1 Change log file for Exim from version 4.21
2 -------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
10 supports it and a size is available (ie. the sending peer gave us one).
12 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
13 years of being deprecated). The replacements are the ACLs
14 acl_smtp_mime and acl_not_smtp_mime.
16 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
17 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
18 or even an in-clear connection were permitted. Now, if the host lookup
19 was dnssec and dane was requested then the host is only used if the
20 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
21 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
22 if one fails this test.
23 This means that a poorly-configured remote DNS will make it incommunicado;
24 but it protects against a DNS-interception attack on it.
26 JH/04 Bug 1810: make continued-use of an open smtp transport connection
27 non-noisy when a race steals the message being considered.
29 JH/05 If main configuration option tls_certificate is unset, generate a
30 self-signed certificate for inbound TLS connections.
32 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
33 in rewrites and routers.
35 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
36 and logged a warning sing 4.83; now they are a configuration file error.
38 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
39 (lacking @domain). Apply the same qualification processing as RCPT.
41 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
43 JH/10 Support ${sha256:} applied to a string (as well as the previous
46 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
47 a cutthrough deliver is pending, as we always want to make a connection.
48 This also avoids re-routing the message when later placing the cutthrough
49 connection after a verify cache hit.
50 Do not update it with the verify result either.
52 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
53 when routing results in more than one destination address.
55 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
56 signing (which inhibits the cutthrough capability). Previously only
57 the presence of an option was tested; now an expansion evaluating as
58 empty is permissible (obviously it should depend only on data available
59 when the cutthrough connection is made).
61 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
62 the relevant preceding SMTP command did not note the pipelining mode.
64 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
65 Previously they were not counted.
67 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
68 as one having no matching records. Previously we deferred the message
69 that needed the lookup.
71 JH/17 Fakereject: previously logged as a norml message arrival "<="; now
72 distinguished as "(=".
74 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
75 for missing MX records. Previously it only worked for missing A records.
77 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
79 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
80 after the data-go-ahead and data-ack. Patch from Jason Betts.
82 JH/21 Bug 1846: Send DMARC forensic reports for reject and quaratine results,
83 even for a "none" policy. Patch from Tony Meyer.
85 JH/22 Fix continued use of a connection for further deliveries. If a port was
86 specified by a router, it must also match for the delivery to be
89 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
90 When one of the recipients of a message was unsuitable for the connection
91 (has no matching addresses), we lost track of needing to mark it
92 deferred. As a result mail would be lost.
94 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
96 JH/25 Decoding ACL controls is now done using a binary search; the source code
97 takes up less space and should be simpler to maintain. Merge the ACL
98 condition decode tables also, with similar effect.
100 JH/26 Fix problem with one_time used on a redirect router which returned the
101 parent address unchanged. A retry would see the parent address marked as
102 delivered, so not attempt the (identical) child. As a result mail would
105 JH/27 Fix a possible security hole, wherein a process operating with the Exim
106 UID can gain a root shell. Credit to http://www.halfdog.net/ for
107 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
110 JH/28 Enable {spool,log} filesystem space and inode checks as default.
111 Main config options check_{log,spool}_{inodes,space} are now
112 100 inodes, 10MB unless set otherwise in the configuration.
114 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
115 Previously it only applied to the main-section connection policy
118 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
120 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
121 by me. Added RFC7919 DH primes as an alternative.
123 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
124 are not in the system include path.
126 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
127 GnuTLS, when a session startup failed (eg because the client disconnected)
128 Exim did stdio operations after fclose. This was exposed by a recent
129 change which nulled out the file handle after the fclose.
131 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
132 signed directly by the cert-signing cert, rather than an intermediate
133 OCSP-signing cert. This is the model used by LetsEncrypt.
135 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
137 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
138 an incoming connection.
140 HS/02 Bug 1802: Do not half-close the connection after sending a request
143 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
144 fallback to "prime256v1".
146 JH/34 SECURITY: Use proper copy of DATA command in error message.
147 Could leak key material. Remotely explaoitable. CVE-2016-9963.
152 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
153 and 3.4.4 - once the server is enabled to respond to an OCSP request
154 it does even when not requested, resulting in a stapling non-aware
155 client dropping the TLS connection.
157 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
158 support variable-length bit vectors. No functional change.
160 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
161 The I= interface field on outgoing lines is now after the H= remote
162 host field, same as incoming lines. There is a separate
163 outgoing_interface log selector which allows you to disable the
166 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
167 If not running log_selector +smtp_connection the mainlog would be held
168 open indefinitely after a "too many connections" event, including to a
169 deleted file after a log rotate. Leave the per net connection logging
170 leaving it open for efficiency as that will be quickly detected by the
171 check on the next write.
173 HS/01 Bug 1671: Fix post transport crash.
174 Processing the wait-<transport> messages could crash the delivery
175 process if the message IDs didn't exist for some reason. When
176 using 'split_spool_directory=yes' the construction of the spool
177 file name failed already, exposing the same netto behaviour.
179 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
180 mime_regex ACL conditions.
182 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
183 to DSN fail messages (bounces): remote IP, remote greeting, remote response
184 to HELO, local diagnostic string.
186 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
187 log line to debug. Even when configured with a tls authenticator many
188 client connections are expected to not authenticate in this way, so
189 an authenticate fail is not an error.
191 HS/02 Add the Exim version string to the process info. This way exiwhat
192 gives some more detail about the running daemon.
194 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
195 matter for fast-change records such as DNSBLs.
197 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
198 retry record, even if constant. There may be multiple transports with
199 different interface settings and the retry behaviour needs to be kept
202 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
204 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
206 JH/10 Bug 840: fix log_defer_output option of pipe transport
208 JH/11 Bug 830: use same host for all RCPTS of a message, even under
209 hosts_randomize. This matters a lot when combined with mua_wrapper.
211 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
212 ${quote_pgsql:<string>} operator.
214 JH/13 Bug 1708: avoid misaligned access in cached lookup.
216 JH/14 Change header file name for freeradius-client. Relevant if compiling
217 with Radius support; from the Gentoo tree and checked under Fedora.
219 JH/15 Bug 1712: Introduce $prdr_requested flag variable
221 JH/16 Bug 1714: Permit an empty string as expansion result for transport
222 option transport_filter, meaning no filtering.
224 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
226 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
227 defaults to "*" (all hosts). The variable is now available when not built
228 with TLS, default unset, mainly to enable keeping the testuite sane.
229 If a server certificate is not supplied (via tls_certificate) an error is
230 logged, and clients will find TLS connections fail on startup. Presumably
231 they will retry in-clear.
232 Packagers of Exim are strongly encouraged to create a server certificate
233 at installation time.
235 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
236 with the $config_file variable.
238 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
239 in transport context, after the attempt, and per-recipient. The latter type
240 is per host attempted. The event data is the error message, and the errno
241 information encodes the lookup type (A vs. MX) used for the (first) host,
242 and the trailing two digits of the smtp 4xx reponse.
244 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
245 to write to mainlog (or rejectlog, paniclog) in the window between file
246 creation and permissions/ownership being changed. Particularly affects
247 installations where exicyclog is run as root, rather than exim user;
248 result is that the running daemon panics and dies.
250 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
252 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
253 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
254 "pri" and "weight". Note that the previous implicit priority given by the
255 list order is no longer honoured.
257 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
260 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
261 by defining SUPPORT_SOCKS.
263 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
264 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
265 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
266 variables are renamed to proxy_{local,external}_{address,port}.
268 JH/27 Move Internationalisation support from Experimental to mainline, enabled
269 for a build by defining SUPPORT_I18N
271 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
272 of the query string, and make ${quote_redis:} do that quoting.
274 JH/29 Move Events support from Experimental to mainline, enabled by default
275 and removable for a build by defining DISABLE_EVENT.
277 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
279 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
280 cached by the daemon.
282 JH/32 Move Redis support from Experimental to mainline, enabled for a build
283 by defining LOOKUP_REDIS. The libhiredis library is required.
285 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
286 keys are given for lookup.
288 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
289 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
290 only supported when built with TLS support. The PolarSSL SHA routines
291 are still used when the TLS library is too old for convenient support.
293 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
294 openssl_options), for security. OpenSSL forces this from version 1.1.0
295 server-side so match that on older versions.
297 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
298 allocation for $value could be released as the expansion processing
299 concluded, but leaving the global pointer active for it.
301 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
302 and to use the domains and local_parts ACL conditions.
304 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
305 incorrectly not doubled on cutthrough transmission, hence seen as a
306 body-termination at the receiving system - resulting in truncated mails.
307 Commonly the sender saw a TCP-level error, and retransmitted the message
308 via the normal store-and-forward channel. This could result in duplicates
309 received - but deduplicating mailstores were liable to retain only the
310 initial truncated version.
312 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
314 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
316 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
317 we're in there, support oversigning also; bug 1309.
319 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
321 HS/04 Add support for keep_environment and add_environment options.
323 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
324 either intentional arithmetic overflow during PRNG, or testing config-
327 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
328 delivery resulted in actual delivery. Cancel cutthrough before DATA
331 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
332 rejects a recipient: pass the reject to the originator.
334 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
335 Many were false-positives and ignorable, but it's worth fixing the
338 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
339 for the new environment-manipulation done at startup. Move the routines
340 from being local to tls.c to being global via the os.c file.
342 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
343 an extract embedded as result-arg for a map, the first arg for extract
344 is unavailable so we cannot tell if this is a numbered or keyed
345 extraction. Accept either.
351 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
354 JH/02 The smtp transport option "multi_domain" is now expanded.
356 JH/03 The smtp transport now requests PRDR by default, if the server offers
359 JH/04 Certificate name checking on server certificates, when exim is a client,
360 is now done by default. The transport option tls_verify_cert_hostnames
361 can be used to disable this per-host. The build option
362 EXPERIMENTAL_CERTNAMES is withdrawn.
364 JH/05 The value of the tls_verify_certificates smtp transport and main options
365 default to the word "system" to access the system default CA bundle.
366 For GnuTLS, only version 3.0.20 or later.
368 JH/06 Verification of the server certificate for a TLS connection is now tried
369 (but not required) by default. The verification status is now logged by
370 default, for both outbound TLS and client-certificate supplying inbound
373 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
376 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
377 Status Notification (bounce) messages are now MIME format per RFC 3464.
378 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
379 under the control of the dsn_advertise_hosts option, and routers may
380 have a dsn_lasthop option.
382 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
383 default, modifiable by a malware= option. The list separator for
384 the options can now be changed in the usual way. Bug 68.
386 JH/10 The smtp_receive_timeout main option is now expanded before use.
388 JH/11 The incoming_interface log option now also enables logging of the
389 local interface on delivery outgoing connections.
391 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
392 if the interface and destination host and port all match.
394 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
397 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
398 Patch from Andrew Lewis.
400 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
401 now supports optional time-restrictions, weighting, and priority
402 modifiers per server. Patch originally by <rommer@active.by>.
404 JH/16 The spamd_address main option now supports a mixed list of local
405 and remote servers. Remote servers can be IPv6 addresses, and
406 specify a port-range.
408 JH/17 Bug 68: The spamd_address main option now supports an optional
409 timeout value per server.
411 JH/18 Bug 1581: Router and transport options headers_add/remove can
412 now have the list separator specified.
414 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
417 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
420 JH/21 Support for the A6 type of dns record is withdrawn.
422 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
423 rather than the verbs used.
425 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
426 from 255 to 1024 chars.
428 JH/24 Verification callouts now attempt to use TLS by default.
430 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
431 are generic router options now. The defaults didn't change.
433 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
434 Original patch from Alexander Shikoff, worked over by JH.
436 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
437 files if ZCAT_COMMAND is not executable.
439 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
441 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
443 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
444 Normally benign, it bites when the pair was led to by a CNAME;
445 modern usage is to not canoicalize the domain to a CNAME target
446 (and we were inconsistent anyway for A-only vs AAAA+A).
448 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
450 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
451 when evaluating $sender_host_dnssec.
453 JH/31 Check the HELO verification lookup for DNSSEC, adding new
454 $sender_helo_dnssec variable.
456 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
458 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
460 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
462 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
463 documented as working, but never had. Support all but $spam_report.
465 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
466 added for tls authenticator.
468 HS/03 Add perl_taintmode main config option
473 TL/01 When running the test suite, the README says that variables such as
474 no_msglog_check are global and can be placed anywhere in a specific
475 test's script, however it was observed that placement needed to be near
476 the beginning for it to behave that way. Changed the runtest perl
477 script to read through the entire script once to detect and set these
478 variables, reset to the beginning of the script, and then run through
479 the script parsing/test process like normal.
481 TL/02 The BSD's have an arc4random API. One of the functions to induce
482 adding randomness was arc4random_stir(), but it has been removed in
483 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
484 function when detected.
486 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
487 cause callback expansion.
489 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
490 syntax errors in an expansion can be treated as a string instead of
491 logging or causing an error, due to the internal use of bool_lax
492 instead of bool when processing it.
494 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
495 server certificates when making smtp deliveries.
497 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
499 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
501 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
503 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
504 Merged patch from Sebastian Wiedenroth.
506 JH/05 Fix results-pipe from transport process. Several recipients, combined
507 with certificate use, exposed issues where response data items split
508 over buffer boundaries were not parsed properly. This eventually
509 resulted in duplicates being sent. This issue only became common enough
510 to notice due to the introduction of conection certificate information,
511 the item size being so much larger. Found and fixed by Wolfgang Breyha.
513 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
514 size buffer was used, resulting in syntax errors when an expansion
517 JH/07 Add support for directories of certificates when compiled with a GnuTLS
518 version 3.3.6 or later.
520 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
521 is EXPERIMENTAL_EVENT, the main-configuration and transport options
522 both become "event_action", the variables become $event_name, $event_data
523 and $event_defer_errno. There is a new variable $verify_mode, usable in
524 routers, transports and related events. The tls:cert event is now also
525 raised for inbound connections, if the main configuration event_action
528 TL/06 In test suite, disable OCSP for old versions of openssl which contained
529 early OCSP support, but no stapling (appears to be less than 1.0.0).
531 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
532 server certificate names available under the smtp transport option
533 "tls_verify_cert_hostname" now do not permit multi-component wildcard
536 JH/10 Time-related extraction expansions from certificates now use the main
537 option "timezone" setting for output formatting, and are consistent
538 between OpenSSL and GnuTLS compilations. Bug 1541.
540 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
541 encoded parameter in the incoming message. Bug 1558.
543 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
544 include certificate info, eximon was claiming there were spoolfile
547 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
549 JH/14 Log delivery-related information more consistently, using the sequence
550 "H=<name> [<ip>]" wherever possible.
552 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
553 are problematic for Debian distribution, omit them from the release
556 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
558 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
561 JH/17 Fix a null-indirection in certextract expansions when a nondefault
562 output list separator was used.
567 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
568 checkers that were complaining about end of non-void function with no
571 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
572 This was a regression intruduced in 4.83 by another bugfix.
574 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
576 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
577 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
583 TF/01 Correctly close the server side of TLS when forking for delivery.
585 When a message was received over SMTP with TLS, Exim failed to clear up
586 the incoming connection properly after forking off the child process to
587 deliver the message. In some situations the subsequent outgoing
588 delivery connection happened to have the same fd number as the incoming
589 connection previously had. Exim would try to use TLS and fail, logging
590 a "Bad file descriptor" error.
592 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
593 utilities have not been installed.
595 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
596 temporary space as the ACL may create new global variables.
598 TL/01 LDAP support uses per connection or global context settings, depending
599 upon the detected version of the libraries at build time.
601 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
602 to extract and use the src ip:port in logging and expansions as if it
603 were a direct connection from the outside internet. PPv2 support was
604 updated based on HAProxy spec change in May 2014.
606 JH/02 Add ${listextract {number}{list}{success}{fail}}.
608 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
609 Properly escape header and check for NULL return.
611 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
614 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
616 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
617 characters in header names, implemented as a verify condition.
618 Contributed by Michael Fischer v. Mollard.
620 TL/05 Rename SPF condition results err_perm and err_temp to standardized
621 results permerror and temperror. Previous values are deprecated but
622 still accepted. In a future release, err_perm and err_temp will be
623 completely removed, which will be a backward incompatibility if the
624 ACL tests for either of these two old results. Patch contributed by
625 user bes-internal on the mailing list.
627 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
629 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
630 selectors, in both main and reject logs.
632 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
635 JH/07 Add malware type "sock" for talking to simple daemon.
637 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
639 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
640 routers/transports under cutthrough routing.
642 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
643 numbers. Touch up "bool" conditional to keep the same definition.
645 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
647 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
649 JH/12 Expand items in router/transport headers_add or headers_remove lists
650 individually rather than the list as a whole. Bug 1452.
652 Required for reasonable handling of multiple headers_ options when
653 they may be empty; requires that headers_remove items with embedded
654 colons must have them doubled (or the list-separator changed).
656 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
657 view the policy declared in the DMARC record. Currently, $dmarc_status
658 is a combined value of both the record presence and the result of the
661 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
663 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
664 dnslookup router and the smtp transport (applying to the forward
667 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
668 of ldap servers used for a specific lookup. Patch provided by Heiko
671 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
672 New variable $lookup_dnssec_authenticated for observability.
674 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
675 Patch submitted by Lars Timman.
677 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
679 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
680 Requires trusted mode and valid format message id, aborts otherwise.
681 Patch contributed by Heiko Schlichting.
683 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
684 certextract with support for various fields. Bug 1358.
686 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
687 is requested by default, modifiable by smtp transport option
690 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
691 operate on certificate variables to give certificate fingerprints
692 Also new ${sha256:cert_variable}.
694 JH/23 The PRDR feature is moved from being Experimental into the mainline.
696 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
697 Christian Aistleitner.
699 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
701 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
702 file. Patch from Wolfgang Breyha.
704 JH/25 Expand the coverage of the delivery $host and $host_address to
705 client authenticators run in verify callout. Bug 1476.
707 JH/26 Port service names are now accepted for tls_on_connect_ports, to
708 align with daemon_smtp_ports. Bug 72.
710 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
711 support and error reporting did not work properly.
713 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
714 and is readable. Patch from Andrew Colin Kissa.
716 TL/14 Enhance documentation of ${run expansion and how it parses the
717 commandline after expansion, particularly in the case when an
718 unquoted variable expansion results in an empty value.
720 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
722 PP/02 Fix internal collision of T_APL on systems which support RFC3123
723 by renaming away from it. Addresses GH issue 15, reported by
726 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
728 TL/15 SECURITY: prevent double expansion in math comparison functions
729 (can expand unsanitized data). Not remotely exploitable.
736 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
738 PP/02 Make -n do something, by making it not do something.
739 When combined with -bP, the name of an option is not output.
741 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
744 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
745 $sender_host_name and config options to manage this, and basic check
748 PP/05 DSCP support for outbound connections and control modifier for inbound.
750 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
751 (Only plugin which currently uses this is kerberos4, which nobody should
752 be using, but we should make it available and other future plugins might
753 conceivably use it, even though it would break NAT; stuff *should* be
754 using channel bindings instead).
756 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
757 name; added for Sendmail compatibility; requires admin caller.
758 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
759 just ignore it); requires trusted caller.
760 Also parse but ignore: -Ac -Am -X<logfile>
763 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
765 TL/02 Add +smtp_confirmation as a default logging option.
767 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
768 Patch by Magnus Holmgren from 2007-02-20.
770 TL/04 Bugzilla 1281 - Spec typo.
771 Bugzilla 1283 - Spec typo.
772 Bugzilla 1290 - Spec grammar fixes.
774 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
776 TL/06 Add Experimental DMARC support using libopendmarc libraries.
778 TL/07 Fix an out of order global option causing a segfault. Reported to dev
779 mailing list by by Dmitry Isaikin.
781 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
783 JH/02 Support "G" suffix to numbers in ${if comparisons.
785 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
787 NM/01 Bugzilla 1197 - Spec typo
788 Bugzilla 1196 - Spec examples corrections
790 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
792 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
793 gnutls_enable_pkcs11, but renamed to more accurately indicate its
796 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
797 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
799 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
800 "acl {{name}{arg}...}", and optional args on acl condition
803 JH/05 Permit multiple router/transport headers_add/remove lines.
805 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
807 JH/07 Avoid using a waiting database for a single-message-only transport.
808 Performance patch from Paul Fisher. Bugzilla 1262.
810 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
813 JH/09 Add $headers_added variable, with content from use of ACL modifier
814 add_header (but not yet added to the message). Bugzilla 199.
816 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
817 Pulled from Bugzilla 817 by Wolfgang Breyha.
819 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
821 (nb: this is the same fix as in Exim 4.80.1)
823 JH/11 Add A= logging on delivery lines, and a client_set_id option on
826 JH/12 Add optional authenticated_sender logging to A= and a log_selector
829 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
831 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
832 advertise SMTP AUTH mechanism to us, instead of a generic
833 protocol violation error. Also, make Exim more robust to bad
834 data from the Dovecot auth socket.
836 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
838 When a queue runner is handling a message, Exim first routes the
839 recipient addresses, during which it prunes them based on the retry
840 hints database. After that it attempts to deliver the message to
841 any remaining recipients. It then updates the hints database using
844 So if a recipient address works intermittently, it can get repeatedly
845 deferred at routing time. The retry hints record remains fresh so the
846 address never reaches the final cutoff time.
848 This is a fairly common occurrence when a user is bumping up against
849 their storage quota. Exim had some logic in its local delivery code
850 to deal with this. However it did not apply to per-recipient defers
851 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
853 This change adds a proper retry rule check during routing so that the
854 final cutoff time is checked against the message's age. We only do
855 this check if there is an address retry record and there is not a
856 domain retry record; this implies that previous attempts to handle
857 the address had the retry_use_local_parts option turned on. We use
858 this as an approximation for the destination being like a local
859 delivery, as in LMTP.
861 I suspect this new check makes the old local delivery cutoff check
862 redundant, but I have not verified this so I left the code in place.
864 TF/02 Correct gecos expansion when From: is a prefix of the username.
866 Test 0254 submits a message to Exim with the header
870 When I ran the test suite under the user fanf2, Exim expanded
871 the header to contain my full name, whereas it should have added
872 a Resent-Sender: header. It erroneously treats any prefix of the
873 username as equal to the username.
875 This change corrects that bug.
877 GF/01 DCC debug and logging tidyup
878 Error conditions log to paniclog rather than rejectlog.
879 Debug lines prefixed by "DCC: " to remove any ambiguity.
881 TF/03 Avoid unnecessary rebuilds of lookup-related code.
883 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
884 Bug spotted by Jeremy Harris; was flawed since initial commit.
885 Would have resulted in OCSP responses post-SNI triggering an Exim
886 NULL dereference and crash.
888 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
890 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
891 Bug detection, analysis and fix by Samuel Thibault.
892 Bugzilla 1331, Debian bug #698092.
894 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
896 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
897 Server implementation by Todd Lyons, client by JH.
898 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
899 config variable "prdr_enable" controls whether the server
900 advertises the facility. If the client requests PRDR a new
901 acl_data_smtp_prdr ACL is called once for each recipient, after
902 the body content is received and before the acl_smtp_data ACL.
903 The client is controlled by bolth of: a hosts_try_prdr option
904 on the smtp transport, and the server advertisement.
905 Default client logging of deliveries and rejections involving
906 PRDR are flagged with the string "PRDR".
908 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
909 fclose(). Diagnosis by Todd Lyons.
911 PP/17 Update configure.default to handle IPv6 localhost better.
912 Patch by Alain Williams (plus minor tweaks).
915 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
916 This is now consistent with GnuTLS, and is now documented: the
917 previous undocumented portable approach to treating the option as
918 unset was to force an expansion failure. That still works, and
919 an empty string is now equivalent.
921 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
922 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
923 not performing validation itself.
925 PP/20 Added force_command boolean option to pipe transport.
926 Patch from Nick Koston, of cPanel Inc.
928 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
931 TF/04 Added udpsend ACL modifer and hexquote expansion operator
933 PP/21 Fix eximon continuous updating with timestamped log-files.
934 Broken in a format-string cleanup in 4.80, missed when I repaired the
935 other false fix of the same issue.
936 Report and fix from Heiko Schlichting.
939 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
940 Report from Prashanth Katuri.
942 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
943 It's SecureTransport, so affects any MacOS clients which use the
944 system-integrated TLS libraries, including email clients.
946 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
947 using a MIME ACL for non-SMTP local injection.
948 Report and assistance in diagnosis by Warren Baker.
950 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
952 JH/16 Fix comparisons for 64b. Bugzilla 1385.
954 TL/09 Add expansion variable $authenticated_fail_id to keep track of
955 last id that failed so it may be referenced in subsequent ACL's.
957 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
960 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
961 ldap library initialization, allowing self-signed CA's to be
962 used. Also properly sets require_cert option later in code by
963 using NULL (global ldap config) instead of ldap handle (per
964 session). Bug diagnosis and testing by alxgomz.
966 TL/12 Enhanced documentation in the ratelimit.pl script provided in
967 the src/util/ subdirectory.
969 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
970 renamed to Transport Post Delivery Action by Jeremy Harris, as
973 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
974 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
975 redis_servers = needs to be configured which will be used by the redis
976 lookup. Patch from Warren Baker, of The Packet Hub.
978 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
980 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
981 hostname or reverse DNS when processing a host list. Used suggestions
982 from multiple comments on this bug.
984 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
986 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
987 Missed a few lines, added it to make the runtest require no keyboard
990 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
991 contains upper case chars. Make router use caseful_local_part.
993 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
994 support when GnuTLS has been built with p11-kit.
1000 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1002 This, or similar/improved, will also be change PP/11 of 4.82.
1008 PP/01 Handle short writes when writing local log-files.
1009 In practice, only affects FreeBSD (8 onwards).
1010 Bugzilla 1053, with thanks to Dmitry Isaikin.
1012 NM/01 Bugzilla 949 - Documentation tweak
1014 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1017 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1019 PP/02 Implemented gsasl authenticator.
1021 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1023 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1024 `pkg-config foo` for cflags/libs.
1026 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1027 with rest of GSASL and with heimdal_gssapi.
1029 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1030 `pkg-config foo` for cflags/libs for the TLS implementation.
1032 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1033 properties get this fed in as external SSF. A number of robustness
1034 and debugging improvements to the cyrus_sasl authenticator.
1036 PP/08 cyrus_sasl server now expands the server_realm option.
1038 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1039 Patch by Jeremy Harris.
1041 PP/10 Added dbmjz lookup type.
1043 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1045 PP/12 MAIL args handles TAB as well as SP, for better interop with
1046 non-compliant senders.
1047 Analysis and variant patch by Todd Lyons.
1049 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1050 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1051 Patch from Dirk Mueller <dmueller@suse.com>
1053 PP/13 tls_peerdn now print-escaped for spool files.
1054 Observed some $tls_peerdn in wild which contained \n, which resulted
1055 in spool file corruption.
1057 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1058 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1059 or write after TLS renegotiation, which otherwise led to messages
1062 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1063 as a tracking header (ie: a signed header comes before the signature).
1064 Patch from Wolfgang Breyha.
1066 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1067 comma-sep list; embedded commas doubled.
1069 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1071 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1073 Report and patch from Dmitry Banschikov.
1075 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
1076 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1077 failures. We appear to now support TLS1.1+ with Exim.
1079 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1080 lets Exim select keys and certificates based upon TLS SNI from client.
1081 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1082 before an outbound SMTP session. New log_selector, +tls_sni.
1084 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1085 NULL dereference. Report and patch from Alun Jones.
1087 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1088 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1089 Not seeing resolver debug output on NetBSD, but suspect this is a
1090 resolver implementation change.
1092 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1093 Left warnings. Added "eximon gdb" invocation mode.
1095 PP/21 Defaulting "accept_8bitmime" to true, not false.
1097 PP/22 Added -bw for inetd wait mode support.
1099 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1100 locate the relevant includes and libraries. Made this the default.
1102 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1103 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1105 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1106 This may cause build issues on older platforms.
1108 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1109 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1110 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1111 Added SNI support via GnuTLS too.
1112 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1114 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1116 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1117 Applied second patch from Janne, implementing suggestion to default
1118 multiple-strings-in-record handling to match SPF spec.
1120 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1122 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1123 read-only, out of scope).
1124 Patch from Wolfgang Breyha, report from Stuart Northfield.
1126 PP/29 Fix three issues highlighted by clang analyser static analysis.
1127 Only crash-plausible issue would require the Cambridge-specific
1128 iplookup router and a misconfiguration.
1129 Report from Marcin Mirosław.
1131 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1133 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1134 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1135 As part of this, removing so much warning spew let me fix some minor
1136 real issues in debug logging.
1138 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1139 assignment on my part. Fixed.
1141 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1142 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1143 Janne Snabb (who went above and beyond: thank you).
1145 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1146 string otherwise requires a connection and a bunch more work and it's
1147 relatively easy to get wrong. Should also expose TLS library linkage
1150 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1151 64-bit ${eval} (JH/03).
1153 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1154 GNU libc to support some of the 64-bit stuff, should not lead to
1155 conflicts. Defined before os.h is pulled in, so if a given platform
1156 needs to override this, it can.
1158 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1159 protection layer was required, which is not implemented.
1160 Bugzilla 1254, patch from Wolfgang Breyha.
1162 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1163 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1164 tls_dhparam take prime identifiers. Also unbreak combination of
1165 OpenSSL+DH_params+TLSSNI.
1167 PP/39 Disable SSLv2 by default in OpenSSL support.
1173 PP/01 Solaris build fix for Oracle's LDAP libraries.
1174 Bugzilla 1109, patch from Stephen Usher.
1176 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1178 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1181 TF/02 Fix a couple more cases where we did not log the error message
1182 when unlink() failed. See also change 4.74-TF/03.
1184 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1185 lock up or crash if it happened to be inside a call to libc when it
1186 got a SIGUSR1 from exiwhat.
1188 The SIGUSR1 handler appends the current process status to the process
1189 log which is later printed by exiwhat. It used to use the general
1190 purpose logging code to do this, but several functions it calls are
1191 not safe for signals.
1193 The new output code in the SIGUSR1 handler is specific to the process
1194 log, and simple enough that it's easy to inspect for signal safety.
1195 Removing some special cases also simplifies the general logging code.
1196 Removing the spurious timestamps from the process log simplifies
1199 TF/04 Improved ratelimit ACL condition.
1201 The /noupdate option has been deprecated in favour of /readonly which
1202 has clearer semantics. The /leaky, /strict, and /readonly update modes
1203 are mutually exclusive. The update mode is no longer included in the
1204 database key; it just determines when the database is updated. (This
1205 means that when you upgrde Exim will forget old rate measurements.)
1207 Exim now checks that the per_* options are used with an update mode that
1208 makes sense for the current ACL. For example, when Exim is processing a
1209 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1210 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1211 must specify per_mail/readonly. If you omit the update mode it defaults to
1212 /leaky where that makes sense (as before) or /readonly where required.
1214 The /noupdate option is now undocumented but still supported for
1215 backwards compatibility. It is equivalent to /readonly except that in
1216 ACLs where /readonly is required you may specify /leaky/noupdate or
1217 /strict/noupdate which are treated the same as /readonly.
1219 A useful new feature is the /count= option. This is a generalization
1220 of the per_byte option, so that you can measure the throughput of other
1221 aggregate values. For example, the per_byte option is now equivalent
1222 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1224 The per_rcpt option has been generalized using the /count= mechanism
1225 (though it's more complicated than the per_byte equivalence). When it is
1226 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1227 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1228 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1229 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1230 Note that using per_rcpt with a non-readonly update mode in more than
1231 one ACL will cause the recipients to be double-counted. (The per_mail
1232 and per_byte options don't have this problem.)
1234 The handling of very low rates has changed slightly. If the computed rate
1235 is less than the event's count (usually one) then this event is the first
1236 after a long gap. In this case the rate is set to the same as this event's
1237 count, so that the first message of a spam run is counted properly.
1239 The major new feature is a mechanism for counting the rate of unique
1240 events. The new per_addr option counts the number of different
1241 recipients that someone has sent messages to in the last time period. It
1242 behaves like per_rcpt if all the recipient addresses are different, but
1243 duplicate recipient addresses do not increase the measured rate. Like
1244 the /count= option this is a general mechanism, so the per_addr option
1245 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1246 example, measure the rate that a client uses different sender addresses
1247 with the options per_mail/unique=$sender_address. There are further
1248 details in the main documentation.
1250 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1252 TF/06 Removed a few PCRE remnants.
1254 TF/07 Automatically extract Exim's version number from tags in the git
1255 repository when doing development or release builds.
1257 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1258 Bugzilla 879. Patch from Paul Fisher.
1260 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1261 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1264 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1266 PP/05 Variable $av_failed, true if the AV scanner deferred.
1267 Bugzilla 1078. Patch from John Horne.
1269 PP/06 Stop make process more reliably on build failure.
1270 Bugzilla 1087. Patch from Heiko Schlittermann.
1272 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1273 Bugzilla 1089. Patch from Heiko Schlittermann.
1275 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1276 Bugzilla 1131. Patch from Holger Weiß.
1278 PP/09 Handle IPv6 addresses with SPF.
1279 Bugzilla 860. Patch from Wolfgang Breyha.
1281 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1283 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1286 PP/11 match_* no longer expand right-hand-side by default.
1287 New compile-time build option, EXPAND_LISTMATCH_RHS.
1288 New expansion conditions, "inlist", "inlisti".
1290 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1292 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1294 PP/14 fix log_write() format string regression from TF/03.
1295 Bugzilla 1152. Patch from Dmitry Isaikin.
1301 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1303 PP/02 Harmonised TLS library version reporting; only show if debugging.
1304 Layout now matches that introduced for other libraries in 4.74 PP/03.
1306 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1308 PP/04 New "dns_use_edns0" global option.
1310 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1313 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1314 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1316 TK/01 Updated PolarSSL code to 0.14.2.
1317 Bugzilla 1097. Patch from Andreas Metzler.
1319 PP/07 Catch divide-by-zero in ${eval:...}.
1320 Fixes bugzilla 1102.
1322 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1325 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1326 format-string attack -- SECURITY: remote arbitrary code execution.
1328 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1329 time unintentionally subject to list matching rules, letting the header
1330 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1331 arbitrary string expansion). This allowed for information disclosure.
1333 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1334 INT_MIN/-1 -- value coerced to INT_MAX.
1340 NM/01 Workround for PCRE version dependency in version reporting
1343 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1344 This fixes portability to compilers other than gcc, notably
1345 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1347 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1348 makefiles for portability to HP-UX and POSIX correctness.
1350 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1351 Also via indented variable definition in the Makefile.
1352 (Debugging by Oliver Heesakkers).
1354 PP/02 Restore caching of spamd results with expanded spamd_address.
1355 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1357 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1358 Improves build reliability. Fix from: Frank Elsner
1360 NM/02 Fix wide character breakage in the rfc2047 coding
1361 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1363 NM/03 Allow underscore in dnslist lookups
1364 Fixes bug 1026. Patch from Graeme Fowler
1366 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1367 Code patches from Adam Ciarcinski of NetBSD.
1369 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1372 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1373 is logged, to avoid truncation. Patch from John Horne.
1375 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1376 Patch from Jakob Hirsch.
1378 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1379 SQL string expansion failure details.
1380 Patch from Andrey Oktyabrski.
1382 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1383 Patch from Simon Arlott.
1385 PP/09 New lookups functionality failed to compile on old gcc which rejects
1386 extern declarations in function scope.
1387 Patch from Oliver Fleischmann
1389 PP/10 Use sig_atomic_t for flags set from signal handlers.
1390 Check getgroups() return and improve debugging.
1391 Fixed developed for diagnosis in bug 927 (which turned out to be
1394 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1395 Patch from Mark Zealey.
1397 PP/12 Bugzilla 1056: Improved spamd server selection.
1398 Patch from Mark Zealey.
1400 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1401 Based on patch from Heiko Schlittermann.
1403 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1404 Patch from Uwe Doering, sign-off by Michael Haardt.
1406 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1407 variable declaration deep within a block. Bug and patch from
1410 PP/15 lookups-Makefile IRIX compatibilty coercion.
1412 PP/16 Make DISABLE_DKIM build knob functional.
1414 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1415 Patch by Simon Arlott
1417 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1418 variable argument macros. Our copy now differs from upstream.
1424 TF/01 Failure to get a lock on a hints database can have serious
1425 consequences so log it to the panic log.
1427 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1428 controlled using the smtp_confirmation log selector.
1430 TF/03 Include the error message when we fail to unlink a spool file.
1432 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1433 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1434 for maintaining out-of-tree patches for some time.
1436 PP/01 Bugzilla 139: Documentation and portability issues.
1437 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1438 Handle per-OS dynamic-module compilation flags.
1440 PP/02 Let /dev/null have normal permissions.
1441 The 4.73 fixes were a little too stringent and complained about the
1442 permissions on /dev/null. Exempt it from some checks.
1443 Reported by Andreas M. Kirchwitz.
1445 PP/03 Report version information for many libraries, including
1446 Exim version information for dynamically loaded libraries. Created
1447 version.h, now support a version extension string for distributors
1448 who patch heavily. Dynamic module ABI change.
1450 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1451 privilege escalation vulnerability whereby the Exim run-time user
1452 can cause root to append content of the attacker's choosing to
1455 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1458 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1459 If dropping privileges for untrusted macros, we disabled normal logging
1460 on the basis that it would fail; for the Exim run-time user, this is not
1461 the case, and it resulted in successful deliveries going unlogged.
1462 Fixed. Reported by Andreas Metzler.
1468 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1469 only prepend for the Resent-* case. Fixes regression introduced in
1470 Exim 4.70 by NM/22 for Bugzilla 607.
1472 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1473 increasing numbers of administrators be bitten by this.
1475 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1477 PP/03 Bugzilla 994: added openssl_options main configuration option.
1479 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1481 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1483 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1485 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1486 without a peer certificate, leading to a segfault because of an
1487 assumption that peers always have certificates. Be a little more
1488 paranoid. Problem reported by Martin Tscholak.
1490 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1491 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1492 NB: ClamAV planning to remove STREAM in "middle of 2010".
1493 CL also introduces -bmalware, various -d+acl logging additions and
1494 more caution in buffer sizes.
1496 PP/09 Implemented reverse_ip expansion operator.
1498 PP/10 Bugzilla 937: provide a "debug" ACL control.
1500 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1502 PP/12 Bugzilla 973: Implement --version.
1504 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1506 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1508 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1510 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1511 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1512 ignore trailing whitespace.
1514 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1516 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1519 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1520 Notification from Dr Andrew Aitchison.
1522 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1523 ExtendedDetectionInfo response format.
1524 Notification from John Horne.
1526 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1529 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1530 XSL and documented dependency on system catalogs, with examples of how
1533 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1536 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1537 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1538 configuration file which is writeable by the Exim user or group.
1540 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1541 of configuration files to cover files specified with the -C option if
1542 they are going to be used with root privileges, not just the default
1545 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1546 option (effectively making it always true).
1548 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1549 files to be used while preserving root privileges.
1551 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1552 that rogue child processes cannot use them.
1554 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1555 run-time user, instead of root.
1557 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1558 Exim run-time user without dropping privileges.
1560 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1561 result string, instead of calling string_vformat() twice with the same
1564 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1565 for other users. Others should always drop root privileges if they use
1566 -C on the command line, even for a whitelisted configure file.
1568 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1570 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1576 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1577 $data_path, and $header_path variables; fixed documentation bugs and
1580 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1581 exipick to access non-standard spools, including the "frozen" queue
1584 NM/01 Bugzilla 965: Support mysql stored procedures.
1585 Patch from Alain Williams
1587 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1589 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1590 Patch from Andreas Metzler
1592 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1593 Patch from Kirill Miazine
1595 NM/05 Bugzilla 671: Added umask to procmail example.
1597 JJ/03 installed exipick 20100323.0, fixing doc bug
1599 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1600 directory. Notification and patch from Dan Rosenberg.
1602 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1604 TK/02 Improve log output when DKIM signing operation fails.
1606 MH/01 Treat the transport option dkim_domain as a colon separated
1607 list, not as a single string, and sign the message with each element,
1608 omitting multiple occurences of the same signer.
1610 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1611 Bugzilla 985, 986. Patch by Simon Arlott
1613 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1614 Patch by Simon Arlott
1616 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1617 MBX locking. Notification from Dan Rosenberg.
1623 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1625 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1627 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1629 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1631 NM/04 Bugzilla 915: Flag broken perl installation during build.
1637 TK/01 Added patch by Johannes Berg that expands the main option
1638 "spamd_address" if it starts with a dollar sign.
1640 TK/02 Write list of recipients to X-Envelope-Sender header when building
1641 the mbox-format spool file for content scanning (suggested by Jakob
1644 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1645 (http://www.dcc-servers.net/) support via dccifd. Activated by
1646 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1648 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1649 by Mark Daniel Reidel <mr@df.eu>.
1651 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1652 When building exim an external PCRE library is now needed -
1653 PCRE is a system library on the majority of modern systems.
1654 See entry on PCRE_LIBS in EDITME file.
1656 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1657 conversation. Added nologin parameter to request.
1658 Patch contributed by Kirill Miazine.
1660 TF/01 Do not log submission mode rewrites if they do not change the address.
1662 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1664 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1665 log files in place. Contributed by Roberto Lima.
1667 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1669 TF/03 Bugzilla 615: When checking the local_parts router precondition
1670 after a local_part_suffix or local_part_prefix option, Exim now
1671 does not use the address's named list lookup cache, since this
1672 contains cached lookups for the whole local part.
1674 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1675 Robert Millan. Documentation is in experimental-spec.txt.
1677 TF/04 Bugzilla 668: Fix parallel build (make -j).
1679 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1681 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1682 Patch provided by Jan Srzednicki.
1684 TF/05 Leading white space used to be stripped from $spam_report which
1685 wrecked the formatting. Now it is preserved.
1687 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1688 that they are available at delivery time.
1690 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1692 TF/08 TLS error reporting now respects the incoming_interface and
1693 incoming_port log selectors.
1695 TF/09 Produce a more useful error message if an SMTP transport's hosts
1696 setting expands to an empty string.
1698 NM/06 Bugzilla 744: EXPN did not work under TLS.
1699 Patch provided by Phil Pennock.
1701 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1702 Patch provided by Richard Godbee.
1704 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1705 acl_smtp_notquit, added index entry.
1707 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1708 Patch provided by Eugene Bujak.
1710 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1711 accept(). Patch provided by Maxim Dounin.
1713 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1714 Patch provided by Phil Pennock.
1716 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1718 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1719 Patch provided by Brad "anomie" Jorsch.
1721 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1722 Patch provided by Dean Brooks.
1724 TK/05 Add native DKIM support (does not depend on external libraries).
1726 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1727 Patch provided by Graeme Fowler.
1729 NM/16 Bugzilla 851: Documentation example syntax fix.
1731 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1733 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1736 NM/19 Bugzilla 745: TLS version reporting.
1737 Patch provided by Phil Pennock.
1739 NM/20 Bugzilla 167: bool: condition support.
1740 Patch provided by Phil Pennock.
1742 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1743 clients. Patch provided by Phil Pennock.
1745 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1746 Patch provided by Brad "anomie" Jorsch.
1748 NM/23 Bugzilla 687: Fix misparses in eximstats.
1749 Patch provided by Heiko Schlittermann.
1751 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1752 Patch provided by Heiko Schlittermann.
1754 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1755 plus update to original patch.
1757 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1759 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1760 Patch provided by David Brownlee.
1762 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1764 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1766 NM/30 Bugzilla 888: TLS documentation bugfixes.
1768 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1770 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1771 Unlike the original bugzilla I have changed all shell scripts in src tree.
1773 NM/33 Bugzilla 898: Transport filter timeout fix.
1774 Patch by Todd Rinaldo.
1776 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1777 Patch by Serge Demonchaux.
1779 NM/35 Bugzilla 39: Base64 decode bug fixes.
1780 Patch by Jakob Hirsch.
1782 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1784 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1786 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1788 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1794 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1795 ALT-N's libdkim that I have put here:
1796 http://duncanthrax.net/exim-experimental/
1798 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1799 were called 'true' and it seems that C99 defines that as a reserved
1800 keyword to be used with 'bool' variable types. That means you could
1801 not include C99-style headers which use bools without triggering
1802 build errors in sieve.c.
1804 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1805 as mailq or other aliases. Changed the --help handling significantly
1806 to do whats expected. exim_usage() emits usage/help information.
1808 SC/01 Added the -bylocaldomain option to eximstats.
1810 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1812 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1814 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1820 PH/01 Another patch from the Sieve maintainer.
1822 PH/02 When an IPv6 address is converted to a string for single-key lookup
1823 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1824 dots are used instead of colons so that keys in lsearch files need not
1825 contain colons. This was done some time before quoting was made available
1826 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1827 (notated using the quote facility) so as to distinguish them from IPv4
1828 keys. This meant that lookups for IP addresses in host lists did not work
1829 for iplsearch lookups.
1831 This has been fixed by arranging for IPv6 addresses to be expressed with
1832 colons if the lookup type is iplsearch. This is not incompatible, because
1833 previously such lookups could never work.
1835 The situation is now rather anomolous, since one *can* have colons in
1836 ordinary lsearch keys. However, making the change in all cases is
1837 incompatible and would probably break a number of configurations.
1839 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1842 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1843 conversion specification without a maximum field width, thereby enabling
1844 a rogue spamd server to cause a buffer overflow. While nobody in their
1845 right mind would setup Exim to query an untrusted spamd server, an
1846 attacker that gains access to a server running spamd could potentially
1847 exploit this vulnerability to run arbitrary code as the Exim user.
1849 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1850 $primary_hostname instead of what libspf2 thinks the hosts name is.
1852 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1853 a directory entry by the name of the lookup key. Previously, if a
1854 symlink pointed to a non-existing file or a file in a directory that
1855 Exim lacked permissions to read, a lookup for a key matching that
1856 symlink would fail. Now it is enough that a matching directory entry
1857 exists, symlink or not. (Bugzilla 503.)
1859 PH/03 The body_linecount and body_zerocount variables are now exported in the
1862 PH/04 Added the $dnslist_matched variable.
1864 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1865 This means they are set thereafter only if the connection becomes
1868 PH/06 Added the client_condition to authenticators so that some can be skipped
1869 by clients under certain conditions.
1871 PH/07 The error message for a badly-placed control=no_multiline_responses left
1872 "_responses" off the end of the name.
1874 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1876 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1877 (without spaces) instead of just copying the configuration text.
1879 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1881 PH/11 Added $max_received_linelength.
1883 PH/12 Added +ignore_defer and +include_defer to host lists.
1885 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1886 way in which PCRE > 7.0 is built.
1888 PH/14 Implemented queue_only_load_latch.
1890 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1891 MAIL command. The effect was to mangle the value on 64-bit systems.
1893 PH/16 Another patch from the Sieve maintainer.
1895 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1897 PH/18 If a system quota error occurred while trying to create the file for
1898 a maildir delivery, the message "Mailbox is full" was not appended to the
1899 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1900 only to a quota excession during the actual writing of the file.
1902 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1903 characters?) which causes problems in log lines. The DN values are now
1904 passed through string_printing() before being added to log lines.
1906 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1907 and InterBase are left for another time.)
1909 PH/21 Added message_body_newlines option.
1911 PH/22 Guard against possible overflow in moan_check_errorcopy().
1913 PH/23 POSIX allows open() to be a macro; guard against that.
1915 PH/24 If the recipient of an error message contained an @ in the local part
1916 (suitably quoted, of course), incorrect values were put in $domain and
1917 $local_part during the evaluation of errors_copy.
1923 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1924 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1927 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1928 issue a MAIL command.
1930 PH/02 In an ACL statement such as
1932 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1934 if a client was not listed at all, or was listed with a value other than
1935 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1936 the condition was not true (as it should be), so access was not denied.
1937 The bug was that the ! inversion was incorrectly passed on to the second
1938 item. This has been fixed.
1940 PH/03 Added additional dnslists conditions == and =& which are different from
1941 = and & when the dns lookup returns more than one IP address.
1943 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1944 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1946 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1947 FSYNC, which compiles an option called disable_fsync that allows for
1948 bypassing fsync(). The documentation is heavily laced with warnings.
1950 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1952 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1953 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1954 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1955 including adding "make clean"; (3) Added -fPIC when compiling the test
1956 dynamically loaded module, to get rid of a warning.
1958 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1959 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1960 = 0s. The bug is otherwise harmless.
1962 PH/07 There was a bug in the dovecot authenticator such that the value of
1963 $auth1 could be overwritten, and so not correctly preserved, after a
1964 successful authentication. This usually meant that the value preserved by
1965 the server_setid option was incorrect.
1967 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1969 PH/09 Installed PCRE release 7.0.
1971 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1972 run for batched SMTP input. It is now run at the start of every message
1973 in the batch. While fixing this I discovered that the process information
1974 (output by running exiwhat) was not always getting set for -bs and -bS
1975 input. This is fixed, and it now also says "batched" for BSMTP.
1977 PH/11 Added control=no_pipelining.
1979 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1980 patch, slightly modified), and move the expansion of helo_data till after
1981 the connection is made in the smtp transport (so it can use these
1984 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1986 PH/14 Added log_selector = +pid.
1988 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1990 PH/16 Add ${if forany and ${if forall.
1992 PH/17 Added dsn_from option to vary the From: line in DSNs.
1994 PH/18 Flush SMTP output before performing a callout, unless control =
1995 no_callout_flush is set.
1997 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1998 was true (the default) a successful delivery failed to delete the retry
1999 item, thus causing premature timeout of the address. The bug is now
2002 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2004 PH/21 Long custom messages for fakedefer and fakereject are now split up
2005 into multiline reponses in the same way that messages for "deny" and
2006 other ACL rejections are.
2008 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2009 with slight modification.
2011 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2012 draft, changing the syntax and factoring some duplicate code".
2014 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2015 for deliveries of the second and subsequent messages over the same SMTP
2018 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2019 ${reduce, with only minor "tidies".
2021 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2023 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2024 expansion side effects.
2026 PH/27 When a message times out after an over-quota error from an Exim-imposed
2027 quota, the bounce message says "mailbox is full". This message was not
2028 being given when it was a system quota that was exceeded. It now should
2031 MH/03 Made $recipients available in local_scan(). local_scan() already has
2032 better access to the recipient list through recipients_list[], but
2033 $recipients can be useful in postmaster-provided expansion strings.
2035 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2036 in the case of a MAIL command with additional options following the
2037 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2038 were accidentally chopped off.
2040 PH/29 SMTP synchronization checks are implemented when a command is read -
2041 there is a check that no more input is waiting when there shouldn't be
2042 any. However, for some commands, a delay in an ACL can mean that it is
2043 some time before the response is written. In this time, more input might
2044 arrive, invalidly. So now there are extra checks after an ACL has run for
2045 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2046 pipelining has not been advertised.
2048 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2050 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2051 returned in the SMTP response when smtp_return_error_details was false.
2052 This has been fixed.
2054 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2055 instead of the C I/O that was originally supplied, because problems were
2056 reported on Solaris.
2058 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2059 Exim which did not show up earlier: it was assuming that a call to
2060 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2061 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2062 was a macro that became an assignment, so it seemed to work. This has
2063 changed to a proper function call with a void return, hence the compile
2064 error. Exim's code has been fixed.
2066 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2069 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2071 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2073 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2074 because it turns out to be too big or there is a timeout) there is no
2075 "Completed" line in the log. When some messages of this type were
2076 selected by exigrep, they were listed as "not completed". Others were
2077 picked up by some special patterns. I have improved the selection
2078 criteria to be more general.
2080 PH/38 The host_find_failed option in the manualroute router can now be set
2081 to "ignore", to completely ignore a host whose IP address cannot be
2082 found. If all hosts are ignored, the behaviour is controlled by the new
2083 host_all_ignored option.
2085 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2086 homing or because of multiple MX records with /mx) generated more than
2087 one IP address, and the following item turned out to be the local host,
2088 all the secondary addresses of the first item were incorrectly removed
2089 from the list, along with the local host and any following hosts (which
2090 is what is supposed to happen).
2092 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2093 whoever called Exim into the -H file. In the case of the daemon it was
2094 behaving confusingly. When first started, it used values for whoever
2095 started the daemon, but after a SIGHUP it used the Exim user (because it
2096 calls itself on a restart). I have changed the code so that it now always
2099 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2100 message are rejected with the same error (e.g. no authentication or bad
2101 sender address), and a DATA command is nevertheless sent (as can happen
2102 with PIPELINING or a stupid MUA), the error message that was given to the
2103 RCPT commands is included in the rejection of the DATA command. This is
2104 intended to be helpful for MUAs that show only the final error to their
2107 PH/42 Another patch from the Sieve maintainer.
2109 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2110 Eximstats - Fixed some broken HTML links and added missing column headers
2112 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2113 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2115 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2117 PH/43 Yet another patch from the Sieve maintainer.
2119 PH/44 I found a way to check for a TCP/IP connection going away before sending
2120 the response to the final '.' that terminates a message, but only in the
2121 case where the client has not sent further data following the '.'
2122 (unfortunately, this is allowed). However, in many cases there won't be
2123 any further data because there won't be any more messages to send. A call
2124 to select() can be used: if it shows that the input is "ready", there is
2125 either input waiting, or the socket has been closed. An attempt to read
2126 the next input character can distinguish the two cases. Previously, Exim
2127 would have sent an OK response which the client would never have see.
2128 This could lead to message repetition. This fix should cure that, at
2129 least in a lot of common cases.
2131 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2132 advertised in response to EHLO.
2138 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2139 fixed by 4.65/MH/01 (is this a record?) are fixed:
2141 (i) An empty string was always treated as zero by the numeric comparison
2142 operators. This behaviour has been restored.
2144 (ii) It is documented that the numeric comparison operators always treat
2145 their arguments as decimal numbers. This was broken in that numbers
2146 starting with 0 were being interpreted as octal.
2148 While fixing these problems I realized that there was another issue that
2149 hadn't been noticed. Values of message_size_limit (both the global option
2150 and the transport option) were treated as octal if they started with 0.
2151 The documentation was vague. These values are now always treated as
2152 decimal, and I will make that clear in the documentation.
2158 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2159 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2162 MH/01 Don't check that the operands of numeric comparison operators are
2163 integers when their expansion is in "skipping" mode (fixes bug
2164 introduced by 4.64-PH/07).
2166 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2167 child addresses, Exim now panics and dies. Previously, because the count
2168 is held in a short int, deliveries were likely to be lost. As such a
2169 large number of recipients for a single message is ridiculous
2170 (performance will be very, very poor), I have chosen to impose a limit
2171 rather than extend the field.
2177 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2178 leftover -K file (the existence of which was triggered by #402).
2179 While we were at it, introduced process PID as part of the -K
2180 filename. This should rule out race conditions when creating
2183 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2184 processing considerably. Previous code took too long for large mails,
2185 triggering a timeout which in turn triggers #401.
2187 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2188 in the DK code in transports.c. sendfile() is not really portable,
2189 hence the _LINUX specificness.
2191 TF/01 In the add_headers option to the mail command in an Exim filter,
2192 there was a bug that Exim would claim a syntax error in any
2193 header after the first one which had an odd number of characters
2196 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2197 callout verification, Exim cached a "reject" for the entire domain. This
2198 is correct for most verifications, but it is not correct for a recipient
2199 verification with use_sender or use_postmaster set, because in that case
2200 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2201 case of MAIL FROM:<> rejection from other early rejections (e.g.
2202 rejection of HELO). When verifying a recipient using a non-null MAIL
2203 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2204 Whatever the result of the callout, the value of the domain cache is
2205 left unchanged (for any other kind of callout, getting as far as trying
2206 RCPT means that the domain itself is ok).
2208 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2211 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2212 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2213 session, the daemon ignores ECONNECT errors and logs others; it now
2214 ignores EPIPE as well.
2216 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2217 (quoted-printable decoding).
2219 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2220 later the small subsequent patch to fix an introduced bug.
2222 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2224 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2226 PH/08 An error is now given if message_size_limit is specified negative.
2228 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2229 to be given (somewhat) arbitrary names.
2231 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2234 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2235 miscellaneous code fixes
2237 PH/10 Added the log_reject_target ACL modifier to specify where to log
2240 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2241 hostname. This is wrong, because it relates to the incoming message (and
2242 probably the interface on which it is arriving) and not to the outgoing
2243 callout (which could be using a different interface). This has been
2244 changed to use the value of the helo_data option from the smtp transport
2245 instead - this is what is used when a message is actually being sent. If
2246 there is no remote transport (possible with a router that sets up host
2247 addresses), $smtp_active_hostname is used.
2249 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2250 tweaks were necessary in order to get it to work (see also 21 below):
2251 (a) The code assumed that strncpy() returns a negative number on buffer
2252 overflow, which isn't the case. Replaced with Exim's string_format()
2254 (b) There were several signed/unsigned issues. I just did the minimum
2255 hacking in of casts. There is scope for a larger refactoring.
2256 (c) The code used strcasecmp() which is not a standard C function.
2257 Replaced with Exim's strcmpic() function.
2258 (d) The code set only $1; it now sets $auth1 as well.
2259 (e) A simple test gave the error "authentication client didn't specify
2260 service in request". It would seem that Dovecot has changed its
2261 interface. Fortunately there's a specification; I followed it and
2262 changed what the client sends and it appears to be working now.
2264 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2267 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2268 address A is aliased to B and C, where B exists and C does not. Without
2269 -v the output is "A verified" because verification stops after a
2270 successful redirection if more than one address is generated. However,
2271 with -v the child addresses are also verified. Exim was outputting "A
2272 failed to verify" and then showing the successful verification for C,
2273 with its parentage. It now outputs "B failed to verify", showing B's
2274 parentage before showing the successful verification of C.
2276 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2277 look up a TXT record in a specific list after matching in a combined
2280 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2281 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2282 they consult the DNS. I had assumed they would set it the way they
2283 wanted; and indeed my experiments on Linux seem to show that in some
2284 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2285 To be on the safe side, however, I have now made the interface to
2286 host_find_byname() similar to host_find_bydns(), with an argument
2287 containing the DNS resolver options. The host_find_byname() function now
2288 sets these options at its start, just as host_find_bydns() does. The smtp
2289 transport options dns_qualify_single and dns_search_parents are passed to
2290 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2291 of host_find_byname() use the default settings of RES_DEFNAMES
2292 (qualify_single) but not RES_DNSRCH (search_parents).
2294 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2295 spool_read_header() do less string testing, by means of a preliminary
2296 switch on the second character of optional "-foo" lines. (This is
2297 overdue, caused by the large number of possibilities that now exist.
2298 Originally there were few.) While I was there, I also converted the
2299 str(n)cmp tests so they don't re-test the leading "-" and the first
2300 character, in the hope this might squeeze out yet more improvement.
2302 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2303 flag allowing group syntax was set by the header_syntax check but not
2304 turned off, possible causing trouble later; (2) The flag was not being
2305 set at all for the header_verify test, causing "group"-style headers to
2306 be rejected. I have now set it in this case, and also caused header_
2307 verify to ignore an empty address taken from a group. While doing this, I
2308 came across some other cases where the code for allowing group syntax
2309 while scanning a header line wasn't quite right (mostly, not resetting
2310 the flag correctly in the right place). These bugs could have caused
2311 trouble for malformed header lines. I hope it is now all correct.
2313 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2314 with the "reply" argument non-NULL. The code, however (which originally
2315 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2316 but it didn't always do it. This confused somebody who was copying the
2317 code for some other use. I have removed all the tests.
2319 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2320 feature that was used to support insecure browsers during the U.S. crypto
2321 embargo. It requires special client support, and Exim is probably the
2322 only MTA that supported it -- and would never use it because real RSA is
2323 always available. This code has been removed, because it had the bad
2324 effect of slowing Exim down by computing (never used) parameters for the
2325 RSA_EXPORT functionality.
2327 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2328 authenticator to fail if there's a tab character in the incoming data
2329 (there should never be unless someone is messing about, as it's supposed
2330 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2331 if the connection is using TLS or if the remote IP is the same as the
2332 local IP, and the "valid-client-cert option" if a client certificate has
2335 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2336 authenticators. This can be used for authorization after authentication
2337 succeeds. (In the case of plaintext, it servers for both authentication
2340 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2341 if any retry times were supplied.
2343 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2344 connection, where there is no client IP address to check. In this
2345 situation, the verify now always succeeds.
2347 PH/25 Applied John Jetmore's -Mset patch.
2349 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2351 PH/27 In a string expansion for a processed (not raw) header when multiple
2352 headers of the same name were present, leading whitespace was being
2353 removed from all of them, but trailing whitespace was being removed only
2354 from the last one. Now trailing whitespace is removed from each header
2355 before concatenation. Completely empty headers in a concatenation (as
2356 before) are ignored.
2358 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2359 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2361 PH/29 [Removed. This was a change that I later backed out, and forgot to
2362 correct the ChangeLog entry (that I had efficiently created) before
2363 committing the later change.]
2365 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2366 address errors (4xx response to RCPT) over the same connection as other
2367 messages routed to the same hosts. Such deliveries are always "forced",
2368 so retry times are not inspected. This resulted in far too many retries
2369 for the affected addresses. The effect occurred only when there were more
2370 hosts than the hosts_max_try setting in the smtp transport when it had
2371 the 4xx errors. Those hosts that it had tried were not added to the list
2372 of hosts for which the message was waiting, so if all were tried, there
2373 was no problem. Two fixes have been applied:
2375 (i) If there are any address or message errors in an SMTP delivery, none
2376 of the hosts (tried or untried) are now added to the list of hosts
2377 for which the message is waiting, so the message should not be a
2378 candidate for sending over the same connection that was used for a
2379 successful delivery of some other message. This seems entirely
2380 reasonable: after all the message is NOT "waiting for some host".
2381 This is so "obvious" that I'm not sure why it wasn't done
2382 previously. Hope I haven't missed anything, but it can't do any
2383 harm, as the worst effect is to miss an optimization.
2385 (ii) If, despite (i), such a delivery is accidentally attempted, the
2386 routing retry time is respected, so at least it doesn't keep
2387 hammering the server.
2389 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2390 in ${readsocket because some servers need this prod.
2392 PH/32 Added some extra debug output when updating a wait-xxx database.
2394 PH/33 The hint "could be header name not terminated by colon", which has been
2395 given for certain expansion errors for a long time, was not being given
2396 for the ${if def:h_colon_omitted{... case.
2398 PH/34 The spec says: "With one important exception, whenever a domain list is
2399 being scanned, $domain contains the subject domain." There was at least
2400 one case where this was not true.
2402 PH/35 The error "getsockname() failed: connection reset by peer" was being
2403 written to the panic log as well as the main log, but it isn't really
2404 panic-worthy as it just means the connection died rather early on. I have
2405 removed the panic log writing for the ECONNRESET error when getsockname()
2408 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2409 runs only) independently of the message's sender address. This meant
2410 that, if the 4xx error was in fact related to the sender, a different
2411 message to the same recipient with a different sender could confuse
2412 things. In particualar, this can happen when sending to a greylisting
2413 server, but other circumstances could also provoke similar problems.
2414 I have changed the default so that the retry time for these errors is now
2415 based a combination of the sender and recipient addresses. This change
2416 can be overridden by setting address_retry_include_sender=false in the
2419 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2420 remote server are returned as part of bounce messages. This was not
2421 happening for LMTP over a pipe (the lmtp transport), but now it is the
2422 same for both kinds of LMTP.
2424 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2425 in header lines that were in fact CNAMEs. This is no longer the case.
2427 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2428 and queue runs started by the daemon processed all messages. This has
2429 been fixed so that -R and -S can now usefully be given with -q<time>.
2431 PH/40 Import PCRE release 6.7 (fixes some bugs).
2433 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2435 PH/42 Give an error if -q is specified more than once.
2437 PH/43 Renamed the variables $interface_address and $interface_port as
2438 $received_ip_address and $received_port, to make it clear that these
2439 values apply to message reception, and not to the outgoing interface when
2440 a message is delivered. (The old names remain recognized, of course.)
2442 PH/44 There was no timeout on the connect() call when using a Unix domain
2443 socket in the ${readsocket expansion. There now is.
2445 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2446 be meaningful with "accept".
2448 SC/01 Eximstats V1.43
2449 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2451 SC/02 Eximstats V1.44
2452 Use a glob alias rather than an array ref in the generated
2453 parser. This improves both readability and performance.
2455 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2456 Collect SpamAssassin and rejection statistics.
2457 Don't display local sender or destination tables unless
2458 there is data to show.
2459 Added average volumes into the top table text output.
2461 SC/04 Eximstats V1.46
2462 Collect data on the number of addresses (recipients)
2463 as well as the number of messages.
2465 SC/05 Eximstats V1.47
2466 Added 'Message too big' to the list of mail rejection
2467 reasons (thanks to Marco Gaiarin).
2469 SC/06 Eximstats V1.48
2470 Mainlog lines which have GMT offsets and are too short to
2471 have a flag are now skipped.
2473 SC/07 Eximstats V1.49 (Alain Williams)
2474 Added the -emptyok flag.
2476 SC/08 Eximstats V1.50
2477 Fixes for obtaining the IP address from reject messages.
2479 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2480 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2481 whitesspace changes from 4.64-PH/27
2483 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2486 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2487 are found, allow negative numbers in numeric criteria)
2489 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2491 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2492 to match changes made in 4.64-PH/43
2494 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2496 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2497 "domain" is actually an IP address, and force a failure. This locks out
2498 those revolvers/nameservers that support "A-for-A" lookups, in
2499 contravention of the specifications.
2501 PH/48 When a host name was looked up from an IP address, and the subsequent
2502 forward lookup of the name timed out, the host name was left in
2503 $sender_host_name, contrary to the specification.
2505 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2506 restricted to single-key lookups, Exim was not diagnosing an error if
2507 * or *@ was used with a query-style lookup.
2509 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2511 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2512 long ago, but noone interested enough thought of it. Let's just say that
2513 the "1.1" means that there are some new functions that weren't there at
2514 some point in the past.
2516 PH/51 Error processing for expansion failure of helo_data from an smtp
2517 transport during callout processing was broken.
2519 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2520 tested/used via the -bh/-bhc/-bs options.
2522 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2523 bug, fixed in subsequent PCRE releases).
2525 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2526 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2528 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2534 SC/01 Use a glob alias rather than an array ref in eximstats generated
2535 parser. This improves both readability and performance.
2537 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2538 Don't display local sender or destination tables in eximstats unless
2539 there is data to show.
2540 Added average volumes into the eximstats top table text output.
2542 SC/03 Collect data on the number of addresses (recipients) as well
2543 as the number of messages in eximstats.
2545 TF/01 Correct an error in the documentation for the redirect router. Exim
2546 does (usually) call initgroups() when daemonizing.
2548 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2549 with consistent privilege compared to when running as a daemon.
2551 TF/03 Note in the spec that $authenticated_id is not set for local
2552 submissions from trusted users.
2554 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2555 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2557 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2558 by adding some example configuration directives to the default
2559 configuration file. A little bit of work is required to uncomment the
2560 directives and define how usernames and passwords are checked, but
2561 there is now a framework to start from.
2563 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2564 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2565 without this. I don't know how relevant this is to other LDAP libraries.
2567 PH/02 Add the verb name to the "unknown ACL verb" error.
2569 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2571 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2573 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2574 directory not expanded when it should be if an expanded home directory
2575 was set for the address (which is overridden by the transport).
2577 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2580 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2581 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2582 because it is too late at that time, and has no effect.
2584 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2585 security issue with \' (bugzilla #107). I could not use the
2586 PQescapeStringConn() function, because it needs a PGconn value as one of
2589 PH/08 When testing addresses using -bt, indicate those final addresses that
2590 are duplicates that would not cause an additional delivery. At least one
2591 person was confused, thinking that -bt output corresponded to deliveries.
2592 (Suppressing duplicates isn't a good idea as you lose the information
2593 about possibly different redirections that led to the duplicates.)
2595 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2596 systems where poll() doesn't work, in particular OS X.
2598 PH/10 Added more information to debugging output for retry time not reached.
2600 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2601 operations in malware.c.
2603 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2606 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2607 syslog_duplication set false, log lines that would normally be written
2608 both the the main log and to the reject log were not written to syslog at
2611 PH/14 In the default configuration, change the use of "message" in ACL warn
2612 statements to "add_header".
2614 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2615 not followed by a command (e.g. "seen endif").
2617 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2618 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2621 PH/17 Added extra conditions to the default value of delay_warning_condition
2625 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2626 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2627 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2630 The Auto-Submitted: and various List- headers are standardised, whereas I
2631 don't think Precedence: ever was.
2633 PH/18 Refactored debugging code in route_finduser() to show more information,
2634 in particular, the error code if getpwnam() issues one.
2636 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2637 This is apparently needed in addition to the PH/07 change above to avoid
2638 any possible encoding problems.
2640 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2641 but not after initializing Perl.
2643 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2644 output them only if debugging. By default they are written stderr,
2645 apparently, which is not desirable.
2647 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2650 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2653 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2655 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2656 authenticated or an ident call has been made. Suppress the default
2657 values for $authenticated_id and $authenticated_sender (but permit -oMai
2658 and -oMas) when testing with -bh.
2660 PH/24 Re-jigged the order of the tests in the default configuration so that the
2661 tests for valid domains and recipients precede the DNS black list and CSA
2662 tests, on the grounds that those ones are more expensive.
2664 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2665 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2666 This bug exists in every version of Exim that I still have, right back to
2669 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2670 However, an attempt to turn on case-sensitivity in a regex key by
2671 including (?-i) didn't work because the subject string was already
2672 lowercased, and the effects were non-intuitive. It turns out that a
2673 one-line patch can be used to allow (?-i) to work as expected.
2679 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2680 other effects) broke the use of negated acl sub-conditions.
2682 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2685 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2686 "Deny" causes Exim to reject the incoming connection with a 554 error.
2687 Unfortunately, if there is a major crisis, such as a disk failure,
2688 tcp-wrappers gives "deny", whereas what one would like would be some
2689 kind of temporary error. A kludge has been added to help with this.
2690 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2691 554 error is used if errno is still zero or contains ENOENT (which occurs
2692 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2695 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2697 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2698 errors. Otherwise a message that provokes a temporary error (when other
2699 messages do not) can cause a whole host to time out.
2701 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2702 addresses were routed directly to files or pipes from a redirect router.
2703 File deliveries just didn't batch; pipe deliveries might have suffered
2706 PH/06 A failure to get a lock for a hints database would erroneously always say
2707 "Failed to get write lock", even when it was really a read lock.
2709 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2710 of 0600. This has been changed to use the value of the lockfile_mode
2711 option (which defaults to 0600).
2713 PH/08 Applied small patch from the Sieve maintainer.
2715 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2716 folder from quota calculations, a direct delivery into this folder messed
2717 up the contents of the maildirsize file. This was because the regex was
2718 used only to exclude .Trash (or whatever) when the size of the mailbox
2719 was calculated. There was no check that a delivery was happening into an
2720 excluded directory. This bug has been fixed by ignoring all quota
2721 processing for deliveries into excluded directories.
2723 PH/10 Added the maildirfolder_create_regex option to appendfile.
2729 PH/01 The code for finding all the local interface addresses on a FreeBSD
2730 system running IPv6 was broken. This may well have applied to all BSD
2731 systems, as well as to others that have similar system calls. The broken
2732 code found IPv4 interfaces correctly, but gave incorrect values for the
2733 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2734 that it would not match correctly against @[] and not recognize the IPv6
2737 PH/02 The ipliteral router was not recognizing addresses of the form user@
2738 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2740 PH/03 Added disable_ipv6.
2742 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2743 decoded form, because it is most often used to construct To: headers
2744 lines in autoreplies, and the decoded form may well be syntactically
2745 invalid. However, $reply_address has leading white space removed, and all
2746 newlines turned into spaces so that the autoreply transport does not
2749 PH/05 If group was specified without a user on a router, and no group or user
2750 was specified on a transport, the group from the router was ignored.
2752 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2753 for visible compile-time settings that can be used to change these
2754 numbers, for those that want even more. Backwards compatibility with old
2755 spool files has been maintained. However, going back to a previous Exim
2756 release will lost any variables that are in spool files.
2758 PH/07 Two small changes when running in the test harness: increase delay when
2759 passing a TCP/IP connection to a new process, in case the original
2760 process has to generate a bounce, and remove special handling of
2761 127.0.0.2 (sic), which is no longer necessary.
2763 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2764 be the same on different OS.
2766 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2769 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2770 whether --show-vars was specified or not
2772 JJ/02 exipick: Added support for new ACL variable spool format introduced
2775 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2776 syntactically invalid From: or Reply-to: line, and a filter used this to
2777 generate an autoreply, and therefore failed to obtain an address for the
2778 autoreply, Exim could try to deliver to a non-existent relative file
2779 name, causing unrelated and misleading errors. What now happens is that
2780 it logs this as a hard delivery error, but does not attempt to create a
2783 PH/11 The exinext utility has a -C option for testing purposes, but although
2784 the given file was scanned by exinext itself; it wasn't being passed on
2785 when Exim was called.
2787 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2788 an end-of-file indication when reading a command response.
2790 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2791 compiled. In many other places in Exim, IPv6 addresses are always
2792 recognized, so I have changed this. It also means that IPv4 domain
2793 literals of the form [IPV4:n.n.n.n] are now always recognized.
2795 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2796 used if the router is not running as root, for example, when verifying at
2797 ACL time, or when using -bh. The debugging output from this situation was
2798 non-existent - all you got was a failure to exec. I have made two
2801 (a) Failures to set uid/gid, the current directory, or a process leader
2802 in a subprocess such as that created by queryprogram now generate
2803 suitable debugging ouput when -d is set.
2805 (b) The queryprogram router detects when it is not running as root,
2806 outputs suitable debugging information if -d is set, and then runs
2807 the subprocess without attempting to change uid/gid.
2809 PH/15 Minor change to Makefile for building test_host (undocumented testing
2812 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2813 additional section of a DNS packet that returns MX or SRV records.
2814 Instead, it always explicitly searches for A/AAAA records. This avoids
2815 major problems that occur when a DNS server includes only records of one
2816 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2817 fixed another bug: if SRV records were looked up and the corresponding
2818 address records were *not* found in the additional section, the port
2819 values from the SRV records were lost.
2821 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2822 using the correct key (the original address) when searching the retry
2823 rules in order to find which one to use for generating the retry hint.
2825 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2826 adding the default one. Similarly, if it contains a Reply-To: header, the
2827 errors_reply_to option, if set, is not used.
2829 PH/19 When calculating a retry time, Exim used to measure the "time since
2830 failure" by looking at the "first failed" field in the retry record. Now
2831 it does not use this if it is later than than the arrival time of the
2832 message. Instead it uses the arrival time. This makes for better
2833 behaviour in cases where some deliveries succeed, thus re-setting the
2834 "first failed" field. An example is a quota failure for a huge message
2835 when small messages continue to be delivered. Without this change, the
2836 "time since failure" will always be short, possible causing more frequent
2837 delivery attempts for the huge message than are intended.
2838 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2840 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2841 $1, $2, $3) because the numerical variables can be reset during some
2842 expansion items (e.g. "match"), thereby losing the authentication data.
2844 PH/21 Make -bV show the size of off_t variables so that the test suite can
2845 decide whether to run tests for quotas > 2G.
2847 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2848 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2849 filecount value is greater than 2G or if a quota value is greater than 2G
2850 on a system where the size of off_t is not greater than 4, a panic error
2853 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2854 never match. The debug and -bh output now contains an explicit error
2855 message indicating a malformed IPv4 address or mask.
2857 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2858 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2859 PH/23 above applies.
2861 PH/25 Do not write to syslog when running in the test harness. The only
2862 occasion when this arises is a failure to open the main or panic logs
2863 (for which there is an explicit test).
2865 PH/26 Added the /no_tell option to "control=freeze".
2867 PH/27 If a host name lookup failed very early in a connection, for example, if
2868 the IP address matched host_lookup and the reverse lookup yielded a name
2869 that did not have a forward lookup, an error message of the form "no IP
2870 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2871 could be logged. Now it outputs the IP address instead of "NULL".
2873 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2874 allows the sender and the authenticated sender to be set when
2875 submitting a message from within Exim. Since child_open_exim() is
2876 documented for local_scan(), the new function should be too.
2878 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2879 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2880 results in an empty string is now treated as unset.
2882 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2884 PH/31 Added sender_verify_fail logging option.
2886 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2887 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2888 tidied the source and removed it altogether.
2890 PH/33 When a queue run was abandoned because the load average was too high, a
2891 log line was always written; now it is written only if the queue_run log
2892 selector is set. In addition, the log line for abandonment now contains
2893 information about the queue run such as the pid. This is always present
2894 in "start" and "stop" lines but was omitted from the "abandon" line.
2896 PH/34 Omit spaces between a header name and the colon in the error message that
2897 is given when verify = headers_syntax fails (if there are lots of them,
2898 the message gets confusing).
2900 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2901 names, as there are now some PTR records that contain slashes. This check
2902 is only to protect against broken name servers that fall over on strange
2903 characters, so the fact that it applies to all lookups doesn't matter.
2905 PH/36 Now that the new test suite is complete, we can remove some of the
2906 special code in Exim that was needed for the old test suite. For example,
2907 sorting DNS records because real resolvers return them in an arbitrary
2908 order. The new test suite's fake resolver always returns records in the
2911 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2912 bounces) except when queue_only is set, to avoid logging races between
2913 the different processes.
2915 PH/38 Panic-die if .include specifies a non-absolute path.
2917 PH/39 A tweak to the "H" retry rule from its user.
2919 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2920 a label. They prevented compilation on older perls.
2922 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2923 a warning to be raised on newish perls.
2925 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2926 on queue. Changes to match documented behaviour of showing count of
2927 messages matching specified criteria.
2929 PH/40 Changed the default ident timeout from 30s to 5s.
2931 PH/41 Added support for the use of login_cap features, on those BSD systems
2932 that have them, for controlling the resources used by pipe deliveries.
2934 PH/42 The content-scanning code uses fopen() to create files in which to put
2935 message data. Previously it was not paying any attention to the mode of
2936 the files. Exim runs with umask(0) because the rest of the code creates
2937 files with open(), and sets the required mode explicitly. Thus, these
2938 files were ending up world-writeable. This was not a big issue, because,
2939 being within the spool directory, they were not world-accessible. I have
2940 created a function called modefopen, which takes an additional mode
2941 argument. It sets umask(777), creates the file, chmods it to the required
2942 mode, then resets the umask. All the relevant calls to fopen() in the
2943 content scanning code have been changed to use this function.
2945 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2946 to 24 hours. This avoids potential overflow problems when processing G
2947 and H retry rules. I suspect nobody ever tinkers with this value.
2949 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2951 PH/45 When the plaintext authenticator is running as a client, the server's
2952 challenges are checked to ensure they are valid base64 strings. By
2953 default, the authentication attempt is cancelled if an invalid string is
2954 received. Setting client_ignore_invalid_base64 true ignores these errors.
2955 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2956 they are received. Thus, the responses can be made to depend on the
2957 challenges. If an invalid string is ignored, an empty string is placed in
2960 PH/46 Messages that are created by the autoreply transport now contains a
2961 References: header, in accordance with RFCs 2822 and 3834.
2963 PH/47 Added authenticated_sender_force to the smtp transport.
2965 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2967 PH/49 Installed latest patch from the Sieve maintainer.
2969 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2970 was also set, the appendfile transport was unnecessarily scanning a
2971 directory of message files (e.g. for maildir delivery) to find the count
2972 of files (along with the size), even though it did not need this
2973 information. It now does the scan only if it needs to find either the
2974 size of the count of files.
2976 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2978 PH/52 Two bugs concerned with error handling when the smtp transport is
2981 (i) Exim was not creating retry information for temporary errors given
2982 for individual recipients after the DATA command when the smtp transport
2983 was used in LMTP mode. This meant that they could be retried too
2984 frequently, and not timed out correctly.
2986 (ii) Exim was setting the flag that allows error details to be returned
2987 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2988 recipients that were returned after the DATA command.
2990 PH/53 This is related to PH/52, but is more general: for any failing address,
2991 when detailed error information was permitted to be returned to the
2992 sender, but the error was temporary, then after the final timeout, only
2993 "retry timeout exceeded" was returned. Now it returns the full error as
2994 well as "retry timeout exceeded".
2996 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2997 do this, and (what is worse) MTAs that accept it.
2999 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3000 will now be deprecated.
3002 PH/56 New os.c-cygwin from the Cygwin maintainer.
3004 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3005 formats (previously only available in exim formats via -bpr, -bpru,
3006 and -bpra. Now also available in native and exiqgrep formats)
3008 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3009 with very large, slow to parse queues
3011 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3013 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3015 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3016 responses to authentication challenges, though it was showing the
3017 challenges; (ii) I've removed the CR characters from the debug output for
3020 PH/58 Allow for the insertion of a newline as well as a space when a string
3021 is turned into more than one encoded-word during RFC 2047 encoding. The
3022 Sieve code now uses this.
3024 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3025 data_4xx, lost_connection, tls_required.
3027 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3028 message was being sent as an SMTP response.
3030 PH/61 Add -l and -k options to exicyclog.
3032 PH/62 When verifying, if an address was redirected to one new address, so that
3033 verification continued, and the new address failed or deferred after
3034 having set something in $address_data, the value of $address_data was not
3035 passed back to the ACL. This was different to the case when no
3036 redirection occurred. The value is now passed back in both cases.
3038 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3039 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3040 use login_cap.h, so on its own it isn't the distinguishing feature. The
3041 new name refers directly to the setclassresources() function.
3043 PH/65 Added configuration files for NetBSD3.
3045 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3047 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3048 is preferred over IPv4.
3050 PH/68 The bounce_return_message and bounce_return_body options were not being
3051 honoured for bounces generated during the reception of non-SMTP messages.
3052 In particular, this applied to messages rejected by the ACL. This bug has
3053 been fixed. However, if bounce_return_message is true and bounce_return_
3054 body is false, the headers that are returned for a non-SMTP message
3055 include only those that have been read before the error was detected.
3056 (In the case of an ACL rejection, they have all been read.)
3058 PH/69 The HTML version of the specification is now built in a directory called
3059 spec_html instead of spec.html, because the latter looks like a path with
3060 a MIME-type, and this confuses some software.
3062 PH/70 Catch two compiler warnings in sieve.c.
3064 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3065 function verify_get_ident() calls ip_connect() to connect a socket, but
3066 if the "connect()" function timed out, ip_connect() used to close the
3067 socket. However, verify_get_ident() also closes the socket later, and in
3068 between Exim writes to the log, which may get opened at this point. When
3069 the socket was closed in ip_connect(), the log could get the same file
3070 descriptor number as the socket. This naturally causes chaos. The fix is
3071 not to close the socket in ip_connect(); the socket should be closed by
3072 the function that creates it. There was only one place in the code where
3073 this was missing, in the iplookup router, which I don't think anybody now
3074 uses, but I've fixed it anyway.
3076 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3077 well as to direct DNS lookups. Otherwise the handling of names in host
3078 lists is inconsistent and therefore confusing.
3084 PH/01 Two changes to the default runtime configuration:
3086 (1) Move the checks for relay_from_hosts and authenticated clients from
3087 after to before the (commented out) DNS black list checks.
3089 (2) Add control=submission to the relay_from_hosts and authenticated
3090 clients checks, on the grounds that messages accepted by these
3091 statements are most likely to be submissions.
3093 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3095 (1) Generate an error if the third argument for the ${prvs expansion is
3098 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3101 (3) Reset the variables that are obtained from the first argument of
3102 ${prvscheck and used in the second argument before leaving the code,
3103 because their memory is reclaimed, so using them afterwards may do
3106 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3107 one (it's much easier than Tom thought :-).
3109 (5) Because of (4), we can now allow for the use of $prvscheck_result
3110 inside the third argument.
3112 PH/03 For some reason, the default setting of PATH when running a command from
3113 a pipe transport was just "/usr/bin". I have changed it to
3116 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3117 anything to be listed in the output from -bV.
3119 PH/05 When a filter generated an autoreply, the entire To: header line was
3120 quoted in the delivery log line, like this:
3122 => >A.N.Other <ano@some.domain> <original@ddress> ...
3124 This has been changed so that it extracts the operative address. There
3125 may be more than one such address. If so, they are comma-separated, like
3128 => >ano@some.domain,ona@other.domain <original@ddress> ...
3130 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3131 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3132 not being looked up in the rDNS to get a host name, Exim was showing the
3133 IP address twice in Received: lines, even though the IP addresses were
3134 identical. For example:
3136 Received: from [1.2.3.4] (helo=[1.2.3.4])
3138 However, if the real host name was known, it was omitting the HELO data
3139 if it matched the actual IP address. This has been tidied up so that it
3140 doesn't show the same IP address twice.
3142 PH/07 When both +timestamp and +memory debugging was on, the value given by
3143 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3144 called by the debug printing, thereby overwriting the timestamp buffer.
3145 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3147 PH/08 When the original message was included in an autoreply transport, it
3148 always said "this is a copy of the message, including all the headers",
3149 even if body_only or headers_only was set. It now gives an appropriate
3152 PH/09 Applied a patch from the Sieve maintainer which:
3154 o fixes some comments
3155 o adds the (disabled) notify extension core
3156 o adds some debug output for the result of if/elsif tests
3157 o points to the current vacation draft in the documentation
3158 and documents the missing references header update
3162 o fixes a bug in processing the envelope test (when testing
3163 multiple envelope elements, the last element determinted the
3166 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3167 Electronic Mail") by including:
3169 Auto-submitted: auto-generated
3171 in the messages that it generates (bounce messages and others, such as
3172 warnings). In the case of bounce messages for non-SMTP mesages, there was
3173 also a typo: it was using "Auto_submitted" (underscore instead of
3174 hyphen). Since every message generated by Exim is necessarily in response
3175 to another message, thes have all been changed to:
3177 Auto-Submitted: auto-replied
3179 in accordance with these statements in the RFC:
3181 The auto-replied keyword:
3183 - SHOULD be used on messages sent in direct response to another
3184 message by an automatic process,
3186 - MUST NOT be used on manually-generated messages,
3188 - MAY be used on Delivery Status Notifications (DSNs) and Message
3189 Disposition Notifications (MDNs),
3191 - MUST NOT be used on messages generated by automatic or periodic
3192 processes, except for messages which are automatic responses to
3195 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3196 to the default Received: header definition.
3198 PH/12 Added log selector acl_warn_skipped (default on).
3200 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3201 variables because (a) they are in the wrong storage pool and (b) even if
3202 they were copied, it wouldn't work properly because of the caching.
3204 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3205 checking when decoding. Apparently there are clients that generate
3206 overlong encoded strings. Why am I not surprised?
3208 PH/15 If the first argument of "${if match_address" was not empty, but did not
3209 contain an "@" character, Exim crashed. Now it writes a panic log message
3210 and treats the condition as false.
3212 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3214 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3215 extension "envelope-auth". The code is finished and in agreement with
3216 other implementations, but there is no documentation so far and in fact,
3217 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3218 not changing the active code.
3220 Print executed "if" and "elsif" statements when debugging is used. This
3221 helps a great deal to understand what a filter does.
3223 Document more things not specified clearly in RFC3028. I had all this
3224 sorted out, when out of a sudden new issues came to my mind. Oops."
3226 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3229 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3230 canonical form (as documented). However, after a host name lookup from
3231 the IP address, check_host() was doing a simple string comparison with
3232 addresses acquired from the DNS when checking that the found name did
3233 have the original IP as one of its addresses. Since any found IPv6
3234 addresses are likely to be in abbreviated form, the comparison could
3235 fail. Luckily, there already exists a function for doing the comparison
3236 by converting both addresses to binary, so now that is used instead of
3237 the text comparison.
3239 PH/20 There was another similar case to PH/19, when a complete host name was
3240 given in a host list; looking up its IP address could give an abbreviated
3241 form, whereas the current host's name might or might not be abbreviated.
3242 The same fix has been applied.
3248 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3249 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3252 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3253 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3255 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3257 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3258 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3259 Also, for "Auto-Submitted", treat anything other than "no" as
3260 non-personal, in accordance with RFC 3834. (Previously it treated
3261 anything starting "auto-" as non-personal.)
3263 TF/01 The control=submission/name=... option had a problem with syntax
3264 errors if the name included a slash character. The /name= option
3265 now slurps the rest of the string, so it can include any characters
3266 but it must come last in the list of options (after /sender_retain
3269 PH/05 Some modifications to the interface to the fake nameserver for the new
3277 TK/01 Added the "success_on_redirect" address verification option. See
3278 NewStuff for rationale and an example.
3280 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3282 PH/02 Patch to exigrep to allow it to work on syslog lines.
3284 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3285 fread() to skip over the body file's header line, because in Cygwin the
3286 header line is locked and is inaccessible.
3288 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3289 co-exist for some time) to make it clear that it is the Exim ID that is
3290 referenced, not the Message-ID: header line.
3292 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3293 string_format() function, because snprintf() does not exist on all
3296 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3297 ${stat: expansion item.
3299 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3300 protocol synchronization error", to keep the pedants happy.
3302 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3303 well as for IRIX systems, when gcc is being used. See the host.c source
3306 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3308 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3311 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3312 transport and to the smtp transport in LMTP mode.
3314 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3316 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3317 run a filter in a subprocess. This could lead to confusion in subsequent
3318 lookups in the parent process. There should also be a search_tidyup() at
3319 the end of the subprocess.
3321 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3322 only if the host matched helo_try_verify_hosts, which caused the
3323 verification to occur when the EHLO/HELO command was issued. The ACL just
3324 tested the remembered result. Now, if a previous verification attempt has
3325 not happened, "verify = helo" does it there and then.
3327 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3329 TK/03 Fix log output including CR from clamd.
3331 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3332 exist provoked a memory error which could cause a segfault.
3334 PH/15 Installed PCRE 6.2
3336 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3338 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3339 of the problem. Specifically, suggested +O2 rather than +O1 for the
3342 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3344 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3345 record, the port was not correctly logged when the outgoing_port log
3346 selector was set (it logged the transort's default port).
3348 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3349 fallback_hosts, and "hosts" in the smtp transport.
3351 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3352 host errors such as "Connection refused".
3354 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3355 authentication with radiusclient 0.4.9:
3357 - Error returned from rc_read_config was caught wrongly
3358 - Username/password not passed on to radius server due to wrong length.
3360 The presumption is that some radiusclient API changes for 4.51/PH/17
3361 were not taken care of correctly. The code is still untested by me (my
3362 Linux distribution still has 0.3.2 of radiusclient), but it was
3363 contributed by a Radius user.
3365 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3366 expanding the "port" option of the smtp transport.
3368 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3369 while reading a MIME header. Thanks to Tom Hughes for a patch.
3371 PH/24 Include config.h inside local_scan.h so that configuration settings are
3374 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3375 that in an ACL for RCPT (for example), you can examine exactly what was
3378 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3379 commands, but it was not correctly comparing the address with the actual
3380 client host address. Thus, it would show the EHLO address in Received:
3381 header lines when this was not necessary.
3383 PH/27 Added the % operator to ${eval:}.
3385 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3386 it should be ignoring failures (because with -C, for example, it has lost
3387 privilege). It wasn't ignoring creation failures other than "already
3390 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3393 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3394 because an input line was too long, either on its own, or by virtue of
3395 too many continuations, the temporary file was not being removed, and the
3396 return code was incorrect.
3398 PH/31 Missing "BOOL" in function definition in filtertest.c.
3400 PH/32 Applied Sieve patches from the maintainer.
3402 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3404 PH/33 Added "verify = not_blind".
3406 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3407 Local/Makefile (with some defaults set). These are used in built scripts
3408 such as exicyclog, but they have never been used in the exim_install
3409 script (though there are many overriding facilities there). I have
3410 arranged that the exim_install script now takes note of these two
3413 PH/35 Installed configuration files for Dragonfly.
3415 PH/36 When a locally submitted message by a trusted user did not contain a
3416 From: header, and the sender address was obtained from -f or from an SMTP
3417 MAIL command, and the trusted user did not use -F to supply a sender
3418 name, $originator_name was incorrectly used when constructing a From:
3419 header. Furthermore, $originator_name was used for submission mode
3420 messages from external hosts without From: headers in a similar way,
3421 which is clearly wrong.
3423 PH/37 Added control=suppress_local_fixups.
3425 PH/38 When log_selector = +received_sender was set, and the addition of the
3426 sender made the log line's construction buffer exactly full, or one byte
3427 less than full, an overflow happened when the terminating "\n" was
3430 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3431 when the result of a list match is failure because a DNS lookup failed.
3433 PH/40 RM_COMMAND is now used in the building process.
3435 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3436 the "build-* directories that it finds.
3438 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3439 address in a domain literal was a prefix of an interface address.
3441 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3442 when verifying a sender address, unless rewrite_headers is false.
3444 PH/44 Wrote a long comment about why errors_to addresses are verified as
3445 recipients, not senders.
3447 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3448 the ratelimit ACL was added.
3450 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3452 PH/46 Added extra information about PostgreSQL errors to the error string.
3454 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3455 avoiding the need to install special zones in a real server. This is
3456 backwards compatible; if it can't find the fake resolver, it drops back.
3457 Thus, both old and new test suites can be run.
3459 TF/02 Added util/ratelimit.pl
3461 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3462 clock is set back in time.
3464 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3465 Candler <B.Candler@pobox.com>.
3467 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3468 OK for addresses that are the result of redirections.
3470 PH/48 A number of further additions for the benefit of the new test suite,
3471 including a fake gethostbyname() that interfaces to the fake DNS resolver
3474 TF/06 The fix for widen_domains has also been applied to qualify_single and
3475 search_parents which are the other dnslookup options that can cause
3478 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3481 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3483 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3484 service" if it encountered an MX record with an empty target hostname.
3485 The message is now "an MX or SRV record indicated no SMTP service".
3487 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3488 if the DNS of the sending site is misconfigured. This is quite a
3489 common situation. This change restores the behaviour of treating a
3490 helo verification defer as a failure.
3492 PH/51 If self=fail was set on a router, the bounce message did not include the
3493 actual error message.
3499 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3501 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3502 command itself ended in error, the underlying message about the transport
3503 filter timeout was being overwritten with the pipe command error. Now the
3504 underlying error message should be appended to the second error message.
3506 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3508 PH/02 Reduce the amount of output that "make" produces by default. Full output
3509 can still be requested.
3511 PH/03 The warning log line about a condition test deferring for a "warn" verb
3512 was being output only once per connection, rather than after each
3513 occurrence (because it was using the same function as for successful
3514 "warn" verbs). This seems wrong, so I have changed it.
3516 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3517 it should not have, which might have caused a crash in the right
3518 circumstances, but probably never did.
3520 PH/04 Installed a modified version of Tony Finch's patch to make submission
3521 mode fix the return path as well as the Sender: header line, and to
3522 add a /name= option so that you can make the user's friendly name appear
3525 TF/03 Added the control = fakedefer ACL modifier.
3527 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3528 Mark Lowes for thorough testing.
3530 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3532 TK/03 Merged latest SRS patch from Miles Wilton.
3534 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3535 with the definition in sysexits.h (which is #included earlier).
3536 Fortunately, Exim does not actually use EX_OK. The code used to try to
3537 preserve the sysexits.h value, by assumimg that macro definitions were
3538 scanned for macro replacements. I have been disabused of this notion,
3539 so now the code just undefines EX_OK before #including unistd.h.
3541 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3542 in the smtp transport. When a block could not be written in a single
3543 write() function, the timeout was being re-applied to each part-write.
3544 This seems wrong - if the receiver was accepting one byte at a time it
3545 would take for ever. The timeout is now adjusted when this happens. It
3546 doesn't have to be particularly precise.
3548 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3549 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3551 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3552 without a domain if the check to <postmaster@domain> fails.
3554 SC/01 Eximstats: added -xls and the ability to specify output files
3555 (patch written by Frank Heydlauf).
3557 SC/02 Eximstats: use FileHandles for outputing results.
3559 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3561 SC/04 Eximstats: fixed display of large numbers with -nvr option
3563 SC/05 Eximstats: fixed merging of reports with empty tables.
3565 SC/06 Eximstats: added the -include_original_destination flag
3567 SC/07 Eximstats: removed tabs and trailing whitespace.
3569 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3571 TK/06 MBOX spool code: Add real "From " MBOX separator line
3572 so the .eml file is really in mbox format (even though
3573 most programs do not really care). Patch from Alex Miller.
3575 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3576 The latter is generated from $received_to and is only set if the
3577 message has one envelope recipient. SA can use these headers,
3578 obviously out-of-the-box. Patch from Alex Miller.
3580 PH/08 The ${def test on a variable was returning false if the variable's
3581 value was "0", contrary to what the specification has always said!
3582 The result should be true unless the variable is empty.
3584 PH/09 The syntax error of a character other than { following "${if
3585 def:variable_name" (after optional whitespace) was not being diagnosed.
3586 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3587 accidental colon was present, for example, could give incorrect results.
3589 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3590 result is used (not including appendfile, where other changes are about
3593 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3594 This involved changing a lot of size variables from int to off_t. It
3595 should work with maildirs and everything.
3597 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3598 spamd dying while we are connected to it.
3600 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3603 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3604 function for BDB 4.3.
3606 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3608 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3609 the sources for PCRE have been re-arranged and more files are now
3612 PH/15 The code I had for printing potentially long long variables in PH/11
3613 above was not the best (it lost precision). The length of off_t variables
3614 is now inspected at build time, and an appropriate printing format (%ld
3615 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3616 to be "long long int" or "long int". This is needed for the internal
3617 formatting function string_vformat().
3619 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3620 the configuration file to be ":syslog", then the script "guesses" where
3621 the logs files are, rather than using the compiled in default. In our
3622 case the guess is not the same as the compiled default, so the script
3623 suddenly stopped working when I started to use syslog. The patch checks
3624 to see if log_file_path is "". If so, it attempts to read it from exim
3625 with no configuration file to get the compiled in version, before it
3626 falls back to the previous guessing code."
3628 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3629 implementing BATV in an Exim configuration. See NewStuff for the gory
3632 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3633 Makefile that are specific to HP-UX.
3635 PH/18 If the "use_postmaster" option was set for a recipient callout together
3636 with the "random" option, the postmaster address was used as the MAIL
3637 FROM address for the random test, but not for the subsequent recipient
3638 test. It is now used for both.
3640 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3641 patch removes a few documentation additions to RFC 3028, because the
3642 latest draft now contains them. It adds the new en;ascii-case comparator
3643 and a new error check for 8bit text in MIME parts. Comparator and
3644 require names are now matched exactly. I enabled the subaddress
3645 extension, but it is not well tested yet (read: it works for me)."
3647 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3648 rework some of the code of TK/09 above to avoid the hardwired use of
3649 "%lld" and "long long". Replaced the call to snprintf() with a call to
3652 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3653 records point to non-existent hosts", "retry timeout exceeded", and
3654 "retry time not reached for any host after a long failure period".
3656 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3657 experimental DomainKeys support:
3659 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3660 (2) On an error such as an illegally used "control", the wrong name for
3661 the control was given.
3663 These problems did NOT occur unless DomainKeys support was compiled.
3665 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3667 PH/24 Added ${if match_ip condition.
3669 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3670 databases so that it will be absolutely obvious if a crash occurs in the
3671 DB library. This is a regular occurrence (often caused by mis-matched
3674 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3675 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3676 picked up on a user's system that detects such things. There doesn't seem
3677 to be a gcc warning option for this - only an attribute that has to be
3678 put on the function's prototype. It seems that in Fedora Core 4 they have
3679 set this on a number of new functions. No doubt there will be more in due
3682 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3683 specify a transport. However, if an address that was verified by such a
3684 router was the subject of a callout, Exim crashed because it tried to
3685 read the rcpt_include_affixes from the non-existent transport. Now it
3686 just assumes that the setting of that option is false. This bug was
3687 introduced by 4.51/PH/31.
3689 PH/28 Changed -d+all to exclude +memory, because that information is very
3690 rarely of interest, but it makes the output a lot bigger. People tend to
3691 do -d+all out of habit.
3693 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3694 code in os-type was giving problems when libc.so lives in lib64, like on
3697 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3698 aren't the modern standard, and it seems that some systems' include files
3699 don't always have them. Exim was already checking for some of the newer
3700 ones like T_AAAA, and defining it itself. I've added checks for all the
3701 record types that Exim uses.
3703 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3704 not automatically generating a new one, as it is supposed to. This
3705 prevented TLS from working. If the file did exist, but contained invalid
3706 data, a new version was generated, as expected. It was only the case of a
3707 non-existent file that was broken.
3709 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3710 with a change in libdomainkeys > 0.64.
3712 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3713 from DNS. If the selector record carries the flag, it now has
3714 precedence over the domain-wide flag.
3716 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3718 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3719 the use of a port name that isn't defined in /etc/services) occurred, the
3720 message was deferred as in a normal delivery, and thus remained on the
3721 spool, instead of being failed because of the mua_wrapper setting. This
3722 is now fixed, and I tidied up some of the mua_wrapper messages at the
3725 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3726 the messages in a hash of arrays rather than using individual hashes.
3727 This is a bit cleaner and results in dramatic memory savings, albeit
3728 at a slight CPU cost.
3730 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3731 as requested by Marc Sherman.
3733 SC/10 Eximstats: added histograms for user specified patterns as requested
3736 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3738 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3739 fopen() in the content-scanning modules that did not already have it.
3745 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3746 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3748 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3750 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3752 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3753 to test Sieve filters that use "vacation".
3755 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3756 that changes the way the GnuTLS parameters are stored in the cache file.
3757 The new format can be generated externally. For backward compatibility,
3758 if the data in the cache doesn't make sense, Exim assumes it has read an
3759 old-format file, and it generates new data and writes a new file. This
3760 means that you can't go back to an older release without removing the
3763 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3764 work if there are any delivery delays because "one_time" forces the
3765 parent to be marked "delivered", so its unseen clone is never tried
3766 again. For this reason, Exim now forbids the simultaneous setting of
3769 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3770 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3771 entry if you want to know the details. The fix, however, neglected to
3772 consider the case when local delivery batching is involved. The test for
3773 "previously delivered" was not happening when checking to see if an
3774 address could be batched with a previous (undelivered) one; under
3775 certain circumstances this could lead to multiple deliveries to the same
3778 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3779 in its include files, and this causes problems building Exim.
3781 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3782 header_syntax) but Exim was just ignoring anything given after a slash.
3783 In particular, this caused confusion with an attempt to use "verify =
3784 reverse_host_lookup/defer_ok". An error is now given when options are
3785 supplied for verify items that do not have them. (Maybe reverse_host_
3786 lookup should have a defer_ok option, but that's a different point.)
3788 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3789 defined by RFC 821) to 2048, because there were problems with some AUTH
3790 commands, and RFC 1869 says the size should be increased for extended
3791 SMTP commands that take arguments.
3793 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3796 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3797 "unknown" error; now it says that the functionality isn't in the binary.
3799 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3800 an address' error message when a string expansion fails (syntax or
3801 whatever). Otherwise the password may appear in the log. Following change
3802 PH/42 below, there is no longer a chance of it appearing in a bounce
3805 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3807 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3808 of its addresses was ever tried. (Bugzilla bug #2.)
3810 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3811 the result incorrectly in the debug output. (It correctly added a newline
3812 to what was transported.)
3814 TF/01 Added $received_time.
3816 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3817 commented out examples of how to interface to a virus scanner and to
3818 SpamAssassin. Also added commented examples of av_scanner and
3819 spamd_address settings.
3821 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3822 and controls are allowed in which ACLs. There were a couple of minor
3823 errors. Some of the entries in the conditions table (which is a table of
3824 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3825 negation of where the condition IS allowed.
3827 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3829 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3830 header file does not have a version number, so I've had to invent a new
3831 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3832 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3833 radiusclient), but it was contributed by a Radius user.
3835 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3836 files or directories, for OpenSSL.
3838 PH/19 When an Exim process that is running as root has to create an Exim log
3839 file, it does so in a subprocess that runs as exim:exim so as to get the
3840 ownership right at creation (otherwise, other Exim processes might see
3841 the file with the wrong ownership). There was no test for failure of this
3842 fork() call, which would lead to the process getting stuck as it waited
3843 for a non-existent subprocess. Forks do occasionally fail when resources
3844 run out. I reviewed all the other calls to fork(); they all seem to check
3847 PH/20 When checking for unexpected SMTP input at connect time (before writing
3848 the banner), Exim was not dealing correctly with a non-positive return
3849 from the read() function. If the client had disconnected by this time,
3850 the result was a log entry for a synchronization error with an empty
3851 string after "input=" when read() returned zero. If read() returned -1
3852 (an event I could not check), uninitialized data bytes were printed.
3853 There were reports of junk text (parts of files, etc) appearing after
3856 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3858 PH/22 Added support for macro redefinition, and (re)definition in between
3859 driver and ACL definitions.
3861 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3862 forgetting to use the resulting value; it was using the unexpanded value.
3864 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3865 hadn't been configured. The fix is from Juergen Kreileder, who
3866 understands it better than I do:
3868 "Here's what I see happening with three configured cyrus_sasl
3869 authenticators configured (plain, login, cram-md5):
3871 On startup auth_cyrus_sasl_init() gets called for each of these.
3872 This means three calls to sasl_listmech() without a specified mech_list.
3873 => SASL tests which mechs of all available mechs actually work
3874 => three warnings about OTP not working
3875 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3877 With the patch, sasl_listmech() also gets called three times. But now
3878 SASL's mech_list option is set to the server_mech specified in the the
3879 authenticator. Or in other words, the answer from sasl_listmech()
3880 gets limited to just the mech you're testing for (which is different
3882 => the return list contains just 'plain' or 'login', 'cram-md5' or
3883 nothing depending on the value of ob->server_mech.
3885 I've just tested the patch: Authentication still works fine,
3886 unavailable mechs specified in the exim configuration are still
3887 caught, and the auth.log warnings about OTP are gone."
3889 PH/25 When debugging is enabled, the contents of the command line are added
3890 to the debugging output, even when log_selector=+arguments is not
3893 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3894 answer is "GNU", and only if the return is "GNU/something" is the answer
3897 PH/27 $acl_verify_message is now set immediately after the failure of a
3898 verification in an ACL, and so is available in subsequent modifiers. In
3899 particular, the message can be preserved by coding like this:
3901 warn !verify = sender
3902 set acl_m0 = $acl_verify_message
3904 Previously, $acl_verify_message was set only while expanding "message"
3905 and "log_message" when a very denied access.
3907 PH/28 Modified OS/os.c-Linux with
3909 -#ifndef OS_LOAD_AVERAGE
3910 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3912 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3913 nomenclature these days.)
3915 PH/29 Installed patch from the Sieve maintainer that adds the options
3916 sieve_useraddress and sieve_subaddress to the redirect router.
3918 PH/30 In these circumstances:
3919 . Two addresses routed to the same list of hosts;
3920 . First host does not offer TLS;
3921 . First host accepts first address;
3922 . First host gives temporary error to second address;
3923 . Second host offers TLS and a TLS session is established;
3924 . Second host accepts second address.
3925 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3926 and peerdn, if requested) that were in fact used only for the second
3929 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3930 attention to any local part prefix or suffix that was matched by the
3931 router that accepted the address. It now behaves in the same way as it
3932 does for delivery: the affixes are removed from the local part unless
3933 rcpt_include_affixes is set on the transport.
3935 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3936 timeout during the DATA phase of an incoming message.
3938 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3939 applied a patch sanctioned by the Sieve maintainer.
3941 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3942 the uid or gid is negative. A case of a negative gid caused this to be
3943 noticed. The fix allows for either to be negative.
3945 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3946 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3949 PH/36 The change PH/12 above was broken. Fixed it.
3951 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3952 the grounds that routing the same address twice would always produce the
3953 same answer. This might have been true once, but it is certainly no
3954 longer true now. Routing a child address may depend on the previous
3955 routing that produced that child. Some complicated redirection strategies
3956 went wrong when messages had multiple recipients, and made Exim's
3957 behaviour dependent on the order in which the addresses were given.
3959 I have moved the duplicate checking until after the routing is complete.
3960 Exim scans the addresses that are assigned to local and remote
3961 transports, and removes any duplicates. This means that more work will be
3962 done, as duplicates will always all be routed, but duplicates are
3963 presumably rare, so I don't expect this is of any significance.
3965 For deliveries to pipes, files, and autoreplies, the duplicate checking
3966 still happens during the routing process, since they are not going to be
3969 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3970 It corrects a timeout issue with spamd. This is Ian's comment: "The
3971 background is that sometimes spamd either never reads data from a
3972 connection it has accepted, or it never writes response data. The exiscan
3973 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3974 blindly assumes that writes won't block so it may never time out."
3976 PH/39 Allow G after quota size as well as K and M.
3978 PH/40 The value set for $authenticated_id in an authenticator may not contain
3979 binary zeroes or newlines because the value is written to log lines and
3980 to spool files. There was no check on this. Now the value is run through
3981 the string_printing() function so that such characters are converted to
3982 printable escape sequences.
3984 PH/41 $message_linecount is a new variable that contains the total number of
3985 lines in the message. Compare $body_linecount, which is the count for the
3988 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3989 bounce and delay warning messages, except in certain special cases, which
3992 (a) An SMTP error message from a remote host;
3993 (b) A message specified in a :fail: redirection;
3994 (c) A message specified in a "fail" command in a system filter;
3995 (d) A message specified in a FAIL return from the queryprogram router;
3996 (e) A message specified by the cannot_route_message router option.
3998 In these cases only, Exim does include the error details in bounce and
3999 warning messages. There are also a few cases where bland messages such
4000 as "unrouteable address" or "local delivery error" are given.
4002 PH/43 $value is now also set for the "else" part of a ${run expansion.
4004 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4005 being worked on, but at least Exim now implements the latest version to
4008 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4009 process to complete was treated as a delivery failure, a timeout while
4010 writing the message to the pipe was logged, but erroneously treated as a
4011 successful delivery. Such timeouts include transport filter timeouts. For
4012 consistency with the overall process timeout, these timeouts are now
4013 treated as errors, giving rise to delivery failures by default. However,
4014 there is now a new Boolean option for the pipe transport called
4015 timeout_defer, which, if set TRUE, converts the failures into defers for
4016 both kinds of timeout. A transport filter timeout is now identified in
4019 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4020 systems where "make" and "gmake" are different, calling "gmake" at top
4021 level broke things. I've arranged for the value of $(MAKE) to be passed
4022 from the Makefile to this script so that it can call the same version of
4026 A note about Exim versions 4.44 and 4.50
4027 ----------------------------------------
4029 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4030 changes of various kinds. As a consequence, a big documentation update was
4031 needed. This delayed the release for rather longer than seemed good, especially
4032 in the light of a couple of (minor) security issues. Therefore, the changes
4033 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4034 release. So 4.44 and 4.50 are in effect two different branches that both start
4037 I have left the 4.50 change log unchanged; it contains all the changes since
4038 4.43. The change log for 4.44 is below; many of its items are identical to
4039 those for 4.50. This seems to be the most sensible way to preserve the
4040 historical information.
4046 1. Minor wording change to the doc/README.SIEVE file.
4048 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4049 computation of the current number of files was incorrect.
4051 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4052 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4055 4. Give more explanation in the error message when the command for a transport
4056 filter fails to execute.
4058 5. There are several places where Exim runs a non-Exim command in a
4059 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4060 was being done only for the command run by the queryprogram router. It is
4061 now done for all such subprocesses. The other cases are: ${run, transport
4062 filters, and the commands run by the lmtp and pipe transports.
4064 6. Added CONFIGURE_GROUP build-time option.
4066 7. Some older OS have a limit of 256 on the maximum number of file
4067 descriptors. Exim was using setrlimit() to set 1000 as a large value
4068 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4069 systems. I've change it so that if it can't get 1000, it tries for 256.
4071 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4072 was an oversight, and furthermore, ever since the addition of extra
4073 controls (e.g. 4.43/32), the checks on when to allow different forms of
4074 "control" were broken. There should now be diagnostics for all cases when a
4075 control that does not make sense is encountered.
4077 9. Added the /retain_sender option to "control=submission".
4079 10. $recipients is now available in the predata ACL (oversight).
4081 11. Tidy the search cache before the fork to do a delivery from a message
4082 received from the command line. Otherwise the child will trigger a lookup
4083 failure and thereby defer the delivery if it tries to use (for example) a
4084 cached ldap connection that the parent has called unbind on.
4086 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4087 of $address_data from the recipient verification was clobbered by the
4088 sender verification.
4090 13. The value of address_data from a sender verification is now available in
4091 $sender_address_data in subsequent conditions in the ACL statement.
4093 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4095 15. Added a new option "connect=<time>" to callout options, to set a different
4098 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4099 was its contents. (It was OK if the option was not defined at all.)
4101 17. A "Completed" log line is now written for messages that are removed from
4102 the spool by the -Mrm option.
4104 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4105 information about exactly what failed.
4107 19. Added -dd to debug only the daemon process.
4109 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4110 handles timeouts, both on the server side and network timeouts. Renamed the
4111 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4113 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4114 It is now set to "smtps".
4116 22. $host_address is now set to the target address during the checking of
4117 ignore_target_hosts.
4119 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4120 being passed; this would have caused $sender_host_name to have been used if
4121 matching the list had actually called for a host name (not very likely,
4122 since this list is usually IP addresses). A host name is now passed as
4125 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4126 code that specifies a non-restarting handler (typically sigaction() in
4127 modern systems) in an attempt to fix a rare and obscure crash bug.
4129 25. Narrowed the window for a race in the daemon that could cause it to ignore
4130 SIGCHLD signals. This is not a major problem, because they are used only to
4131 wake it up if nothing else does.
4133 26. A malformed maildirsize file could cause Exim to calculate negative values
4134 for the mailbox size or file count. Odd effects could occur as a result.
4135 The maildirsize information is now recalculated if the size or filecount
4138 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4139 support for a long time. Removed HAVE_SYS_VFS_H.
4141 28. Installed the latest version of exipick from John Jetmore.
4143 29. In an address list, if the pattern was not a regular expression, an empty
4144 subject address (from a bounce message) matched only if the pattern was an
4145 empty string. Non-empty patterns were not even tested. This was the wrong
4146 because it is perfectly reasonable to use an empty address as part of a
4147 database query. An empty address is now tested by patterns that are
4148 lookups. However, all the other forms of pattern expect the subject to
4149 contain a local part and a domain, and therefore, for them, an empty
4150 address still always fails if the pattern is not itself empty.
4152 30. Exim went into a mad DNS loop when attempting to do a callout where the
4153 host was specified on an smtp transport, and looking it up yielded more
4154 than one IP address.
4156 31. Re-factored the code for checking spool and log partition space into a
4157 function that finds that data and another that does the check. The former
4158 is then used to implement four new variables: $spool_space, $log_space,
4159 $spool_inodes, and $log_inodes.
4161 32. The RFC2047 encoding function was originally intended for short strings
4162 such as real names; it was not keeping to the 75-character limit for
4163 encoded words that the RFC imposes. It now respects the limit, and
4164 generates multiple encoded words if necessary. To be on the safe side, I
4165 have increased the buffer size for the ${rfc2047: expansion operator from
4168 33. It is now permitted to omit both strings after an "if" condition; if the
4169 condition is true, the result is "true". As before, when the second string
4170 is omitted, a false condition yields an empty string. This makes it less
4171 cumbersome to write custom ACL and router conditions.
4173 34. Failure to deliver a bounce message always caused it to be frozen, even if
4174 there was an errors_to setting on the router. The errors_to setting is now
4177 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4178 canonical form (fully expanded) before being placed in
4179 $sender_host_address.
4181 36. The table in the code that translates DNS record types into text (T_A to
4182 "A" for instance) was missing entries for NS and CNAME. It is just possible
4183 that this could have caused confusion if both these types were looked up
4184 for the same domain, because the text type is used as part of Exim's
4185 per-process caching. But the chance of anyone hitting this buglet seems
4188 37. The dnsdb lookup has been extended in a number of ways.
4190 (1) There is a new type, "zns", which walks up the domain tree until it
4191 finds some nameserver records. It should be used with care.
4193 (2) There is a new type, "mxh", which is like "mx" except that it returns
4194 just the host names, not the priorities.
4196 (3) It is now possible to give a list of domains (or IP addresses) to be
4197 looked up. The behaviour when one of the lookups defers can be
4198 controlled by a keyword.
4200 (4) It is now possible to specify the separator character for use when
4201 multiple records are returned.
4203 38. The dnslists ACL condition has been extended: it is now possible to supply
4204 a list of IP addresses and/or domains to be looked up in a particular DNS
4207 39. Added log_selector=+queue_time_overall.
4209 40. When running the queue in the test harness, wait just a tad after forking a
4210 delivery process, to get repeatability of debugging output.
4212 41. Include certificate and key file names in error message when GnuTLS fails
4213 to set them up, because the GnuTLS error message doesn't include the name
4214 of the failing file when there is a problem reading it.
4216 42. Allow both -bf and -bF in the same test run.
4218 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4220 44. The "Exiscan patch" is now merged into the mainline Exim source.
4222 45. Sometimes the final signoff response after QUIT could fail to get
4223 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4224 < 0 before doing a fflush(). This bug looks as though it goes back to the
4225 introduction of TLS in release 3.20, but "sometimes" must have been rare
4226 because the tests only now provoked it.
4228 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4229 (this can affect the format of dates).
4231 47. exim_tidydb, when checking for the continued existence of a message for
4232 which it has found a message-specific retry record, was not finding
4233 messages that were in split spool directories. Consequently, it was
4234 deleting retry records that should have stayed in existence.
4236 48. Steve fixed some bugs in eximstats.
4238 49. The SPA authentication driver was not abandoning authentication and moving
4239 on to the next authenticator when an expansion was forced to fail,
4240 contradicting the general specification for all authenticators. Instead it
4241 was generating a temporary error. It now behaves as specified.
4243 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4244 (the order specifies the preference for clients). The order is now AES256,
4245 AES128, 3DES, ARCFOUR128.
4247 51. Small patch to Sieve code - explicitly set From: when generating an
4250 52. Exim crashed if a remote delivery caused a very long error message to be
4251 recorded - for instance if somebody sent an entire SpamAssassin report back
4252 as a large number of 550 error lines. This bug was coincidentally fixed by
4253 increasing the size of one of Exim's internal buffers (big_buffer) that
4254 happened as part of the Exiscan merge. However, to be on the safe side, I
4255 have made the code more robust (and fixed the comments that describe what
4258 53. Now that there can be additional text after "Completed" in log lines (if
4259 the queue_time_overall log selector is set), a one-byte patch to exigrep
4260 was needed to allow it to recognize "Completed" as not the last thing in
4263 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4264 patch that reportedly fixes this has been added. I am not expert enough to
4265 create a test for it. This is what the patch creator wrote:
4267 "I found a little strange behaviour of ldap code when working with
4268 Windows 2003 AD Domain, where users was placed in more than one
4269 Organization Units. When I tried to give exim partial DN, the exit code
4270 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4271 But simultaneously result of request was absolutely normal ldap result,
4272 so I produce this patch..."
4274 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4275 REFERENCE, so I have modified the code to exclude the patch when that macro
4278 55. Some experimental protocols are using DNS PTR records for new purposes. The
4279 keys for these records are domain names, not reversed IP addresses. The
4280 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4281 leaves it alone. Component reversal etc. now happens only for IP addresses.
4284 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4286 57. Double the size of the debug message buffer (to 2048) so that more of very
4287 long debug lines gets shown.
4289 58. The exicyclog utility now does better if the number of log files to keep
4290 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4292 59. Two changes related to the smtp_active_hostname option:
4294 (1) $smtp_active_hostname is now available as a variable.
4295 (2) The default for smtp_banner uses $smtp_active_hostname instead
4296 of $primary_hostname.
4298 60. The host_aton() function is supposed to be passed a string that is known
4299 to be a valid IP address. However, in the case of IPv6 addresses, it was
4300 not checking this. This is a hostage to fortune. Exim now panics and dies
4301 if the condition is not met. A case was found where this could be provoked
4302 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4303 components; fortuitously, this particular loophole had already been fixed
4304 by change 4.50/55 above.
4306 If there are any other similar loopholes, the new check in host_aton()
4307 itself should stop them being exploited. The report I received stated that
4308 data on the command line could provoke the exploit when Exim was running as
4309 exim, but did not say which command line option was involved. All I could
4310 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4311 running as the user.
4314 61. There was a buffer overflow vulnerability in the SPA authentication code
4315 (which came originally from the Samba project). I have added a test to the
4316 spa_base64_to_bits() function which I hope fixes it.
4319 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4320 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4322 63. The daemon start-up calls getloadavg() while still root for those OS that
4323 need the first call to be done as root, but it missed one case: when
4324 deliver_queue_load_max is set with deliver_drop_privilege. This is
4325 necessary for the benefit of the queue runner, because there is no re-exec
4326 when deliver_drop_privilege is set.
4328 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4329 This has been fixed.
4331 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4332 was in use, was not putting the data itself into the right store pool;
4333 consequently, it could be overwritten for a subsequent message in the same
4334 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4337 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4339 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4340 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4341 the function were treating the return as a boolean value, which happened to
4342 work because 0=false and not-0=true, but is not correct code.
4344 68. The host_aton() function was not handling scoped IPv6 addresses (those
4345 with, for example, "%eth0" on the end) correctly.
4347 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4348 negated items (that is, ~something) in unsigned ints. Some compilers
4349 apparently mutter when there is no cast.
4351 70. If an address verification called from an ACL failed, and did not produce a
4352 user-specific message (i.e. there was only a "system" message), nothing was
4353 put in $acl_verify_message. In this situation, it now puts the system
4356 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4357 session; change 4.31/43 added the unwanted input to the log line - except
4358 that it did not do this in the start of session case. It now does.
4360 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4361 This is wrong and can cause the other end to generate a synchronization
4362 error if it is another Exim or anything else that does the synchronization
4363 check. A QUIT command is no longer sent after a timeout.
4365 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4366 during host lookups.
4368 74. The defer_ok option of callout verification was not working if it was used
4369 when verifying addresses in header lines, that is, for this case:
4371 verify = header_sender/callout=defer_ok
4373 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4374 those file descriptors could be used for SMTP connections. If anything
4375 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4376 could be sent to the SMTP client, causing chaos. The daemon now opens
4377 stdin, stdout, and stderr to /dev/null when it puts itself into the
4380 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4381 log by default. The user can override this with suitable Perl magic.
4383 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4384 the log message when discard triggers, was not working for the DATA ACL or
4385 for the non-SMTP ACL.
4387 78. Error message wording change in sieve.c.
4389 79. If smtp_accept_max_per_host was set, the number of connections could be
4390 restricted to fewer than expected, because the daemon was trying to set up
4391 a new connection before checking whether the processes handling previous
4392 connections had finished. The check for completed processes is now done
4393 earlier. On busy systems, this bug wouldn't be noticed because something
4394 else would have woken the daemon, and it would have reaped the completed
4397 80. If a message was submitted locally by a user whose login name contained one
4398 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4399 It caused a spool format error. I have fixed the spool reading code. A
4400 related problem was that the "from" clause in the Received: line became
4401 illegal because of the space(s). It is now covered by ${quote_local_part.
4403 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4406 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4408 83. Patch from Sieve maintainer to fix unterminated string problem in
4409 "vacation" handling.
4411 84. Some minor changes to the Linux configuration files to help with other
4412 OS variants using glibc.
4414 85. One more patch for Sieve to update vacation handling to latest spec.
4417 ----------------------------------------------------
4418 See the note above about the 4.44 and 4.50 releases.
4419 ----------------------------------------------------
4425 1. Change 4.43/35 introduced a bug that caused file counts to be
4426 incorrectly computed when quota_filecount was set in an appendfile
4429 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4430 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4433 3. Give more explanation in the error message when the command for a transport
4434 filter fails to execute.
4436 4. There are several places where Exim runs a non-Exim command in a
4437 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4438 was being done only for the command run by the queryprogram router. It is
4439 now done for all such subprocesses. The other cases are: ${run, transport
4440 filters, and the commands run by the lmtp and pipe transports.
4442 5. Some older OS have a limit of 256 on the maximum number of file
4443 descriptors. Exim was using setrlimit() to set 1000 as a large value
4444 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4445 systems. I've change it so that if it can't get 1000, it tries for 256.
4447 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4448 was an oversight, and furthermore, ever since the addition of extra
4449 controls (e.g. 4.43/32), the checks on when to allow different forms of
4450 "control" were broken. There should now be diagnostics for all cases when a
4451 control that does not make sense is encountered.
4453 7. $recipients is now available in the predata ACL (oversight).
4455 8. Tidy the search cache before the fork to do a delivery from a message
4456 received from the command line. Otherwise the child will trigger a lookup
4457 failure and thereby defer the delivery if it tries to use (for example) a
4458 cached ldap connection that the parent has called unbind on.
4460 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4461 of $address_data from the recipient verification was clobbered by the
4462 sender verification.
4464 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4465 was its contents. (It was OK if the option was not defined at all.)
4467 11. A "Completed" log line is now written for messages that are removed from
4468 the spool by the -Mrm option.
4470 12. $host_address is now set to the target address during the checking of
4471 ignore_target_hosts.
4473 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4474 being passed; this would have caused $sender_host_name to have been used if
4475 matching the list had actually called for a host name (not very likely,
4476 since this list is usually IP addresses). A host name is now passed as
4479 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4480 code that specifies a non-restarting handler (typically sigaction() in
4481 modern systems) in an attempt to fix a rare and obscure crash bug.
4483 15. Narrowed the window for a race in the daemon that could cause it to ignore
4484 SIGCHLD signals. This is not a major problem, because they are used only to
4485 wake it up if nothing else does.
4487 16. A malformed maildirsize file could cause Exim to calculate negative values
4488 for the mailbox size or file count. Odd effects could occur as a result.
4489 The maildirsize information is now recalculated if the size or filecount
4492 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4493 support for a long time. Removed HAVE_SYS_VFS_H.
4495 18. Updated exipick to current release from John Jetmore.
4497 19. Allow an empty sender to be matched against a lookup in an address list.
4498 Previously the only cases considered were a regular expression, or an
4501 20. Exim went into a mad DNS lookup loop when doing a callout where the
4502 host was specified on the transport, if the DNS lookup yielded more than
4505 21. The RFC2047 encoding function was originally intended for short strings
4506 such as real names; it was not keeping to the 75-character limit for
4507 encoded words that the RFC imposes. It now respects the limit, and
4508 generates multiple encoded words if necessary. To be on the safe side, I
4509 have increased the buffer size for the ${rfc2047: expansion operator from
4512 22. Failure to deliver a bounce message always caused it to be frozen, even if
4513 there was an errors_to setting on the router. The errors_to setting is now
4516 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4517 canonical form (fully expanded) before being placed in
4518 $sender_host_address.
4520 24. Updated eximstats to version 1.33
4522 25. Include certificate and key file names in error message when GnuTLS fails
4523 to set them up, because the GnuTLS error message doesn't include the name
4524 of the failing file when there is a problem reading it.
4526 26. Expand error message when OpenSSL has problems setting up cert/key files.
4529 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4530 (this can affect the format of dates).
4532 28. exim_tidydb, when checking for the continued existence of a message for
4533 which it has found a message-specific retry record, was not finding
4534 messages that were in split spool directories. Consequently, it was
4535 deleting retry records that should have stayed in existence.
4537 29. eximstats updated to version 1.35
4538 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4539 1.35 - bugfix such that pie charts by volume are generated correctly
4541 30. The SPA authentication driver was not abandoning authentication and moving
4542 on to the next authenticator when an expansion was forced to fail,
4543 contradicting the general specification for all authenticators. Instead it
4544 was generating a temporary error. It now behaves as specified.
4546 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4547 (the order specifies the preference for clients). The order is now AES256,
4548 AES128, 3DES, ARCFOUR128.
4550 31. Small patch to Sieve code - explicitly set From: when generating an
4553 32. Exim crashed if a remote delivery caused a very long error message to be
4554 recorded - for instance if somebody sent an entire SpamAssassin report back
4555 as a large number of 550 error lines. This bug was coincidentally fixed by
4556 increasing the size of one of Exim's internal buffers (big_buffer) that
4557 happened as part of the Exiscan merge. However, to be on the safe side, I
4558 have made the code more robust (and fixed the comments that describe what
4561 33. Some experimental protocols are using DNS PTR records for new purposes. The
4562 keys for these records are domain names, not reversed IP addresses. The
4563 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4564 leaves it alone. Component reversal etc. now happens only for IP addresses.
4567 34. The host_aton() function is supposed to be passed a string that is known
4568 to be a valid IP address. However, in the case of IPv6 addresses, it was
4569 not checking this. This is a hostage to fortune. Exim now panics and dies
4570 if the condition is not met. A case was found where this could be provoked
4571 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4572 components; fortuitously, this particular loophole had already been fixed
4573 by change 4.50/55 or 4.44/33 above.
4575 If there are any other similar loopholes, the new check in host_aton()
4576 itself should stop them being exploited. The report I received stated that
4577 data on the command line could provoke the exploit when Exim was running as
4578 exim, but did not say which command line option was involved. All I could
4579 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4580 running as the user.
4583 35. There was a buffer overflow vulnerability in the SPA authentication code
4584 (which came originally from the Samba project). I have added a test to the
4585 spa_base64_to_bits() function which I hope fixes it.
4588 36. The daemon start-up calls getloadavg() while still root for those OS that
4589 need the first call to be done as root, but it missed one case: when
4590 deliver_queue_load_max is set with deliver_drop_privilege. This is
4591 necessary for the benefit of the queue runner, because there is no re-exec
4592 when deliver_drop_privilege is set.
4594 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4595 was in use, was not putting the data itself into the right store pool;
4596 consequently, it could be overwritten for a subsequent message in the same
4597 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4600 38. Sometimes the final signoff response after QUIT could fail to get
4601 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4602 < 0 before doing a fflush(). This bug looks as though it goes back to the
4603 introduction of TLS in release 3.20, but "sometimes" must have been rare
4604 because the tests only now provoked it.
4610 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4611 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4612 Now it returns an integer. A number of calls were still expecting a T/F
4613 return. Fortuitously, in all cases, the tests worked in OK situations,
4614 which is the norm. However, things would have gone wrong on any write
4615 failures on the smtp file descriptor. This function is used when sending
4616 messages over SMTP and also when doing verify callouts.
4618 2. When Exim is called to do synchronous delivery of a locally submitted
4619 message (the -odf or -odi options), it no longer closes stderr before doing
4622 3. Implemented the mua_wrapper option.
4624 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4626 5. Implemented the functions header_remove(), header_testname(),
4627 header_add_at_position(), and receive_remove_recipient(), and exported them
4630 6. If an ACL "warn" statement specified the addition of headers, Exim already
4631 inserted X-ACL-Warn: at the start if there was no header name. However, it
4632 was not making this test for the second and subsequent header lines if
4633 there were newlines in the string. This meant that an invalid header could
4634 be inserted if Exim was badly configured.
4636 7. Allow an ACL "warn" statement to add header lines at the start or after all
4637 the Received: headers, as well as at the end.
4639 8. Added the rcpt_4xx retry error code.
4641 9. Added postmaster_mailfrom=xxx to callout verification option.
4643 10. Added mailfrom=xxxx to the callout verification option, for verify=
4646 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4647 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4649 12. Inserted some casts to stop certain compilers warning when using pointer
4650 differences as field lengths or precisions in printf-type calls (mostly
4651 affecting debugging statements).
4653 13. Added optional readline() support for -be (dynamically loaded).
4655 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4656 same clock tick as a message's arrival, so that its received time was the
4657 same as the "first fail" time on the retry record, and that message
4658 remained on the queue past the ultimate address timeout, every queue runner
4659 would try a delivery (because it was past the ultimate address timeout) but
4660 after another failure, the ultimate address timeout, which should have then
4661 bounced the address, did not kick in. This was a "< instead of <=" error;
4662 in most cases the first failure would have been in the next clock tick
4663 after the received time, and all would be well.
4665 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4666 being recognized when the domain list was tested by the match_domain
4667 condition in an expansion string.
4669 16. Added the ${str2b64: operator.
4671 17. Exim was always calling setrlimit() to set a large limit for the number of
4672 processes, without checking whether the existing limit was already
4673 adequate. (It did check for the limit on file descriptors.) Furthermore,
4674 errors from getrlimit() and setrlimit() were being ignored. Now they are
4675 logged to the main and panic logs, but Exim does carry on, to try to do its
4676 job under whatever limits there are.
4678 18. Imported PCRE 5.0.
4680 19. Trivial typo in log message " temporarily refused connection" (the leading
4683 20. If the log selector return_path_on_delivery was set and an address was
4684 redirected to /dev/null, the delivery process crashed because it assumed
4685 that a return path would always be set for a "successful" delivery. In this
4686 case, the whole delivery is bypassed as an optimization, and therefore no
4689 21. Internal re-arrangement: the function for sending a challenge and reading
4690 a response while authentication was assuming a zero-terminated challenge
4691 string. It's now changed to take a pointer and a length, to allow for
4692 binary data in such strings.
4694 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4696 23. Exim was not respecting finduser_retries when seeking the login of the
4697 uid under which it was called; it was always trying 10 times. (The default
4698 setting of finduser_retries is zero.) Also, it was sleeping after the final
4699 failure, which is pointless.
4701 24. Implemented tls_on_connect_ports.
4703 25. Implemented acl_smtp_predata.
4705 26. If the domain in control=submission is set empty, Exim assumes that the
4706 authenticated id is a complete email address when it generates From: or
4707 Sender: header lines.
4709 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4710 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4711 chown and chgrp in /bin and hostname in /usr/bin.
4713 28. Exim was keeping the "process log" file open after each use, just as it
4714 does for the main log. This opens the possibility of it remaining open for
4715 long periods when the USR1 signal hits a daemon. Occasional processlog
4716 errors were reported, that could have been caused by this. Anyway, it seems
4717 much more sensible not to leave this file open at all, so that is what now
4720 29. The long-running daemon process does not normally write to the log once it
4721 has entered its main loop, and it closes the log before doing so. This is
4722 so that log files can straightforwardly be renamed and moved. However,
4723 there are a couple of unusual error situations where the daemon does write
4724 log entries, and I had neglected to close the log afterwards.
4726 30. The text of an SMTP error response that was received during a remote
4727 delivery was being truncated at 512 bytes. This is too short for some of
4728 the long messages that one sometimes sees. I've increased the limit to
4731 31. It is now possible to make retry rules that apply only when a message has a
4732 specific sender, in particular, an empty sender.
4734 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4735 it possible to be selective about when SMTP synchronization is enforced.
4737 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4739 32. Implemented hosts_connection_nolog.
4741 33. Added an ACL for QUIT.
4743 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4746 35. Added mailbox_size and mailbox_filecount to appendfile.
4748 36. Added control = no_multiline_responses to ACLs.
4750 37. There was a bug in the logic of the code that waits for the clock to tick
4751 in the case where the clock went backwards by a substantial amount such
4752 that the microsecond fraction of "now" was more than the microsecond
4753 fraction of "then" (but the whole seconds number was less).
4755 38. Added support for the libradius Radius client library this is found on
4756 FreeBSD (previously only the radiusclient library was supported).
4762 1. When certain lookups returned multiple values in the form name=value, the
4763 quoting of the values was not always being done properly. Specifically:
4764 (a) If the value started with a double quote, but contained no whitespace,
4766 (b) If the value contained whitespace other than a space character (i.e.
4767 tabs or newlines or carriage returns) it was not quoted.
4768 This fix has been applied to the mysql and pgsql lookups by writing a
4769 separate quoting function and calling it from the lookup code. The fix
4770 should probably also be applied to nisplus, ibase and oracle lookups, but
4771 since I cannot test any of those, I have not disturbed their existing code.
4773 2. A hit in the callout cache for a specific address caused a log line with no
4774 reason for rejecting RCPT. Now it says "Previous (cached) callout
4775 verification failure".
4777 3. There was an off-by-one bug in the queryprogram router. An over-long
4778 return line was truncated at 256 instead of 255 characters, thereby
4779 overflowing its buffer with the terminating zero. As well as fixing this, I
4780 have increased the buffer size to 1024 (and made a note to document this).
4782 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4783 when Exim is waiting for an SMTP response from a remote server, Exim
4784 restarts its select() call on the socket, thereby resetting its timeout.
4785 This is not a problem when such interrupts are rare. Somebody set up a cron
4786 job to run exiwhat every 2 minutes, which is less than the normal select()
4787 timeout (5 or 10 minutes). This meant that the select() timeout never
4788 kicked in because it was always reset. I have fixed this by comparing the
4789 time when an interrupt arrives with the time at the start of the first call
4790 to select(). If more time than the timeout has elapsed, the interrupt is
4791 treated as a timeout.
4793 5. Some internal re-factoring in preparation for the addition of Sieve
4794 extensions (by MH). In particular, the "personal" test is moved to a
4795 separate function, and given an option for scanning Cc: and Bcc: (which is
4796 not set for Exim filters).
4798 6. When Exim created an email address using the login of the caller as the
4799 local part (e.g. when creating a From: or Sender: header line), it was not
4800 quoting the local part when it contained special characters such as @.
4802 7. Installed new OpenBSD configuration files.
4804 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4805 try to make them clearer.
4807 9. Callout options, other than the timeout value, were being ignored when
4808 verifying sender addresses in header lines. For example, when using
4810 verify = header_sender/callout=no_cache
4812 the cache was (incorrectly) being used.
4814 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4815 only the Cygwin environment.
4817 11. When return_path_on_delivery was set as a log selector, if different remote
4818 addresses in the same message used different return paths and parallel
4819 remote delivery occurred, the wrong values would sometimes be logged.
4820 (Whenever a remote delivery process finished, the return path value from
4821 the most recently started remote delivery process was logged.)
4823 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4824 lines when AUTH and/or TLS are in use. This is the "received protocol"
4825 field. Exim used to use "asmtp" for authenticated SMTP, without any
4826 indication (in the protocol name) for TLS use. Now it follows the RFC and
4827 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4828 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4829 names appear in log lines as well as in Received: header lines.
4831 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4832 extensions, and comparison tests, and to fix some bugs.
4834 14. Changes to the "personal" filter test:
4836 (1) The test was buggy in that it was just doing the equivalent of
4837 "contains" tests on header lines. For example, if a user's address was
4838 anne@some.where, the "personal" test would incorrectly be true for
4840 To: susanne@some.where
4842 This test is now done by extracting each address from the header in turn,
4843 and checking the entire address. Other tests that are part of "personal"
4844 are now done using regular expressions (for example, to check local parts
4845 of addresses in From: header lines).
4847 (2) The list of non-personal local parts in From: addresses has been
4848 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4849 taken from the Sieve specification recommendations.
4851 (3) If the message contains any header line starting with "List-" it is
4852 treated as non-personal.
4854 (4) The test for "circular" in the Subject: header line has been removed
4855 because it now seems ill-conceived.
4857 15. Minor typos in src/EDITME comments corrected.
4859 16. Installed latest exipick from John Jetmore.
4861 17. If headers_add on a router specified a text string that was too long for
4862 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4863 of string_sprintf() is now avoided.
4865 18. $message_body_size was not set (it was always zero) when running the DATA
4866 ACL and the local_scan() function.
4868 19. For the "mail" command in an Exim filter, no default was being set for
4869 the once_repeat time, causing a random time value to be used if "once" was
4870 specified. (If the value happened to be <= 0, no repeat happened.) The
4871 default is now 0s, meaning "never repeat". The "vacation" command was OK
4872 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4873 (I found it when inspecting the code).
4875 20. There is now an overall timeout for performing a callout verification. It
4876 defaults to 4 times the callout timeout, which applies to individual SMTP
4877 commands during the callout. The overall timeout applies when there is more
4878 than one host that can be tried. The timeout is checked before trying the
4879 next host. This prevents very long delays if there are a large number of
4880 hosts and all are timing out (e.g. when the network connections are timing
4881 out). The value of the overall timeout can be changed by specifying an
4882 additional sub-option for "callout", called "maxwait". For example:
4884 verify = sender/callout=5s,maxwait=20s
4886 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4887 to the end before writing, but this should make it even safer).
4889 22. Exim was forgetting that it had advertised PIPELINING for the second and
4890 subsequent messages on an SMTP connection. It was also not resetting its
4891 memory on STARTTLS and an internal HELO.
4893 23. When Exim logs an SMTP synchronization error within a session, it now
4894 records whether PIPELINING has been advertised or not.
4896 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4897 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4898 rather than long int.
4900 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4902 26. Added the never_mail option to autoreply.
4908 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4909 crash if the getsockname() call failed; this can happen if a connection is
4910 closed very soon after it is established. The problem was simply in the
4911 order in which certain operations were done, causing Exim to try to write
4912 to the SMTP stream before it had set up the file descriptor. The bug has
4913 been fixed by making things happen in the correct order.
4919 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4920 before the connection was closed, thus losing the rejection response.
4922 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4923 some early Solaris releases, but causes trouble in current releases where
4924 socklen_t is defined.
4926 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4929 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4932 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4933 that releasing the top of it at the end releases what was used for sub-
4934 expansions (unless the block got too big). However, discard this block if
4935 the first thing is a variable or header, so that we can use its block when
4936 it is dynamic (useful for very large $message_headers, for example).
4938 6. Lookups now cache *every* query, not just the most recent. A new, separate
4939 store pool is used for this. It can be recovered when all lookup caches are
4940 flushed. Lookups now release memory at the end of their result strings.
4941 This has involved some general refactoring of the lookup sources.
4943 7. Some code has been added to the store_xxx() functions to reduce the amount
4944 of flapping under certain conditions.
4946 8. log_incoming_interface used to affect only the <= reception log lines. Now
4947 it causes the local interface and port to be added to several more SMTP log
4948 lines, for example "SMTP connection from", and rejection lines.
4950 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4952 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4954 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4955 could be overwritten at the end of the current message (or the start of a
4956 new message if it was set in a HELO ACL). The value is now preserved for
4957 the duration of the SMTP connection.
4959 12. If a transport had a headers_rewrite setting, and a matching header line
4960 contained an unqualified address, that address was qualified, even if it
4961 did not match any rewriting rules. The underlying bug was that the values
4962 of the flags that permit the existence of unqualified sender and recipient
4963 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4964 non-local messages, and by -bnq for local messages) were not being
4965 preserved with the message after it was received.
4967 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4968 "next input=" as part of the text comprising the host identity instead of
4969 the correct text. The code was using the same buffer for two different
4970 strings. However, depending on which order the printing function evaluated
4971 its arguments, the bug did not always show up. Under Linux, for example, my
4972 test suite worked just fine.
4974 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4975 doesn't work with some older versions of Perl. It has been changed to "my",
4976 which in any case is probably the better facility to use.
4978 15. A really picky compiler found some instances of statements for creating
4979 error messages that either had too many or two few arguments for the format
4982 16. The size of the buffer for calls to the DNS resolver has been increased
4983 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4984 for addresses that have a lot of PTR records. This alleviates a problem; it
4985 does not fully solve it.
4987 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4988 buffer now truncates the list and logs the incident, which is the same
4989 action as happens when Exim is looking up a host name and its aliases.
4990 Previously in this situation something unpredictable would happen;
4991 sometimes it was "internal error: store_reset failed".
4993 18. If a server dropped the connection unexpectedly when an Exim client was
4994 using GnuTLS and trying to read a response, the client delivery process
4995 crashed while trying to generate an error log message.
4997 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4998 string, for example:
5000 warn message = H1: something\nH2: something
5002 the text was added as a single header line from Exim's point of view
5003 though it ended up OK in the delivered message. However, searching for the
5004 second and subsequent header lines using $h_h2: did not work. This has been
5005 fixed. Similarly, if a system filter added multiple headers in this way,
5006 the routers could not see them.
5008 20. Expanded the error message when iplsearch is called with an invalid key to
5009 suggest using net-iplsearch in a host list.
5011 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5012 ACLs are no longer actually imposed (and a message to that effect is
5015 22. If a "gecos" field in a passwd entry contained escaped characters, in
5016 particular, if it contained a \" sequence, Exim got it wrong when building
5017 a From: or a Sender: header from that name. A second bug also caused
5018 incorrect handling when an unquoted " was present following a character
5019 that needed quoting.
5021 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5022 was not being matched caselessly.
5024 24. Arranged for all hyphens in the exim.8 source to be escaped with
5027 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5028 properly. Recipient callout cache records were still being keyed to include
5029 the sender, even when use_sender was set false. This led to far more
5030 callouts that were necessary. The sender is no longer included in the key
5031 when use_sender is false.
5033 26. Added "control = submission" modifier to ACLs.
5035 27. Added the ${base62d: operator to decode base 62 numbers.
5037 28. dnsdb lookups can now access SRV records.
5039 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5040 the configuration file.
5042 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5043 (-v) mode. This makes the output for a verbose queue run more intelligible.
5045 31. Added a use_postmaster feature to recipient callouts.
5047 32. Added the $body_zerocount variable, containing the number of binary zero
5048 bytes in the message body.
5050 33. The time of last modification of the "new" subdirectory is now used as the
5051 "mailbox time last read" when there is a quota error for a maildir
5054 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5056 35. Added +ignore_unknown as a special item in host lists.
5058 36. Code for decoding IPv6 addresses in host lists is now included, even if
5059 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5060 address was recognized as an IP address, but was then not correctly decoded
5061 into binary, causing unexpected and incorrect effects when compared with
5068 1. Very minor rewording of debugging text in manualroute to say "list of
5069 hosts" instead of "hostlist".
5071 2. If verify=header_syntax was set, and a header line with an unqualified
5072 address (no domain) and a large number of spaces between the end of the
5073 name and the colon was received, the reception process suffered a buffer
5074 overflow, and (when I tested it) crashed. This was caused by some obsolete
5075 code that should have been removed. The fix is to remove it!
5077 3. When running in the test harness, delay a bit after writing a bounce
5078 message to get a bit more predictability in the log output.
5080 4. Added a call to search_tidyup() just before forking a reception process. In
5081 theory, someone could use a lookup in the expansion of smtp_accept_max_
5082 per_host which, without the tidyup, could leave open a database connection.
5084 5. Added the variables $recipient_data and $sender_data which get set from a
5085 lookup success in an ACL "recipients" or "senders" condition, or a router
5086 "senders" option, similar to $domain_data and $local_part_data.
5088 6. Moved the writing of debug_print from before to after the "senders" test
5091 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5092 problems for message scanning, either using a data ACL, or using
5093 local_scan() because the Received: header was not generated till after they
5094 were called (in order to set the time as the time of reception completion).
5095 I have revised the way this works. The header is now generated after the
5096 body is received, but before the ACL or local_scan() are called. After they
5097 are run, the timestamp in the header is updated.
5103 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5104 before starting a queue runner without re-exec. This happened only when
5105 deliver_drop_privilege was set or when the Exim user was set to root. The
5106 effect of the bug was that timeouts during subsequent deliveries caused
5107 crashes instead of being properly handled. The handler is now left at its
5108 default (and expected) setting.
5110 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5111 message, again when deliver_drop_privilege is set or Exim is run as root.
5112 The bug described in (1) was not present in this case, but the tidying up
5113 of the other signals was missing. I have made the two cases consistent.
5115 3. The ignore_target_hosts setting on a manualroute router was being ignored
5116 for hosts that were looked up using the /MX notation.
5118 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5121 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5122 operated on the sender address. After changing the $sender_address to <>
5123 for the sender address verify, Exim was re-instated it as the original
5124 (before rewriting) address, but remembering that it had rewritten it, so it
5125 wasn't rewriting it again. This bug also had the effect of breaking the
5126 sender address verification caching when the sender address was rewritten.
5128 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5129 This has been changed so that if the ip literal address matches
5130 ignore_target_hosts, the router declines.
5132 7. Added expansion conditions match_domain, match_address, and match_local_
5133 part (NOT match_host).
5135 8. The placeholder for the Received: header didn't have a length field set.
5137 9. Added code to Exim itself and to exim_lock to test for a specific race
5138 condition that could lead to file corruption when using MBX delivery. The
5139 issue is with the lockfile that is created in /tmp. If this file is removed
5140 after a process has opened it but before that process has acquired a lock,
5141 there is the potential for a second process to recreate the file and also
5142 acquire a lock. This could lead to two Exim processes writing to the file
5143 at the same time. The added code performs the same test as UW imapd; it
5144 checks after acquiring the lock that its file descriptor still refers to
5145 the same named file.
5147 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5148 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5151 11. Added the smtp_active_hostname option. If used, this will typically be made
5152 to depend on the incoming interface address. Because $interface_address is
5153 not set up until the daemon has forked a reception process, error responses
5154 that can happen earlier (such as "too many connections") no longer contain
5157 12. If an expansion in a condition on a "warn" statement fails because a lookup
5158 defers, the "warn" statement is abandoned, and the next ACL statement is
5159 processed. Previously this caused the whole ACL to be aborted.
5161 13. Added the iplsearch lookup type.
5163 14. Added ident_timeout as a log selector.
5165 15. Added tls_certificate_verified as a log selector.
5167 16. Added a global option tls_require_ciphers (compare the smtp transport
5168 option of the same name). This controls incoming TLS connections.
5170 17. I finally figured out how to make tls_require_ciphers do a similar thing
5171 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5172 before starting the TLS session.
5174 18. Tabs are now shown as \t in -bP output.
5176 19. If the log selector return_path_on_delivery was set, Exim crashed when
5177 bouncing a message because it had too many Received: header lines.
5179 20. If two routers both had headers_remove settings, and the first one included
5180 a superfluous trailing colon, the final name in the first list and the
5181 first name in the second list were incorrectly joined into one item (with a
5182 colon in the middle).
5188 1. Added -C and -D options to the exinext utility, mainly to make it easier
5189 to include in the automated testing, but these could be helpful when
5190 multiple configurations are in use.
5192 2. The exinext utility was not formatting the output nicely when there was
5193 an alternate port involved in the retry record key, nor when there was a
5194 message id as well (for retries that were specific to a specific message
5195 and a specific host). It was also confused by IPv6 addresses, because of
5196 the additional colons they contain. I have fixed the IPv4 problem, and
5197 patched it up to do a reasonable job for IPv6.
5199 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5200 delivery, the log line now contains "pipelined" if PIPELINING was used.
5202 4. An SMTP transport process used to panic and die if the bind() call to set
5203 an explicit outgoing interface failed. This has been changed; it is now
5204 treated in the same way as a connect() failure.
5206 5. A reference to $sender_host_name in the part of a conditional expansion
5207 that was being skipped was still causing a DNS lookup. This no longer
5210 6. The def: expansion condition was not recognizing references to header lines
5211 that used bh_ and bheader_.
5213 7. Added the _cache feature to named lists.
5215 8. The code for checking quota_filecount in the appendfile transport was
5216 allowing one more file than it should have been.
5218 9. For compatibility with Sendmail, the command line option
5226 and sets the incoming protocol and host name (for trusted callers). The
5227 host name and its colon can be omitted when only the protocol is to be set.
5228 Note the Exim already has two private options, -pd and -ps, that refer to
5229 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5230 "s", but I don't think that's a major issue.
5232 10. A number of refactoring changes to the code, none of which should affect
5235 (a) The number of logging options was getting close to filling up the
5236 32-bit word that was used as a bit map. I have split them into two classes:
5237 those that are passed in the argument to log_write(), and those that are
5238 only ever tested independently outside of that function. These are now in
5239 separate 32-bit words, so there is plenty of room for expansion again.
5240 There is no change in the user interface or the logging behaviour.
5242 (b) When building, for example, log lines, the code previously used a
5243 macro that called string_cat() twice, in order to add two strings. This is
5244 not really sufficiently general. Furthermore, there was one instance where
5245 it was actually wrong because one of the argument was used twice, and in
5246 one call a function was used. (As it happened, calling the function twice
5247 did not affect the overall behaviour.) The macro has been replaced by a
5248 function that can join an arbitrary number of extra strings onto a growing
5251 (c) The code for expansion conditions now uses a table and a binary chop
5252 instead of a serial search (which was left over from when there were very
5253 few conditions). Also, it now recognizes conditions like "pam" even when
5254 the relevant support is not compiled in: a suitably worded error message is
5255 given if an attempt is made to use such a condition.
5257 11. Added ${time_interval:xxxxx}.
5259 12. A bug was causing one of the ddress fields not to be passed back correctly
5260 from remote delivery subprocesses. The field in question was not being
5261 subsequently used, so this caused to problems in practice.
5263 13. Added new log selectors queue_time and deliver_time.
5265 14. Might have fixed a bug in maildirsizefile handling that threw up
5266 "unexpected character" debug warnings, and recalculated the data
5267 unnecessarily. In any case, I expanded the warning message to give more
5270 15. Added the message "Restricted characters in address" to the statements in
5271 the default ACL that block characters like @ and % in local parts.
5273 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5274 Three changes have been made:
5276 (a) There was a serious bug; a negative response to MAIL caused the whole
5277 recipient domain to be cached as invalid, thereby blocking all messages
5278 to all local parts at the same domain, from all senders. This bug has
5279 been fixed. The domain is no longer cached after a negative response to
5280 MAIL if the sender used is not empty.
5282 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5285 (c) A new callout option, "use_sender" has been added for people who want
5286 the modified behaviour.
5292 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5295 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5296 indeed breaks things for older releases.
5298 3. Added additional logging to the case where there is a problem reading data
5299 from a filter that is running in a subprocess using a pipe, in order to
5300 try to track down a specific problem.
5302 4. Testing facility fudge: when running in the test harness and attempting
5303 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5304 getting "No route to host". Convert this to a timeout.
5306 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5309 6. Some OS don't have socklen_t but use size_t instead. This affects the
5310 fifth argument of getsockopt() amongst other things. This is now
5311 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5312 can be set for individual OS. I have set it for SunOS5, OSF1, and
5313 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5314 some earlier ones do not.
5316 7. Change 4.30/15 was not doing the test caselessly.
5318 8. The standard form for an IPv6 address literal was being rejected by address
5319 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5320 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5321 this, as well as the form without the "IPv6" on the front (but only when
5322 address literals are enabled, of course).
5324 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5326 10. Exim crashed if a message with an empty sender address specified by -f
5327 encountered a router with an errors_to setting. This could be provoked only
5328 by a command such as
5332 where an empty string was supplied; "<>" did not hit this bug.
5334 11. Installed PCRE release 4.5.
5336 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5337 remained set. It is now erased.
5339 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5340 times from message ids (which are base 36 rather than the normal 62).
5342 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5343 were being counted as actual protocol errors, and logged if the log
5344 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5345 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5346 and DATA following a set of rejected RCPTs do not count as protocol errors.
5347 In other words, Exim assumes they were pipelined, though this may not
5348 actually be the case. Of course, in all cases the client gets an
5349 appropriate error code.
5351 15. If a lookup fails in an ACL condition, a message about the failure may
5352 be available; it is used if testing the ACL cannot continue, because most
5353 such messages specify what the cause of the deferral is. However, some
5354 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5355 that caused an old message to be retained and used if a later statement
5356 caused a defer, replacing the real cause of the deferral.
5358 16. If an IP address had so many PTR records that the DNS lookup buffer
5359 was not large enough to hold them, Exim could crash while trying to process
5360 the truncated data. It now detects and logs this case.
5362 17. Further to 4.21/58, another change has been made: if (and only if) the
5363 first line of a message (the first header line) ends with CRLF, a bare LF
5364 in a subsequent header line has a space inserted after it, so as not to
5365 terminate the header.
5367 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5368 unnecessarily, used atoi() instead of strtol(), and didn't check the
5369 termination when getting file sizes from file names by regex.
5371 19. Completely re-implemented the support for maildirsize files, in the light
5372 of a number of problems with the previous contributed implementation
5373 (4.30/29). In particular:
5375 . If the quota is zero, the maildirsize file is maintained, but no quota is
5378 . If the maildir directory does not exist, it is created before any attempt
5379 to write a maildirsize file.
5381 . The quota value in the file is just a cache; if the quota is changed in
5382 the transport, the new value overrides.
5384 . A regular expression is available for excluding directories from the
5387 20. The autoreply transport checks the characters in options that define the
5388 message's headers; it allows continued headers, but it was checking with
5389 isspace() after an embedded newline instead of explicitly looking for a
5392 21. If all the "regular" hosts to which an address was routed had passed their
5393 expiry times, and had not reached their retry times, the address was
5394 bounced, even if fallback hosts were defined. Now Exim should go on to try
5397 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5398 equivalent code in the SMTP transport. Some hosts send humungous responses
5399 to HELO/EHLO, more than 1024 it seems.
5401 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5402 gives compiler warnings in some environments. I've now done it "properly",
5405 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5406 (because of problems with the built-in one) was declared to return uschar *
5407 instead of char *, causing compiler failure.
5409 25. Fixed a file descriptor leak when processing alias/forward files.
5411 26. Fixed a minor format string issue in dbfn.c.
5413 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5415 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5416 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5419 29. When a sender address is verified, it is cached, to save repeating the test
5420 when there is more than one recipient in a message. However, when the
5421 verification involves a callout, it is possible for different callout
5422 options to be set for different recipients. It is too complicated to keep
5423 track of this in the cache, so now Exim always runs a verification when a
5424 callout is required, relying on the callout cache for the optimization.
5425 The overhead is duplication of the address routing, but this should not be
5428 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5429 to be verified with callout=postmaster, and the main callout worked but the
5430 postmaster check failed, the verification correctly failed. However, if a
5431 subsequent RCPT command asked for sender verification *without* the
5432 postmaster check, incorrect caching caused this verification also to fail,
5435 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5436 it was not caching the DNS options (qualify_single, search_parents) that
5437 were used when the lookup failed. A subsequent lookup with different
5438 options therefore always gave the same answer, though there were cases
5439 where it should not have. (Example: a "domains = !$mx_any" option on a
5440 dnslookup router: the "domains" option is always processed without any
5441 widening, but the router might have qualify_single set.) Now Exim uses the
5442 cached value only when the same options are set.
5444 32. Added John Jetmore's "exipick" utility to the distribution.
5446 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5447 than a timeout (e.g. a certificate is required, and is not provided), an
5448 Exim server now closes the connection immediately. Previously it waited for
5449 the client to close - but if the client is SSL, it seems that they each
5450 wait for each other, leading to a delay before one of them times out.
5452 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5453 maintained 0.8.x compatibility because I don't think many are using it, and
5454 it is clearly obsolete.
5456 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5459 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5460 client certificate was expired. A simple patch fixes this, though I don't
5461 understand the full logic of why the verify callback is called multiple
5464 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5465 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5466 which causes problems with some clients (such as the Certicom SSL Plus
5467 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5468 disables the coutermeasure allowing Eudora to connect."
5470 38. Exim was not checking that a write() to a log file succeeded. This could
5471 lead to Bad Things if a log got too big, in particular if it hit a file
5472 size limit. Exim now panics and dies if it cannot write to a log file, just
5473 as it does if it cannot open a log file.
5475 39. Modified OS/Makefile-Linux so that it now contains
5477 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5479 The two -D definitions ensure that Exim is compiled with large file
5480 support, which makes it possible to handle log files that are bigger than
5483 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5484 instance) a domain was checked against a named list that involved a lookup,
5485 causing $domain_data to be set, then another domain was checked against the
5486 same list, then the first domain was re-checked, the value of $domain_data
5487 after the final check could be wrong. In particular, if the second check
5488 failed, it could be set empty. This bug probably also applied to
5491 41. The strip_trailing_dot option was not being applied to the address given
5492 with the -f command-line option.
5494 42. The code for reading a message's header from the spool was incrementing
5495 $received_count, but never initializing it. This meant that the value was
5496 incorrect (doubled) while delivering a message in the same process in which
5497 it was received. In the most common configuration of Exim, this never
5498 happens - a fresh exec is done - but it can happen when
5499 deliver_drop_privilege is set.
5501 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5502 it now includes up to 150 characters of the unexpected data in the log
5505 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5506 and building data strings. The size of both of these buffers was 10 000
5507 bytes - far larger than anybody would *ever* want, thought I. Needless to
5508 say, somebody hit the limit. I have increased the maximum line length to
5509 20 000 and the maximum data length of concatenated lines to 100 000. I have
5510 also fixed two bugs, because there was no checking on these buffers. Tsk,
5511 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5512 buffer is too small.
5514 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5515 lsearch lookups. Now it does.
5517 46. When parsing a route_list item in a manualroute router, a fixed-length
5518 buffer was used for the list of hosts. I made this 1024 bytes long,
5519 thinking that nobody would ever have a list of hosts that long. Wrong.
5520 Somebody had a whole pile of complicated expansion conditions, and the
5521 string was silently truncated, leading to an expansion error. It turns out
5522 that it is easier to change to an unlimited length (owing to other changes
5523 that have happened since this code was originally written) than to build
5524 structure for giving a limitation error. The length of the item that
5525 expands into the list of hosts is now unlimited.
5527 47. The lsearch lookup could not handle data where the length of text line was
5528 more than 4095 characters. Such lines were truncated, leading to shortened
5529 data being returned. It should now handle lines of any length.
5531 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5532 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5535 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5536 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5537 no longer happens in comment lines. A list of replacements is now placed
5538 at the head of all of the source files, except those whose only change is
5539 to replace PERL_COMMAND in the very first #! line.
5541 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5542 messages on the queue, with a bottom-up merge sort, using code contributed
5543 by Michael Haardt. This should make operations like -bp somewhat faster on
5544 large queues. It won't affect queue runners, except when queue_run_in_order
5547 51. Installed eximstats 1.31 in the distribution.
5549 52. Added support for SRV lookups to the dnslookup router.
5551 53. If an ACL referred to $message_body or $message_body_end, the value was not
5552 reset for any messages that followed in the same SMTP session.
5554 54. The store-handling optimization for building very long strings was not
5555 differentiating between the different store pools. I don't think this
5556 actually made any difference in practice, but I've tidied it.
5558 55. While running the routers to verify a sender address, $sender_address
5559 was still set to the sender address. This is wrong, because when routing to
5560 send a bounce to the sender, it would be empty. Therefore, I have changed
5561 it so that, while verifying a sender address, $sender_address is set to <>.
5562 (There is no change to what happens when verifying a recipient address.)
5564 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5565 target A or AAAA records (if not already returned) without resetting the
5566 qualify_single or search_parents options of the DNS resolver. These are
5567 inappropriate in this case because the targets of MX and SRV records must
5568 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5569 target that, when qualified, matched something in the local domain. These
5570 two options are now turned off when doing these lookups.
5572 57. It seems that at least some releases of Reiserfs (which does not have the
5573 concept of a fixed number of inodes) returns zero and not -1 for the
5574 number of available inodes. This interacted badly with check_spool_inodes,
5575 which assumed that -1 was the "no such thing" setting. What I have done is
5576 to check that the total number of inodes is greater than zero before doing
5577 the test of how many are available.
5579 58. When a "warn" ACL statement has a log_message modifier, the message is
5580 remembered, and not repeated. This is to avoid a lot of repetition when a
5581 message has many recipients that cause the same warning to be written.
5582 Howewer, Exim was preserving the list of already written lines for an
5583 entire SMTP session, which doesn't seem right. The memory is now reset if a
5584 new message is started.
5586 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5587 debugging output unless log_rewrite was also set.
5589 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5590 of (int)(handle) when we know that handle contains (void *)(-1).
5592 61. The Exim daemon panic-logs an error return when it closes the incoming
5593 connection. However "connection reset by peer" seems to be common, and
5594 isn't really an error worthy of noting specially, so that particular error
5597 62. When Exim is trying to find all the local interfaces, it used to panic and
5598 die if the ioctl to get the interface flags failed. However, it seems that
5599 on at least one OS (Solaris 9) it is possible to have an interface that is
5600 included in the list of interfaces, but for which you get a failure error
5601 for this call. This happens when the interface is not "plumbed" into a
5602 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5603 failure of the "get flags" call assumes that the interface is down.
5605 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5606 makes life easier for people who are doing arithmetic on fields extracted
5607 from dates, where you often get leading zeros that should not be
5608 interpreted as octal.
5610 64. Added qualify_domain to the redirect router, to override the global
5613 65. If a pathologically long header line contained very many addresses (the
5614 report of this problem mentioned 10 000) and each of them was rewritten,
5615 Exim could use up a very large amount of memory. (It kept on making new
5616 copies of the header line as it rewrote, and never released the old ones.)
5617 At the expense of a bit more processing, the header rewriting function has
5618 been changed so that it no longer eats memory in this way.
5620 66. The generation of the Received: header has been moved from the time that a
5621 message starts to be received, to the time that it finishes. The timestamp
5622 in the Received: header should now be very close to that of the <= log
5623 line. There are two side-effects of this change:
5625 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5626 logged header lines no longer include the local Received: line, because
5627 it has not yet been created. The same applies to a copy of the message
5628 that is returned to a non-SMTP sender when a message is rejected.
5630 (b) When a filter file is tested using -bf, no additional Received: header
5631 is added to the test message. After some thought, I decided that this
5634 This change does not affect the value of $received_for. It is still set
5635 after address rewriting, but before local_scan() is called.
5637 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5639 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5640 gave an unhelpful panic error message, and a defer error. I have managed to
5641 change this behaviour so that it now rejects any supplied certificate,
5642 which seems right, as the list of acceptable certificates is empty.
5644 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5645 gave an unhelpful defer error. I have not managed to make this reject any
5646 supplied certificates, but the error message it gives is "no certificate
5647 supplied", which is not helpful.
5649 70. exigrep's output now also includes lines that are not associated with any
5650 message, but which match the given pattern. Implemented by a patch from
5651 Martin Sluka, which also tidied up the Perl a bit.
5653 71. Recipient callout verification, like sender verification, was using <> in
5654 the MAIL FROM command. This isn't really the right thing, since the actual
5655 sender may affect whether the remote host accepts the recipient or not. I
5656 have changed it to use the actual sender in the callout; this means that
5657 the cache record is now keyed on a recipient/sender pair, not just the
5658 recipient address. There doesn't seem to be a real danger of callout loops,
5659 since a callout by the remote host to check the sender would use <>.
5660 [SEE ABOVE: changed after hitting problems.]
5662 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5663 temporary errors. However, in the case of such a code being given after
5664 the end of a data transmission (i.e. after ".") Exim was failing to write
5665 a retry record for the message. (Yes, there was some broken host that was
5666 actually sending 8xx at this point.)
5668 73. An unknown lookup type in a host list could cause Exim to panic-die when
5669 the list was checked. (An example that provoked this was putting <; in the
5670 middle of a list instead of at the start.) If this happened during a DATA
5671 ACL check, a -D file could be left lying around. This kind of configuration
5672 error no longer causes Exim to die; instead it causes a defer errror. The
5673 incident is still logged to the main and panic logs.
5675 74. Buglet left over from Exim 3 conversion. The message "too many messages
5676 in one connection" was written to the rejectlog but not the mainlog, except
5677 when address rewriting (yes!) was being logged.
5679 75. Added write_rejectlog option.
5681 76. When a system filter was run not as root (that is, when system_filter_user
5682 was set), the values of the $n variables were not being returned to the
5683 main process; thus, they were not subsequently available in the $sn
5686 77. Added +return_path_on_delivery log selector.
5688 78. A connection timeout was being treated differently from recipients deferred
5689 when testing hosts_max_try with a message that was older than the host's
5690 retry timeout. (The host should not be counted, thus allowing all hosts to
5691 be tried at least once before bouncing.) This may have been the cause of an
5692 occasionally reported bug whereby a message would remain on the queue
5693 longer than the retry timeout, but would be bounced if a delivery was
5694 forced. I say "may" because I never totally pinned down the problem;
5695 setting up timeout/retry tests is difficult. See also the next item.
5697 79. The ultimate address timeout was not being applied to errors that involved
5698 a combination of host plus message (for example, a timeout on a MAIL
5699 command). When an address resolved to a number of possible hosts, and they
5700 were not all tried for each delivery (e.g. because of hosts_max_try), a
5701 message could remain on the queue longer than the retry timeout.
5703 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5706 81. Fixed an obscure SMTP outgoing bug which required at least the following
5707 conditions: (a) there was another message waiting for the same server;
5708 (b) the server returned 5xx to all RCPT commands in the first message so
5709 that the message was not completed; (c) the server dropped the connection
5710 or gave a negative response to the RSET that Exim sends to abort the
5711 transaction. The observed case was a dropped connection after DATA that had
5712 been sent in pipelining mode. That is, the server had advertised PIPELINING
5713 but was not implementing it correctly. The effect of the bug was incorrect
5714 behaviour, such as trying another host, and this could lead to a crash.
5720 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5721 and daemon.c were passed as pointers to ints; they should have been
5722 pointers to socklen_t variables (which are typically unsigned ints).
5724 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5727 3. Fixed a really odd bug that affected only the testing scheme; patching a
5728 certain fixed string in the binary changed the value of another string that
5729 happened to be identical to the end of the original first string.
5731 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5732 name", it returns that address as the IP address. On some operating
5733 systems (e.g. Solaris), it also passes back the IP address string as the
5734 "host name". However, on others (e.g. Linux), it passes back an empty
5735 string. Exim wasn't checking for this, and was changing the host name to an
5736 empty string, assuming it had been canonicized.
5738 5. Although rare, it is permitted to have more than one PTR record for a given
5739 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5740 all the names associated with an address, because they do in Solaris.
5741 However, it seems that they do not in Linux for data that comes from the
5742 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5743 I found this out when I moved to a new Linux workstation and tried to run
5744 the Exim test suite.
5746 To get round this problem I have changed the code so that it now does its
5747 own call to the DNS to look up PTR records when searching for a host name.
5748 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5749 addresses that are only in /etc/hosts are still found.
5751 This behaviour is, however, controlled by an option called host_lookup_
5752 order, which defaults to "bydns:byaddr". If people want to use the other
5753 order, or indeed, just use one or the other means of lookup, they can
5754 specify it in this variable.
5756 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5757 some operating systems, this comes back from gethostbyaddr() as an empty
5758 string, and this is what Exim used to test for. However, it seems that in
5759 other systems, "." is yielded. Exim now tests for this case too.
5761 7. The values of check_spool_space and check_log_space are now held internally
5762 as a number of kilobytes instead of an absolute number of bytes. If a
5763 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5764 kilobyte. This means that much larger values can be stored.
5766 8. Exim monitor: an attempt to get the action menu when not actually pointing
5767 at a message produces an empty menu entitled "No message selected". This
5768 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5769 no entries in it ("Shell widget menu has zero width and/or height"). So I
5770 have added a single, blank menu entry in this case.
5772 9. Added ${quote_local_part.
5774 10. MIME decoding is now applied to the contents of Subject: header lines when
5777 11. Now that a reference to $sender_host_address automatically causes a reverse
5778 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5779 host lookup before query-style lookups in lists that might use this
5780 variable. This has therefore been abolished, and the "net-" prefix is no
5781 longer necessary for query-style lookups.
5783 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5784 have been a typo for LFLAGS, so it has been changed.
5786 13. The install script calls Exim with "-C /dev/null" in order to find the
5787 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5788 to be output. Howeve, since Exim outputs its version number before the
5789 error, it didn't break the script. It just looked ugly. I fixed this by
5790 always allowing "-C /dev/null" if the caller is root.
5792 14. Ignore overlarge ACL variable number when reading spool file - insurance
5793 against a later release with more variables having written the file.
5795 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5796 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5797 this, as well as the form without the "IPv6" on the front.
5799 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5800 OS/Makefile-Darwin file.
5802 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5803 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5805 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5806 variable. However, some OS use ints for inodes. I've added cast to long int
5807 to get rid of the compiler warning.
5809 19. I had forgotten to lock out "/../" in configuration file names when
5810 ALT_CONFIG_PREFIX was set.
5812 20. Routers used for verification do not need to specify transports. However,
5813 if such a router generated a host list, and callout was configured, Exim
5814 crashed, because it could not find a port number from the (non-existent)
5815 transport. It now assumes port 25 in this circumstance.
5817 21. Added the -t option to exigrep.
5819 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5820 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5821 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5822 them is set, LOOKUP_LSEARCH is forced.
5824 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5826 24. Added sender and host information to the "rejected by local_scan()" log
5827 line; previously there was no indication of these.
5829 25. Added .include_if_exists.
5831 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5832 Linux. It turns out that not all file systems support this. Apparently some
5833 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5834 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5835 not supported on the file descriptor, is now ignored when Exim is trying to
5836 sync a directory. This applies only to Linux.
5838 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5840 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5841 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5844 29. Added support for maildirsize files from supplied patch (modified a bit).
5846 30. The use of :fail: followed by an empty string could lead Exim to respond to
5847 sender verification failures with (e.g.):
5849 550 Verification failed for <xxx>
5850 550 Sender verify failed
5852 where the first response line was missing the '-' that indicates it is not
5853 the final line of the response.
5855 31. The loop for finding the name of the user that called Exim had a hardwired
5856 limit of 10; it now uses the value of finduser_retries, which is used for
5857 all other user lookups.
5859 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5862 33. Exim was neglecting to zero errno before one call of strtol() when
5863 expanding a string and expecting an integer value. On some systems this
5864 resulted in spurious "integer overflow" errors. Also, it was casting the
5865 result into an int without checking.
5867 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5868 did not work. The code looks as if it has *never* worked, though it appears
5869 to have been documented since at least releast 1.62. I have made it work.
5871 35. The "timeout_DNS" error in retry rules, also documented since at least
5872 1.62, also never worked. As it isn't clear exactly what this means, and
5873 clearly it isn't a major issue, I have abolished the feature by treating it
5874 as "timeout", and writing a warning to the main and panic logs.
5876 36. The display of retry rules for -brt wasn't always showing the error code
5879 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5880 timeout_connect_A, timeout_connect_MX.
5882 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5883 to the empty sender.
5885 39. The daemon was not analysing the content of -oX till after it had closed
5886 stderr and disconnected from the controlling terminal. This meant that any
5887 syntax errors were only noted on the panic log, and the return code from
5888 the command was 0. By re-arranging the code a little, I've made the
5889 decoding happen first, so such errors now appear on stderr, and the return
5890 code is 1. However, the actual setting up of the sockets still happens in
5891 the disconnected process, so errors there are still only recorded on the
5894 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5895 connections (as happens on some IP stacks) was logged at start up time as
5896 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5897 it from "IPv6 and IPv4", which means that two separate sockets are being
5900 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5901 says whether AF_INET or AF_INET6 was passed as an argument.
5903 42. Exiwhat output was messed up when time zones were included in log
5906 43. Exiwhat now gives more information about the daemon's listening ports,
5907 and whether -tls-on-connect was used.
5909 44. The "port" option of the smtp transport is now expanded.
5911 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5912 silently ignored. Now an error message is written to the main and panic
5915 46. There's a new ACL modifier called "logwrite" which writes to a log file
5916 as soon as it is encountered.
5918 47. Added $local_user_uid and $local_user_gid at routing time.
5920 48. Exim crashed when trying to verify a sender address that was being
5923 49. Exim was recognizing only a space character after ".include". It now also
5924 recognizes a tab character.
5926 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5927 extracting the relevant information from the specification. The man page no
5928 longer contains scrambled data for the -d option, and I've added a section
5929 at the front about calling Exim under different names.
5931 51. Added "extra_headers" argument to the "mail" command in filter files.
5933 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5936 53. Installed eximstats 1.29.
5938 54. Added transport_filter_timeout as a generic transport option.
5940 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5941 Cc: header lines. This was required by RFC 822, but it not required by RFC
5944 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5945 incoming messages that did not have them. Now it does so only if the
5946 message originates locally, that is, if there is no associated remote host
5947 address. When Resent- header lines are present, this applies to the Resent-
5948 lines rather than the non-Resent- lines.
5950 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5951 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5953 58. Messages for configuration errors now include the name of the main
5954 configuration file - useful now that there may be more than one file in a
5955 list (.included file names were always shown).
5957 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5958 for those rare installations that do not start the daemon as root or run it
5959 setuid root. I've cut out the call to initgroups() if the daemon is not
5962 60. The Exim user and group can now be bound into the binary as text strings
5963 that are looked up at the start of Exim's processing.
5965 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5967 62. Added $mailstore_basename variable.
5969 63. Installed patch to sieve.c from Michael Haardt.
5971 64. When Exim failed to open the panic log after failing to open the main log,
5972 the original message it was trying to log was written to stderr and debug
5973 output, but if they were not available (the usual case in production), it
5974 was lost. Now it is written to syslog before the two lines that record the
5975 failures to open the logs.
5977 65. Users' Exim filters run in subprocesses under the user's uid. It is
5978 possible for a "deliver" command or an alias in a "personal" command to
5979 provoke an address rewrite. If logging of address rewriting is configured,
5980 this fails because the process is not running as root or exim. There may be
5981 a better way of dealing with this, but for the moment (because 4.30 needs
5982 to be released), I have disabled address rewrite logging when running a
5983 filter in a non-root, non-exim process.
5989 1. The buildconfig auxiliary program wasn't quoting the value set for
5990 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5991 not defined. This bug was masked in 4.22 by the effect that was fixed in
5994 2. Some messages that were rejected after a message id was allocated were
5995 shown as "incomplete" by exigrep. It no longer does this for messages that
5996 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5998 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5999 have allow_domain_literals set, the ID did not get logged in the <= line.
6000 Domain literals are now always recognized in Message-ID: header lines.
6002 4. The first argument for a ${extract expansion item is the key name or field
6003 number. Leading and trailing spaces in this item were not being ignored,
6004 causing some misleading effects.
6006 5. When deliver_drop_privilege was set, single queue runner processes started
6007 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6008 same command in the process it spins off) were not dropping privilege.
6010 6. When the daemon running as "exim" started a queue runner, it always
6011 re-executed Exim in the spun-off process. This is a waste of effort when
6012 deliver_drop_privilege is set. The new process now just calls the
6013 queue-runner function directly.
6019 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6022 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6023 ignored. Though the use of -oP was forcing the writing of a pid file, it
6024 was always written to the default place.
6026 3. If the message "no IP address found for host xxxx" is generated during
6027 incoming verification, it is now followed by identification of the incoming
6028 connection (so you can more easily find what provoked it).
6030 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6032 5. Added some features to "harden" Exim a bit more against certain attacks:
6034 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6035 be put in Local/Makefile. This is like the never_users runtime option,
6036 but it cannot be overridden. The default setting is "root".
6038 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6039 prefix string with which any file named in a -C command line option
6042 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6043 is retained for -C and -D only if the caller of Exim is root. Without
6044 it, the exim user may also use -C and -D and retain privilege.
6046 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6047 command line option is disabled.
6049 6. Macro names set by the -D option must start with an upper case letter, just
6050 like macro names defined in the configuration file.
6052 7. Added "dereference=" facility to LDAP.
6054 8. Two instances of the typo "uknown" in the source files are fixed.
6056 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6057 the Configure-Makefile script screwed up while processing it.
6059 10. Incorporated PCRE 4.4.
6061 11. The SMTP synchronization check was not operating right at the start of an
6062 SMTP session. For example, it could not catch a HELO sent before the client
6063 waited for the greeting. There is now a check for outstanding input at the
6064 point when the greeting is written. Because of the duplex, asynchronous
6065 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6066 way, but not yet received, when the check is performed.
6068 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6069 on TCP/IP sockets, because this apparently causes some broken clients to
6072 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6073 unchanged) from the Cygwin maintainer.
6075 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6076 was supported instead of testing for mbx. Effectively a typo.
6078 15. The spa authenticator server code was not checking that the input it
6079 received was valid base64.
6081 16. The debug output line for the "set" modifier in ACLs was not showing the
6082 name of the variable that was being set.
6084 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6086 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6087 lookup on its own. Something else was needed to trigger the lookup. For
6088 example, a match in host_lookup or the need for a host name in a host list.
6089 Now, if $sender_host_name is referenced and the host name has not yet been
6090 looked up, a lookup is performed. If the lookup fails, the variable remains
6091 empty, and $host_lookup_failed is set to "1".
6093 19. Added "eqi" as a case-independent comparison operator.
6095 20. The saslauthd authentication condition could segfault if neither service
6096 nor realm was specified.
6098 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6099 error message that was logged was misleading, and incoming SMTP
6100 connections were dropped. The message is now more accurate, and temporary
6101 errors are given to SMTP connections.
6103 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6104 (except RSET) with a 421 error, until QUIT is received. However, it was
6105 failing to send a response to QUIT.
6107 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6108 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6109 ACL is now run after the helo_try_verify_hosts code.
6111 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6112 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6113 case-independent, so other case variants are also recognized). Apparently
6114 some systems use these upper case variants.
6116 25. If more than two messages were waiting for the same host, and a transport
6117 filter was specified for the transport, Exim sent two messages over the
6118 same TCP/IP connection, and then failed with "socket operation on non-
6119 socket" when it tried to send the third.
6121 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6123 27. The extern definition of crypt16() in expand.c was not being excluded when
6124 the OS had its own crypt16() function.
6126 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6127 as a preferred synonym for return_size_limit, both as an option and as an
6130 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6132 30. Changed the default configuration ACL to relax the local part checking rule
6133 for addresses that are not in any local domains. For these addresses,
6134 slashes and pipe symbols are allowed within local parts, but the sequence
6135 /../ is explicitly forbidden.
6137 31. SPA server authentication was not clearing the challenge buffer before
6140 32. log_message in a "warn" ACL statement was writing to the reject log as
6141 well as to the main log, which contradicts the documentation and doesn't
6142 seem right (because no rejection is happening). So I have stopped it.
6144 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6145 However, I am unable to do any testing of this.
6147 34. Fixed an infelicity in the appendfile transport. When checking directories
6148 for a mailbox, to see if any needed to be created, it was accidentally
6149 using path names with one or more superfluous leading slashes; tracing
6150 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6152 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6153 contents are added to the log line that is written for every discarded
6154 recipient. (Previously a log_message setting was ignored.)
6156 36. The ${quote: operator now quotes the string if it is empty.
6158 37. The install script runs exim in order to find its version number. If for
6159 some reason other than non-existence or emptiness, which it checks, it
6160 could not run './exim', it was installing it with an empty version number,
6161 i.e. as "exim-". This error state is now caught, and the installation is
6164 38. An argument was missing from the function that creates an error message
6165 when Exim fails to connect to the socket for saslauthd authentication.
6166 This could cause Exim to crash, or give a corrupted message.
6168 39. Added isip, isip4, and isip6 to ${if conditions.
6170 40. The ACL variables $acl_xx are now saved with the message, and can be
6171 accessed later in routers, transports, and filters.
6173 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6174 strings in the file are not string-expanded.
6176 42. If a MAIL command specified a SIZE value that was too large to fit into an
6177 int variable, the check against message_size_limit failed. Such values are
6178 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6179 day this will have to be increased, but I don't think I want to be around
6180 when emails are that large.
6187 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6188 iconv() is not standard in FreeBSD.
6190 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6191 IPv6 enabled. The observed symptom was a segmentation fault on return from
6192 the function os_common_find_running_interfaces() in src/os.c.
6194 3. In the check_special_case() function in daemon.c I had used "errno" as an
6195 argument name, which causes warnings on some systems. This was basically a
6196 typo, since it was named "eno" in the comments!
6198 4. The code that waits for the clock to tick (at a resolution of some fraction
6199 of a second) so as to ensure message-id uniqueness was always waiting for
6200 at least one whole tick, when it could have waited for less. [This is
6201 almost certainly not relevant at current processor speeds, where it is
6202 unlikely to ever wait at all. But we try to future-proof.]
6204 5. The function that sleeps for a time interval that includes fractions of a
6205 second contained a race. It did not block SIGALRM between setting the
6206 timer, and suspending (a couple of lines later). If the interval was short
6207 and the sigsuspend() was delayed until after it had expired, the suspension
6208 never ended. On busy systems this could lead to processes getting stuck for
6211 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6212 process, before it forks any delivery processes. The open lookup caching
6213 mechanism meant that the open file or database connection was passed into
6214 the delivery process. The problem was that delivery processes always tidy
6215 up cached lookup data. This could cause a problem for the next delivery
6216 process started by the queue runner, because the external queue runner
6217 process does not know about the closure. So the next delivery process
6218 still has data in the lookup cache. In the case of a file lookup, there was
6219 no problem because closing a file descriptor in a subprocess doesn't affect
6220 the parent. However, if the lookup was caching a connection to a database,
6221 the connection was closed, and the second delivery process was likely to
6222 see errors such as "PGSQL: query failed: server closed the connection
6223 unexpectedly". The problem has been fixed by closing all cached lookups
6224 in a queue runner before running a delivery process.
6226 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6227 seem to have the "const" qualifier which it has on other OS. I've
6230 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6231 *of the same type* (client or server) with the same public name that an
6232 error should be diagnosed.
6234 9. When Exim looked up a host name for an IP address, but failed to find the
6235 original IP address when looking up the host name (a safety check), it
6236 output the message "<ip address> does not match any IP for NULL", which was
6237 confusing, to say the least. The bug was that the host name should have
6238 appeared instead of "NULL".
6240 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6241 user that is built into the binary, and the -C or -D options is used, root
6242 privilege is dropped before the configuration file is read. In addition,
6243 logging is switched to stderr instead of the normal log files. If the
6244 configuration then re-defines the Exim user, the unprivileged environment
6245 is probably not what is expected, so Exim logs a panic warning message (but
6248 However, if deliver_drop_privilege is set, the unprivileged state may well
6249 be exactly what is intended, so the warning has been cut out in that case,
6250 and Exim is allowed to try to write to its normal log files.
6256 1. smtp_return_error_details was not giving details for temporary sender
6257 or receiver verification errors.
6259 2. Diagnose a configuration error if two authenticators have the same public
6262 3. Exim used not to create the message log file for a message until the first
6263 delivery attempt. This could be confusing when incoming messages were held
6264 for policy or load reasons. The message log file is now created at the time
6265 the message is received, and an initial "Received" line is written to it.
6267 4. The automatically generated man page for command line options had a minor
6268 bug that caused no ill effects; however, a more serious problem was that
6269 the procedure for building the man page automatically didn't always
6270 operate. Consequently, release 4.20 contains an out-of-date version. This
6271 shouldn't happen again.
6273 5. When building Exim with embedded Perl support, the script that builds the
6274 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6275 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6277 6. The freeze_tell option was not being used for messages that were frozen on
6278 arrival, either by an ACL or by local_scan().
6280 7. Added the smtp_incomplete_transaction log selector.
6282 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6283 was accepting AUTH without a new EHLO.
6285 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6286 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6289 10. Logging of TCP/IP connections (when configured) now happens in the main
6290 daemon process instead of the child process, so that the TCP/IP connection
6291 count is more accurate (but it can never be perfect).
6293 11. The use of "drop" in a nested ACL was not being handled correctly in the
6294 outer ACL. Now, if condition failure induced by the nested "drop" causes
6295 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6296 or "require"), the connection is dropped.
6298 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6299 that yield "discard" can now be used with an "accept" or a "discard" verb,
6300 but an error is generated for any others (because I can't see a useful way
6301 to define what should happen).
6303 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6304 are now processed in the same way as lines in the Exim configuration file.
6305 In particular, continuation lines are supported.
6307 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6309 15. Added -ti meaning -t -i.
6311 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6312 domains, and warn by logging if others are found.
6314 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6315 returned from GIFCONF when Exim is trying to find the list of interfaces on
6316 a host. The code in os.c has been modified to copy each ifreq to an aligned
6317 structure in all cases.
6319 Also, in some cases, the returned ifreq's were being copied to a 'struct
6320 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6321 means the last couple of bytes of an IPv6 address could be chopped if the
6322 ifreq contained only a normal sockaddr (14 bytes storage).
6324 18. Named domain lists were not supported in the hosts_treat_as_local option.
6325 An entry such as +xxxx was not recognized, and was treated as a literal
6328 19. Ensure that header lines added by a DATA ACL are included in the reject log
6329 if the ACL subsequently rejects the message.
6331 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6332 MD5 (which is deprecated).
6334 21. When testing a filter file using -bf, Exim was writing a message when it
6335 took the sender from a "From " line in the message, but it was not doing so
6336 when it took $return_path from a Return-Path: header line. It now does.
6338 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6339 with a valid header line field name (a series of printing characters
6340 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6342 23. Changed "disc" in the source to "disk" to conform to the documentation and
6343 the book and for uniformity.
6345 24. Ignore Sendmail's -Ooption=value command line item.
6347 25. When execve() failed while trying to run a command in a pipe transport,
6348 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6349 could be confused with a return value of 69 from the command itself. This
6350 has been changed to 127, the value the shell returns if it is asked to run
6351 a non-existent command. The wording for the related log line suggests a
6352 non-existent command as the problem.
6354 26. If received_header_text expands to an empty string, do not add a Received:
6355 header line to the message. (Well, it adds a token one on the spool, but
6356 marks it "old" so that it doesn't get used or transmitted.)
6358 27. Installed eximstats 1.28 (addition of -nt option).
6360 28. There was no check for failure on the call to getsockname() in the daemon
6361 code. This can fail if there is a shortage of resources on the system, with
6362 ENOMEM, for example. A temporary error is now given on failure.
6364 29. Contrary to the C standard, it seems that in some environments, the
6365 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6366 program. Exim now does this explicitly; it affects the formatting of
6367 timestamps using strftime().
6369 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6370 complaints. I've now initialized all the variables, to avoid this.
6372 32. Header lines added by a system filter were not being "seen" during
6373 transport-time rewrites.
6375 33. The info_callback() function passed to OpenSSL is set up with type void
6376 (*)(SSL *, int, int), as described somewhere. However, when calling the
6377 function (actually a macro) that sets it up, the type void(*)() is
6378 expected. I've put in a cast to prevent warnings from picky compilers.
6380 34. If a DNS black list lookup found a CNAME record, but there were no A
6381 records associated with the domain it pointed at, Exim crashed.
6383 35. If a DNS black list lookup returned more than one A record, Exim ignored
6384 all but the first. It now scans all returned addresses if a particular IP
6385 value is being sought. In this situation, the contents of the
6386 $dnslist_value variable are a list of all the addresses, separated by a
6389 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6390 to accept a host name and all its aliases if the forward lookup for any of
6391 them yielded the IP address of the incoming connection. Now it accepts only
6392 those names whose forward lookup yields the correct IP address. Any other
6393 names are discarded. This closes a loophole whereby a rogue DNS
6394 administrator could create reverse DNS records to break through a
6395 wildcarded host restriction in an ACL.
6397 37. If a user filter or a system filter that ran in a subprocess used any of
6398 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6399 the wrong values were passed to the pipe command ($thisaddress had the
6400 value of $0, $0 had the value of $1, etc). This bug was introduced by
6401 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6403 38. Improved the line breaking for long SMTP error messages from ACLs.
6404 Previously, if there was no break point between 40 and 75 characters, Exim
6405 left the rest of the message alone. Two changes have been made: (a) I've
6406 reduced the minimum length to 35 characters; (b) if it can't find a break
6407 point between 35 and 75 characters, it looks ahead and uses the first one
6408 that it finds. This may give the occasional overlong line, but at least the
6409 remaining text gets split now.
6411 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6412 file descriptors might be low, and that setting 1000 would always raise it.
6413 It turns out that in some environments, the limit is already over 1000 and
6414 that lowering it causes trouble. So now Exim takes care not to decrease it.
6416 40. When delivering a message, the value of $return_path is set to $sender_
6417 address at the start of routing (routers may change the value). By an
6418 oversight, this default was not being set up when an address was tested by
6419 -bt or -bv, which affected the outcome if any router or filter referred to
6422 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6423 log only when "message" or "log_message" are set. However, if one of the
6424 conditions was an address verification, or a call to a nested ACL, the
6425 messages generated by the underlying test were being passed through. This
6426 no longer happens. The underlying message is available in $acl_verify_
6427 message for both "message" and "log_message" expansions, so it can be
6428 passed through if needed.
6430 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6431 new expansion $bh_ to give the encoded byte string without charset
6432 translation. Translation happens only if iconv() is available; HAVE_ICONV
6433 indicates this at build time. HEADERS_CHARSET gives the charset to
6434 translate to; headers_charset can change it in the configuration, and
6435 "headers charset" can change it in an individual filter file.
6437 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6438 that creates RFC 2047 encoded "words" labels them as that charset instead
6439 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6440 expansion operator; (ii) when Exim creates a From: line for a local
6441 message; (iii) when a header line is rewritten to include a "phrase" part.
6443 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6444 buggy, causing it to skip the first lines of messages whose message ID
6445 ended in 'D'. This would not have bitten before Exim release 4.14, because
6446 message IDs were unlikely to end in 'D' before then. The effect was to have
6447 incorrect size information for certain domains.
6449 45. #include "config.h" was missing at the start of the crypt16.c module. This
6450 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6453 46. If there was a timeout during a "random" callout check, Exim treated it as
6454 a failure of the random address, and carried on sending RSET and the real
6455 address. If the delay was just some slowness somewhere, the response to the
6456 original RCPT would be taken as a response to RSET and so on, causing
6457 mayhem of various kinds.
6459 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6460 when I implemented it. It didn't allow for the fact that some option values
6461 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6462 the right test for positive values.
6464 48. Domain names in DNS records are case-independent. Exim always looks them up
6465 in lower case. Some resolvers return domain names in exactly the case they
6466 appear in the zone file, that is, they may contain uppercase letters. Not
6467 all resolvers do this - some return always lower case. Exim was treating a
6468 change of case by a resolver as a change of domain, similar to a widening
6469 of a domain abbreviation. This triggered its re-routing code and so it was
6470 trying to route what was effectively the same domain again. This normally
6471 caused routing to fail (because the router wouldn't handle the domain
6472 twice). Now Exim checks for this case specially, and just changes the
6473 casing of the domain that it ultimately uses when it transmits the message
6476 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6479 50. If a filter generated a file delivery with a non-absolute name (possible if
6480 no home directory exists for the router), the forbid_file option was not
6483 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6484 the existing equality matching.
6486 52. Exim was using ints instead of ino_t variables in some places where it was
6487 dealing with inode numbers.
6489 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6490 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6491 called TMPDIR, and if it finds it is different, it changes its value.
6493 54. The smtp_printf() function is now made available to local_scan() so
6494 additional output lines can be written before returning. There is also an
6495 smtp_fflush() function to enable the detection of a dropped connection.
6496 The variables smtp_input and smtp_batched_input are exported to
6499 55. Changed the default runtime configuration: the message "Unknown user"
6500 has been removed from the ACL, and instead placed on the localuser router,
6501 using the cannot_route_message feature. This means that any verification
6502 failures that generate their own messages won't get overridden. Similarly,
6503 the "Unrouteable address" message that was in the ACL for unverifiable
6504 relay addresses has also been removed.
6506 56. Added hosts_avoid_esmtp to the smtp transport.
6508 57. The exicyclog script was not checking for the esoteric option
6509 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6510 will work only if exicyclog is run under the appropriate euid.
6512 58. Following a discussion on the list, the rules by which Exim recognises line
6513 endings on incoming messages have been changed. The -dropcr and drop_cr
6514 options are now no-ops, retained only for backwards compatibility. The
6515 following line terminators are recognized: LF CRLF CR. However, special
6516 processing applies to CR:
6518 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6519 nor a local message in the state where . is a terminator.
6521 (ii) If a bare CR is encountered in a header line, an extra space is added
6522 after the line terminator so as not to end the header. The reasoning
6523 behind this is that bare CRs in header lines are most likely either
6524 to be mistakes, or people trying to play silly games.
6526 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6527 was being incorrectly given as 18 bytes larger than it should have been.
6528 This is a VOB (very old bug).
6530 60. This may never have affected anything current, but just in case it has:
6531 When the local host is found other than at the start of a list of hosts,
6532 the local host, those with the same MX, and any that follow, are discarded.
6533 When the list in question was part of a longer list of hosts, the following
6534 hosts (not currently being processed) were also being discarded. This no
6535 longer happens. I'm not sure if this situation could ever has previously
6538 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6541 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6542 header. This is something that is recommended in a new Internet Draft, and
6543 is something that is documented as being done by Sendmail. There are two
6544 possible values. For messages generated by the autoreply transport, Exim
6547 Auto-Submitted: auto-replied
6549 whereas for all other generated messages (e.g. bounces) it adds
6551 Auto-Submitted: auto-generated
6553 63. The "personal" condition in filters now includes a test for the
6554 Auto-Submitted: header. If it contains the string "auto-" the message it
6555 not considered personal.
6557 64. Added rcpt_include_affixes as a generic transport option.
6559 65. Added queue_only_override (default true).
6561 66. Added the syslog_duplication option.
6563 67. If what should have been the first header line of a message consisted of
6564 a space followed by a colon, Exim was mis-interpreting it as a header line.
6565 It isn't of course - it is syntactically invalid and should therefore be
6566 treated as the start of the message body. The misbehaviour could have
6567 caused a number of strange effects, including loss of data in subsequent
6568 header lines, and spool format errors.
6570 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6571 client host had authenticated. This control can now be exercised by an ACL
6572 for more flexibility.
6574 69. By default, callouts do not happen when testing with -bh. There is now a
6575 variant, -bhc, which does actually run the callout code, including
6576 consulting and updating the callout cache.
6578 70. Added support for saslauthd authentication, courtesy of Alexander
6581 71. If statvfs() failed on the spool or log directories while checking their
6582 size for availability, Exim confusingly gave the error "space shortage".
6583 Furthermore, in debugging mode it crashed with a floating point exception.
6584 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6585 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6586 serious problem, Exim now writes to the main and panic logs when this
6587 happens, with details of the failure. It then refuses to accept the
6588 incoming message, giving the message "spool directory problem" or "log
6589 directory problem" with a 421 code for SMTP messages.
6591 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6592 0, 1, and 2 exist, because some OS complain for execs without them (see
6593 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6594 descriptors. However, the code omitted to check that the open succeeded,
6595 causing mysterious errors if for some reason the permissions on /dev/null
6596 got screwed. Now Exim writes a message to the main and panic logs, and
6597 bombs out if it can't open /dev/null.
6599 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6600 interact so that it is all more flexible. It is supposed to remain
6601 backwards compatible. Also added extra_local_interfaces.
6603 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6604 to bomb out with an assertion failure - to the client this appears as a
6605 connection drop. This problem occurs in the part of the code that was taken
6606 from the Samba project. Fortunately, the assertion is in a very simple
6607 function, so I have fixed this by reproducing the function inline in the
6608 one place where it is called, and arranging for authentication to fail
6609 instead of killing the process with assert().
6611 75. The SPA client code was not working when the server requested OEM rather
6612 than Unicode encoding.
6614 76. Added code to make require_files with a specific uid setting more usable in
6615 the case where statting the file as root fails - usually a non-root-mounted
6616 NFS file system. When this happens and the failure is EACCES, Exim now
6617 forks a subprocess and does the per-uid checking as the relevant uid.
6619 77. Added process_log_path.
6621 78. If log_file_path was not explicitly set, a setting of check_log_space or
6622 check_log_inodes was ignored.
6624 79. If a space check for the spool or log partitions fails, the incident is now
6625 logged. Of course, in the latter case the data may get lost...
6627 80. Added the %p formatting code to string_format() so that it can be used to
6628 print addresses in debug_print(). Adjusted all the address printing in the
6629 debugging in store.c to use %p rather than %d.
6631 81. There was a concern that a line of code in smtp_in.c could overflow a
6632 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6633 initially expressed, the concern was not well-founded, because trailing
6634 spaces are removed early. However, if the trailing spaces were followed by
6635 a NULL, they did not get removed, so the overflow was possible. Two fixes
6638 (a) I re-wrote the offending code in a cleaner fashion.
6639 (b) If an incoming SMTP command contains a NULL character, it is rejected
6642 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6643 runs initgroups(), so that if the Exim user is in any additional groups,
6644 they will be used during message reception.
6650 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob