
5580 included in the list of interfaces, but for which you get a failure error
5581 for this call. This happens when the interface is not "plumbed" into a
5582 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5583 failure of the "get flags" call assumes that the interface is down.
5585 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5586 makes life easier for people who are doing arithmetic on fields extracted
5587 from dates, where you often get leading zeros that should not be
5588 interpreted as octal.
5590 64. Added qualify_domain to the redirect router, to override the global
5593 65. If a pathologically long header line contained very many addresses (the
5594 report of this problem mentioned 10 000) and each of them was rewritten,
5595 Exim could use up a very large amount of memory. (It kept on making new
5596 copies of the header line as it rewrote, and never released the old ones.)
5597 At the expense of a bit more processing, the header rewriting function has
5598 been changed so that it no longer eats memory in this way.
5600 66. The generation of the Received: header has been moved from the time that a
5601 message starts to be received, to the time that it finishes. The timestamp
5602 in the Received: header should now be very close to that of the <= log
5603 line. There are two side-effects of this change:
5605 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5606 logged header lines no longer include the local Received: line, because
5607 it has not yet been created. The same applies to a copy of the message
5608 that is returned to a non-SMTP sender when a message is rejected.
5610 (b) When a filter file is tested using -bf, no additional Received: header
5611 is added to the test message. After some thought, I decided that this
5614 This change does not affect the value of $received_for. It is still set
5615 after address rewriting, but before local_scan() is called.
5617 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5619 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5620 gave an unhelpful panic error message, and a defer error. I have managed to
5621 change this behaviour so that it now rejects any supplied certificate,
5622 which seems right, as the list of acceptable certificates is empty.
5624 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5625 gave an unhelpful defer error. I have not managed to make this reject any
5626 supplied certificates, but the error message it gives is "no certificate
5627 supplied", which is not helpful.
5629 70. exigrep's output now also includes lines that are not associated with any
5630 message, but which match the given pattern. Implemented by a patch from
5631 Martin Sluka, which also tidied up the Perl a bit.
5633 71. Recipient callout verification, like sender verification, was using <> in
5634 the MAIL FROM command. This isn't really the right thing, since the actual
5635 sender may affect whether the remote host accepts the recipient or not. I
5636 have changed it to use the actual sender in the callout; this means that
5637 the cache record is now keyed on a recipient/sender pair, not just the
5638 recipient address. There doesn't seem to be a real danger of callout loops,
5639 since a callout by the remote host to check the sender would use <>.
5640 [SEE ABOVE: changed after hitting problems.]
5642 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5643 temporary errors. However, in the case of such a code being given after
5644 the end of a data transmission (i.e. after ".") Exim was failing to write
5645 a retry record for the message. (Yes, there was some broken host that was
5646 actually sending 8xx at this point.)
5648 73. An unknown lookup type in a host list could cause Exim to panic-die when
5649 the list was checked. (An example that provoked this was putting <; in the
5650 middle of a list instead of at the start.) If this happened during a DATA
5651 ACL check, a -D file could be left lying around. This kind of configuration
5652 error no longer causes Exim to die; instead it causes a defer errror. The
5653 incident is still logged to the main and panic logs.
5655 74. Buglet left over from Exim 3 conversion. The message "too many messages
5656 in one connection" was written to the rejectlog but not the mainlog, except
5657 when address rewriting (yes!) was being logged.
5659 75. Added write_rejectlog option.
5661 76. When a system filter was run not as root (that is, when system_filter_user
5662 was set), the values of the $n variables were not being returned to the
5663 main process; thus, they were not subsequently available in the $sn
5666 77. Added +return_path_on_delivery log selector.
5668 78. A connection timeout was being treated differently from recipients deferred
5669 when testing hosts_max_try with a message that was older than the host's
5670 retry timeout. (The host should not be counted, thus allowing all hosts to
5671 be tried at least once before bouncing.) This may have been the cause of an
5672 occasionally reported bug whereby a message would remain on the queue
5673 longer than the retry timeout, but would be bounced if a delivery was
5674 forced. I say "may" because I never totally pinned down the problem;
5675 setting up timeout/retry tests is difficult. See also the next item.
5677 79. The ultimate address timeout was not being applied to errors that involved
5678 a combination of host plus message (for example, a timeout on a MAIL
5679 command). When an address resolved to a number of possible hosts, and they
5680 were not all tried for each delivery (e.g. because of hosts_max_try), a
5681 message could remain on the queue longer than the retry timeout.
5683 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5686 81. Fixed an obscure SMTP outgoing bug which required at least the following
5687 conditions: (a) there was another message waiting for the same server;
5688 (b) the server returned 5xx to all RCPT commands in the first message so
5689 that the message was not completed; (c) the server dropped the connection
5690 or gave a negative response to the RSET that Exim sends to abort the
5691 transaction. The observed case was a dropped connection after DATA that had
5692 been sent in pipelining mode. That is, the server had advertised PIPELINING
5693 but was not implementing it correctly. The effect of the bug was incorrect
5694 behaviour, such as trying another host, and this could lead to a crash.
5700 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5701 and daemon.c were passed as pointers to ints; they should have been
5702 pointers to socklen_t variables (which are typically unsigned ints).
5704 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5707 3. Fixed a really odd bug that affected only the testing scheme; patching a
5708 certain fixed string in the binary changed the value of another string that
5709 happened to be identical to the end of the original first string.
5711 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5712 name", it returns that address as the IP address. On some operating
5713 systems (e.g. Solaris), it also passes back the IP address string as the
5714 "host name". However, on others (e.g. Linux), it passes back an empty
5715 string. Exim wasn't checking for this, and was changing the host name to an
5716 empty string, assuming it had been canonicized.
5718 5. Although rare, it is permitted to have more than one PTR record for a given
5719 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5720 all the names associated with an address, because they do in Solaris.
5721 However, it seems that they do not in Linux for data that comes from the
5722 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5723 I found this out when I moved to a new Linux workstation and tried to run
5724 the Exim test suite.
5726 To get round this problem I have changed the code so that it now does its
5727 own call to the DNS to look up PTR records when searching for a host name.
5728 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5729 addresses that are only in /etc/hosts are still found.
5731 This behaviour is, however, controlled by an option called host_lookup_
5732 order, which defaults to "bydns:byaddr". If people want to use the other
5733 order, or indeed, just use one or the other means of lookup, they can
5734 specify it in this variable.
5736 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5737 some operating systems, this comes back from gethostbyaddr() as an empty
5738 string, and this is what Exim used to test for. However, it seems that in
5739 other systems, "." is yielded. Exim now tests for this case too.
5741 7. The values of check_spool_space and check_log_space are now held internally
5742 as a number of kilobytes instead of an absolute number of bytes. If a
5743 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5744 kilobyte. This means that much larger values can be stored.
5746 8. Exim monitor: an attempt to get the action menu when not actually pointing
5747 at a message produces an empty menu entitled "No message selected". This
5748 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5749 no entries in it ("Shell widget menu has zero width and/or height"). So I
5750 have added a single, blank menu entry in this case.
5752 9. Added ${quote_local_part.
5754 10. MIME decoding is now applied to the contents of Subject: header lines when
5757 11. Now that a reference to $sender_host_address automatically causes a reverse
5758 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5759 host lookup before query-style lookups in lists that might use this
5760 variable. This has therefore been abolished, and the "net-" prefix is no
5761 longer necessary for query-style lookups.
5763 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5764 have been a typo for LFLAGS, so it has been changed.
5766 13. The install script calls Exim with "-C /dev/null" in order to find the
5767 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5768 to be output. Howeve, since Exim outputs its version number before the
5769 error, it didn't break the script. It just looked ugly. I fixed this by
5770 always allowing "-C /dev/null" if the caller is root.
5772 14. Ignore overlarge ACL variable number when reading spool file - insurance
5773 against a later release with more variables having written the file.
5775 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5776 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5777 this, as well as the form without the "IPv6" on the front.
5779 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5780 OS/Makefile-Darwin file.
5782 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5783 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5785 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5786 variable. However, some OS use ints for inodes. I've added cast to long int
5787 to get rid of the compiler warning.
5789 19. I had forgotten to lock out "/../" in configuration file names when
5790 ALT_CONFIG_PREFIX was set.
5792 20. Routers used for verification do not need to specify transports. However,
5793 if such a router generated a host list, and callout was configured, Exim
5794 crashed, because it could not find a port number from the (non-existent)
5795 transport. It now assumes port 25 in this circumstance.
5797 21. Added the -t option to exigrep.
5799 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5800 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5801 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5802 them is set, LOOKUP_LSEARCH is forced.
5804 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5806 24. Added sender and host information to the "rejected by local_scan()" log
5807 line; previously there was no indication of these.
5809 25. Added .include_if_exists.
5811 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5812 Linux. It turns out that not all file systems support this. Apparently some
5813 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5814 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5815 not supported on the file descriptor, is now ignored when Exim is trying to
5816 sync a directory. This applies only to Linux.
5818 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5820 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5821 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5824 29. Added support for maildirsize files from supplied patch (modified a bit).
5826 30. The use of :fail: followed by an empty string could lead Exim to respond to
5827 sender verification failures with (e.g.):
5829 550 Verification failed for <xxx>
5830 550 Sender verify failed
5832 where the first response line was missing the '-' that indicates it is not
5833 the final line of the response.
5835 31. The loop for finding the name of the user that called Exim had a hardwired
5836 limit of 10; it now uses the value of finduser_retries, which is used for
5837 all other user lookups.
5839 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5842 33. Exim was neglecting to zero errno before one call of strtol() when
5843 expanding a string and expecting an integer value. On some systems this
5844 resulted in spurious "integer overflow" errors. Also, it was casting the
5845 result into an int without checking.
5847 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5848 did not work. The code looks as if it has *never* worked, though it appears
5849 to have been documented since at least releast 1.62. I have made it work.
5851 35. The "timeout_DNS" error in retry rules, also documented since at least
5852 1.62, also never worked. As it isn't clear exactly what this means, and
5853 clearly it isn't a major issue, I have abolished the feature by treating it
5854 as "timeout", and writing a warning to the main and panic logs.
5856 36. The display of retry rules for -brt wasn't always showing the error code
5859 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5860 timeout_connect_A, timeout_connect_MX.
5862 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5863 to the empty sender.
5865 39. The daemon was not analysing the content of -oX till after it had closed
5866 stderr and disconnected from the controlling terminal. This meant that any
5867 syntax errors were only noted on the panic log, and the return code from
5868 the command was 0. By re-arranging the code a little, I've made the
5869 decoding happen first, so such errors now appear on stderr, and the return
5870 code is 1. However, the actual setting up of the sockets still happens in
5871 the disconnected process, so errors there are still only recorded on the
5874 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5875 connections (as happens on some IP stacks) was logged at start up time as
5876 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5877 it from "IPv6 and IPv4", which means that two separate sockets are being
5880 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5881 says whether AF_INET or AF_INET6 was passed as an argument.
5883 42. Exiwhat output was messed up when time zones were included in log
5886 43. Exiwhat now gives more information about the daemon's listening ports,
5887 and whether -tls-on-connect was used.
5889 44. The "port" option of the smtp transport is now expanded.
5891 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5892 silently ignored. Now an error message is written to the main and panic
5895 46. There's a new ACL modifier called "logwrite" which writes to a log file
5896 as soon as it is encountered.
5898 47. Added $local_user_uid and $local_user_gid at routing time.
5900 48. Exim crashed when trying to verify a sender address that was being
5903 49. Exim was recognizing only a space character after ".include". It now also
5904 recognizes a tab character.
5906 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5907 extracting the relevant information from the specification. The man page no
5908 longer contains scrambled data for the -d option, and I've added a section
5909 at the front about calling Exim under different names.
5911 51. Added "extra_headers" argument to the "mail" command in filter files.
5913 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5916 53. Installed eximstats 1.29.
5918 54. Added transport_filter_timeout as a generic transport option.
5920 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5921 Cc: header lines. This was required by RFC 822, but it not required by RFC
5924 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5925 incoming messages that did not have them. Now it does so only if the
5926 message originates locally, that is, if there is no associated remote host
5927 address. When Resent- header lines are present, this applies to the Resent-
5928 lines rather than the non-Resent- lines.
5930 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5931 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5933 58. Messages for configuration errors now include the name of the main
5934 configuration file - useful now that there may be more than one file in a
5935 list (.included file names were always shown).
5937 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5938 for those rare installations that do not start the daemon as root or run it
5939 setuid root. I've cut out the call to initgroups() if the daemon is not
5942 60. The Exim user and group can now be bound into the binary as text strings
5943 that are looked up at the start of Exim's processing.
5945 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5947 62. Added $mailstore_basename variable.
5949 63. Installed patch to sieve.c from Michael Haardt.
5951 64. When Exim failed to open the panic log after failing to open the main log,
5952 the original message it was trying to log was written to stderr and debug
5953 output, but if they were not available (the usual case in production), it
5954 was lost. Now it is written to syslog before the two lines that record the
5955 failures to open the logs.
5957 65. Users' Exim filters run in subprocesses under the user's uid. It is
5958 possible for a "deliver" command or an alias in a "personal" command to
5959 provoke an address rewrite. If logging of address rewriting is configured,
5960 this fails because the process is not running as root or exim. There may be
5961 a better way of dealing with this, but for the moment (because 4.30 needs
5962 to be released), I have disabled address rewrite logging when running a
5963 filter in a non-root, non-exim process.
5969 1. The buildconfig auxiliary program wasn't quoting the value set for
5970 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5971 not defined. This bug was masked in 4.22 by the effect that was fixed in
5974 2. Some messages that were rejected after a message id was allocated were
5975 shown as "incomplete" by exigrep. It no longer does this for messages that
5976 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5978 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5979 have allow_domain_literals set, the ID did not get logged in the <= line.
5980 Domain literals are now always recognized in Message-ID: header lines.
5982 4. The first argument for a ${extract expansion item is the key name or field
5983 number. Leading and trailing spaces in this item were not being ignored,
5984 causing some misleading effects.
5986 5. When deliver_drop_privilege was set, single queue runner processes started
5987 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5988 same command in the process it spins off) were not dropping privilege.
5990 6. When the daemon running as "exim" started a queue runner, it always
5991 re-executed Exim in the spun-off process. This is a waste of effort when
5992 deliver_drop_privilege is set. The new process now just calls the
5993 queue-runner function directly.
5999 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6002 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6003 ignored. Though the use of -oP was forcing the writing of a pid file, it
6004 was always written to the default place.
6006 3. If the message "no IP address found for host xxxx" is generated during
6007 incoming verification, it is now followed by identification of the incoming
6008 connection (so you can more easily find what provoked it).
6010 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6012 5. Added some features to "harden" Exim a bit more against certain attacks:
6014 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6015 be put in Local/Makefile. This is like the never_users runtime option,
6016 but it cannot be overridden. The default setting is "root".
6018 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6019 prefix string with which any file named in a -C command line option
6022 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6023 is retained for -C and -D only if the caller of Exim is root. Without
6024 it, the exim user may also use -C and -D and retain privilege.
6026 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6027 command line option is disabled.
6029 6. Macro names set by the -D option must start with an upper case letter, just
6030 like macro names defined in the configuration file.
6032 7. Added "dereference=" facility to LDAP.
6034 8. Two instances of the typo "uknown" in the source files are fixed.
6036 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6037 the Configure-Makefile script screwed up while processing it.
6039 10. Incorporated PCRE 4.4.
6041 11. The SMTP synchronization check was not operating right at the start of an
6042 SMTP session. For example, it could not catch a HELO sent before the client
6043 waited for the greeting. There is now a check for outstanding input at the
6044 point when the greeting is written. Because of the duplex, asynchronous
6045 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6046 way, but not yet received, when the check is performed.
6048 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6049 on TCP/IP sockets, because this apparently causes some broken clients to
6052 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6053 unchanged) from the Cygwin maintainer.
6055 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6056 was supported instead of testing for mbx. Effectively a typo.
6058 15. The spa authenticator server code was not checking that the input it
6059 received was valid base64.
6061 16. The debug output line for the "set" modifier in ACLs was not showing the
6062 name of the variable that was being set.
6064 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6066 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6067 lookup on its own. Something else was needed to trigger the lookup. For
6068 example, a match in host_lookup or the need for a host name in a host list.
6069 Now, if $sender_host_name is referenced and the host name has not yet been
6070 looked up, a lookup is performed. If the lookup fails, the variable remains
6071 empty, and $host_lookup_failed is set to "1".
6073 19. Added "eqi" as a case-independent comparison operator.
6075 20. The saslauthd authentication condition could segfault if neither service
6076 nor realm was specified.
6078 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6079 error message that was logged was misleading, and incoming SMTP
6080 connections were dropped. The message is now more accurate, and temporary
6081 errors are given to SMTP connections.
6083 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6084 (except RSET) with a 421 error, until QUIT is received. However, it was
6085 failing to send a response to QUIT.
6087 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6088 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6089 ACL is now run after the helo_try_verify_hosts code.
6091 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6092 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6093 case-independent, so other case variants are also recognized). Apparently
6094 some systems use these upper case variants.
6096 25. If more than two messages were waiting for the same host, and a transport
6097 filter was specified for the transport, Exim sent two messages over the
6098 same TCP/IP connection, and then failed with "socket operation on non-
6099 socket" when it tried to send the third.
6101 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6103 27. The extern definition of crypt16() in expand.c was not being excluded when
6104 the OS had its own crypt16() function.
6106 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6107 as a preferred synonym for return_size_limit, both as an option and as an
6110 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6112 30. Changed the default configuration ACL to relax the local part checking rule
6113 for addresses that are not in any local domains. For these addresses,
6114 slashes and pipe symbols are allowed within local parts, but the sequence
6115 /../ is explicitly forbidden.
6117 31. SPA server authentication was not clearing the challenge buffer before
6120 32. log_message in a "warn" ACL statement was writing to the reject log as
6121 well as to the main log, which contradicts the documentation and doesn't
6122 seem right (because no rejection is happening). So I have stopped it.
6124 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6125 However, I am unable to do any testing of this.
6127 34. Fixed an infelicity in the appendfile transport. When checking directories
6128 for a mailbox, to see if any needed to be created, it was accidentally
6129 using path names with one or more superfluous leading slashes; tracing
6130 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6132 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6133 contents are added to the log line that is written for every discarded
6134 recipient. (Previously a log_message setting was ignored.)
6136 36. The ${quote: operator now quotes the string if it is empty.
6138 37. The install script runs exim in order to find its version number. If for
6139 some reason other than non-existence or emptiness, which it checks, it
6140 could not run './exim', it was installing it with an empty version number,
6141 i.e. as "exim-". This error state is now caught, and the installation is
6144 38. An argument was missing from the function that creates an error message
6145 when Exim fails to connect to the socket for saslauthd authentication.
6146 This could cause Exim to crash, or give a corrupted message.
6148 39. Added isip, isip4, and isip6 to ${if conditions.
6150 40. The ACL variables $acl_xx are now saved with the message, and can be
6151 accessed later in routers, transports, and filters.
6153 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6154 strings in the file are not string-expanded.
6156 42. If a MAIL command specified a SIZE value that was too large to fit into an
6157 int variable, the check against message_size_limit failed. Such values are
6158 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6159 day this will have to be increased, but I don't think I want to be around
6160 when emails are that large.
6167 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6168 iconv() is not standard in FreeBSD.
6170 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6171 IPv6 enabled. The observed symptom was a segmentation fault on return from
6172 the function os_common_find_running_interfaces() in src/os.c.
6174 3. In the check_special_case() function in daemon.c I had used "errno" as an
6175 argument name, which causes warnings on some systems. This was basically a
6176 typo, since it was named "eno" in the comments!
6178 4. The code that waits for the clock to tick (at a resolution of some fraction
6179 of a second) so as to ensure message-id uniqueness was always waiting for
6180 at least one whole tick, when it could have waited for less. [This is
6181 almost certainly not relevant at current processor speeds, where it is
6182 unlikely to ever wait at all. But we try to future-proof.]
6184 5. The function that sleeps for a time interval that includes fractions of a
6185 second contained a race. It did not block SIGALRM between setting the
6186 timer, and suspending (a couple of lines later). If the interval was short
6187 and the sigsuspend() was delayed until after it had expired, the suspension
6188 never ended. On busy systems this could lead to processes getting stuck for
6191 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6192 process, before it forks any delivery processes. The open lookup caching
6193 mechanism meant that the open file or database connection was passed into
6194 the delivery process. The problem was that delivery processes always tidy
6195 up cached lookup data. This could cause a problem for the next delivery
6196 process started by the queue runner, because the external queue runner
6197 process does not know about the closure. So the next delivery process
6198 still has data in the lookup cache. In the case of a file lookup, there was
6199 no problem because closing a file descriptor in a subprocess doesn't affect
6200 the parent. However, if the lookup was caching a connection to a database,
6201 the connection was closed, and the second delivery process was likely to
6202 see errors such as "PGSQL: query failed: server closed the connection
6203 unexpectedly". The problem has been fixed by closing all cached lookups
6204 in a queue runner before running a delivery process.
6206 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6207 seem to have the "const" qualifier which it has on other OS. I've
6210 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6211 *of the same type* (client or server) with the same public name that an
6212 error should be diagnosed.
6214 9. When Exim looked up a host name for an IP address, but failed to find the
6215 original IP address when looking up the host name (a safety check), it
6216 output the message "<ip address> does not match any IP for NULL", which was
6217 confusing, to say the least. The bug was that the host name should have
6218 appeared instead of "NULL".
6220 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6221 user that is built into the binary, and the -C or -D options is used, root
6222 privilege is dropped before the configuration file is read. In addition,
6223 logging is switched to stderr instead of the normal log files. If the
6224 configuration then re-defines the Exim user, the unprivileged environment
6225 is probably not what is expected, so Exim logs a panic warning message (but
6228 However, if deliver_drop_privilege is set, the unprivileged state may well
6229 be exactly what is intended, so the warning has been cut out in that case,
6230 and Exim is allowed to try to write to its normal log files.
6236 1. smtp_return_error_details was not giving details for temporary sender
6237 or receiver verification errors.
6239 2. Diagnose a configuration error if two authenticators have the same public
6242 3. Exim used not to create the message log file for a message until the first
6243 delivery attempt. This could be confusing when incoming messages were held
6244 for policy or load reasons. The message log file is now created at the time
6245 the message is received, and an initial "Received" line is written to it.
6247 4. The automatically generated man page for command line options had a minor
6248 bug that caused no ill effects; however, a more serious problem was that
6249 the procedure for building the man page automatically didn't always
6250 operate. Consequently, release 4.20 contains an out-of-date version. This
6251 shouldn't happen again.
6253 5. When building Exim with embedded Perl support, the script that builds the
6254 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6255 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6257 6. The freeze_tell option was not being used for messages that were frozen on
6258 arrival, either by an ACL or by local_scan().
6260 7. Added the smtp_incomplete_transaction log selector.
6262 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6263 was accepting AUTH without a new EHLO.
6265 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6266 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6269 10. Logging of TCP/IP connections (when configured) now happens in the main
6270 daemon process instead of the child process, so that the TCP/IP connection
6271 count is more accurate (but it can never be perfect).
6273 11. The use of "drop" in a nested ACL was not being handled correctly in the
6274 outer ACL. Now, if condition failure induced by the nested "drop" causes
6275 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6276 or "require"), the connection is dropped.
6278 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6279 that yield "discard" can now be used with an "accept" or a "discard" verb,
6280 but an error is generated for any others (because I can't see a useful way
6281 to define what should happen).
6283 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6284 are now processed in the same way as lines in the Exim configuration file.
6285 In particular, continuation lines are supported.
6287 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6289 15. Added -ti meaning -t -i.
6291 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6292 domains, and warn by logging if others are found.
6294 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6295 returned from GIFCONF when Exim is trying to find the list of interfaces on
6296 a host. The code in os.c has been modified to copy each ifreq to an aligned
6297 structure in all cases.
6299 Also, in some cases, the returned ifreq's were being copied to a 'struct
6300 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6301 means the last couple of bytes of an IPv6 address could be chopped if the
6302 ifreq contained only a normal sockaddr (14 bytes storage).
6304 18. Named domain lists were not supported in the hosts_treat_as_local option.
6305 An entry such as +xxxx was not recognized, and was treated as a literal
6308 19. Ensure that header lines added by a DATA ACL are included in the reject log
6309 if the ACL subsequently rejects the message.
6311 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6312 MD5 (which is deprecated).
6314 21. When testing a filter file using -bf, Exim was writing a message when it
6315 took the sender from a "From " line in the message, but it was not doing so
6316 when it took $return_path from a Return-Path: header line. It now does.
6318 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6319 with a valid header line field name (a series of printing characters
6320 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6322 23. Changed "disc" in the source to "disk" to conform to the documentation and
6323 the book and for uniformity.
6325 24. Ignore Sendmail's -Ooption=value command line item.
6327 25. When execve() failed while trying to run a command in a pipe transport,
6328 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6329 could be confused with a return value of 69 from the command itself. This
6330 has been changed to 127, the value the shell returns if it is asked to run
6331 a non-existent command. The wording for the related log line suggests a
6332 non-existent command as the problem.
6334 26. If received_header_text expands to an empty string, do not add a Received:
6335 header line to the message. (Well, it adds a token one on the spool, but
6336 marks it "old" so that it doesn't get used or transmitted.)
6338 27. Installed eximstats 1.28 (addition of -nt option).
6340 28. There was no check for failure on the call to getsockname() in the daemon
6341 code. This can fail if there is a shortage of resources on the system, with
6342 ENOMEM, for example. A temporary error is now given on failure.
6344 29. Contrary to the C standard, it seems that in some environments, the
6345 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6346 program. Exim now does this explicitly; it affects the formatting of
6347 timestamps using strftime().
6349 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6350 complaints. I've now initialized all the variables, to avoid this.
6352 32. Header lines added by a system filter were not being "seen" during
6353 transport-time rewrites.
6355 33. The info_callback() function passed to OpenSSL is set up with type void
6356 (*)(SSL *, int, int), as described somewhere. However, when calling the
6357 function (actually a macro) that sets it up, the type void(*)() is
6358 expected. I've put in a cast to prevent warnings from picky compilers.
6360 34. If a DNS black list lookup found a CNAME record, but there were no A
6361 records associated with the domain it pointed at, Exim crashed.
6363 35. If a DNS black list lookup returned more than one A record, Exim ignored
6364 all but the first. It now scans all returned addresses if a particular IP
6365 value is being sought. In this situation, the contents of the
6366 $dnslist_value variable are a list of all the addresses, separated by a
6369 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6370 to accept a host name and all its aliases if the forward lookup for any of
6371 them yielded the IP address of the incoming connection. Now it accepts only
6372 those names whose forward lookup yields the correct IP address. Any other
6373 names are discarded. This closes a loophole whereby a rogue DNS
6374 administrator could create reverse DNS records to break through a
6375 wildcarded host restriction in an ACL.
6377 37. If a user filter or a system filter that ran in a subprocess used any of
6378 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6379 the wrong values were passed to the pipe command ($thisaddress had the
6380 value of $0, $0 had the value of $1, etc). This bug was introduced by
6381 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6383 38. Improved the line breaking for long SMTP error messages from ACLs.
6384 Previously, if there was no break point between 40 and 75 characters, Exim
6385 left the rest of the message alone. Two changes have been made: (a) I've
6386 reduced the minimum length to 35 characters; (b) if it can't find a break
6387 point between 35 and 75 characters, it looks ahead and uses the first one
6388 that it finds. This may give the occasional overlong line, but at least the
6389 remaining text gets split now.
6391 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6392 file descriptors might be low, and that setting 1000 would always raise it.
6393 It turns out that in some environments, the limit is already over 1000 and
6394 that lowering it causes trouble. So now Exim takes care not to decrease it.
6396 40. When delivering a message, the value of $return_path is set to $sender_
6397 address at the start of routing (routers may change the value). By an
6398 oversight, this default was not being set up when an address was tested by
6399 -bt or -bv, which affected the outcome if any router or filter referred to
6402 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6403 log only when "message" or "log_message" are set. However, if one of the
6404 conditions was an address verification, or a call to a nested ACL, the
6405 messages generated by the underlying test were being passed through. This
6406 no longer happens. The underlying message is available in $acl_verify_
6407 message for both "message" and "log_message" expansions, so it can be
6408 passed through if needed.
6410 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6411 new expansion $bh_ to give the encoded byte string without charset
6412 translation. Translation happens only if iconv() is available; HAVE_ICONV
6413 indicates this at build time. HEADERS_CHARSET gives the charset to
6414 translate to; headers_charset can change it in the configuration, and
6415 "headers charset" can change it in an individual filter file.
6417 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6418 that creates RFC 2047 encoded "words" labels them as that charset instead
6419 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6420 expansion operator; (ii) when Exim creates a From: line for a local
6421 message; (iii) when a header line is rewritten to include a "phrase" part.
6423 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6424 buggy, causing it to skip the first lines of messages whose message ID
6425 ended in 'D'. This would not have bitten before Exim release 4.14, because
6426 message IDs were unlikely to end in 'D' before then. The effect was to have
6427 incorrect size information for certain domains.
6429 45. #include "config.h" was missing at the start of the crypt16.c module. This
6430 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6433 46. If there was a timeout during a "random" callout check, Exim treated it as
6434 a failure of the random address, and carried on sending RSET and the real
6435 address. If the delay was just some slowness somewhere, the response to the
6436 original RCPT would be taken as a response to RSET and so on, causing
6437 mayhem of various kinds.
6439 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6440 when I implemented it. It didn't allow for the fact that some option values
6441 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6442 the right test for positive values.
6444 48. Domain names in DNS records are case-independent. Exim always looks them up
6445 in lower case. Some resolvers return domain names in exactly the case they
6446 appear in the zone file, that is, they may contain uppercase letters. Not
6447 all resolvers do this - some return always lower case. Exim was treating a
6448 change of case by a resolver as a change of domain, similar to a widening
6449 of a domain abbreviation. This triggered its re-routing code and so it was
6450 trying to route what was effectively the same domain again. This normally
6451 caused routing to fail (because the router wouldn't handle the domain
6452 twice). Now Exim checks for this case specially, and just changes the
6453 casing of the domain that it ultimately uses when it transmits the message
6456 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6459 50. If a filter generated a file delivery with a non-absolute name (possible if
6460 no home directory exists for the router), the forbid_file option was not
6463 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6464 the existing equality matching.
6466 52. Exim was using ints instead of ino_t variables in some places where it was
6467 dealing with inode numbers.
6469 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6470 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6471 called TMPDIR, and if it finds it is different, it changes its value.
6473 54. The smtp_printf() function is now made available to local_scan() so
6474 additional output lines can be written before returning. There is also an
6475 smtp_fflush() function to enable the detection of a dropped connection.
6476 The variables smtp_input and smtp_batched_input are exported to
6479 55. Changed the default runtime configuration: the message "Unknown user"
6480 has been removed from the ACL, and instead placed on the localuser router,
6481 using the cannot_route_message feature. This means that any verification
6482 failures that generate their own messages won't get overridden. Similarly,
6483 the "Unrouteable address" message that was in the ACL for unverifiable
6484 relay addresses has also been removed.
6486 56. Added hosts_avoid_esmtp to the smtp transport.
6488 57. The exicyclog script was not checking for the esoteric option
6489 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6490 will work only if exicyclog is run under the appropriate euid.
6492 58. Following a discussion on the list, the rules by which Exim recognises line
6493 endings on incoming messages have been changed. The -dropcr and drop_cr
6494 options are now no-ops, retained only for backwards compatibility. The
6495 following line terminators are recognized: LF CRLF CR. However, special
6496 processing applies to CR:
6498 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6499 nor a local message in the state where . is a terminator.
6501 (ii) If a bare CR is encountered in a header line, an extra space is added
6502 after the line terminator so as not to end the header. The reasoning
6503 behind this is that bare CRs in header lines are most likely either
6504 to be mistakes, or people trying to play silly games.
6506 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6507 was being incorrectly given as 18 bytes larger than it should have been.
6508 This is a VOB (very old bug).
6510 60. This may never have affected anything current, but just in case it has:
6511 When the local host is found other than at the start of a list of hosts,
6512 the local host, those with the same MX, and any that follow, are discarded.
6513 When the list in question was part of a longer list of hosts, the following
6514 hosts (not currently being processed) were also being discarded. This no
6515 longer happens. I'm not sure if this situation could ever has previously
6518 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6521 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6522 header. This is something that is recommended in a new Internet Draft, and
6523 is something that is documented as being done by Sendmail. There are two
6524 possible values. For messages generated by the autoreply transport, Exim
6527 Auto-Submitted: auto-replied
6529 whereas for all other generated messages (e.g. bounces) it adds
6531 Auto-Submitted: auto-generated
6533 63. The "personal" condition in filters now includes a test for the
6534 Auto-Submitted: header. If it contains the string "auto-" the message it
6535 not considered personal.
6537 64. Added rcpt_include_affixes as a generic transport option.
6539 65. Added queue_only_override (default true).
6541 66. Added the syslog_duplication option.
6543 67. If what should have been the first header line of a message consisted of
6544 a space followed by a colon, Exim was mis-interpreting it as a header line.
6545 It isn't of course - it is syntactically invalid and should therefore be
6546 treated as the start of the message body. The misbehaviour could have
6547 caused a number of strange effects, including loss of data in subsequent
6548 header lines, and spool format errors.
6550 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6551 client host had authenticated. This control can now be exercised by an ACL
6552 for more flexibility.
6554 69. By default, callouts do not happen when testing with -bh. There is now a
6555 variant, -bhc, which does actually run the callout code, including
6556 consulting and updating the callout cache.
6558 70. Added support for saslauthd authentication, courtesy of Alexander
6561 71. If statvfs() failed on the spool or log directories while checking their
6562 size for availability, Exim confusingly gave the error "space shortage".
6563 Furthermore, in debugging mode it crashed with a floating point exception.
6564 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6565 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6566 serious problem, Exim now writes to the main and panic logs when this
6567 happens, with details of the failure. It then refuses to accept the
6568 incoming message, giving the message "spool directory problem" or "log
6569 directory problem" with a 421 code for SMTP messages.
6571 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6572 0, 1, and 2 exist, because some OS complain for execs without them (see
6573 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6574 descriptors. However, the code omitted to check that the open succeeded,
6575 causing mysterious errors if for some reason the permissions on /dev/null
6576 got screwed. Now Exim writes a message to the main and panic logs, and
6577 bombs out if it can't open /dev/null.
6579 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6580 interact so that it is all more flexible. It is supposed to remain
6581 backwards compatible. Also added extra_local_interfaces.
6583 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6584 to bomb out with an assertion failure - to the client this appears as a
6585 connection drop. This problem occurs in the part of the code that was taken
6586 from the Samba project. Fortunately, the assertion is in a very simple
6587 function, so I have fixed this by reproducing the function inline in the
6588 one place where it is called, and arranging for authentication to fail
6589 instead of killing the process with assert().
6591 75. The SPA client code was not working when the server requested OEM rather
6592 than Unicode encoding.
6594 76. Added code to make require_files with a specific uid setting more usable in
6595 the case where statting the file as root fails - usually a non-root-mounted
6596 NFS file system. When this happens and the failure is EACCES, Exim now
6597 forks a subprocess and does the per-uid checking as the relevant uid.
6599 77. Added process_log_path.
6601 78. If log_file_path was not explicitly set, a setting of check_log_space or
6602 check_log_inodes was ignored.
6604 79. If a space check for the spool or log partitions fails, the incident is now
6605 logged. Of course, in the latter case the data may get lost...
6607 80. Added the %p formatting code to string_format() so that it can be used to
6608 print addresses in debug_print(). Adjusted all the address printing in the
6609 debugging in store.c to use %p rather than %d.
6611 81. There was a concern that a line of code in smtp_in.c could overflow a
6612 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6613 initially expressed, the concern was not well-founded, because trailing
6614 spaces are removed early. However, if the trailing spaces were followed by
6615 a NULL, they did not get removed, so the overflow was possible. Two fixes
6618 (a) I re-wrote the offending code in a cleaner fashion.
6619 (b) If an incoming SMTP command contains a NULL character, it is rejected
6622 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6623 runs initgroups(), so that if the Exim user is in any additional groups,
6624 they will be used during message reception.
6630 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob