1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
12 than -2003 did; needs libidn2 in addition to libidn.
14 JH/02 The path option on a pipe transport is now expanded before use.
16 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
17 Patch provided by "Björn", documentation fix added too.
19 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
20 missing a wire-to-host endian conversion.
22 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
23 close after a BDAT command line could be taken as a following command,
24 giving a synch failure. Fix by only checking for synch immediately
25 before acknowledging the chunk.
27 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
28 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
29 macro. Patches provided by Josh Soref.
31 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
32 Previously we did not; the RFC seems ambiguous and VRFY is not listed
33 by IANA as a service extension. However, John Klensin suggests that we
36 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
37 the dkim code may be unix-mode line endings rather than smtp wire-format
38 CRLF, so prepend a CR to any bare LF.
40 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
41 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
43 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
44 the first were themselves being wrongly included in the feed into dkim
45 processing; with most chunk sizes in use this resulted in an incorrect
46 body hash calculated value.
48 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
49 DKIM signature block, for verification. Although advised against by
50 standards it is specifically not ruled illegal.
52 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
54 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
55 missing a body hash (the bh= tag).
57 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
58 It seems that HAProxy sends the Proxy Protocol information in clear and
59 only then does a TLS startup, so do the same.
61 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
62 TCP connections (such as for Spamd) unless the daemon successfully set
63 Fast Open mode on its listening sockets. This fixes breakage seen on
64 too-old kernels or those not configured for Fast Open, at the cost of
65 requiring both directions being enabled for TFO, and TFO never being used
66 by non-daemon-related Exim processes.
68 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
69 endings, at least on the first header line. Try to canonify any that get
70 past that check, despite the cost.
72 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
73 now limited to an arbitrary five deep, while parsing addresses with the
74 strip_excess_angle_brackets option enabled.
76 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
77 instead leave the unprompted TLS handshake in socket buffer for the
78 TLS library to consume.
80 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
86 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
87 supports it and a size is available (ie. the sending peer gave us one).
89 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
90 years of being deprecated). The replacements are the ACLs
91 acl_smtp_mime and acl_not_smtp_mime.
93 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
94 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
95 or even an in-clear connection were permitted. Now, if the host lookup
96 was dnssec and dane was requested then the host is only used if the
97 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
98 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
99 if one fails this test.
100 This means that a poorly-configured remote DNS will make it incommunicado;
101 but it protects against a DNS-interception attack on it.
103 JH/04 Bug 1810: make continued-use of an open smtp transport connection
104 non-noisy when a race steals the message being considered.
106 JH/05 If main configuration option tls_certificate is unset, generate a
107 self-signed certificate for inbound TLS connections.
109 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
110 in rewrites and routers.
112 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
113 and logged a warning sing 4.83; now they are a configuration file error.
115 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
116 (lacking @domain). Apply the same qualification processing as RCPT.
118 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
120 JH/10 Support ${sha256:} applied to a string (as well as the previous
123 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
124 a cutthrough deliver is pending, as we always want to make a connection.
125 This also avoids re-routing the message when later placing the cutthrough
126 connection after a verify cache hit.
127 Do not update it with the verify result either.
129 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
130 when routing results in more than one destination address.
132 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
133 signing (which inhibits the cutthrough capability). Previously only
134 the presence of an option was tested; now an expansion evaluating as
135 empty is permissible (obviously it should depend only on data available
136 when the cutthrough connection is made).
138 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
139 the relevant preceding SMTP command did not note the pipelining mode.
141 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
142 Previously they were not counted.
144 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
145 as one having no matching records. Previously we deferred the message
146 that needed the lookup.
148 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
149 distinguished as "(=".
151 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
152 for missing MX records. Previously it only worked for missing A records.
154 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
156 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
157 after the data-go-ahead and data-ack. Patch from Jason Betts.
159 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
160 even for a "none" policy. Patch from Tony Meyer.
162 JH/22 Fix continued use of a connection for further deliveries. If a port was
163 specified by a router, it must also match for the delivery to be
166 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
167 When one of the recipients of a message was unsuitable for the connection
168 (has no matching addresses), we lost track of needing to mark it
169 deferred. As a result mail would be lost.
171 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
173 JH/25 Decoding ACL controls is now done using a binary search; the source code
174 takes up less space and should be simpler to maintain. Merge the ACL
175 condition decode tables also, with similar effect.
177 JH/26 Fix problem with one_time used on a redirect router which returned the
178 parent address unchanged. A retry would see the parent address marked as
179 delivered, so not attempt the (identical) child. As a result mail would
182 JH/27 Fix a possible security hole, wherein a process operating with the Exim
183 UID can gain a root shell. Credit to http://www.halfdog.net/ for
184 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
187 JH/28 Enable {spool,log} filesystem space and inode checks as default.
188 Main config options check_{log,spool}_{inodes,space} are now
189 100 inodes, 10MB unless set otherwise in the configuration.
191 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
192 Previously it only applied to the main-section connection policy
195 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
197 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
198 by me. Added RFC7919 DH primes as an alternative.
200 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
201 are not in the system include path.
203 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
204 GnuTLS, when a session startup failed (eg because the client disconnected)
205 Exim did stdio operations after fclose. This was exposed by a recent
206 change which nulled out the file handle after the fclose.
208 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
209 signed directly by the cert-signing cert, rather than an intermediate
210 OCSP-signing cert. This is the model used by LetsEncrypt.
212 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
214 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
215 an incoming connection.
217 HS/02 Bug 1802: Do not half-close the connection after sending a request
220 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
221 fallback to "prime256v1".
223 JH/34 SECURITY: Use proper copy of DATA command in error message.
224 Could leak key material. Remotely exploitable. CVE-2016-9963.
230 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
231 and 3.4.4 - once the server is enabled to respond to an OCSP request
232 it does even when not requested, resulting in a stapling non-aware
233 client dropping the TLS connection.
235 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
236 support variable-length bit vectors. No functional change.
238 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
239 The I= interface field on outgoing lines is now after the H= remote
240 host field, same as incoming lines. There is a separate
241 outgoing_interface log selector which allows you to disable the
244 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
245 If not running log_selector +smtp_connection the mainlog would be held
246 open indefinitely after a "too many connections" event, including to a
247 deleted file after a log rotate. Leave the per net connection logging
248 leaving it open for efficiency as that will be quickly detected by the
249 check on the next write.
251 HS/01 Bug 1671: Fix post transport crash.
252 Processing the wait-<transport> messages could crash the delivery
253 process if the message IDs didn't exist for some reason. When
254 using 'split_spool_directory=yes' the construction of the spool
255 file name failed already, exposing the same netto behaviour.
257 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
258 mime_regex ACL conditions.
260 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
261 to DSN fail messages (bounces): remote IP, remote greeting, remote response
262 to HELO, local diagnostic string.
264 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
265 log line to debug. Even when configured with a tls authenticator many
266 client connections are expected to not authenticate in this way, so
267 an authenticate fail is not an error.
269 HS/02 Add the Exim version string to the process info. This way exiwhat
270 gives some more detail about the running daemon.
272 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
273 matter for fast-change records such as DNSBLs.
275 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
276 retry record, even if constant. There may be multiple transports with
277 different interface settings and the retry behaviour needs to be kept
280 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
282 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
284 JH/10 Bug 840: fix log_defer_output option of pipe transport
286 JH/11 Bug 830: use same host for all RCPTS of a message, even under
287 hosts_randomize. This matters a lot when combined with mua_wrapper.
289 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
290 ${quote_pgsql:<string>} operator.
292 JH/13 Bug 1708: avoid misaligned access in cached lookup.
294 JH/14 Change header file name for freeradius-client. Relevant if compiling
295 with Radius support; from the Gentoo tree and checked under Fedora.
297 JH/15 Bug 1712: Introduce $prdr_requested flag variable
299 JH/16 Bug 1714: Permit an empty string as expansion result for transport
300 option transport_filter, meaning no filtering.
302 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
304 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
305 defaults to "*" (all hosts). The variable is now available when not built
306 with TLS, default unset, mainly to enable keeping the testsuite sane.
307 If a server certificate is not supplied (via tls_certificate) an error is
308 logged, and clients will find TLS connections fail on startup. Presumably
309 they will retry in-clear.
310 Packagers of Exim are strongly encouraged to create a server certificate
311 at installation time.
313 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
314 with the $config_file variable.
316 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
317 in transport context, after the attempt, and per-recipient. The latter type
318 is per host attempted. The event data is the error message, and the errno
319 information encodes the lookup type (A vs. MX) used for the (first) host,
320 and the trailing two digits of the smtp 4xx response.
322 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
323 to write to mainlog (or rejectlog, paniclog) in the window between file
324 creation and permissions/ownership being changed. Particularly affects
325 installations where exicyclog is run as root, rather than exim user;
326 result is that the running daemon panics and dies.
328 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
330 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
331 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
332 "pri" and "weight". Note that the previous implicit priority given by the
333 list order is no longer honoured.
335 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
338 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
339 by defining SUPPORT_SOCKS.
341 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
342 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
343 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
344 variables are renamed to proxy_{local,external}_{address,port}.
346 JH/27 Move Internationalisation support from Experimental to mainline, enabled
347 for a build by defining SUPPORT_I18N
349 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
350 of the query string, and make ${quote_redis:} do that quoting.
352 JH/29 Move Events support from Experimental to mainline, enabled by default
353 and removable for a build by defining DISABLE_EVENT.
355 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
357 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
358 cached by the daemon.
360 JH/32 Move Redis support from Experimental to mainline, enabled for a build
361 by defining LOOKUP_REDIS. The libhiredis library is required.
363 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
364 keys are given for lookup.
366 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
367 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
368 only supported when built with TLS support. The PolarSSL SHA routines
369 are still used when the TLS library is too old for convenient support.
371 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
372 openssl_options), for security. OpenSSL forces this from version 1.1.0
373 server-side so match that on older versions.
375 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
376 allocation for $value could be released as the expansion processing
377 concluded, but leaving the global pointer active for it.
379 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
380 and to use the domains and local_parts ACL conditions.
382 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
383 incorrectly not doubled on cutthrough transmission, hence seen as a
384 body-termination at the receiving system - resulting in truncated mails.
385 Commonly the sender saw a TCP-level error, and retransmitted the message
386 via the normal store-and-forward channel. This could result in duplicates
387 received - but deduplicating mailstores were liable to retain only the
388 initial truncated version.
390 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
392 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
394 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
395 we're in there, support oversigning also; bug 1309.
397 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
399 HS/04 Add support for keep_environment and add_environment options.
401 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
402 either intentional arithmetic overflow during PRNG, or testing config-
405 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
406 delivery resulted in actual delivery. Cancel cutthrough before DATA
409 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
410 rejects a recipient: pass the reject to the originator.
412 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
413 Many were false-positives and ignorable, but it's worth fixing the
416 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
417 for the new environment-manipulation done at startup. Move the routines
418 from being local to tls.c to being global via the os.c file.
420 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
421 an extract embedded as result-arg for a map, the first arg for extract
422 is unavailable so we cannot tell if this is a numbered or keyed
423 extraction. Accept either.
429 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
432 JH/02 The smtp transport option "multi_domain" is now expanded.
434 JH/03 The smtp transport now requests PRDR by default, if the server offers
437 JH/04 Certificate name checking on server certificates, when exim is a client,
438 is now done by default. The transport option tls_verify_cert_hostnames
439 can be used to disable this per-host. The build option
440 EXPERIMENTAL_CERTNAMES is withdrawn.
442 JH/05 The value of the tls_verify_certificates smtp transport and main options
443 default to the word "system" to access the system default CA bundle.
444 For GnuTLS, only version 3.0.20 or later.
446 JH/06 Verification of the server certificate for a TLS connection is now tried
447 (but not required) by default. The verification status is now logged by
448 default, for both outbound TLS and client-certificate supplying inbound
451 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
454 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
455 Status Notification (bounce) messages are now MIME format per RFC 3464.
456 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
457 under the control of the dsn_advertise_hosts option, and routers may
458 have a dsn_lasthop option.
460 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
461 default, modifiable by a malware= option. The list separator for
462 the options can now be changed in the usual way. Bug 68.
464 JH/10 The smtp_receive_timeout main option is now expanded before use.
466 JH/11 The incoming_interface log option now also enables logging of the
467 local interface on delivery outgoing connections.
469 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
470 if the interface and destination host and port all match.
472 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
475 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
476 Patch from Andrew Lewis.
478 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
479 now supports optional time-restrictions, weighting, and priority
480 modifiers per server. Patch originally by <rommer@active.by>.
482 JH/16 The spamd_address main option now supports a mixed list of local
483 and remote servers. Remote servers can be IPv6 addresses, and
484 specify a port-range.
486 JH/17 Bug 68: The spamd_address main option now supports an optional
487 timeout value per server.
489 JH/18 Bug 1581: Router and transport options headers_add/remove can
490 now have the list separator specified.
492 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
495 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
498 JH/21 Support for the A6 type of dns record is withdrawn.
500 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
501 rather than the verbs used.
503 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
504 from 255 to 1024 chars.
506 JH/24 Verification callouts now attempt to use TLS by default.
508 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
509 are generic router options now. The defaults didn't change.
511 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
512 Original patch from Alexander Shikoff, worked over by JH.
514 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
515 files if ZCAT_COMMAND is not executable.
517 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
519 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
521 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
522 Normally benign, it bites when the pair was led to by a CNAME;
523 modern usage is to not canonicalize the domain to a CNAME target
524 (and we were inconsistent anyway for A-only vs AAAA+A).
526 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
528 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
529 when evaluating $sender_host_dnssec.
531 JH/31 Check the HELO verification lookup for DNSSEC, adding new
532 $sender_helo_dnssec variable.
534 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
536 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
538 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
540 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
541 documented as working, but never had. Support all but $spam_report.
543 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
544 added for tls authenticator.
546 HS/03 Add perl_taintmode main config option
552 TL/01 When running the test suite, the README says that variables such as
553 no_msglog_check are global and can be placed anywhere in a specific
554 test's script, however it was observed that placement needed to be near
555 the beginning for it to behave that way. Changed the runtest perl
556 script to read through the entire script once to detect and set these
557 variables, reset to the beginning of the script, and then run through
558 the script parsing/test process like normal.
560 TL/02 The BSD's have an arc4random API. One of the functions to induce
561 adding randomness was arc4random_stir(), but it has been removed in
562 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
563 function when detected.
565 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
566 cause callback expansion.
568 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
569 syntax errors in an expansion can be treated as a string instead of
570 logging or causing an error, due to the internal use of bool_lax
571 instead of bool when processing it.
573 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
574 server certificates when making smtp deliveries.
576 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
578 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
580 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
582 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
583 Merged patch from Sebastian Wiedenroth.
585 JH/05 Fix results-pipe from transport process. Several recipients, combined
586 with certificate use, exposed issues where response data items split
587 over buffer boundaries were not parsed properly. This eventually
588 resulted in duplicates being sent. This issue only became common enough
589 to notice due to the introduction of connection certificate information,
590 the item size being so much larger. Found and fixed by Wolfgang Breyha.
592 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
593 size buffer was used, resulting in syntax errors when an expansion
596 JH/07 Add support for directories of certificates when compiled with a GnuTLS
597 version 3.3.6 or later.
599 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
600 is EXPERIMENTAL_EVENT, the main-configuration and transport options
601 both become "event_action", the variables become $event_name, $event_data
602 and $event_defer_errno. There is a new variable $verify_mode, usable in
603 routers, transports and related events. The tls:cert event is now also
604 raised for inbound connections, if the main configuration event_action
607 TL/06 In test suite, disable OCSP for old versions of openssl which contained
608 early OCSP support, but no stapling (appears to be less than 1.0.0).
610 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
611 server certificate names available under the smtp transport option
612 "tls_verify_cert_hostname" now do not permit multi-component wildcard
615 JH/10 Time-related extraction expansions from certificates now use the main
616 option "timezone" setting for output formatting, and are consistent
617 between OpenSSL and GnuTLS compilations. Bug 1541.
619 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
620 encoded parameter in the incoming message. Bug 1558.
622 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
623 include certificate info, eximon was claiming there were spoolfile
626 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
628 JH/14 Log delivery-related information more consistently, using the sequence
629 "H=<name> [<ip>]" wherever possible.
631 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
632 are problematic for Debian distribution, omit them from the release
635 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
637 JH/16 Fix string representation of time values on 64bit time_t architectures.
640 JH/17 Fix a null-indirection in certextract expansions when a nondefault
641 output list separator was used.
646 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
647 checkers that were complaining about end of non-void function with no
650 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
651 This was a regression introduced in 4.83 by another bugfix.
653 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
655 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
656 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
662 TF/01 Correctly close the server side of TLS when forking for delivery.
664 When a message was received over SMTP with TLS, Exim failed to clear up
665 the incoming connection properly after forking off the child process to
666 deliver the message. In some situations the subsequent outgoing
667 delivery connection happened to have the same fd number as the incoming
668 connection previously had. Exim would try to use TLS and fail, logging
669 a "Bad file descriptor" error.
671 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
672 utilities have not been installed.
674 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
675 temporary space as the ACL may create new global variables.
677 TL/01 LDAP support uses per connection or global context settings, depending
678 upon the detected version of the libraries at build time.
680 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
681 to extract and use the src ip:port in logging and expansions as if it
682 were a direct connection from the outside internet. PPv2 support was
683 updated based on HAProxy spec change in May 2014.
685 JH/02 Add ${listextract {number}{list}{success}{fail}}.
687 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
688 Properly escape header and check for NULL return.
690 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
693 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
695 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
696 characters in header names, implemented as a verify condition.
697 Contributed by Michael Fischer v. Mollard.
699 TL/05 Rename SPF condition results err_perm and err_temp to standardized
700 results permerror and temperror. Previous values are deprecated but
701 still accepted. In a future release, err_perm and err_temp will be
702 completely removed, which will be a backward incompatibility if the
703 ACL tests for either of these two old results. Patch contributed by
704 user bes-internal on the mailing list.
706 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
708 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
709 selectors, in both main and reject logs.
711 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
714 JH/07 Add malware type "sock" for talking to simple daemon.
716 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
718 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
719 routers/transports under cutthrough routing.
721 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
722 numbers. Touch up "bool" conditional to keep the same definition.
724 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
726 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
728 JH/12 Expand items in router/transport headers_add or headers_remove lists
729 individually rather than the list as a whole. Bug 1452.
731 Required for reasonable handling of multiple headers_ options when
732 they may be empty; requires that headers_remove items with embedded
733 colons must have them doubled (or the list-separator changed).
735 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
736 view the policy declared in the DMARC record. Currently, $dmarc_status
737 is a combined value of both the record presence and the result of the
740 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
742 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
743 dnslookup router and the smtp transport (applying to the forward
746 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
747 of ldap servers used for a specific lookup. Patch provided by Heiko
750 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
751 New variable $lookup_dnssec_authenticated for observability.
753 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
754 Patch submitted by Lars Timman.
756 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
758 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
759 Requires trusted mode and valid format message id, aborts otherwise.
760 Patch contributed by Heiko Schlichting.
762 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
763 certextract with support for various fields. Bug 1358.
765 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
766 is requested by default, modifiable by smtp transport option
769 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
770 operate on certificate variables to give certificate fingerprints
771 Also new ${sha256:cert_variable}.
773 JH/23 The PRDR feature is moved from being Experimental into the mainline.
775 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
776 Christian Aistleitner.
778 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
780 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
781 file. Patch from Wolfgang Breyha.
783 JH/25 Expand the coverage of the delivery $host and $host_address to
784 client authenticators run in verify callout. Bug 1476.
786 JH/26 Port service names are now accepted for tls_on_connect_ports, to
787 align with daemon_smtp_ports. Bug 72.
789 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
790 support and error reporting did not work properly.
792 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
793 and is readable. Patch from Andrew Colin Kissa.
795 TL/14 Enhance documentation of ${run expansion and how it parses the
796 commandline after expansion, particularly in the case when an
797 unquoted variable expansion results in an empty value.
799 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
801 PP/02 Fix internal collision of T_APL on systems which support RFC3123
802 by renaming away from it. Addresses GH issue 15, reported by
805 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
807 TL/15 SECURITY: prevent double expansion in math comparison functions
808 (can expand unsanitized data). Not remotely exploitable.
815 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
817 PP/02 Make -n do something, by making it not do something.
818 When combined with -bP, the name of an option is not output.
820 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
823 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
824 $sender_host_name and config options to manage this, and basic check
827 PP/05 DSCP support for outbound connections and control modifier for inbound.
829 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
830 (Only plugin which currently uses this is kerberos4, which nobody should
831 be using, but we should make it available and other future plugins might
832 conceivably use it, even though it would break NAT; stuff *should* be
833 using channel bindings instead).
835 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
836 name; added for Sendmail compatibility; requires admin caller.
837 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
838 just ignore it); requires trusted caller.
839 Also parse but ignore: -Ac -Am -X<logfile>
842 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
844 TL/02 Add +smtp_confirmation as a default logging option.
846 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
847 Patch by Magnus Holmgren from 2007-02-20.
849 TL/04 Bugzilla 1281 - Spec typo.
850 Bugzilla 1283 - Spec typo.
851 Bugzilla 1290 - Spec grammar fixes.
853 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
855 TL/06 Add Experimental DMARC support using libopendmarc libraries.
857 TL/07 Fix an out of order global option causing a segfault. Reported to dev
858 mailing list by by Dmitry Isaikin.
860 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
862 JH/02 Support "G" suffix to numbers in ${if comparisons.
864 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
866 NM/01 Bugzilla 1197 - Spec typo
867 Bugzilla 1196 - Spec examples corrections
869 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
871 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
872 gnutls_enable_pkcs11, but renamed to more accurately indicate its
875 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
876 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
878 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
879 "acl {{name}{arg}...}", and optional args on acl condition
882 JH/05 Permit multiple router/transport headers_add/remove lines.
884 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
886 JH/07 Avoid using a waiting database for a single-message-only transport.
887 Performance patch from Paul Fisher. Bugzilla 1262.
889 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
892 JH/09 Add $headers_added variable, with content from use of ACL modifier
893 add_header (but not yet added to the message). Bugzilla 199.
895 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
896 Pulled from Bugzilla 817 by Wolfgang Breyha.
898 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
900 (nb: this is the same fix as in Exim 4.80.1)
902 JH/11 Add A= logging on delivery lines, and a client_set_id option on
905 JH/12 Add optional authenticated_sender logging to A= and a log_selector
908 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
910 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
911 advertise SMTP AUTH mechanism to us, instead of a generic
912 protocol violation error. Also, make Exim more robust to bad
913 data from the Dovecot auth socket.
915 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
917 When a queue runner is handling a message, Exim first routes the
918 recipient addresses, during which it prunes them based on the retry
919 hints database. After that it attempts to deliver the message to
920 any remaining recipients. It then updates the hints database using
923 So if a recipient address works intermittently, it can get repeatedly
924 deferred at routing time. The retry hints record remains fresh so the
925 address never reaches the final cutoff time.
927 This is a fairly common occurrence when a user is bumping up against
928 their storage quota. Exim had some logic in its local delivery code
929 to deal with this. However it did not apply to per-recipient defers
930 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
932 This change adds a proper retry rule check during routing so that the
933 final cutoff time is checked against the message's age. We only do
934 this check if there is an address retry record and there is not a
935 domain retry record; this implies that previous attempts to handle
936 the address had the retry_use_local_parts option turned on. We use
937 this as an approximation for the destination being like a local
938 delivery, as in LMTP.
940 I suspect this new check makes the old local delivery cutoff check
941 redundant, but I have not verified this so I left the code in place.
943 TF/02 Correct gecos expansion when From: is a prefix of the username.
945 Test 0254 submits a message to Exim with the header
949 When I ran the test suite under the user fanf2, Exim expanded
950 the header to contain my full name, whereas it should have added
951 a Resent-Sender: header. It erroneously treats any prefix of the
952 username as equal to the username.
954 This change corrects that bug.
956 GF/01 DCC debug and logging tidyup
957 Error conditions log to paniclog rather than rejectlog.
958 Debug lines prefixed by "DCC: " to remove any ambiguity.
960 TF/03 Avoid unnecessary rebuilds of lookup-related code.
962 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
963 Bug spotted by Jeremy Harris; was flawed since initial commit.
964 Would have resulted in OCSP responses post-SNI triggering an Exim
965 NULL dereference and crash.
967 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
969 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
970 Bug detection, analysis and fix by Samuel Thibault.
971 Bugzilla 1331, Debian bug #698092.
973 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
975 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
976 Server implementation by Todd Lyons, client by JH.
977 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
978 config variable "prdr_enable" controls whether the server
979 advertises the facility. If the client requests PRDR a new
980 acl_data_smtp_prdr ACL is called once for each recipient, after
981 the body content is received and before the acl_smtp_data ACL.
982 The client is controlled by both of: a hosts_try_prdr option
983 on the smtp transport, and the server advertisement.
984 Default client logging of deliveries and rejections involving
985 PRDR are flagged with the string "PRDR".
987 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
988 fclose(). Diagnosis by Todd Lyons.
990 PP/17 Update configure.default to handle IPv6 localhost better.
991 Patch by Alain Williams (plus minor tweaks).
994 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
995 This is now consistent with GnuTLS, and is now documented: the
996 previous undocumented portable approach to treating the option as
997 unset was to force an expansion failure. That still works, and
998 an empty string is now equivalent.
1000 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1001 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1002 not performing validation itself.
1004 PP/20 Added force_command boolean option to pipe transport.
1005 Patch from Nick Koston, of cPanel Inc.
1007 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1010 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1012 PP/21 Fix eximon continuous updating with timestamped log-files.
1013 Broken in a format-string cleanup in 4.80, missed when I repaired the
1014 other false fix of the same issue.
1015 Report and fix from Heiko Schlichting.
1018 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1019 Report from Prashanth Katuri.
1021 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1022 It's SecureTransport, so affects any MacOS clients which use the
1023 system-integrated TLS libraries, including email clients.
1025 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1026 using a MIME ACL for non-SMTP local injection.
1027 Report and assistance in diagnosis by Warren Baker.
1029 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1031 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1033 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1034 last id that failed so it may be referenced in subsequent ACL's.
1036 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1039 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1040 ldap library initialization, allowing self-signed CA's to be
1041 used. Also properly sets require_cert option later in code by
1042 using NULL (global ldap config) instead of ldap handle (per
1043 session). Bug diagnosis and testing by alxgomz.
1045 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1046 the src/util/ subdirectory.
1048 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1049 renamed to Transport Post Delivery Action by Jeremy Harris, as
1052 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1053 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1054 redis_servers = needs to be configured which will be used by the redis
1055 lookup. Patch from Warren Baker, of The Packet Hub.
1057 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1059 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1060 hostname or reverse DNS when processing a host list. Used suggestions
1061 from multiple comments on this bug.
1063 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1065 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1066 Missed a few lines, added it to make the runtest require no keyboard
1069 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1070 contains upper case chars. Make router use caseful_local_part.
1072 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1073 support when GnuTLS has been built with p11-kit.
1079 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1081 This, or similar/improved, will also be change PP/11 of 4.82.
1087 PP/01 Handle short writes when writing local log-files.
1088 In practice, only affects FreeBSD (8 onwards).
1089 Bugzilla 1053, with thanks to Dmitry Isaikin.
1091 NM/01 Bugzilla 949 - Documentation tweak
1093 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1096 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1098 PP/02 Implemented gsasl authenticator.
1100 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1102 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1103 `pkg-config foo` for cflags/libs.
1105 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1106 with rest of GSASL and with heimdal_gssapi.
1108 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1109 `pkg-config foo` for cflags/libs for the TLS implementation.
1111 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1112 properties get this fed in as external SSF. A number of robustness
1113 and debugging improvements to the cyrus_sasl authenticator.
1115 PP/08 cyrus_sasl server now expands the server_realm option.
1117 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1118 Patch by Jeremy Harris.
1120 PP/10 Added dbmjz lookup type.
1122 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1124 PP/12 MAIL args handles TAB as well as SP, for better interop with
1125 non-compliant senders.
1126 Analysis and variant patch by Todd Lyons.
1128 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1129 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1130 Patch from Dirk Mueller <dmueller@suse.com>
1132 PP/13 tls_peerdn now print-escaped for spool files.
1133 Observed some $tls_peerdn in wild which contained \n, which resulted
1134 in spool file corruption.
1136 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1137 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1138 or write after TLS renegotiation, which otherwise led to messages
1141 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1142 as a tracking header (ie: a signed header comes before the signature).
1143 Patch from Wolfgang Breyha.
1145 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1146 comma-sep list; embedded commas doubled.
1148 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1150 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1152 Report and patch from Dmitry Banschikov.
1154 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1155 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1156 failures. We appear to now support TLS1.1+ with Exim.
1158 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1159 lets Exim select keys and certificates based upon TLS SNI from client.
1160 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1161 before an outbound SMTP session. New log_selector, +tls_sni.
1163 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1164 NULL dereference. Report and patch from Alun Jones.
1166 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1167 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1168 Not seeing resolver debug output on NetBSD, but suspect this is a
1169 resolver implementation change.
1171 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1172 Left warnings. Added "eximon gdb" invocation mode.
1174 PP/21 Defaulting "accept_8bitmime" to true, not false.
1176 PP/22 Added -bw for inetd wait mode support.
1178 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1179 locate the relevant includes and libraries. Made this the default.
1181 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1182 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1184 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1185 This may cause build issues on older platforms.
1187 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1188 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1189 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1190 Added SNI support via GnuTLS too.
1191 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1193 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1195 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1196 Applied second patch from Janne, implementing suggestion to default
1197 multiple-strings-in-record handling to match SPF spec.
1199 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1201 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1202 read-only, out of scope).
1203 Patch from Wolfgang Breyha, report from Stuart Northfield.
1205 PP/29 Fix three issues highlighted by clang analyser static analysis.
1206 Only crash-plausible issue would require the Cambridge-specific
1207 iplookup router and a misconfiguration.
1208 Report from Marcin Mirosław.
1210 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1212 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1213 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1214 As part of this, removing so much warning spew let me fix some minor
1215 real issues in debug logging.
1217 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1218 assignment on my part. Fixed.
1220 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1221 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1222 Janne Snabb (who went above and beyond: thank you).
1224 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1225 string otherwise requires a connection and a bunch more work and it's
1226 relatively easy to get wrong. Should also expose TLS library linkage
1229 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1230 64-bit ${eval} (JH/03).
1232 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1233 GNU libc to support some of the 64-bit stuff, should not lead to
1234 conflicts. Defined before os.h is pulled in, so if a given platform
1235 needs to override this, it can.
1237 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1238 protection layer was required, which is not implemented.
1239 Bugzilla 1254, patch from Wolfgang Breyha.
1241 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1242 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1243 tls_dhparam take prime identifiers. Also unbreak combination of
1244 OpenSSL+DH_params+TLSSNI.
1246 PP/39 Disable SSLv2 by default in OpenSSL support.
1252 PP/01 Solaris build fix for Oracle's LDAP libraries.
1253 Bugzilla 1109, patch from Stephen Usher.
1255 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1257 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1260 TF/02 Fix a couple more cases where we did not log the error message
1261 when unlink() failed. See also change 4.74-TF/03.
1263 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1264 lock up or crash if it happened to be inside a call to libc when it
1265 got a SIGUSR1 from exiwhat.
1267 The SIGUSR1 handler appends the current process status to the process
1268 log which is later printed by exiwhat. It used to use the general
1269 purpose logging code to do this, but several functions it calls are
1270 not safe for signals.
1272 The new output code in the SIGUSR1 handler is specific to the process
1273 log, and simple enough that it's easy to inspect for signal safety.
1274 Removing some special cases also simplifies the general logging code.
1275 Removing the spurious timestamps from the process log simplifies
1278 TF/04 Improved ratelimit ACL condition.
1280 The /noupdate option has been deprecated in favour of /readonly which
1281 has clearer semantics. The /leaky, /strict, and /readonly update modes
1282 are mutually exclusive. The update mode is no longer included in the
1283 database key; it just determines when the database is updated. (This
1284 means that when you upgrade Exim will forget old rate measurements.)
1286 Exim now checks that the per_* options are used with an update mode that
1287 makes sense for the current ACL. For example, when Exim is processing a
1288 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1289 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1290 must specify per_mail/readonly. If you omit the update mode it defaults to
1291 /leaky where that makes sense (as before) or /readonly where required.
1293 The /noupdate option is now undocumented but still supported for
1294 backwards compatibility. It is equivalent to /readonly except that in
1295 ACLs where /readonly is required you may specify /leaky/noupdate or
1296 /strict/noupdate which are treated the same as /readonly.
1298 A useful new feature is the /count= option. This is a generalization
1299 of the per_byte option, so that you can measure the throughput of other
1300 aggregate values. For example, the per_byte option is now equivalent
1301 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1303 The per_rcpt option has been generalized using the /count= mechanism
1304 (though it's more complicated than the per_byte equivalence). When it is
1305 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1306 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1307 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1308 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1309 Note that using per_rcpt with a non-readonly update mode in more than
1310 one ACL will cause the recipients to be double-counted. (The per_mail
1311 and per_byte options don't have this problem.)
1313 The handling of very low rates has changed slightly. If the computed rate
1314 is less than the event's count (usually one) then this event is the first
1315 after a long gap. In this case the rate is set to the same as this event's
1316 count, so that the first message of a spam run is counted properly.
1318 The major new feature is a mechanism for counting the rate of unique
1319 events. The new per_addr option counts the number of different
1320 recipients that someone has sent messages to in the last time period. It
1321 behaves like per_rcpt if all the recipient addresses are different, but
1322 duplicate recipient addresses do not increase the measured rate. Like
1323 the /count= option this is a general mechanism, so the per_addr option
1324 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1325 example, measure the rate that a client uses different sender addresses
1326 with the options per_mail/unique=$sender_address. There are further
1327 details in the main documentation.
1329 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1331 TF/06 Removed a few PCRE remnants.
1333 TF/07 Automatically extract Exim's version number from tags in the git
1334 repository when doing development or release builds.
1336 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1337 Bugzilla 879. Patch from Paul Fisher.
1339 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1340 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1343 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1345 PP/05 Variable $av_failed, true if the AV scanner deferred.
1346 Bugzilla 1078. Patch from John Horne.
1348 PP/06 Stop make process more reliably on build failure.
1349 Bugzilla 1087. Patch from Heiko Schlittermann.
1351 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1352 Bugzilla 1089. Patch from Heiko Schlittermann.
1354 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1355 Bugzilla 1131. Patch from Holger Weiß.
1357 PP/09 Handle IPv6 addresses with SPF.
1358 Bugzilla 860. Patch from Wolfgang Breyha.
1360 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1362 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1365 PP/11 match_* no longer expand right-hand-side by default.
1366 New compile-time build option, EXPAND_LISTMATCH_RHS.
1367 New expansion conditions, "inlist", "inlisti".
1369 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1371 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1373 PP/14 fix log_write() format string regression from TF/03.
1374 Bugzilla 1152. Patch from Dmitry Isaikin.
1380 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1382 PP/02 Harmonised TLS library version reporting; only show if debugging.
1383 Layout now matches that introduced for other libraries in 4.74 PP/03.
1385 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1387 PP/04 New "dns_use_edns0" global option.
1389 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1392 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1393 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1395 TK/01 Updated PolarSSL code to 0.14.2.
1396 Bugzilla 1097. Patch from Andreas Metzler.
1398 PP/07 Catch divide-by-zero in ${eval:...}.
1399 Fixes bugzilla 1102.
1401 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1404 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1405 format-string attack -- SECURITY: remote arbitrary code execution.
1407 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1408 time unintentionally subject to list matching rules, letting the header
1409 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1410 arbitrary string expansion). This allowed for information disclosure.
1412 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1413 INT_MIN/-1 -- value coerced to INT_MAX.
1419 NM/01 Workaround for PCRE version dependency in version reporting
1422 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1423 This fixes portability to compilers other than gcc, notably
1424 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1426 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1427 makefiles for portability to HP-UX and POSIX correctness.
1429 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1430 Also via indented variable definition in the Makefile.
1431 (Debugging by Oliver Heesakkers).
1433 PP/02 Restore caching of spamd results with expanded spamd_address.
1434 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1436 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1437 Improves build reliability. Fix from: Frank Elsner
1439 NM/02 Fix wide character breakage in the rfc2047 coding
1440 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1442 NM/03 Allow underscore in dnslist lookups
1443 Fixes bug 1026. Patch from Graeme Fowler
1445 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1446 Code patches from Adam Ciarcinski of NetBSD.
1448 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1451 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1452 is logged, to avoid truncation. Patch from John Horne.
1454 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1455 Patch from Jakob Hirsch.
1457 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1458 SQL string expansion failure details.
1459 Patch from Andrey Oktyabrski.
1461 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1462 Patch from Simon Arlott.
1464 PP/09 New lookups functionality failed to compile on old gcc which rejects
1465 extern declarations in function scope.
1466 Patch from Oliver Fleischmann
1468 PP/10 Use sig_atomic_t for flags set from signal handlers.
1469 Check getgroups() return and improve debugging.
1470 Fixed developed for diagnosis in bug 927 (which turned out to be
1473 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1474 Patch from Mark Zealey.
1476 PP/12 Bugzilla 1056: Improved spamd server selection.
1477 Patch from Mark Zealey.
1479 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1480 Based on patch from Heiko Schlittermann.
1482 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1483 Patch from Uwe Doering, sign-off by Michael Haardt.
1485 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1486 variable declaration deep within a block. Bug and patch from
1489 PP/15 lookups-Makefile IRIX compatibility coercion.
1491 PP/16 Make DISABLE_DKIM build knob functional.
1493 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1494 Patch by Simon Arlott
1496 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1497 variable argument macros. Our copy now differs from upstream.
1503 TF/01 Failure to get a lock on a hints database can have serious
1504 consequences so log it to the panic log.
1506 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1507 controlled using the smtp_confirmation log selector.
1509 TF/03 Include the error message when we fail to unlink a spool file.
1511 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1512 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1513 for maintaining out-of-tree patches for some time.
1515 PP/01 Bugzilla 139: Documentation and portability issues.
1516 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1517 Handle per-OS dynamic-module compilation flags.
1519 PP/02 Let /dev/null have normal permissions.
1520 The 4.73 fixes were a little too stringent and complained about the
1521 permissions on /dev/null. Exempt it from some checks.
1522 Reported by Andreas M. Kirchwitz.
1524 PP/03 Report version information for many libraries, including
1525 Exim version information for dynamically loaded libraries. Created
1526 version.h, now support a version extension string for distributors
1527 who patch heavily. Dynamic module ABI change.
1529 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1530 privilege escalation vulnerability whereby the Exim run-time user
1531 can cause root to append content of the attacker's choosing to
1534 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1537 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1538 If dropping privileges for untrusted macros, we disabled normal logging
1539 on the basis that it would fail; for the Exim run-time user, this is not
1540 the case, and it resulted in successful deliveries going unlogged.
1541 Fixed. Reported by Andreas Metzler.
1547 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1548 only prepend for the Resent-* case. Fixes regression introduced in
1549 Exim 4.70 by NM/22 for Bugzilla 607.
1551 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1552 increasing numbers of administrators be bitten by this.
1554 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1556 PP/03 Bugzilla 994: added openssl_options main configuration option.
1558 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1560 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1562 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1564 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1565 without a peer certificate, leading to a segfault because of an
1566 assumption that peers always have certificates. Be a little more
1567 paranoid. Problem reported by Martin Tscholak.
1569 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1570 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1571 NB: ClamAV planning to remove STREAM in "middle of 2010".
1572 CL also introduces -bmalware, various -d+acl logging additions and
1573 more caution in buffer sizes.
1575 PP/09 Implemented reverse_ip expansion operator.
1577 PP/10 Bugzilla 937: provide a "debug" ACL control.
1579 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1581 PP/12 Bugzilla 973: Implement --version.
1583 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1585 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1587 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1589 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1590 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1591 ignore trailing whitespace.
1593 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1595 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1598 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1599 Notification from Dr Andrew Aitchison.
1601 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1602 ExtendedDetectionInfo response format.
1603 Notification from John Horne.
1605 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1608 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1609 XSL and documented dependency on system catalogs, with examples of how
1612 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1615 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1616 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1617 configuration file which is writeable by the Exim user or group.
1619 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1620 of configuration files to cover files specified with the -C option if
1621 they are going to be used with root privileges, not just the default
1624 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1625 option (effectively making it always true).
1627 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1628 files to be used while preserving root privileges.
1630 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1631 that rogue child processes cannot use them.
1633 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1634 run-time user, instead of root.
1636 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1637 Exim run-time user without dropping privileges.
1639 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1640 result string, instead of calling string_vformat() twice with the same
1643 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1644 for other users. Others should always drop root privileges if they use
1645 -C on the command line, even for a whitelisted configure file.
1647 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1649 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1655 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1656 $data_path, and $header_path variables; fixed documentation bugs and
1659 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1660 exipick to access non-standard spools, including the "frozen" queue
1663 NM/01 Bugzilla 965: Support mysql stored procedures.
1664 Patch from Alain Williams
1666 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1668 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1669 Patch from Andreas Metzler
1671 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1672 Patch from Kirill Miazine
1674 NM/05 Bugzilla 671: Added umask to procmail example.
1676 JJ/03 installed exipick 20100323.0, fixing doc bug
1678 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1679 directory. Notification and patch from Dan Rosenberg.
1681 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1683 TK/02 Improve log output when DKIM signing operation fails.
1685 MH/01 Treat the transport option dkim_domain as a colon separated
1686 list, not as a single string, and sign the message with each element,
1687 omitting multiple occurences of the same signer.
1689 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1690 Bugzilla 985, 986. Patch by Simon Arlott
1692 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1693 Patch by Simon Arlott
1695 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1696 MBX locking. Notification from Dan Rosenberg.
1702 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1704 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1706 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1708 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1710 NM/04 Bugzilla 915: Flag broken perl installation during build.
1716 TK/01 Added patch by Johannes Berg that expands the main option
1717 "spamd_address" if it starts with a dollar sign.
1719 TK/02 Write list of recipients to X-Envelope-Sender header when building
1720 the mbox-format spool file for content scanning (suggested by Jakob
1723 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1724 (http://www.dcc-servers.net/) support via dccifd. Activated by
1725 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1727 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1728 by Mark Daniel Reidel <mr@df.eu>.
1730 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1731 When building exim an external PCRE library is now needed -
1732 PCRE is a system library on the majority of modern systems.
1733 See entry on PCRE_LIBS in EDITME file.
1735 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1736 conversation. Added nologin parameter to request.
1737 Patch contributed by Kirill Miazine.
1739 TF/01 Do not log submission mode rewrites if they do not change the address.
1741 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1743 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1744 log files in place. Contributed by Roberto Lima.
1746 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1748 TF/03 Bugzilla 615: When checking the local_parts router precondition
1749 after a local_part_suffix or local_part_prefix option, Exim now
1750 does not use the address's named list lookup cache, since this
1751 contains cached lookups for the whole local part.
1753 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1754 Robert Millan. Documentation is in experimental-spec.txt.
1756 TF/04 Bugzilla 668: Fix parallel build (make -j).
1758 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1760 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1761 Patch provided by Jan Srzednicki.
1763 TF/05 Leading white space used to be stripped from $spam_report which
1764 wrecked the formatting. Now it is preserved.
1766 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1767 that they are available at delivery time.
1769 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1771 TF/08 TLS error reporting now respects the incoming_interface and
1772 incoming_port log selectors.
1774 TF/09 Produce a more useful error message if an SMTP transport's hosts
1775 setting expands to an empty string.
1777 NM/06 Bugzilla 744: EXPN did not work under TLS.
1778 Patch provided by Phil Pennock.
1780 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1781 Patch provided by Richard Godbee.
1783 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1784 acl_smtp_notquit, added index entry.
1786 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1787 Patch provided by Eugene Bujak.
1789 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1790 accept(). Patch provided by Maxim Dounin.
1792 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1793 Patch provided by Phil Pennock.
1795 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1797 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1798 Patch provided by Brad "anomie" Jorsch.
1800 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1801 Patch provided by Dean Brooks.
1803 TK/05 Add native DKIM support (does not depend on external libraries).
1805 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1806 Patch provided by Graeme Fowler.
1808 NM/16 Bugzilla 851: Documentation example syntax fix.
1810 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1812 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1815 NM/19 Bugzilla 745: TLS version reporting.
1816 Patch provided by Phil Pennock.
1818 NM/20 Bugzilla 167: bool: condition support.
1819 Patch provided by Phil Pennock.
1821 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1822 clients. Patch provided by Phil Pennock.
1824 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1825 Patch provided by Brad "anomie" Jorsch.
1827 NM/23 Bugzilla 687: Fix misparses in eximstats.
1828 Patch provided by Heiko Schlittermann.
1830 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1831 Patch provided by Heiko Schlittermann.
1833 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1834 plus update to original patch.
1836 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1838 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1839 Patch provided by David Brownlee.
1841 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1843 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1845 NM/30 Bugzilla 888: TLS documentation bugfixes.
1847 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1849 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1850 Unlike the original bugzilla I have changed all shell scripts in src tree.
1852 NM/33 Bugzilla 898: Transport filter timeout fix.
1853 Patch by Todd Rinaldo.
1855 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1856 Patch by Serge Demonchaux.
1858 NM/35 Bugzilla 39: Base64 decode bug fixes.
1859 Patch by Jakob Hirsch.
1861 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1863 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1865 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1867 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1873 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1874 ALT-N's libdkim that I have put here:
1875 http://duncanthrax.net/exim-experimental/
1877 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1878 were called 'true' and it seems that C99 defines that as a reserved
1879 keyword to be used with 'bool' variable types. That means you could
1880 not include C99-style headers which use bools without triggering
1881 build errors in sieve.c.
1883 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1884 as mailq or other aliases. Changed the --help handling significantly
1885 to do whats expected. exim_usage() emits usage/help information.
1887 SC/01 Added the -bylocaldomain option to eximstats.
1889 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1891 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1893 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1899 PH/01 Another patch from the Sieve maintainer.
1901 PH/02 When an IPv6 address is converted to a string for single-key lookup
1902 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1903 dots are used instead of colons so that keys in lsearch files need not
1904 contain colons. This was done some time before quoting was made available
1905 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1906 (notated using the quote facility) so as to distinguish them from IPv4
1907 keys. This meant that lookups for IP addresses in host lists did not work
1908 for iplsearch lookups.
1910 This has been fixed by arranging for IPv6 addresses to be expressed with
1911 colons if the lookup type is iplsearch. This is not incompatible, because
1912 previously such lookups could never work.
1914 The situation is now rather anomalous, since one *can* have colons in
1915 ordinary lsearch keys. However, making the change in all cases is
1916 incompatible and would probably break a number of configurations.
1918 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1921 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1922 conversion specification without a maximum field width, thereby enabling
1923 a rogue spamd server to cause a buffer overflow. While nobody in their
1924 right mind would setup Exim to query an untrusted spamd server, an
1925 attacker that gains access to a server running spamd could potentially
1926 exploit this vulnerability to run arbitrary code as the Exim user.
1928 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1929 $primary_hostname instead of what libspf2 thinks the hosts name is.
1931 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1932 a directory entry by the name of the lookup key. Previously, if a
1933 symlink pointed to a non-existing file or a file in a directory that
1934 Exim lacked permissions to read, a lookup for a key matching that
1935 symlink would fail. Now it is enough that a matching directory entry
1936 exists, symlink or not. (Bugzilla 503.)
1938 PH/03 The body_linecount and body_zerocount variables are now exported in the
1941 PH/04 Added the $dnslist_matched variable.
1943 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1944 This means they are set thereafter only if the connection becomes
1947 PH/06 Added the client_condition to authenticators so that some can be skipped
1948 by clients under certain conditions.
1950 PH/07 The error message for a badly-placed control=no_multiline_responses left
1951 "_responses" off the end of the name.
1953 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1955 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1956 (without spaces) instead of just copying the configuration text.
1958 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1960 PH/11 Added $max_received_linelength.
1962 PH/12 Added +ignore_defer and +include_defer to host lists.
1964 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1965 way in which PCRE > 7.0 is built.
1967 PH/14 Implemented queue_only_load_latch.
1969 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1970 MAIL command. The effect was to mangle the value on 64-bit systems.
1972 PH/16 Another patch from the Sieve maintainer.
1974 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1976 PH/18 If a system quota error occurred while trying to create the file for
1977 a maildir delivery, the message "Mailbox is full" was not appended to the
1978 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1979 only to a quota excession during the actual writing of the file.
1981 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1982 characters?) which causes problems in log lines. The DN values are now
1983 passed through string_printing() before being added to log lines.
1985 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1986 and InterBase are left for another time.)
1988 PH/21 Added message_body_newlines option.
1990 PH/22 Guard against possible overflow in moan_check_errorcopy().
1992 PH/23 POSIX allows open() to be a macro; guard against that.
1994 PH/24 If the recipient of an error message contained an @ in the local part
1995 (suitably quoted, of course), incorrect values were put in $domain and
1996 $local_part during the evaluation of errors_copy.
2002 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2003 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2006 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2007 issue a MAIL command.
2009 PH/02 In an ACL statement such as
2011 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2013 if a client was not listed at all, or was listed with a value other than
2014 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2015 the condition was not true (as it should be), so access was not denied.
2016 The bug was that the ! inversion was incorrectly passed on to the second
2017 item. This has been fixed.
2019 PH/03 Added additional dnslists conditions == and =& which are different from
2020 = and & when the dns lookup returns more than one IP address.
2022 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2023 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2025 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2026 FSYNC, which compiles an option called disable_fsync that allows for
2027 bypassing fsync(). The documentation is heavily laced with warnings.
2029 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2031 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2032 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2033 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2034 including adding "make clean"; (3) Added -fPIC when compiling the test
2035 dynamically loaded module, to get rid of a warning.
2037 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2038 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2039 = 0s. The bug is otherwise harmless.
2041 PH/07 There was a bug in the dovecot authenticator such that the value of
2042 $auth1 could be overwritten, and so not correctly preserved, after a
2043 successful authentication. This usually meant that the value preserved by
2044 the server_setid option was incorrect.
2046 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2048 PH/09 Installed PCRE release 7.0.
2050 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2051 run for batched SMTP input. It is now run at the start of every message
2052 in the batch. While fixing this I discovered that the process information
2053 (output by running exiwhat) was not always getting set for -bs and -bS
2054 input. This is fixed, and it now also says "batched" for BSMTP.
2056 PH/11 Added control=no_pipelining.
2058 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2059 patch, slightly modified), and move the expansion of helo_data till after
2060 the connection is made in the smtp transport (so it can use these
2063 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2065 PH/14 Added log_selector = +pid.
2067 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2069 PH/16 Add ${if forany and ${if forall.
2071 PH/17 Added dsn_from option to vary the From: line in DSNs.
2073 PH/18 Flush SMTP output before performing a callout, unless control =
2074 no_callout_flush is set.
2076 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2077 was true (the default) a successful delivery failed to delete the retry
2078 item, thus causing premature timeout of the address. The bug is now
2081 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2083 PH/21 Long custom messages for fakedefer and fakereject are now split up
2084 into multiline responses in the same way that messages for "deny" and
2085 other ACL rejections are.
2087 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2088 with slight modification.
2090 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2091 draft, changing the syntax and factoring some duplicate code".
2093 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2094 for deliveries of the second and subsequent messages over the same SMTP
2097 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2098 ${reduce, with only minor "tidies".
2100 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2102 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2103 expansion side effects.
2105 PH/27 When a message times out after an over-quota error from an Exim-imposed
2106 quota, the bounce message says "mailbox is full". This message was not
2107 being given when it was a system quota that was exceeded. It now should
2110 MH/03 Made $recipients available in local_scan(). local_scan() already has
2111 better access to the recipient list through recipients_list[], but
2112 $recipients can be useful in postmaster-provided expansion strings.
2114 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2115 in the case of a MAIL command with additional options following the
2116 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2117 were accidentally chopped off.
2119 PH/29 SMTP synchronization checks are implemented when a command is read -
2120 there is a check that no more input is waiting when there shouldn't be
2121 any. However, for some commands, a delay in an ACL can mean that it is
2122 some time before the response is written. In this time, more input might
2123 arrive, invalidly. So now there are extra checks after an ACL has run for
2124 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2125 pipelining has not been advertised.
2127 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2129 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2130 returned in the SMTP response when smtp_return_error_details was false.
2131 This has been fixed.
2133 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2134 instead of the C I/O that was originally supplied, because problems were
2135 reported on Solaris.
2137 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2138 Exim which did not show up earlier: it was assuming that a call to
2139 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2140 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2141 was a macro that became an assignment, so it seemed to work. This has
2142 changed to a proper function call with a void return, hence the compile
2143 error. Exim's code has been fixed.
2145 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2148 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2150 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2152 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2153 because it turns out to be too big or there is a timeout) there is no
2154 "Completed" line in the log. When some messages of this type were
2155 selected by exigrep, they were listed as "not completed". Others were
2156 picked up by some special patterns. I have improved the selection
2157 criteria to be more general.
2159 PH/38 The host_find_failed option in the manualroute router can now be set
2160 to "ignore", to completely ignore a host whose IP address cannot be
2161 found. If all hosts are ignored, the behaviour is controlled by the new
2162 host_all_ignored option.
2164 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2165 homing or because of multiple MX records with /mx) generated more than
2166 one IP address, and the following item turned out to be the local host,
2167 all the secondary addresses of the first item were incorrectly removed
2168 from the list, along with the local host and any following hosts (which
2169 is what is supposed to happen).
2171 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2172 whoever called Exim into the -H file. In the case of the daemon it was
2173 behaving confusingly. When first started, it used values for whoever
2174 started the daemon, but after a SIGHUP it used the Exim user (because it
2175 calls itself on a restart). I have changed the code so that it now always
2178 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2179 message are rejected with the same error (e.g. no authentication or bad
2180 sender address), and a DATA command is nevertheless sent (as can happen
2181 with PIPELINING or a stupid MUA), the error message that was given to the
2182 RCPT commands is included in the rejection of the DATA command. This is
2183 intended to be helpful for MUAs that show only the final error to their
2186 PH/42 Another patch from the Sieve maintainer.
2188 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2189 Eximstats - Fixed some broken HTML links and added missing column headers
2191 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2192 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2194 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2196 PH/43 Yet another patch from the Sieve maintainer.
2198 PH/44 I found a way to check for a TCP/IP connection going away before sending
2199 the response to the final '.' that terminates a message, but only in the
2200 case where the client has not sent further data following the '.'
2201 (unfortunately, this is allowed). However, in many cases there won't be
2202 any further data because there won't be any more messages to send. A call
2203 to select() can be used: if it shows that the input is "ready", there is
2204 either input waiting, or the socket has been closed. An attempt to read
2205 the next input character can distinguish the two cases. Previously, Exim
2206 would have sent an OK response which the client would never have see.
2207 This could lead to message repetition. This fix should cure that, at
2208 least in a lot of common cases.
2210 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2211 advertised in response to EHLO.
2217 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2218 fixed by 4.65/MH/01 (is this a record?) are fixed:
2220 (i) An empty string was always treated as zero by the numeric comparison
2221 operators. This behaviour has been restored.
2223 (ii) It is documented that the numeric comparison operators always treat
2224 their arguments as decimal numbers. This was broken in that numbers
2225 starting with 0 were being interpreted as octal.
2227 While fixing these problems I realized that there was another issue that
2228 hadn't been noticed. Values of message_size_limit (both the global option
2229 and the transport option) were treated as octal if they started with 0.
2230 The documentation was vague. These values are now always treated as
2231 decimal, and I will make that clear in the documentation.
2237 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2238 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2241 MH/01 Don't check that the operands of numeric comparison operators are
2242 integers when their expansion is in "skipping" mode (fixes bug
2243 introduced by 4.64-PH/07).
2245 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2246 child addresses, Exim now panics and dies. Previously, because the count
2247 is held in a short int, deliveries were likely to be lost. As such a
2248 large number of recipients for a single message is ridiculous
2249 (performance will be very, very poor), I have chosen to impose a limit
2250 rather than extend the field.
2256 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2257 leftover -K file (the existence of which was triggered by #402).
2258 While we were at it, introduced process PID as part of the -K
2259 filename. This should rule out race conditions when creating
2262 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2263 processing considerably. Previous code took too long for large mails,
2264 triggering a timeout which in turn triggers #401.
2266 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2267 in the DK code in transports.c. sendfile() is not really portable,
2268 hence the _LINUX specificness.
2270 TF/01 In the add_headers option to the mail command in an Exim filter,
2271 there was a bug that Exim would claim a syntax error in any
2272 header after the first one which had an odd number of characters
2275 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2276 callout verification, Exim cached a "reject" for the entire domain. This
2277 is correct for most verifications, but it is not correct for a recipient
2278 verification with use_sender or use_postmaster set, because in that case
2279 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2280 case of MAIL FROM:<> rejection from other early rejections (e.g.
2281 rejection of HELO). When verifying a recipient using a non-null MAIL
2282 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2283 Whatever the result of the callout, the value of the domain cache is
2284 left unchanged (for any other kind of callout, getting as far as trying
2285 RCPT means that the domain itself is ok).
2287 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2290 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2291 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2292 session, the daemon ignores ECONNECT errors and logs others; it now
2293 ignores EPIPE as well.
2295 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2296 (quoted-printable decoding).
2298 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2299 later the small subsequent patch to fix an introduced bug.
2301 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2303 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2305 PH/08 An error is now given if message_size_limit is specified negative.
2307 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2308 to be given (somewhat) arbitrary names.
2310 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2313 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2314 miscellaneous code fixes
2316 PH/10 Added the log_reject_target ACL modifier to specify where to log
2319 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2320 hostname. This is wrong, because it relates to the incoming message (and
2321 probably the interface on which it is arriving) and not to the outgoing
2322 callout (which could be using a different interface). This has been
2323 changed to use the value of the helo_data option from the smtp transport
2324 instead - this is what is used when a message is actually being sent. If
2325 there is no remote transport (possible with a router that sets up host
2326 addresses), $smtp_active_hostname is used.
2328 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2329 tweaks were necessary in order to get it to work (see also 21 below):
2330 (a) The code assumed that strncpy() returns a negative number on buffer
2331 overflow, which isn't the case. Replaced with Exim's string_format()
2333 (b) There were several signed/unsigned issues. I just did the minimum
2334 hacking in of casts. There is scope for a larger refactoring.
2335 (c) The code used strcasecmp() which is not a standard C function.
2336 Replaced with Exim's strcmpic() function.
2337 (d) The code set only $1; it now sets $auth1 as well.
2338 (e) A simple test gave the error "authentication client didn't specify
2339 service in request". It would seem that Dovecot has changed its
2340 interface. Fortunately there's a specification; I followed it and
2341 changed what the client sends and it appears to be working now.
2343 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2346 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2347 address A is aliased to B and C, where B exists and C does not. Without
2348 -v the output is "A verified" because verification stops after a
2349 successful redirection if more than one address is generated. However,
2350 with -v the child addresses are also verified. Exim was outputting "A
2351 failed to verify" and then showing the successful verification for C,
2352 with its parentage. It now outputs "B failed to verify", showing B's
2353 parentage before showing the successful verification of C.
2355 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2356 look up a TXT record in a specific list after matching in a combined
2359 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2360 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2361 they consult the DNS. I had assumed they would set it the way they
2362 wanted; and indeed my experiments on Linux seem to show that in some
2363 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2364 To be on the safe side, however, I have now made the interface to
2365 host_find_byname() similar to host_find_bydns(), with an argument
2366 containing the DNS resolver options. The host_find_byname() function now
2367 sets these options at its start, just as host_find_bydns() does. The smtp
2368 transport options dns_qualify_single and dns_search_parents are passed to
2369 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2370 of host_find_byname() use the default settings of RES_DEFNAMES
2371 (qualify_single) but not RES_DNSRCH (search_parents).
2373 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2374 spool_read_header() do less string testing, by means of a preliminary
2375 switch on the second character of optional "-foo" lines. (This is
2376 overdue, caused by the large number of possibilities that now exist.
2377 Originally there were few.) While I was there, I also converted the
2378 str(n)cmp tests so they don't re-test the leading "-" and the first
2379 character, in the hope this might squeeze out yet more improvement.
2381 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2382 flag allowing group syntax was set by the header_syntax check but not
2383 turned off, possible causing trouble later; (2) The flag was not being
2384 set at all for the header_verify test, causing "group"-style headers to
2385 be rejected. I have now set it in this case, and also caused header_
2386 verify to ignore an empty address taken from a group. While doing this, I
2387 came across some other cases where the code for allowing group syntax
2388 while scanning a header line wasn't quite right (mostly, not resetting
2389 the flag correctly in the right place). These bugs could have caused
2390 trouble for malformed header lines. I hope it is now all correct.
2392 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2393 with the "reply" argument non-NULL. The code, however (which originally
2394 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2395 but it didn't always do it. This confused somebody who was copying the
2396 code for some other use. I have removed all the tests.
2398 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2399 feature that was used to support insecure browsers during the U.S. crypto
2400 embargo. It requires special client support, and Exim is probably the
2401 only MTA that supported it -- and would never use it because real RSA is
2402 always available. This code has been removed, because it had the bad
2403 effect of slowing Exim down by computing (never used) parameters for the
2404 RSA_EXPORT functionality.
2406 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2407 authenticator to fail if there's a tab character in the incoming data
2408 (there should never be unless someone is messing about, as it's supposed
2409 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2410 if the connection is using TLS or if the remote IP is the same as the
2411 local IP, and the "valid-client-cert option" if a client certificate has
2414 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2415 authenticators. This can be used for authorization after authentication
2416 succeeds. (In the case of plaintext, it servers for both authentication
2419 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2420 if any retry times were supplied.
2422 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2423 connection, where there is no client IP address to check. In this
2424 situation, the verify now always succeeds.
2426 PH/25 Applied John Jetmore's -Mset patch.
2428 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2430 PH/27 In a string expansion for a processed (not raw) header when multiple
2431 headers of the same name were present, leading whitespace was being
2432 removed from all of them, but trailing whitespace was being removed only
2433 from the last one. Now trailing whitespace is removed from each header
2434 before concatenation. Completely empty headers in a concatenation (as
2435 before) are ignored.
2437 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2438 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2440 PH/29 [Removed. This was a change that I later backed out, and forgot to
2441 correct the ChangeLog entry (that I had efficiently created) before
2442 committing the later change.]
2444 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2445 address errors (4xx response to RCPT) over the same connection as other
2446 messages routed to the same hosts. Such deliveries are always "forced",
2447 so retry times are not inspected. This resulted in far too many retries
2448 for the affected addresses. The effect occurred only when there were more
2449 hosts than the hosts_max_try setting in the smtp transport when it had
2450 the 4xx errors. Those hosts that it had tried were not added to the list
2451 of hosts for which the message was waiting, so if all were tried, there
2452 was no problem. Two fixes have been applied:
2454 (i) If there are any address or message errors in an SMTP delivery, none
2455 of the hosts (tried or untried) are now added to the list of hosts
2456 for which the message is waiting, so the message should not be a
2457 candidate for sending over the same connection that was used for a
2458 successful delivery of some other message. This seems entirely
2459 reasonable: after all the message is NOT "waiting for some host".
2460 This is so "obvious" that I'm not sure why it wasn't done
2461 previously. Hope I haven't missed anything, but it can't do any
2462 harm, as the worst effect is to miss an optimization.
2464 (ii) If, despite (i), such a delivery is accidentally attempted, the
2465 routing retry time is respected, so at least it doesn't keep
2466 hammering the server.
2468 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2469 in ${readsocket because some servers need this prod.
2471 PH/32 Added some extra debug output when updating a wait-xxx database.
2473 PH/33 The hint "could be header name not terminated by colon", which has been
2474 given for certain expansion errors for a long time, was not being given
2475 for the ${if def:h_colon_omitted{... case.
2477 PH/34 The spec says: "With one important exception, whenever a domain list is
2478 being scanned, $domain contains the subject domain." There was at least
2479 one case where this was not true.
2481 PH/35 The error "getsockname() failed: connection reset by peer" was being
2482 written to the panic log as well as the main log, but it isn't really
2483 panic-worthy as it just means the connection died rather early on. I have
2484 removed the panic log writing for the ECONNRESET error when getsockname()
2487 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2488 runs only) independently of the message's sender address. This meant
2489 that, if the 4xx error was in fact related to the sender, a different
2490 message to the same recipient with a different sender could confuse
2491 things. In particular, this can happen when sending to a greylisting
2492 server, but other circumstances could also provoke similar problems.
2493 I have changed the default so that the retry time for these errors is now
2494 based a combination of the sender and recipient addresses. This change
2495 can be overridden by setting address_retry_include_sender=false in the
2498 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2499 remote server are returned as part of bounce messages. This was not
2500 happening for LMTP over a pipe (the lmtp transport), but now it is the
2501 same for both kinds of LMTP.
2503 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2504 in header lines that were in fact CNAMEs. This is no longer the case.
2506 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2507 and queue runs started by the daemon processed all messages. This has
2508 been fixed so that -R and -S can now usefully be given with -q<time>.
2510 PH/40 Import PCRE release 6.7 (fixes some bugs).
2512 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2514 PH/42 Give an error if -q is specified more than once.
2516 PH/43 Renamed the variables $interface_address and $interface_port as
2517 $received_ip_address and $received_port, to make it clear that these
2518 values apply to message reception, and not to the outgoing interface when
2519 a message is delivered. (The old names remain recognized, of course.)
2521 PH/44 There was no timeout on the connect() call when using a Unix domain
2522 socket in the ${readsocket expansion. There now is.
2524 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2525 be meaningful with "accept".
2527 SC/01 Eximstats V1.43
2528 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2530 SC/02 Eximstats V1.44
2531 Use a glob alias rather than an array ref in the generated
2532 parser. This improves both readability and performance.
2534 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2535 Collect SpamAssassin and rejection statistics.
2536 Don't display local sender or destination tables unless
2537 there is data to show.
2538 Added average volumes into the top table text output.
2540 SC/04 Eximstats V1.46
2541 Collect data on the number of addresses (recipients)
2542 as well as the number of messages.
2544 SC/05 Eximstats V1.47
2545 Added 'Message too big' to the list of mail rejection
2546 reasons (thanks to Marco Gaiarin).
2548 SC/06 Eximstats V1.48
2549 Mainlog lines which have GMT offsets and are too short to
2550 have a flag are now skipped.
2552 SC/07 Eximstats V1.49 (Alain Williams)
2553 Added the -emptyok flag.
2555 SC/08 Eximstats V1.50
2556 Fixes for obtaining the IP address from reject messages.
2558 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2559 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2560 whitespace changes from 4.64-PH/27
2562 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2565 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2566 are found, allow negative numbers in numeric criteria)
2568 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2570 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2571 to match changes made in 4.64-PH/43
2573 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2575 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2576 "domain" is actually an IP address, and force a failure. This locks out
2577 those revolvers/nameservers that support "A-for-A" lookups, in
2578 contravention of the specifications.
2580 PH/48 When a host name was looked up from an IP address, and the subsequent
2581 forward lookup of the name timed out, the host name was left in
2582 $sender_host_name, contrary to the specification.
2584 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2585 restricted to single-key lookups, Exim was not diagnosing an error if
2586 * or *@ was used with a query-style lookup.
2588 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2590 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2591 long ago, but noone interested enough thought of it. Let's just say that
2592 the "1.1" means that there are some new functions that weren't there at
2593 some point in the past.
2595 PH/51 Error processing for expansion failure of helo_data from an smtp
2596 transport during callout processing was broken.
2598 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2599 tested/used via the -bh/-bhc/-bs options.
2601 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2602 bug, fixed in subsequent PCRE releases).
2604 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2605 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2607 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2613 SC/01 Use a glob alias rather than an array ref in eximstats generated
2614 parser. This improves both readability and performance.
2616 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2617 Don't display local sender or destination tables in eximstats unless
2618 there is data to show.
2619 Added average volumes into the eximstats top table text output.
2621 SC/03 Collect data on the number of addresses (recipients) as well
2622 as the number of messages in eximstats.
2624 TF/01 Correct an error in the documentation for the redirect router. Exim
2625 does (usually) call initgroups() when daemonizing.
2627 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2628 with consistent privilege compared to when running as a daemon.
2630 TF/03 Note in the spec that $authenticated_id is not set for local
2631 submissions from trusted users.
2633 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2634 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2636 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2637 by adding some example configuration directives to the default
2638 configuration file. A little bit of work is required to uncomment the
2639 directives and define how usernames and passwords are checked, but
2640 there is now a framework to start from.
2642 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2643 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2644 without this. I don't know how relevant this is to other LDAP libraries.
2646 PH/02 Add the verb name to the "unknown ACL verb" error.
2648 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2650 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2652 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2653 directory not expanded when it should be if an expanded home directory
2654 was set for the address (which is overridden by the transport).
2656 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2659 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2660 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2661 because it is too late at that time, and has no effect.
2663 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2664 security issue with \' (bugzilla #107). I could not use the
2665 PQescapeStringConn() function, because it needs a PGconn value as one of
2668 PH/08 When testing addresses using -bt, indicate those final addresses that
2669 are duplicates that would not cause an additional delivery. At least one
2670 person was confused, thinking that -bt output corresponded to deliveries.
2671 (Suppressing duplicates isn't a good idea as you lose the information
2672 about possibly different redirections that led to the duplicates.)
2674 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2675 systems where poll() doesn't work, in particular OS X.
2677 PH/10 Added more information to debugging output for retry time not reached.
2679 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2680 operations in malware.c.
2682 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2685 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2686 syslog_duplication set false, log lines that would normally be written
2687 both the the main log and to the reject log were not written to syslog at
2690 PH/14 In the default configuration, change the use of "message" in ACL warn
2691 statements to "add_header".
2693 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2694 not followed by a command (e.g. "seen endif").
2696 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2697 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2700 PH/17 Added extra conditions to the default value of delay_warning_condition
2704 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2705 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2706 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2709 The Auto-Submitted: and various List- headers are standardised, whereas I
2710 don't think Precedence: ever was.
2712 PH/18 Refactored debugging code in route_finduser() to show more information,
2713 in particular, the error code if getpwnam() issues one.
2715 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2716 This is apparently needed in addition to the PH/07 change above to avoid
2717 any possible encoding problems.
2719 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2720 but not after initializing Perl.
2722 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2723 output them only if debugging. By default they are written stderr,
2724 apparently, which is not desirable.
2726 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2729 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2732 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2734 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2735 authenticated or an ident call has been made. Suppress the default
2736 values for $authenticated_id and $authenticated_sender (but permit -oMai
2737 and -oMas) when testing with -bh.
2739 PH/24 Re-jigged the order of the tests in the default configuration so that the
2740 tests for valid domains and recipients precede the DNS black list and CSA
2741 tests, on the grounds that those ones are more expensive.
2743 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2744 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2745 This bug exists in every version of Exim that I still have, right back to
2748 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2749 However, an attempt to turn on case-sensitivity in a regex key by
2750 including (?-i) didn't work because the subject string was already
2751 lowercased, and the effects were non-intuitive. It turns out that a
2752 one-line patch can be used to allow (?-i) to work as expected.
2758 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2759 other effects) broke the use of negated acl sub-conditions.
2761 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2764 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2765 "Deny" causes Exim to reject the incoming connection with a 554 error.
2766 Unfortunately, if there is a major crisis, such as a disk failure,
2767 tcp-wrappers gives "deny", whereas what one would like would be some
2768 kind of temporary error. A kludge has been added to help with this.
2769 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2770 554 error is used if errno is still zero or contains ENOENT (which occurs
2771 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2774 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2776 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2777 errors. Otherwise a message that provokes a temporary error (when other
2778 messages do not) can cause a whole host to time out.
2780 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2781 addresses were routed directly to files or pipes from a redirect router.
2782 File deliveries just didn't batch; pipe deliveries might have suffered
2785 PH/06 A failure to get a lock for a hints database would erroneously always say
2786 "Failed to get write lock", even when it was really a read lock.
2788 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2789 of 0600. This has been changed to use the value of the lockfile_mode
2790 option (which defaults to 0600).
2792 PH/08 Applied small patch from the Sieve maintainer.
2794 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2795 folder from quota calculations, a direct delivery into this folder messed
2796 up the contents of the maildirsize file. This was because the regex was
2797 used only to exclude .Trash (or whatever) when the size of the mailbox
2798 was calculated. There was no check that a delivery was happening into an
2799 excluded directory. This bug has been fixed by ignoring all quota
2800 processing for deliveries into excluded directories.
2802 PH/10 Added the maildirfolder_create_regex option to appendfile.
2808 PH/01 The code for finding all the local interface addresses on a FreeBSD
2809 system running IPv6 was broken. This may well have applied to all BSD
2810 systems, as well as to others that have similar system calls. The broken
2811 code found IPv4 interfaces correctly, but gave incorrect values for the
2812 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2813 that it would not match correctly against @[] and not recognize the IPv6
2816 PH/02 The ipliteral router was not recognizing addresses of the form user@
2817 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2819 PH/03 Added disable_ipv6.
2821 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2822 decoded form, because it is most often used to construct To: headers
2823 lines in autoreplies, and the decoded form may well be syntactically
2824 invalid. However, $reply_address has leading white space removed, and all
2825 newlines turned into spaces so that the autoreply transport does not
2828 PH/05 If group was specified without a user on a router, and no group or user
2829 was specified on a transport, the group from the router was ignored.
2831 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2832 for visible compile-time settings that can be used to change these
2833 numbers, for those that want even more. Backwards compatibility with old
2834 spool files has been maintained. However, going back to a previous Exim
2835 release will lost any variables that are in spool files.
2837 PH/07 Two small changes when running in the test harness: increase delay when
2838 passing a TCP/IP connection to a new process, in case the original
2839 process has to generate a bounce, and remove special handling of
2840 127.0.0.2 (sic), which is no longer necessary.
2842 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2843 be the same on different OS.
2845 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2848 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2849 whether --show-vars was specified or not
2851 JJ/02 exipick: Added support for new ACL variable spool format introduced
2854 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2855 syntactically invalid From: or Reply-to: line, and a filter used this to
2856 generate an autoreply, and therefore failed to obtain an address for the
2857 autoreply, Exim could try to deliver to a non-existent relative file
2858 name, causing unrelated and misleading errors. What now happens is that
2859 it logs this as a hard delivery error, but does not attempt to create a
2862 PH/11 The exinext utility has a -C option for testing purposes, but although
2863 the given file was scanned by exinext itself; it wasn't being passed on
2864 when Exim was called.
2866 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2867 an end-of-file indication when reading a command response.
2869 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2870 compiled. In many other places in Exim, IPv6 addresses are always
2871 recognized, so I have changed this. It also means that IPv4 domain
2872 literals of the form [IPV4:n.n.n.n] are now always recognized.
2874 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2875 used if the router is not running as root, for example, when verifying at
2876 ACL time, or when using -bh. The debugging output from this situation was
2877 non-existent - all you got was a failure to exec. I have made two
2880 (a) Failures to set uid/gid, the current directory, or a process leader
2881 in a subprocess such as that created by queryprogram now generate
2882 suitable debugging output when -d is set.
2884 (b) The queryprogram router detects when it is not running as root,
2885 outputs suitable debugging information if -d is set, and then runs
2886 the subprocess without attempting to change uid/gid.
2888 PH/15 Minor change to Makefile for building test_host (undocumented testing
2891 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2892 additional section of a DNS packet that returns MX or SRV records.
2893 Instead, it always explicitly searches for A/AAAA records. This avoids
2894 major problems that occur when a DNS server includes only records of one
2895 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2896 fixed another bug: if SRV records were looked up and the corresponding
2897 address records were *not* found in the additional section, the port
2898 values from the SRV records were lost.
2900 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2901 using the correct key (the original address) when searching the retry
2902 rules in order to find which one to use for generating the retry hint.
2904 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2905 adding the default one. Similarly, if it contains a Reply-To: header, the
2906 errors_reply_to option, if set, is not used.
2908 PH/19 When calculating a retry time, Exim used to measure the "time since
2909 failure" by looking at the "first failed" field in the retry record. Now
2910 it does not use this if it is later than than the arrival time of the
2911 message. Instead it uses the arrival time. This makes for better
2912 behaviour in cases where some deliveries succeed, thus re-setting the
2913 "first failed" field. An example is a quota failure for a huge message
2914 when small messages continue to be delivered. Without this change, the
2915 "time since failure" will always be short, possible causing more frequent
2916 delivery attempts for the huge message than are intended.
2917 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2919 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2920 $1, $2, $3) because the numerical variables can be reset during some
2921 expansion items (e.g. "match"), thereby losing the authentication data.
2923 PH/21 Make -bV show the size of off_t variables so that the test suite can
2924 decide whether to run tests for quotas > 2G.
2926 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2927 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2928 filecount value is greater than 2G or if a quota value is greater than 2G
2929 on a system where the size of off_t is not greater than 4, a panic error
2932 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2933 never match. The debug and -bh output now contains an explicit error
2934 message indicating a malformed IPv4 address or mask.
2936 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2937 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2938 PH/23 above applies.
2940 PH/25 Do not write to syslog when running in the test harness. The only
2941 occasion when this arises is a failure to open the main or panic logs
2942 (for which there is an explicit test).
2944 PH/26 Added the /no_tell option to "control=freeze".
2946 PH/27 If a host name lookup failed very early in a connection, for example, if
2947 the IP address matched host_lookup and the reverse lookup yielded a name
2948 that did not have a forward lookup, an error message of the form "no IP
2949 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2950 could be logged. Now it outputs the IP address instead of "NULL".
2952 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2953 allows the sender and the authenticated sender to be set when
2954 submitting a message from within Exim. Since child_open_exim() is
2955 documented for local_scan(), the new function should be too.
2957 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2958 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2959 results in an empty string is now treated as unset.
2961 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2963 PH/31 Added sender_verify_fail logging option.
2965 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2966 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2967 tidied the source and removed it altogether.
2969 PH/33 When a queue run was abandoned because the load average was too high, a
2970 log line was always written; now it is written only if the queue_run log
2971 selector is set. In addition, the log line for abandonment now contains
2972 information about the queue run such as the pid. This is always present
2973 in "start" and "stop" lines but was omitted from the "abandon" line.
2975 PH/34 Omit spaces between a header name and the colon in the error message that
2976 is given when verify = headers_syntax fails (if there are lots of them,
2977 the message gets confusing).
2979 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2980 names, as there are now some PTR records that contain slashes. This check
2981 is only to protect against broken name servers that fall over on strange
2982 characters, so the fact that it applies to all lookups doesn't matter.
2984 PH/36 Now that the new test suite is complete, we can remove some of the
2985 special code in Exim that was needed for the old test suite. For example,
2986 sorting DNS records because real resolvers return them in an arbitrary
2987 order. The new test suite's fake resolver always returns records in the
2990 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2991 bounces) except when queue_only is set, to avoid logging races between
2992 the different processes.
2994 PH/38 Panic-die if .include specifies a non-absolute path.
2996 PH/39 A tweak to the "H" retry rule from its user.
2998 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2999 a label. They prevented compilation on older perls.
3001 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3002 a warning to be raised on newish perls.
3004 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3005 on queue. Changes to match documented behaviour of showing count of
3006 messages matching specified criteria.
3008 PH/40 Changed the default ident timeout from 30s to 5s.
3010 PH/41 Added support for the use of login_cap features, on those BSD systems
3011 that have them, for controlling the resources used by pipe deliveries.
3013 PH/42 The content-scanning code uses fopen() to create files in which to put
3014 message data. Previously it was not paying any attention to the mode of
3015 the files. Exim runs with umask(0) because the rest of the code creates
3016 files with open(), and sets the required mode explicitly. Thus, these
3017 files were ending up world-writeable. This was not a big issue, because,
3018 being within the spool directory, they were not world-accessible. I have
3019 created a function called modefopen, which takes an additional mode
3020 argument. It sets umask(777), creates the file, chmods it to the required
3021 mode, then resets the umask. All the relevant calls to fopen() in the
3022 content scanning code have been changed to use this function.
3024 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3025 to 24 hours. This avoids potential overflow problems when processing G
3026 and H retry rules. I suspect nobody ever tinkers with this value.
3028 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3030 PH/45 When the plaintext authenticator is running as a client, the server's
3031 challenges are checked to ensure they are valid base64 strings. By
3032 default, the authentication attempt is cancelled if an invalid string is
3033 received. Setting client_ignore_invalid_base64 true ignores these errors.
3034 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3035 they are received. Thus, the responses can be made to depend on the
3036 challenges. If an invalid string is ignored, an empty string is placed in
3039 PH/46 Messages that are created by the autoreply transport now contains a
3040 References: header, in accordance with RFCs 2822 and 3834.
3042 PH/47 Added authenticated_sender_force to the smtp transport.
3044 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3046 PH/49 Installed latest patch from the Sieve maintainer.
3048 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3049 was also set, the appendfile transport was unnecessarily scanning a
3050 directory of message files (e.g. for maildir delivery) to find the count
3051 of files (along with the size), even though it did not need this
3052 information. It now does the scan only if it needs to find either the
3053 size of the count of files.
3055 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3057 PH/52 Two bugs concerned with error handling when the smtp transport is
3060 (i) Exim was not creating retry information for temporary errors given
3061 for individual recipients after the DATA command when the smtp transport
3062 was used in LMTP mode. This meant that they could be retried too
3063 frequently, and not timed out correctly.
3065 (ii) Exim was setting the flag that allows error details to be returned
3066 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3067 recipients that were returned after the DATA command.
3069 PH/53 This is related to PH/52, but is more general: for any failing address,
3070 when detailed error information was permitted to be returned to the
3071 sender, but the error was temporary, then after the final timeout, only
3072 "retry timeout exceeded" was returned. Now it returns the full error as
3073 well as "retry timeout exceeded".
3075 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3076 do this, and (what is worse) MTAs that accept it.
3078 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3079 will now be deprecated.
3081 PH/56 New os.c-cygwin from the Cygwin maintainer.
3083 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3084 formats (previously only available in exim formats via -bpr, -bpru,
3085 and -bpra. Now also available in native and exiqgrep formats)
3087 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3088 with very large, slow to parse queues
3090 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3092 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3094 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3095 responses to authentication challenges, though it was showing the
3096 challenges; (ii) I've removed the CR characters from the debug output for
3099 PH/58 Allow for the insertion of a newline as well as a space when a string
3100 is turned into more than one encoded-word during RFC 2047 encoding. The
3101 Sieve code now uses this.
3103 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3104 data_4xx, lost_connection, tls_required.
3106 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3107 message was being sent as an SMTP response.
3109 PH/61 Add -l and -k options to exicyclog.
3111 PH/62 When verifying, if an address was redirected to one new address, so that
3112 verification continued, and the new address failed or deferred after
3113 having set something in $address_data, the value of $address_data was not
3114 passed back to the ACL. This was different to the case when no
3115 redirection occurred. The value is now passed back in both cases.
3117 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3118 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3119 use login_cap.h, so on its own it isn't the distinguishing feature. The
3120 new name refers directly to the setclassresources() function.
3122 PH/65 Added configuration files for NetBSD3.
3124 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3126 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3127 is preferred over IPv4.
3129 PH/68 The bounce_return_message and bounce_return_body options were not being
3130 honoured for bounces generated during the reception of non-SMTP messages.
3131 In particular, this applied to messages rejected by the ACL. This bug has
3132 been fixed. However, if bounce_return_message is true and bounce_return_
3133 body is false, the headers that are returned for a non-SMTP message
3134 include only those that have been read before the error was detected.
3135 (In the case of an ACL rejection, they have all been read.)
3137 PH/69 The HTML version of the specification is now built in a directory called
3138 spec_html instead of spec.html, because the latter looks like a path with
3139 a MIME-type, and this confuses some software.
3141 PH/70 Catch two compiler warnings in sieve.c.
3143 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3144 function verify_get_ident() calls ip_connect() to connect a socket, but
3145 if the "connect()" function timed out, ip_connect() used to close the
3146 socket. However, verify_get_ident() also closes the socket later, and in
3147 between Exim writes to the log, which may get opened at this point. When
3148 the socket was closed in ip_connect(), the log could get the same file
3149 descriptor number as the socket. This naturally causes chaos. The fix is
3150 not to close the socket in ip_connect(); the socket should be closed by
3151 the function that creates it. There was only one place in the code where
3152 this was missing, in the iplookup router, which I don't think anybody now
3153 uses, but I've fixed it anyway.
3155 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3156 well as to direct DNS lookups. Otherwise the handling of names in host
3157 lists is inconsistent and therefore confusing.
3163 PH/01 Two changes to the default runtime configuration:
3165 (1) Move the checks for relay_from_hosts and authenticated clients from
3166 after to before the (commented out) DNS black list checks.
3168 (2) Add control=submission to the relay_from_hosts and authenticated
3169 clients checks, on the grounds that messages accepted by these
3170 statements are most likely to be submissions.
3172 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3174 (1) Generate an error if the third argument for the ${prvs expansion is
3177 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3180 (3) Reset the variables that are obtained from the first argument of
3181 ${prvscheck and used in the second argument before leaving the code,
3182 because their memory is reclaimed, so using them afterwards may do
3185 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3186 one (it's much easier than Tom thought :-).
3188 (5) Because of (4), we can now allow for the use of $prvscheck_result
3189 inside the third argument.
3191 PH/03 For some reason, the default setting of PATH when running a command from
3192 a pipe transport was just "/usr/bin". I have changed it to
3195 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3196 anything to be listed in the output from -bV.
3198 PH/05 When a filter generated an autoreply, the entire To: header line was
3199 quoted in the delivery log line, like this:
3201 => >A.N.Other <ano@some.domain> <original@ddress> ...
3203 This has been changed so that it extracts the operative address. There
3204 may be more than one such address. If so, they are comma-separated, like
3207 => >ano@some.domain,ona@other.domain <original@ddress> ...
3209 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3210 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3211 not being looked up in the rDNS to get a host name, Exim was showing the
3212 IP address twice in Received: lines, even though the IP addresses were
3213 identical. For example:
3215 Received: from [1.2.3.4] (helo=[1.2.3.4])
3217 However, if the real host name was known, it was omitting the HELO data
3218 if it matched the actual IP address. This has been tidied up so that it
3219 doesn't show the same IP address twice.
3221 PH/07 When both +timestamp and +memory debugging was on, the value given by
3222 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3223 called by the debug printing, thereby overwriting the timestamp buffer.
3224 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3226 PH/08 When the original message was included in an autoreply transport, it
3227 always said "this is a copy of the message, including all the headers",
3228 even if body_only or headers_only was set. It now gives an appropriate
3231 PH/09 Applied a patch from the Sieve maintainer which:
3233 o fixes some comments
3234 o adds the (disabled) notify extension core
3235 o adds some debug output for the result of if/elsif tests
3236 o points to the current vacation draft in the documentation
3237 and documents the missing references header update
3241 o fixes a bug in processing the envelope test (when testing
3242 multiple envelope elements, the last element determined the
3245 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3246 Electronic Mail") by including:
3248 Auto-submitted: auto-generated
3250 in the messages that it generates (bounce messages and others, such as
3251 warnings). In the case of bounce messages for non-SMTP messages, there was
3252 also a typo: it was using "Auto_submitted" (underscore instead of
3253 hyphen). Since every message generated by Exim is necessarily in response
3254 to another message, thes have all been changed to:
3256 Auto-Submitted: auto-replied
3258 in accordance with these statements in the RFC:
3260 The auto-replied keyword:
3262 - SHOULD be used on messages sent in direct response to another
3263 message by an automatic process,
3265 - MUST NOT be used on manually-generated messages,
3267 - MAY be used on Delivery Status Notifications (DSNs) and Message
3268 Disposition Notifications (MDNs),
3270 - MUST NOT be used on messages generated by automatic or periodic
3271 processes, except for messages which are automatic responses to
3274 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3275 to the default Received: header definition.
3277 PH/12 Added log selector acl_warn_skipped (default on).
3279 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3280 variables because (a) they are in the wrong storage pool and (b) even if
3281 they were copied, it wouldn't work properly because of the caching.
3283 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3284 checking when decoding. Apparently there are clients that generate
3285 overlong encoded strings. Why am I not surprised?
3287 PH/15 If the first argument of "${if match_address" was not empty, but did not
3288 contain an "@" character, Exim crashed. Now it writes a panic log message
3289 and treats the condition as false.
3291 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3293 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3294 extension "envelope-auth". The code is finished and in agreement with
3295 other implementations, but there is no documentation so far and in fact,
3296 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3297 not changing the active code.
3299 Print executed "if" and "elsif" statements when debugging is used. This
3300 helps a great deal to understand what a filter does.
3302 Document more things not specified clearly in RFC3028. I had all this
3303 sorted out, when out of a sudden new issues came to my mind. Oops."
3305 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3308 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3309 canonical form (as documented). However, after a host name lookup from
3310 the IP address, check_host() was doing a simple string comparison with
3311 addresses acquired from the DNS when checking that the found name did
3312 have the original IP as one of its addresses. Since any found IPv6
3313 addresses are likely to be in abbreviated form, the comparison could
3314 fail. Luckily, there already exists a function for doing the comparison
3315 by converting both addresses to binary, so now that is used instead of
3316 the text comparison.
3318 PH/20 There was another similar case to PH/19, when a complete host name was
3319 given in a host list; looking up its IP address could give an abbreviated
3320 form, whereas the current host's name might or might not be abbreviated.
3321 The same fix has been applied.
3327 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3328 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3331 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3332 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3334 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3336 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3337 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3338 Also, for "Auto-Submitted", treat anything other than "no" as
3339 non-personal, in accordance with RFC 3834. (Previously it treated
3340 anything starting "auto-" as non-personal.)
3342 TF/01 The control=submission/name=... option had a problem with syntax
3343 errors if the name included a slash character. The /name= option
3344 now slurps the rest of the string, so it can include any characters
3345 but it must come last in the list of options (after /sender_retain
3348 PH/05 Some modifications to the interface to the fake nameserver for the new
3356 TK/01 Added the "success_on_redirect" address verification option. See
3357 NewStuff for rationale and an example.
3359 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3361 PH/02 Patch to exigrep to allow it to work on syslog lines.
3363 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3364 fread() to skip over the body file's header line, because in Cygwin the
3365 header line is locked and is inaccessible.
3367 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3368 co-exist for some time) to make it clear that it is the Exim ID that is
3369 referenced, not the Message-ID: header line.
3371 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3372 string_format() function, because snprintf() does not exist on all
3375 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3376 ${stat: expansion item.
3378 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3379 protocol synchronization error", to keep the pedants happy.
3381 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3382 well as for IRIX systems, when gcc is being used. See the host.c source
3385 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3387 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3390 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3391 transport and to the smtp transport in LMTP mode.
3393 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3395 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3396 run a filter in a subprocess. This could lead to confusion in subsequent
3397 lookups in the parent process. There should also be a search_tidyup() at
3398 the end of the subprocess.
3400 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3401 only if the host matched helo_try_verify_hosts, which caused the
3402 verification to occur when the EHLO/HELO command was issued. The ACL just
3403 tested the remembered result. Now, if a previous verification attempt has
3404 not happened, "verify = helo" does it there and then.
3406 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3408 TK/03 Fix log output including CR from clamd.
3410 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3411 exist provoked a memory error which could cause a segfault.
3413 PH/15 Installed PCRE 6.2
3415 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3417 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3418 of the problem. Specifically, suggested +O2 rather than +O1 for the
3421 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3423 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3424 record, the port was not correctly logged when the outgoing_port log
3425 selector was set (it logged the transort's default port).
3427 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3428 fallback_hosts, and "hosts" in the smtp transport.
3430 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3431 host errors such as "Connection refused".
3433 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3434 authentication with radiusclient 0.4.9:
3436 - Error returned from rc_read_config was caught wrongly
3437 - Username/password not passed on to radius server due to wrong length.
3439 The presumption is that some radiusclient API changes for 4.51/PH/17
3440 were not taken care of correctly. The code is still untested by me (my
3441 Linux distribution still has 0.3.2 of radiusclient), but it was
3442 contributed by a Radius user.
3444 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3445 expanding the "port" option of the smtp transport.
3447 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3448 while reading a MIME header. Thanks to Tom Hughes for a patch.
3450 PH/24 Include config.h inside local_scan.h so that configuration settings are
3453 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3454 that in an ACL for RCPT (for example), you can examine exactly what was
3457 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3458 commands, but it was not correctly comparing the address with the actual
3459 client host address. Thus, it would show the EHLO address in Received:
3460 header lines when this was not necessary.
3462 PH/27 Added the % operator to ${eval:}.
3464 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3465 it should be ignoring failures (because with -C, for example, it has lost
3466 privilege). It wasn't ignoring creation failures other than "already
3469 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3472 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3473 because an input line was too long, either on its own, or by virtue of
3474 too many continuations, the temporary file was not being removed, and the
3475 return code was incorrect.
3477 PH/31 Missing "BOOL" in function definition in filtertest.c.
3479 PH/32 Applied Sieve patches from the maintainer.
3481 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3483 PH/33 Added "verify = not_blind".
3485 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3486 Local/Makefile (with some defaults set). These are used in built scripts
3487 such as exicyclog, but they have never been used in the exim_install
3488 script (though there are many overriding facilities there). I have
3489 arranged that the exim_install script now takes note of these two
3492 PH/35 Installed configuration files for Dragonfly.
3494 PH/36 When a locally submitted message by a trusted user did not contain a
3495 From: header, and the sender address was obtained from -f or from an SMTP
3496 MAIL command, and the trusted user did not use -F to supply a sender
3497 name, $originator_name was incorrectly used when constructing a From:
3498 header. Furthermore, $originator_name was used for submission mode
3499 messages from external hosts without From: headers in a similar way,
3500 which is clearly wrong.
3502 PH/37 Added control=suppress_local_fixups.
3504 PH/38 When log_selector = +received_sender was set, and the addition of the
3505 sender made the log line's construction buffer exactly full, or one byte
3506 less than full, an overflow happened when the terminating "\n" was
3509 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3510 when the result of a list match is failure because a DNS lookup failed.
3512 PH/40 RM_COMMAND is now used in the building process.
3514 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3515 the "build-* directories that it finds.
3517 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3518 address in a domain literal was a prefix of an interface address.
3520 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3521 when verifying a sender address, unless rewrite_headers is false.
3523 PH/44 Wrote a long comment about why errors_to addresses are verified as
3524 recipients, not senders.
3526 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3527 the ratelimit ACL was added.
3529 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3531 PH/46 Added extra information about PostgreSQL errors to the error string.
3533 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3534 avoiding the need to install special zones in a real server. This is
3535 backwards compatible; if it can't find the fake resolver, it drops back.
3536 Thus, both old and new test suites can be run.
3538 TF/02 Added util/ratelimit.pl
3540 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3541 clock is set back in time.
3543 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3544 Candler <B.Candler@pobox.com>.
3546 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3547 OK for addresses that are the result of redirections.
3549 PH/48 A number of further additions for the benefit of the new test suite,
3550 including a fake gethostbyname() that interfaces to the fake DNS resolver
3553 TF/06 The fix for widen_domains has also been applied to qualify_single and
3554 search_parents which are the other dnslookup options that can cause
3557 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3560 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3562 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3563 service" if it encountered an MX record with an empty target hostname.
3564 The message is now "an MX or SRV record indicated no SMTP service".
3566 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3567 if the DNS of the sending site is misconfigured. This is quite a
3568 common situation. This change restores the behaviour of treating a
3569 helo verification defer as a failure.
3571 PH/51 If self=fail was set on a router, the bounce message did not include the
3572 actual error message.
3578 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3580 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3581 command itself ended in error, the underlying message about the transport
3582 filter timeout was being overwritten with the pipe command error. Now the
3583 underlying error message should be appended to the second error message.
3585 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3587 PH/02 Reduce the amount of output that "make" produces by default. Full output
3588 can still be requested.
3590 PH/03 The warning log line about a condition test deferring for a "warn" verb
3591 was being output only once per connection, rather than after each
3592 occurrence (because it was using the same function as for successful
3593 "warn" verbs). This seems wrong, so I have changed it.
3595 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3596 it should not have, which might have caused a crash in the right
3597 circumstances, but probably never did.
3599 PH/04 Installed a modified version of Tony Finch's patch to make submission
3600 mode fix the return path as well as the Sender: header line, and to
3601 add a /name= option so that you can make the user's friendly name appear
3604 TF/03 Added the control = fakedefer ACL modifier.
3606 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3607 Mark Lowes for thorough testing.
3609 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3611 TK/03 Merged latest SRS patch from Miles Wilton.
3613 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3614 with the definition in sysexits.h (which is #included earlier).
3615 Fortunately, Exim does not actually use EX_OK. The code used to try to
3616 preserve the sysexits.h value, by assuming that macro definitions were
3617 scanned for macro replacements. I have been disabused of this notion,
3618 so now the code just undefines EX_OK before #including unistd.h.
3620 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3621 in the smtp transport. When a block could not be written in a single
3622 write() function, the timeout was being re-applied to each part-write.
3623 This seems wrong - if the receiver was accepting one byte at a time it
3624 would take for ever. The timeout is now adjusted when this happens. It
3625 doesn't have to be particularly precise.
3627 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3628 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3630 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3631 without a domain if the check to <postmaster@domain> fails.
3633 SC/01 Eximstats: added -xls and the ability to specify output files
3634 (patch written by Frank Heydlauf).
3636 SC/02 Eximstats: use FileHandles for outputting results.
3638 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3640 SC/04 Eximstats: fixed display of large numbers with -nvr option
3642 SC/05 Eximstats: fixed merging of reports with empty tables.
3644 SC/06 Eximstats: added the -include_original_destination flag
3646 SC/07 Eximstats: removed tabs and trailing whitespace.
3648 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3650 TK/06 MBOX spool code: Add real "From " MBOX separator line
3651 so the .eml file is really in mbox format (even though
3652 most programs do not really care). Patch from Alex Miller.
3654 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3655 The latter is generated from $received_to and is only set if the
3656 message has one envelope recipient. SA can use these headers,
3657 obviously out-of-the-box. Patch from Alex Miller.
3659 PH/08 The ${def test on a variable was returning false if the variable's
3660 value was "0", contrary to what the specification has always said!
3661 The result should be true unless the variable is empty.
3663 PH/09 The syntax error of a character other than { following "${if
3664 def:variable_name" (after optional whitespace) was not being diagnosed.
3665 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3666 accidental colon was present, for example, could give incorrect results.
3668 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3669 result is used (not including appendfile, where other changes are about
3672 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3673 This involved changing a lot of size variables from int to off_t. It
3674 should work with maildirs and everything.
3676 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3677 spamd dying while we are connected to it.
3679 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3682 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3683 function for BDB 4.3.
3685 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3687 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3688 the sources for PCRE have been re-arranged and more files are now
3691 PH/15 The code I had for printing potentially long long variables in PH/11
3692 above was not the best (it lost precision). The length of off_t variables
3693 is now inspected at build time, and an appropriate printing format (%ld
3694 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3695 to be "long long int" or "long int". This is needed for the internal
3696 formatting function string_vformat().
3698 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3699 the configuration file to be ":syslog", then the script "guesses" where
3700 the logs files are, rather than using the compiled in default. In our
3701 case the guess is not the same as the compiled default, so the script
3702 suddenly stopped working when I started to use syslog. The patch checks
3703 to see if log_file_path is "". If so, it attempts to read it from exim
3704 with no configuration file to get the compiled in version, before it
3705 falls back to the previous guessing code."
3707 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3708 implementing BATV in an Exim configuration. See NewStuff for the gory
3711 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3712 Makefile that are specific to HP-UX.
3714 PH/18 If the "use_postmaster" option was set for a recipient callout together
3715 with the "random" option, the postmaster address was used as the MAIL
3716 FROM address for the random test, but not for the subsequent recipient
3717 test. It is now used for both.
3719 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3720 patch removes a few documentation additions to RFC 3028, because the
3721 latest draft now contains them. It adds the new en;ascii-case comparator
3722 and a new error check for 8bit text in MIME parts. Comparator and
3723 require names are now matched exactly. I enabled the subaddress
3724 extension, but it is not well tested yet (read: it works for me)."
3726 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3727 rework some of the code of TK/09 above to avoid the hardwired use of
3728 "%lld" and "long long". Replaced the call to snprintf() with a call to
3731 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3732 records point to non-existent hosts", "retry timeout exceeded", and
3733 "retry time not reached for any host after a long failure period".
3735 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3736 experimental DomainKeys support:
3738 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3739 (2) On an error such as an illegally used "control", the wrong name for
3740 the control was given.
3742 These problems did NOT occur unless DomainKeys support was compiled.
3744 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3746 PH/24 Added ${if match_ip condition.
3748 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3749 databases so that it will be absolutely obvious if a crash occurs in the
3750 DB library. This is a regular occurrence (often caused by mis-matched
3753 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3754 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3755 picked up on a user's system that detects such things. There doesn't seem
3756 to be a gcc warning option for this - only an attribute that has to be
3757 put on the function's prototype. It seems that in Fedora Core 4 they have
3758 set this on a number of new functions. No doubt there will be more in due
3761 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3762 specify a transport. However, if an address that was verified by such a
3763 router was the subject of a callout, Exim crashed because it tried to
3764 read the rcpt_include_affixes from the non-existent transport. Now it
3765 just assumes that the setting of that option is false. This bug was
3766 introduced by 4.51/PH/31.
3768 PH/28 Changed -d+all to exclude +memory, because that information is very
3769 rarely of interest, but it makes the output a lot bigger. People tend to
3770 do -d+all out of habit.
3772 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3773 code in os-type was giving problems when libc.so lives in lib64, like on
3776 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3777 aren't the modern standard, and it seems that some systems' include files
3778 don't always have them. Exim was already checking for some of the newer
3779 ones like T_AAAA, and defining it itself. I've added checks for all the
3780 record types that Exim uses.
3782 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3783 not automatically generating a new one, as it is supposed to. This
3784 prevented TLS from working. If the file did exist, but contained invalid
3785 data, a new version was generated, as expected. It was only the case of a
3786 non-existent file that was broken.
3788 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3789 with a change in libdomainkeys > 0.64.
3791 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3792 from DNS. If the selector record carries the flag, it now has
3793 precedence over the domain-wide flag.
3795 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3797 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3798 the use of a port name that isn't defined in /etc/services) occurred, the
3799 message was deferred as in a normal delivery, and thus remained on the
3800 spool, instead of being failed because of the mua_wrapper setting. This
3801 is now fixed, and I tidied up some of the mua_wrapper messages at the
3804 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3805 the messages in a hash of arrays rather than using individual hashes.
3806 This is a bit cleaner and results in dramatic memory savings, albeit
3807 at a slight CPU cost.
3809 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3810 as requested by Marc Sherman.
3812 SC/10 Eximstats: added histograms for user specified patterns as requested
3815 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3817 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3818 fopen() in the content-scanning modules that did not already have it.
3824 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3825 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3827 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3829 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3831 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3832 to test Sieve filters that use "vacation".
3834 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3835 that changes the way the GnuTLS parameters are stored in the cache file.
3836 The new format can be generated externally. For backward compatibility,
3837 if the data in the cache doesn't make sense, Exim assumes it has read an
3838 old-format file, and it generates new data and writes a new file. This
3839 means that you can't go back to an older release without removing the
3842 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3843 work if there are any delivery delays because "one_time" forces the
3844 parent to be marked "delivered", so its unseen clone is never tried
3845 again. For this reason, Exim now forbids the simultaneous setting of
3848 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3849 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3850 entry if you want to know the details. The fix, however, neglected to
3851 consider the case when local delivery batching is involved. The test for
3852 "previously delivered" was not happening when checking to see if an
3853 address could be batched with a previous (undelivered) one; under
3854 certain circumstances this could lead to multiple deliveries to the same
3857 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3858 in its include files, and this causes problems building Exim.
3860 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3861 header_syntax) but Exim was just ignoring anything given after a slash.
3862 In particular, this caused confusion with an attempt to use "verify =
3863 reverse_host_lookup/defer_ok". An error is now given when options are
3864 supplied for verify items that do not have them. (Maybe reverse_host_
3865 lookup should have a defer_ok option, but that's a different point.)
3867 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3868 defined by RFC 821) to 2048, because there were problems with some AUTH
3869 commands, and RFC 1869 says the size should be increased for extended
3870 SMTP commands that take arguments.
3872 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3875 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3876 "unknown" error; now it says that the functionality isn't in the binary.
3878 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3879 an address' error message when a string expansion fails (syntax or
3880 whatever). Otherwise the password may appear in the log. Following change
3881 PH/42 below, there is no longer a chance of it appearing in a bounce
3884 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3886 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3887 of its addresses was ever tried. (Bugzilla bug #2.)
3889 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3890 the result incorrectly in the debug output. (It correctly added a newline
3891 to what was transported.)
3893 TF/01 Added $received_time.
3895 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3896 commented out examples of how to interface to a virus scanner and to
3897 SpamAssassin. Also added commented examples of av_scanner and
3898 spamd_address settings.
3900 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3901 and controls are allowed in which ACLs. There were a couple of minor
3902 errors. Some of the entries in the conditions table (which is a table of
3903 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3904 negation of where the condition IS allowed.
3906 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3908 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3909 header file does not have a version number, so I've had to invent a new
3910 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3911 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3912 radiusclient), but it was contributed by a Radius user.
3914 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3915 files or directories, for OpenSSL.
3917 PH/19 When an Exim process that is running as root has to create an Exim log
3918 file, it does so in a subprocess that runs as exim:exim so as to get the
3919 ownership right at creation (otherwise, other Exim processes might see
3920 the file with the wrong ownership). There was no test for failure of this
3921 fork() call, which would lead to the process getting stuck as it waited
3922 for a non-existent subprocess. Forks do occasionally fail when resources
3923 run out. I reviewed all the other calls to fork(); they all seem to check
3926 PH/20 When checking for unexpected SMTP input at connect time (before writing
3927 the banner), Exim was not dealing correctly with a non-positive return
3928 from the read() function. If the client had disconnected by this time,
3929 the result was a log entry for a synchronization error with an empty
3930 string after "input=" when read() returned zero. If read() returned -1
3931 (an event I could not check), uninitialized data bytes were printed.
3932 There were reports of junk text (parts of files, etc) appearing after
3935 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3937 PH/22 Added support for macro redefinition, and (re)definition in between
3938 driver and ACL definitions.
3940 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3941 forgetting to use the resulting value; it was using the unexpanded value.
3943 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3944 hadn't been configured. The fix is from Juergen Kreileder, who
3945 understands it better than I do:
3947 "Here's what I see happening with three configured cyrus_sasl
3948 authenticators configured (plain, login, cram-md5):
3950 On startup auth_cyrus_sasl_init() gets called for each of these.
3951 This means three calls to sasl_listmech() without a specified mech_list.
3952 => SASL tests which mechs of all available mechs actually work
3953 => three warnings about OTP not working
3954 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3956 With the patch, sasl_listmech() also gets called three times. But now
3957 SASL's mech_list option is set to the server_mech specified in the the
3958 authenticator. Or in other words, the answer from sasl_listmech()
3959 gets limited to just the mech you're testing for (which is different
3961 => the return list contains just 'plain' or 'login', 'cram-md5' or
3962 nothing depending on the value of ob->server_mech.
3964 I've just tested the patch: Authentication still works fine,
3965 unavailable mechs specified in the exim configuration are still
3966 caught, and the auth.log warnings about OTP are gone."
3968 PH/25 When debugging is enabled, the contents of the command line are added
3969 to the debugging output, even when log_selector=+arguments is not
3972 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3973 answer is "GNU", and only if the return is "GNU/something" is the answer
3976 PH/27 $acl_verify_message is now set immediately after the failure of a
3977 verification in an ACL, and so is available in subsequent modifiers. In
3978 particular, the message can be preserved by coding like this:
3980 warn !verify = sender
3981 set acl_m0 = $acl_verify_message
3983 Previously, $acl_verify_message was set only while expanding "message"
3984 and "log_message" when a very denied access.
3986 PH/28 Modified OS/os.c-Linux with
3988 -#ifndef OS_LOAD_AVERAGE
3989 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3991 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3992 nomenclature these days.)
3994 PH/29 Installed patch from the Sieve maintainer that adds the options
3995 sieve_useraddress and sieve_subaddress to the redirect router.
3997 PH/30 In these circumstances:
3998 . Two addresses routed to the same list of hosts;
3999 . First host does not offer TLS;
4000 . First host accepts first address;
4001 . First host gives temporary error to second address;
4002 . Second host offers TLS and a TLS session is established;
4003 . Second host accepts second address.
4004 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4005 and peerdn, if requested) that were in fact used only for the second
4008 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4009 attention to any local part prefix or suffix that was matched by the
4010 router that accepted the address. It now behaves in the same way as it
4011 does for delivery: the affixes are removed from the local part unless
4012 rcpt_include_affixes is set on the transport.
4014 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4015 timeout during the DATA phase of an incoming message.
4017 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4018 applied a patch sanctioned by the Sieve maintainer.
4020 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4021 the uid or gid is negative. A case of a negative gid caused this to be
4022 noticed. The fix allows for either to be negative.
4024 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4025 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4028 PH/36 The change PH/12 above was broken. Fixed it.
4030 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4031 the grounds that routing the same address twice would always produce the
4032 same answer. This might have been true once, but it is certainly no
4033 longer true now. Routing a child address may depend on the previous
4034 routing that produced that child. Some complicated redirection strategies
4035 went wrong when messages had multiple recipients, and made Exim's
4036 behaviour dependent on the order in which the addresses were given.
4038 I have moved the duplicate checking until after the routing is complete.
4039 Exim scans the addresses that are assigned to local and remote
4040 transports, and removes any duplicates. This means that more work will be
4041 done, as duplicates will always all be routed, but duplicates are
4042 presumably rare, so I don't expect this is of any significance.
4044 For deliveries to pipes, files, and autoreplies, the duplicate checking
4045 still happens during the routing process, since they are not going to be
4048 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4049 It corrects a timeout issue with spamd. This is Ian's comment: "The
4050 background is that sometimes spamd either never reads data from a
4051 connection it has accepted, or it never writes response data. The exiscan
4052 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4053 blindly assumes that writes won't block so it may never time out."
4055 PH/39 Allow G after quota size as well as K and M.
4057 PH/40 The value set for $authenticated_id in an authenticator may not contain
4058 binary zeroes or newlines because the value is written to log lines and
4059 to spool files. There was no check on this. Now the value is run through
4060 the string_printing() function so that such characters are converted to
4061 printable escape sequences.
4063 PH/41 $message_linecount is a new variable that contains the total number of
4064 lines in the message. Compare $body_linecount, which is the count for the
4067 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4068 bounce and delay warning messages, except in certain special cases, which
4071 (a) An SMTP error message from a remote host;
4072 (b) A message specified in a :fail: redirection;
4073 (c) A message specified in a "fail" command in a system filter;
4074 (d) A message specified in a FAIL return from the queryprogram router;
4075 (e) A message specified by the cannot_route_message router option.
4077 In these cases only, Exim does include the error details in bounce and
4078 warning messages. There are also a few cases where bland messages such
4079 as "unrouteable address" or "local delivery error" are given.
4081 PH/43 $value is now also set for the "else" part of a ${run expansion.
4083 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4084 being worked on, but at least Exim now implements the latest version to
4087 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4088 process to complete was treated as a delivery failure, a timeout while
4089 writing the message to the pipe was logged, but erroneously treated as a
4090 successful delivery. Such timeouts include transport filter timeouts. For
4091 consistency with the overall process timeout, these timeouts are now
4092 treated as errors, giving rise to delivery failures by default. However,
4093 there is now a new Boolean option for the pipe transport called
4094 timeout_defer, which, if set TRUE, converts the failures into defers for
4095 both kinds of timeout. A transport filter timeout is now identified in
4098 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4099 systems where "make" and "gmake" are different, calling "gmake" at top
4100 level broke things. I've arranged for the value of $(MAKE) to be passed
4101 from the Makefile to this script so that it can call the same version of
4105 A note about Exim versions 4.44 and 4.50
4106 ----------------------------------------
4108 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4109 changes of various kinds. As a consequence, a big documentation update was
4110 needed. This delayed the release for rather longer than seemed good, especially
4111 in the light of a couple of (minor) security issues. Therefore, the changes
4112 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4113 release. So 4.44 and 4.50 are in effect two different branches that both start
4116 I have left the 4.50 change log unchanged; it contains all the changes since
4117 4.43. The change log for 4.44 is below; many of its items are identical to
4118 those for 4.50. This seems to be the most sensible way to preserve the
4119 historical information.
4125 1. Minor wording change to the doc/README.SIEVE file.
4127 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4128 computation of the current number of files was incorrect.
4130 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4131 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4134 4. Give more explanation in the error message when the command for a transport
4135 filter fails to execute.
4137 5. There are several places where Exim runs a non-Exim command in a
4138 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4139 was being done only for the command run by the queryprogram router. It is
4140 now done for all such subprocesses. The other cases are: ${run, transport
4141 filters, and the commands run by the lmtp and pipe transports.
4143 6. Added CONFIGURE_GROUP build-time option.
4145 7. Some older OS have a limit of 256 on the maximum number of file
4146 descriptors. Exim was using setrlimit() to set 1000 as a large value
4147 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4148 systems. I've change it so that if it can't get 1000, it tries for 256.
4150 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4151 was an oversight, and furthermore, ever since the addition of extra
4152 controls (e.g. 4.43/32), the checks on when to allow different forms of
4153 "control" were broken. There should now be diagnostics for all cases when a
4154 control that does not make sense is encountered.
4156 9. Added the /retain_sender option to "control=submission".
4158 10. $recipients is now available in the predata ACL (oversight).
4160 11. Tidy the search cache before the fork to do a delivery from a message
4161 received from the command line. Otherwise the child will trigger a lookup
4162 failure and thereby defer the delivery if it tries to use (for example) a
4163 cached ldap connection that the parent has called unbind on.
4165 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4166 of $address_data from the recipient verification was clobbered by the
4167 sender verification.
4169 13. The value of address_data from a sender verification is now available in
4170 $sender_address_data in subsequent conditions in the ACL statement.
4172 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4174 15. Added a new option "connect=<time>" to callout options, to set a different
4177 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4178 was its contents. (It was OK if the option was not defined at all.)
4180 17. A "Completed" log line is now written for messages that are removed from
4181 the spool by the -Mrm option.
4183 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4184 information about exactly what failed.
4186 19. Added -dd to debug only the daemon process.
4188 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4189 handles timeouts, both on the server side and network timeouts. Renamed the
4190 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4192 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4193 It is now set to "smtps".
4195 22. $host_address is now set to the target address during the checking of
4196 ignore_target_hosts.
4198 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4199 being passed; this would have caused $sender_host_name to have been used if
4200 matching the list had actually called for a host name (not very likely,
4201 since this list is usually IP addresses). A host name is now passed as
4204 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4205 code that specifies a non-restarting handler (typically sigaction() in
4206 modern systems) in an attempt to fix a rare and obscure crash bug.
4208 25. Narrowed the window for a race in the daemon that could cause it to ignore
4209 SIGCHLD signals. This is not a major problem, because they are used only to
4210 wake it up if nothing else does.
4212 26. A malformed maildirsize file could cause Exim to calculate negative values
4213 for the mailbox size or file count. Odd effects could occur as a result.
4214 The maildirsize information is now recalculated if the size or filecount
4217 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4218 support for a long time. Removed HAVE_SYS_VFS_H.
4220 28. Installed the latest version of exipick from John Jetmore.
4222 29. In an address list, if the pattern was not a regular expression, an empty
4223 subject address (from a bounce message) matched only if the pattern was an
4224 empty string. Non-empty patterns were not even tested. This was the wrong
4225 because it is perfectly reasonable to use an empty address as part of a
4226 database query. An empty address is now tested by patterns that are
4227 lookups. However, all the other forms of pattern expect the subject to
4228 contain a local part and a domain, and therefore, for them, an empty
4229 address still always fails if the pattern is not itself empty.
4231 30. Exim went into a mad DNS loop when attempting to do a callout where the
4232 host was specified on an smtp transport, and looking it up yielded more
4233 than one IP address.
4235 31. Re-factored the code for checking spool and log partition space into a
4236 function that finds that data and another that does the check. The former
4237 is then used to implement four new variables: $spool_space, $log_space,
4238 $spool_inodes, and $log_inodes.
4240 32. The RFC2047 encoding function was originally intended for short strings
4241 such as real names; it was not keeping to the 75-character limit for
4242 encoded words that the RFC imposes. It now respects the limit, and
4243 generates multiple encoded words if necessary. To be on the safe side, I
4244 have increased the buffer size for the ${rfc2047: expansion operator from
4247 33. It is now permitted to omit both strings after an "if" condition; if the
4248 condition is true, the result is "true". As before, when the second string
4249 is omitted, a false condition yields an empty string. This makes it less
4250 cumbersome to write custom ACL and router conditions.
4252 34. Failure to deliver a bounce message always caused it to be frozen, even if
4253 there was an errors_to setting on the router. The errors_to setting is now
4256 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4257 canonical form (fully expanded) before being placed in
4258 $sender_host_address.
4260 36. The table in the code that translates DNS record types into text (T_A to
4261 "A" for instance) was missing entries for NS and CNAME. It is just possible
4262 that this could have caused confusion if both these types were looked up
4263 for the same domain, because the text type is used as part of Exim's
4264 per-process caching. But the chance of anyone hitting this buglet seems
4267 37. The dnsdb lookup has been extended in a number of ways.
4269 (1) There is a new type, "zns", which walks up the domain tree until it
4270 finds some nameserver records. It should be used with care.
4272 (2) There is a new type, "mxh", which is like "mx" except that it returns
4273 just the host names, not the priorities.
4275 (3) It is now possible to give a list of domains (or IP addresses) to be
4276 looked up. The behaviour when one of the lookups defers can be
4277 controlled by a keyword.
4279 (4) It is now possible to specify the separator character for use when
4280 multiple records are returned.
4282 38. The dnslists ACL condition has been extended: it is now possible to supply
4283 a list of IP addresses and/or domains to be looked up in a particular DNS
4286 39. Added log_selector=+queue_time_overall.
4288 40. When running the queue in the test harness, wait just a tad after forking a
4289 delivery process, to get repeatability of debugging output.
4291 41. Include certificate and key file names in error message when GnuTLS fails
4292 to set them up, because the GnuTLS error message doesn't include the name
4293 of the failing file when there is a problem reading it.
4295 42. Allow both -bf and -bF in the same test run.
4297 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4299 44. The "Exiscan patch" is now merged into the mainline Exim source.
4301 45. Sometimes the final signoff response after QUIT could fail to get
4302 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4303 < 0 before doing a fflush(). This bug looks as though it goes back to the
4304 introduction of TLS in release 3.20, but "sometimes" must have been rare
4305 because the tests only now provoked it.
4307 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4308 (this can affect the format of dates).
4310 47. exim_tidydb, when checking for the continued existence of a message for
4311 which it has found a message-specific retry record, was not finding
4312 messages that were in split spool directories. Consequently, it was
4313 deleting retry records that should have stayed in existence.
4315 48. Steve fixed some bugs in eximstats.
4317 49. The SPA authentication driver was not abandoning authentication and moving
4318 on to the next authenticator when an expansion was forced to fail,
4319 contradicting the general specification for all authenticators. Instead it
4320 was generating a temporary error. It now behaves as specified.
4322 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4323 (the order specifies the preference for clients). The order is now AES256,
4324 AES128, 3DES, ARCFOUR128.
4326 51. Small patch to Sieve code - explicitly set From: when generating an
4329 52. Exim crashed if a remote delivery caused a very long error message to be
4330 recorded - for instance if somebody sent an entire SpamAssassin report back
4331 as a large number of 550 error lines. This bug was coincidentally fixed by
4332 increasing the size of one of Exim's internal buffers (big_buffer) that
4333 happened as part of the Exiscan merge. However, to be on the safe side, I
4334 have made the code more robust (and fixed the comments that describe what
4337 53. Now that there can be additional text after "Completed" in log lines (if
4338 the queue_time_overall log selector is set), a one-byte patch to exigrep
4339 was needed to allow it to recognize "Completed" as not the last thing in
4342 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4343 patch that reportedly fixes this has been added. I am not expert enough to
4344 create a test for it. This is what the patch creator wrote:
4346 "I found a little strange behaviour of ldap code when working with
4347 Windows 2003 AD Domain, where users was placed in more than one
4348 Organization Units. When I tried to give exim partial DN, the exit code
4349 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4350 But simultaneously result of request was absolutely normal ldap result,
4351 so I produce this patch..."
4353 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4354 REFERENCE, so I have modified the code to exclude the patch when that macro
4357 55. Some experimental protocols are using DNS PTR records for new purposes. The
4358 keys for these records are domain names, not reversed IP addresses. The
4359 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4360 leaves it alone. Component reversal etc. now happens only for IP addresses.
4363 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4365 57. Double the size of the debug message buffer (to 2048) so that more of very
4366 long debug lines gets shown.
4368 58. The exicyclog utility now does better if the number of log files to keep
4369 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4371 59. Two changes related to the smtp_active_hostname option:
4373 (1) $smtp_active_hostname is now available as a variable.
4374 (2) The default for smtp_banner uses $smtp_active_hostname instead
4375 of $primary_hostname.
4377 60. The host_aton() function is supposed to be passed a string that is known
4378 to be a valid IP address. However, in the case of IPv6 addresses, it was
4379 not checking this. This is a hostage to fortune. Exim now panics and dies
4380 if the condition is not met. A case was found where this could be provoked
4381 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4382 components; fortuitously, this particular loophole had already been fixed
4383 by change 4.50/55 above.
4385 If there are any other similar loopholes, the new check in host_aton()
4386 itself should stop them being exploited. The report I received stated that
4387 data on the command line could provoke the exploit when Exim was running as
4388 exim, but did not say which command line option was involved. All I could
4389 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4390 running as the user.
4393 61. There was a buffer overflow vulnerability in the SPA authentication code
4394 (which came originally from the Samba project). I have added a test to the
4395 spa_base64_to_bits() function which I hope fixes it.
4398 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4399 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4401 63. The daemon start-up calls getloadavg() while still root for those OS that
4402 need the first call to be done as root, but it missed one case: when
4403 deliver_queue_load_max is set with deliver_drop_privilege. This is
4404 necessary for the benefit of the queue runner, because there is no re-exec
4405 when deliver_drop_privilege is set.
4407 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4408 This has been fixed.
4410 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4411 was in use, was not putting the data itself into the right store pool;
4412 consequently, it could be overwritten for a subsequent message in the same
4413 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4416 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4418 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4419 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4420 the function were treating the return as a boolean value, which happened to
4421 work because 0=false and not-0=true, but is not correct code.
4423 68. The host_aton() function was not handling scoped IPv6 addresses (those
4424 with, for example, "%eth0" on the end) correctly.
4426 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4427 negated items (that is, ~something) in unsigned ints. Some compilers
4428 apparently mutter when there is no cast.
4430 70. If an address verification called from an ACL failed, and did not produce a
4431 user-specific message (i.e. there was only a "system" message), nothing was
4432 put in $acl_verify_message. In this situation, it now puts the system
4435 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4436 session; change 4.31/43 added the unwanted input to the log line - except
4437 that it did not do this in the start of session case. It now does.
4439 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4440 This is wrong and can cause the other end to generate a synchronization
4441 error if it is another Exim or anything else that does the synchronization
4442 check. A QUIT command is no longer sent after a timeout.
4444 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4445 during host lookups.
4447 74. The defer_ok option of callout verification was not working if it was used
4448 when verifying addresses in header lines, that is, for this case:
4450 verify = header_sender/callout=defer_ok
4452 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4453 those file descriptors could be used for SMTP connections. If anything
4454 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4455 could be sent to the SMTP client, causing chaos. The daemon now opens
4456 stdin, stdout, and stderr to /dev/null when it puts itself into the
4459 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4460 log by default. The user can override this with suitable Perl magic.
4462 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4463 the log message when discard triggers, was not working for the DATA ACL or
4464 for the non-SMTP ACL.
4466 78. Error message wording change in sieve.c.
4468 79. If smtp_accept_max_per_host was set, the number of connections could be
4469 restricted to fewer than expected, because the daemon was trying to set up
4470 a new connection before checking whether the processes handling previous
4471 connections had finished. The check for completed processes is now done
4472 earlier. On busy systems, this bug wouldn't be noticed because something
4473 else would have woken the daemon, and it would have reaped the completed
4476 80. If a message was submitted locally by a user whose login name contained one
4477 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4478 It caused a spool format error. I have fixed the spool reading code. A
4479 related problem was that the "from" clause in the Received: line became
4480 illegal because of the space(s). It is now covered by ${quote_local_part.
4482 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4485 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4487 83. Patch from Sieve maintainer to fix unterminated string problem in
4488 "vacation" handling.
4490 84. Some minor changes to the Linux configuration files to help with other
4491 OS variants using glibc.
4493 85. One more patch for Sieve to update vacation handling to latest spec.
4496 ----------------------------------------------------
4497 See the note above about the 4.44 and 4.50 releases.
4498 ----------------------------------------------------
4504 1. Change 4.43/35 introduced a bug that caused file counts to be
4505 incorrectly computed when quota_filecount was set in an appendfile
4508 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4509 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4512 3. Give more explanation in the error message when the command for a transport
4513 filter fails to execute.
4515 4. There are several places where Exim runs a non-Exim command in a
4516 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4517 was being done only for the command run by the queryprogram router. It is
4518 now done for all such subprocesses. The other cases are: ${run, transport
4519 filters, and the commands run by the lmtp and pipe transports.
4521 5. Some older OS have a limit of 256 on the maximum number of file
4522 descriptors. Exim was using setrlimit() to set 1000 as a large value
4523 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4524 systems. I've change it so that if it can't get 1000, it tries for 256.
4526 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4527 was an oversight, and furthermore, ever since the addition of extra
4528 controls (e.g. 4.43/32), the checks on when to allow different forms of
4529 "control" were broken. There should now be diagnostics for all cases when a
4530 control that does not make sense is encountered.
4532 7. $recipients is now available in the predata ACL (oversight).
4534 8. Tidy the search cache before the fork to do a delivery from a message
4535 received from the command line. Otherwise the child will trigger a lookup
4536 failure and thereby defer the delivery if it tries to use (for example) a
4537 cached ldap connection that the parent has called unbind on.
4539 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4540 of $address_data from the recipient verification was clobbered by the
4541 sender verification.
4543 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4544 was its contents. (It was OK if the option was not defined at all.)
4546 11. A "Completed" log line is now written for messages that are removed from
4547 the spool by the -Mrm option.
4549 12. $host_address is now set to the target address during the checking of
4550 ignore_target_hosts.
4552 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4553 being passed; this would have caused $sender_host_name to have been used if
4554 matching the list had actually called for a host name (not very likely,
4555 since this list is usually IP addresses). A host name is now passed as
4558 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4559 code that specifies a non-restarting handler (typically sigaction() in
4560 modern systems) in an attempt to fix a rare and obscure crash bug.
4562 15. Narrowed the window for a race in the daemon that could cause it to ignore
4563 SIGCHLD signals. This is not a major problem, because they are used only to
4564 wake it up if nothing else does.
4566 16. A malformed maildirsize file could cause Exim to calculate negative values
4567 for the mailbox size or file count. Odd effects could occur as a result.
4568 The maildirsize information is now recalculated if the size or filecount
4571 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4572 support for a long time. Removed HAVE_SYS_VFS_H.
4574 18. Updated exipick to current release from John Jetmore.
4576 19. Allow an empty sender to be matched against a lookup in an address list.
4577 Previously the only cases considered were a regular expression, or an
4580 20. Exim went into a mad DNS lookup loop when doing a callout where the
4581 host was specified on the transport, if the DNS lookup yielded more than
4584 21. The RFC2047 encoding function was originally intended for short strings
4585 such as real names; it was not keeping to the 75-character limit for
4586 encoded words that the RFC imposes. It now respects the limit, and
4587 generates multiple encoded words if necessary. To be on the safe side, I
4588 have increased the buffer size for the ${rfc2047: expansion operator from
4591 22. Failure to deliver a bounce message always caused it to be frozen, even if
4592 there was an errors_to setting on the router. The errors_to setting is now
4595 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4596 canonical form (fully expanded) before being placed in
4597 $sender_host_address.
4599 24. Updated eximstats to version 1.33
4601 25. Include certificate and key file names in error message when GnuTLS fails
4602 to set them up, because the GnuTLS error message doesn't include the name
4603 of the failing file when there is a problem reading it.
4605 26. Expand error message when OpenSSL has problems setting up cert/key files.
4608 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4609 (this can affect the format of dates).
4611 28. exim_tidydb, when checking for the continued existence of a message for
4612 which it has found a message-specific retry record, was not finding
4613 messages that were in split spool directories. Consequently, it was
4614 deleting retry records that should have stayed in existence.
4616 29. eximstats updated to version 1.35
4617 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4618 1.35 - bugfix such that pie charts by volume are generated correctly
4620 30. The SPA authentication driver was not abandoning authentication and moving
4621 on to the next authenticator when an expansion was forced to fail,
4622 contradicting the general specification for all authenticators. Instead it
4623 was generating a temporary error. It now behaves as specified.
4625 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4626 (the order specifies the preference for clients). The order is now AES256,
4627 AES128, 3DES, ARCFOUR128.
4629 31. Small patch to Sieve code - explicitly set From: when generating an
4632 32. Exim crashed if a remote delivery caused a very long error message to be
4633 recorded - for instance if somebody sent an entire SpamAssassin report back
4634 as a large number of 550 error lines. This bug was coincidentally fixed by
4635 increasing the size of one of Exim's internal buffers (big_buffer) that
4636 happened as part of the Exiscan merge. However, to be on the safe side, I
4637 have made the code more robust (and fixed the comments that describe what
4640 33. Some experimental protocols are using DNS PTR records for new purposes. The
4641 keys for these records are domain names, not reversed IP addresses. The
4642 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4643 leaves it alone. Component reversal etc. now happens only for IP addresses.
4646 34. The host_aton() function is supposed to be passed a string that is known
4647 to be a valid IP address. However, in the case of IPv6 addresses, it was
4648 not checking this. This is a hostage to fortune. Exim now panics and dies
4649 if the condition is not met. A case was found where this could be provoked
4650 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4651 components; fortuitously, this particular loophole had already been fixed
4652 by change 4.50/55 or 4.44/33 above.
4654 If there are any other similar loopholes, the new check in host_aton()
4655 itself should stop them being exploited. The report I received stated that
4656 data on the command line could provoke the exploit when Exim was running as
4657 exim, but did not say which command line option was involved. All I could
4658 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4659 running as the user.
4662 35. There was a buffer overflow vulnerability in the SPA authentication code
4663 (which came originally from the Samba project). I have added a test to the
4664 spa_base64_to_bits() function which I hope fixes it.
4667 36. The daemon start-up calls getloadavg() while still root for those OS that
4668 need the first call to be done as root, but it missed one case: when
4669 deliver_queue_load_max is set with deliver_drop_privilege. This is
4670 necessary for the benefit of the queue runner, because there is no re-exec
4671 when deliver_drop_privilege is set.
4673 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4674 was in use, was not putting the data itself into the right store pool;
4675 consequently, it could be overwritten for a subsequent message in the same
4676 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4679 38. Sometimes the final signoff response after QUIT could fail to get
4680 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4681 < 0 before doing a fflush(). This bug looks as though it goes back to the
4682 introduction of TLS in release 3.20, but "sometimes" must have been rare
4683 because the tests only now provoked it.
4689 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4690 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4691 Now it returns an integer. A number of calls were still expecting a T/F
4692 return. Fortuitously, in all cases, the tests worked in OK situations,
4693 which is the norm. However, things would have gone wrong on any write
4694 failures on the smtp file descriptor. This function is used when sending
4695 messages over SMTP and also when doing verify callouts.
4697 2. When Exim is called to do synchronous delivery of a locally submitted
4698 message (the -odf or -odi options), it no longer closes stderr before doing
4701 3. Implemented the mua_wrapper option.
4703 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4705 5. Implemented the functions header_remove(), header_testname(),
4706 header_add_at_position(), and receive_remove_recipient(), and exported them
4709 6. If an ACL "warn" statement specified the addition of headers, Exim already
4710 inserted X-ACL-Warn: at the start if there was no header name. However, it
4711 was not making this test for the second and subsequent header lines if
4712 there were newlines in the string. This meant that an invalid header could
4713 be inserted if Exim was badly configured.
4715 7. Allow an ACL "warn" statement to add header lines at the start or after all
4716 the Received: headers, as well as at the end.
4718 8. Added the rcpt_4xx retry error code.
4720 9. Added postmaster_mailfrom=xxx to callout verification option.
4722 10. Added mailfrom=xxxx to the callout verification option, for verify=
4725 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4726 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4728 12. Inserted some casts to stop certain compilers warning when using pointer
4729 differences as field lengths or precisions in printf-type calls (mostly
4730 affecting debugging statements).
4732 13. Added optional readline() support for -be (dynamically loaded).
4734 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4735 same clock tick as a message's arrival, so that its received time was the
4736 same as the "first fail" time on the retry record, and that message
4737 remained on the queue past the ultimate address timeout, every queue runner
4738 would try a delivery (because it was past the ultimate address timeout) but
4739 after another failure, the ultimate address timeout, which should have then
4740 bounced the address, did not kick in. This was a "< instead of <=" error;
4741 in most cases the first failure would have been in the next clock tick
4742 after the received time, and all would be well.
4744 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4745 being recognized when the domain list was tested by the match_domain
4746 condition in an expansion string.
4748 16. Added the ${str2b64: operator.
4750 17. Exim was always calling setrlimit() to set a large limit for the number of
4751 processes, without checking whether the existing limit was already
4752 adequate. (It did check for the limit on file descriptors.) Furthermore,
4753 errors from getrlimit() and setrlimit() were being ignored. Now they are
4754 logged to the main and panic logs, but Exim does carry on, to try to do its
4755 job under whatever limits there are.
4757 18. Imported PCRE 5.0.
4759 19. Trivial typo in log message " temporarily refused connection" (the leading
4762 20. If the log selector return_path_on_delivery was set and an address was
4763 redirected to /dev/null, the delivery process crashed because it assumed
4764 that a return path would always be set for a "successful" delivery. In this
4765 case, the whole delivery is bypassed as an optimization, and therefore no
4768 21. Internal re-arrangement: the function for sending a challenge and reading
4769 a response while authentication was assuming a zero-terminated challenge
4770 string. It's now changed to take a pointer and a length, to allow for
4771 binary data in such strings.
4773 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4775 23. Exim was not respecting finduser_retries when seeking the login of the
4776 uid under which it was called; it was always trying 10 times. (The default
4777 setting of finduser_retries is zero.) Also, it was sleeping after the final
4778 failure, which is pointless.
4780 24. Implemented tls_on_connect_ports.
4782 25. Implemented acl_smtp_predata.
4784 26. If the domain in control=submission is set empty, Exim assumes that the
4785 authenticated id is a complete email address when it generates From: or
4786 Sender: header lines.
4788 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4789 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4790 chown and chgrp in /bin and hostname in /usr/bin.
4792 28. Exim was keeping the "process log" file open after each use, just as it
4793 does for the main log. This opens the possibility of it remaining open for
4794 long periods when the USR1 signal hits a daemon. Occasional processlog
4795 errors were reported, that could have been caused by this. Anyway, it seems
4796 much more sensible not to leave this file open at all, so that is what now
4799 29. The long-running daemon process does not normally write to the log once it
4800 has entered its main loop, and it closes the log before doing so. This is
4801 so that log files can straightforwardly be renamed and moved. However,
4802 there are a couple of unusual error situations where the daemon does write
4803 log entries, and I had neglected to close the log afterwards.
4805 30. The text of an SMTP error response that was received during a remote
4806 delivery was being truncated at 512 bytes. This is too short for some of
4807 the long messages that one sometimes sees. I've increased the limit to
4810 31. It is now possible to make retry rules that apply only when a message has a
4811 specific sender, in particular, an empty sender.
4813 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4814 it possible to be selective about when SMTP synchronization is enforced.
4816 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4818 32. Implemented hosts_connection_nolog.
4820 33. Added an ACL for QUIT.
4822 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4825 35. Added mailbox_size and mailbox_filecount to appendfile.
4827 36. Added control = no_multiline_responses to ACLs.
4829 37. There was a bug in the logic of the code that waits for the clock to tick
4830 in the case where the clock went backwards by a substantial amount such
4831 that the microsecond fraction of "now" was more than the microsecond
4832 fraction of "then" (but the whole seconds number was less).
4834 38. Added support for the libradius Radius client library this is found on
4835 FreeBSD (previously only the radiusclient library was supported).
4841 1. When certain lookups returned multiple values in the form name=value, the
4842 quoting of the values was not always being done properly. Specifically:
4843 (a) If the value started with a double quote, but contained no whitespace,
4845 (b) If the value contained whitespace other than a space character (i.e.
4846 tabs or newlines or carriage returns) it was not quoted.
4847 This fix has been applied to the mysql and pgsql lookups by writing a
4848 separate quoting function and calling it from the lookup code. The fix
4849 should probably also be applied to nisplus, ibase and oracle lookups, but
4850 since I cannot test any of those, I have not disturbed their existing code.
4852 2. A hit in the callout cache for a specific address caused a log line with no
4853 reason for rejecting RCPT. Now it says "Previous (cached) callout
4854 verification failure".
4856 3. There was an off-by-one bug in the queryprogram router. An over-long
4857 return line was truncated at 256 instead of 255 characters, thereby
4858 overflowing its buffer with the terminating zero. As well as fixing this, I
4859 have increased the buffer size to 1024 (and made a note to document this).
4861 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4862 when Exim is waiting for an SMTP response from a remote server, Exim
4863 restarts its select() call on the socket, thereby resetting its timeout.
4864 This is not a problem when such interrupts are rare. Somebody set up a cron
4865 job to run exiwhat every 2 minutes, which is less than the normal select()
4866 timeout (5 or 10 minutes). This meant that the select() timeout never
4867 kicked in because it was always reset. I have fixed this by comparing the
4868 time when an interrupt arrives with the time at the start of the first call
4869 to select(). If more time than the timeout has elapsed, the interrupt is
4870 treated as a timeout.
4872 5. Some internal re-factoring in preparation for the addition of Sieve
4873 extensions (by MH). In particular, the "personal" test is moved to a
4874 separate function, and given an option for scanning Cc: and Bcc: (which is
4875 not set for Exim filters).
4877 6. When Exim created an email address using the login of the caller as the
4878 local part (e.g. when creating a From: or Sender: header line), it was not
4879 quoting the local part when it contained special characters such as @.
4881 7. Installed new OpenBSD configuration files.
4883 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4884 try to make them clearer.
4886 9. Callout options, other than the timeout value, were being ignored when
4887 verifying sender addresses in header lines. For example, when using
4889 verify = header_sender/callout=no_cache
4891 the cache was (incorrectly) being used.
4893 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4894 only the Cygwin environment.
4896 11. When return_path_on_delivery was set as a log selector, if different remote
4897 addresses in the same message used different return paths and parallel
4898 remote delivery occurred, the wrong values would sometimes be logged.
4899 (Whenever a remote delivery process finished, the return path value from
4900 the most recently started remote delivery process was logged.)
4902 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4903 lines when AUTH and/or TLS are in use. This is the "received protocol"
4904 field. Exim used to use "asmtp" for authenticated SMTP, without any
4905 indication (in the protocol name) for TLS use. Now it follows the RFC and
4906 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4907 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4908 names appear in log lines as well as in Received: header lines.
4910 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4911 extensions, and comparison tests, and to fix some bugs.
4913 14. Changes to the "personal" filter test:
4915 (1) The test was buggy in that it was just doing the equivalent of
4916 "contains" tests on header lines. For example, if a user's address was
4917 anne@some.where, the "personal" test would incorrectly be true for
4919 To: susanne@some.where
4921 This test is now done by extracting each address from the header in turn,
4922 and checking the entire address. Other tests that are part of "personal"
4923 are now done using regular expressions (for example, to check local parts
4924 of addresses in From: header lines).
4926 (2) The list of non-personal local parts in From: addresses has been
4927 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4928 taken from the Sieve specification recommendations.
4930 (3) If the message contains any header line starting with "List-" it is
4931 treated as non-personal.
4933 (4) The test for "circular" in the Subject: header line has been removed
4934 because it now seems ill-conceived.
4936 15. Minor typos in src/EDITME comments corrected.
4938 16. Installed latest exipick from John Jetmore.
4940 17. If headers_add on a router specified a text string that was too long for
4941 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4942 of string_sprintf() is now avoided.
4944 18. $message_body_size was not set (it was always zero) when running the DATA
4945 ACL and the local_scan() function.
4947 19. For the "mail" command in an Exim filter, no default was being set for
4948 the once_repeat time, causing a random time value to be used if "once" was
4949 specified. (If the value happened to be <= 0, no repeat happened.) The
4950 default is now 0s, meaning "never repeat". The "vacation" command was OK
4951 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4952 (I found it when inspecting the code).
4954 20. There is now an overall timeout for performing a callout verification. It
4955 defaults to 4 times the callout timeout, which applies to individual SMTP
4956 commands during the callout. The overall timeout applies when there is more
4957 than one host that can be tried. The timeout is checked before trying the
4958 next host. This prevents very long delays if there are a large number of
4959 hosts and all are timing out (e.g. when the network connections are timing
4960 out). The value of the overall timeout can be changed by specifying an
4961 additional sub-option for "callout", called "maxwait". For example:
4963 verify = sender/callout=5s,maxwait=20s
4965 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4966 to the end before writing, but this should make it even safer).
4968 22. Exim was forgetting that it had advertised PIPELINING for the second and
4969 subsequent messages on an SMTP connection. It was also not resetting its
4970 memory on STARTTLS and an internal HELO.
4972 23. When Exim logs an SMTP synchronization error within a session, it now
4973 records whether PIPELINING has been advertised or not.
4975 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4976 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4977 rather than long int.
4979 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4981 26. Added the never_mail option to autoreply.
4987 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4988 crash if the getsockname() call failed; this can happen if a connection is
4989 closed very soon after it is established. The problem was simply in the
4990 order in which certain operations were done, causing Exim to try to write
4991 to the SMTP stream before it had set up the file descriptor. The bug has
4992 been fixed by making things happen in the correct order.
4998 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4999 before the connection was closed, thus losing the rejection response.
5001 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5002 some early Solaris releases, but causes trouble in current releases where
5003 socklen_t is defined.
5005 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5008 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5011 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5012 that releasing the top of it at the end releases what was used for sub-
5013 expansions (unless the block got too big). However, discard this block if
5014 the first thing is a variable or header, so that we can use its block when
5015 it is dynamic (useful for very large $message_headers, for example).
5017 6. Lookups now cache *every* query, not just the most recent. A new, separate
5018 store pool is used for this. It can be recovered when all lookup caches are
5019 flushed. Lookups now release memory at the end of their result strings.
5020 This has involved some general refactoring of the lookup sources.
5022 7. Some code has been added to the store_xxx() functions to reduce the amount
5023 of flapping under certain conditions.
5025 8. log_incoming_interface used to affect only the <= reception log lines. Now
5026 it causes the local interface and port to be added to several more SMTP log
5027 lines, for example "SMTP connection from", and rejection lines.
5029 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5031 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5033 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5034 could be overwritten at the end of the current message (or the start of a
5035 new message if it was set in a HELO ACL). The value is now preserved for
5036 the duration of the SMTP connection.
5038 12. If a transport had a headers_rewrite setting, and a matching header line
5039 contained an unqualified address, that address was qualified, even if it
5040 did not match any rewriting rules. The underlying bug was that the values
5041 of the flags that permit the existence of unqualified sender and recipient
5042 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5043 non-local messages, and by -bnq for local messages) were not being
5044 preserved with the message after it was received.
5046 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5047 "next input=" as part of the text comprising the host identity instead of
5048 the correct text. The code was using the same buffer for two different
5049 strings. However, depending on which order the printing function evaluated
5050 its arguments, the bug did not always show up. Under Linux, for example, my
5051 test suite worked just fine.
5053 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5054 doesn't work with some older versions of Perl. It has been changed to "my",
5055 which in any case is probably the better facility to use.
5057 15. A really picky compiler found some instances of statements for creating
5058 error messages that either had too many or two few arguments for the format
5061 16. The size of the buffer for calls to the DNS resolver has been increased
5062 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5063 for addresses that have a lot of PTR records. This alleviates a problem; it
5064 does not fully solve it.
5066 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5067 buffer now truncates the list and logs the incident, which is the same
5068 action as happens when Exim is looking up a host name and its aliases.
5069 Previously in this situation something unpredictable would happen;
5070 sometimes it was "internal error: store_reset failed".
5072 18. If a server dropped the connection unexpectedly when an Exim client was
5073 using GnuTLS and trying to read a response, the client delivery process
5074 crashed while trying to generate an error log message.
5076 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5077 string, for example:
5079 warn message = H1: something\nH2: something
5081 the text was added as a single header line from Exim's point of view
5082 though it ended up OK in the delivered message. However, searching for the
5083 second and subsequent header lines using $h_h2: did not work. This has been
5084 fixed. Similarly, if a system filter added multiple headers in this way,
5085 the routers could not see them.
5087 20. Expanded the error message when iplsearch is called with an invalid key to
5088 suggest using net-iplsearch in a host list.
5090 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5091 ACLs are no longer actually imposed (and a message to that effect is
5094 22. If a "gecos" field in a passwd entry contained escaped characters, in
5095 particular, if it contained a \" sequence, Exim got it wrong when building
5096 a From: or a Sender: header from that name. A second bug also caused
5097 incorrect handling when an unquoted " was present following a character
5098 that needed quoting.
5100 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5101 was not being matched caselessly.
5103 24. Arranged for all hyphens in the exim.8 source to be escaped with
5106 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5107 properly. Recipient callout cache records were still being keyed to include
5108 the sender, even when use_sender was set false. This led to far more
5109 callouts that were necessary. The sender is no longer included in the key
5110 when use_sender is false.
5112 26. Added "control = submission" modifier to ACLs.
5114 27. Added the ${base62d: operator to decode base 62 numbers.
5116 28. dnsdb lookups can now access SRV records.
5118 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5119 the configuration file.
5121 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5122 (-v) mode. This makes the output for a verbose queue run more intelligible.
5124 31. Added a use_postmaster feature to recipient callouts.
5126 32. Added the $body_zerocount variable, containing the number of binary zero
5127 bytes in the message body.
5129 33. The time of last modification of the "new" subdirectory is now used as the
5130 "mailbox time last read" when there is a quota error for a maildir
5133 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5135 35. Added +ignore_unknown as a special item in host lists.
5137 36. Code for decoding IPv6 addresses in host lists is now included, even if
5138 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5139 address was recognized as an IP address, but was then not correctly decoded
5140 into binary, causing unexpected and incorrect effects when compared with
5147 1. Very minor rewording of debugging text in manualroute to say "list of
5148 hosts" instead of "hostlist".
5150 2. If verify=header_syntax was set, and a header line with an unqualified
5151 address (no domain) and a large number of spaces between the end of the
5152 name and the colon was received, the reception process suffered a buffer
5153 overflow, and (when I tested it) crashed. This was caused by some obsolete
5154 code that should have been removed. The fix is to remove it!
5156 3. When running in the test harness, delay a bit after writing a bounce
5157 message to get a bit more predictability in the log output.
5159 4. Added a call to search_tidyup() just before forking a reception process. In
5160 theory, someone could use a lookup in the expansion of smtp_accept_max_
5161 per_host which, without the tidyup, could leave open a database connection.
5163 5. Added the variables $recipient_data and $sender_data which get set from a
5164 lookup success in an ACL "recipients" or "senders" condition, or a router
5165 "senders" option, similar to $domain_data and $local_part_data.
5167 6. Moved the writing of debug_print from before to after the "senders" test
5170 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5171 problems for message scanning, either using a data ACL, or using
5172 local_scan() because the Received: header was not generated till after they
5173 were called (in order to set the time as the time of reception completion).
5174 I have revised the way this works. The header is now generated after the
5175 body is received, but before the ACL or local_scan() are called. After they
5176 are run, the timestamp in the header is updated.
5182 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5183 before starting a queue runner without re-exec. This happened only when
5184 deliver_drop_privilege was set or when the Exim user was set to root. The
5185 effect of the bug was that timeouts during subsequent deliveries caused
5186 crashes instead of being properly handled. The handler is now left at its
5187 default (and expected) setting.
5189 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5190 message, again when deliver_drop_privilege is set or Exim is run as root.
5191 The bug described in (1) was not present in this case, but the tidying up
5192 of the other signals was missing. I have made the two cases consistent.
5194 3. The ignore_target_hosts setting on a manualroute router was being ignored
5195 for hosts that were looked up using the /MX notation.
5197 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5200 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5201 operated on the sender address. After changing the $sender_address to <>
5202 for the sender address verify, Exim was re-instated it as the original
5203 (before rewriting) address, but remembering that it had rewritten it, so it
5204 wasn't rewriting it again. This bug also had the effect of breaking the
5205 sender address verification caching when the sender address was rewritten.
5207 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5208 This has been changed so that if the ip literal address matches
5209 ignore_target_hosts, the router declines.
5211 7. Added expansion conditions match_domain, match_address, and match_local_
5212 part (NOT match_host).
5214 8. The placeholder for the Received: header didn't have a length field set.
5216 9. Added code to Exim itself and to exim_lock to test for a specific race
5217 condition that could lead to file corruption when using MBX delivery. The
5218 issue is with the lockfile that is created in /tmp. If this file is removed
5219 after a process has opened it but before that process has acquired a lock,
5220 there is the potential for a second process to recreate the file and also
5221 acquire a lock. This could lead to two Exim processes writing to the file
5222 at the same time. The added code performs the same test as UW imapd; it
5223 checks after acquiring the lock that its file descriptor still refers to
5224 the same named file.
5226 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5227 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5230 11. Added the smtp_active_hostname option. If used, this will typically be made
5231 to depend on the incoming interface address. Because $interface_address is
5232 not set up until the daemon has forked a reception process, error responses
5233 that can happen earlier (such as "too many connections") no longer contain
5236 12. If an expansion in a condition on a "warn" statement fails because a lookup
5237 defers, the "warn" statement is abandoned, and the next ACL statement is
5238 processed. Previously this caused the whole ACL to be aborted.
5240 13. Added the iplsearch lookup type.
5242 14. Added ident_timeout as a log selector.
5244 15. Added tls_certificate_verified as a log selector.
5246 16. Added a global option tls_require_ciphers (compare the smtp transport
5247 option of the same name). This controls incoming TLS connections.
5249 17. I finally figured out how to make tls_require_ciphers do a similar thing
5250 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5251 before starting the TLS session.
5253 18. Tabs are now shown as \t in -bP output.
5255 19. If the log selector return_path_on_delivery was set, Exim crashed when
5256 bouncing a message because it had too many Received: header lines.
5258 20. If two routers both had headers_remove settings, and the first one included
5259 a superfluous trailing colon, the final name in the first list and the
5260 first name in the second list were incorrectly joined into one item (with a
5261 colon in the middle).
5267 1. Added -C and -D options to the exinext utility, mainly to make it easier
5268 to include in the automated testing, but these could be helpful when
5269 multiple configurations are in use.
5271 2. The exinext utility was not formatting the output nicely when there was
5272 an alternate port involved in the retry record key, nor when there was a
5273 message id as well (for retries that were specific to a specific message
5274 and a specific host). It was also confused by IPv6 addresses, because of
5275 the additional colons they contain. I have fixed the IPv4 problem, and
5276 patched it up to do a reasonable job for IPv6.
5278 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5279 delivery, the log line now contains "pipelined" if PIPELINING was used.
5281 4. An SMTP transport process used to panic and die if the bind() call to set
5282 an explicit outgoing interface failed. This has been changed; it is now
5283 treated in the same way as a connect() failure.
5285 5. A reference to $sender_host_name in the part of a conditional expansion
5286 that was being skipped was still causing a DNS lookup. This no longer
5289 6. The def: expansion condition was not recognizing references to header lines
5290 that used bh_ and bheader_.
5292 7. Added the _cache feature to named lists.
5294 8. The code for checking quota_filecount in the appendfile transport was
5295 allowing one more file than it should have been.
5297 9. For compatibility with Sendmail, the command line option
5305 and sets the incoming protocol and host name (for trusted callers). The
5306 host name and its colon can be omitted when only the protocol is to be set.
5307 Note the Exim already has two private options, -pd and -ps, that refer to
5308 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5309 "s", but I don't think that's a major issue.
5311 10. A number of refactoring changes to the code, none of which should affect
5314 (a) The number of logging options was getting close to filling up the
5315 32-bit word that was used as a bit map. I have split them into two classes:
5316 those that are passed in the argument to log_write(), and those that are
5317 only ever tested independently outside of that function. These are now in
5318 separate 32-bit words, so there is plenty of room for expansion again.
5319 There is no change in the user interface or the logging behaviour.
5321 (b) When building, for example, log lines, the code previously used a
5322 macro that called string_cat() twice, in order to add two strings. This is
5323 not really sufficiently general. Furthermore, there was one instance where
5324 it was actually wrong because one of the argument was used twice, and in
5325 one call a function was used. (As it happened, calling the function twice
5326 did not affect the overall behaviour.) The macro has been replaced by a
5327 function that can join an arbitrary number of extra strings onto a growing
5330 (c) The code for expansion conditions now uses a table and a binary chop
5331 instead of a serial search (which was left over from when there were very
5332 few conditions). Also, it now recognizes conditions like "pam" even when
5333 the relevant support is not compiled in: a suitably worded error message is
5334 given if an attempt is made to use such a condition.
5336 11. Added ${time_interval:xxxxx}.
5338 12. A bug was causing one of the ddress fields not to be passed back correctly
5339 from remote delivery subprocesses. The field in question was not being
5340 subsequently used, so this caused to problems in practice.
5342 13. Added new log selectors queue_time and deliver_time.
5344 14. Might have fixed a bug in maildirsizefile handling that threw up
5345 "unexpected character" debug warnings, and recalculated the data
5346 unnecessarily. In any case, I expanded the warning message to give more
5349 15. Added the message "Restricted characters in address" to the statements in
5350 the default ACL that block characters like @ and % in local parts.
5352 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5353 Three changes have been made:
5355 (a) There was a serious bug; a negative response to MAIL caused the whole
5356 recipient domain to be cached as invalid, thereby blocking all messages
5357 to all local parts at the same domain, from all senders. This bug has
5358 been fixed. The domain is no longer cached after a negative response to
5359 MAIL if the sender used is not empty.
5361 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5364 (c) A new callout option, "use_sender" has been added for people who want
5365 the modified behaviour.
5371 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5374 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5375 indeed breaks things for older releases.
5377 3. Added additional logging to the case where there is a problem reading data
5378 from a filter that is running in a subprocess using a pipe, in order to
5379 try to track down a specific problem.
5381 4. Testing facility fudge: when running in the test harness and attempting
5382 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5383 getting "No route to host". Convert this to a timeout.
5385 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5388 6. Some OS don't have socklen_t but use size_t instead. This affects the
5389 fifth argument of getsockopt() amongst other things. This is now
5390 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5391 can be set for individual OS. I have set it for SunOS5, OSF1, and
5392 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5393 some earlier ones do not.
5395 7. Change 4.30/15 was not doing the test caselessly.
5397 8. The standard form for an IPv6 address literal was being rejected by address
5398 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5399 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5400 this, as well as the form without the "IPv6" on the front (but only when
5401 address literals are enabled, of course).
5403 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5405 10. Exim crashed if a message with an empty sender address specified by -f
5406 encountered a router with an errors_to setting. This could be provoked only
5407 by a command such as
5411 where an empty string was supplied; "<>" did not hit this bug.
5413 11. Installed PCRE release 4.5.
5415 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5416 remained set. It is now erased.
5418 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5419 times from message ids (which are base 36 rather than the normal 62).
5421 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5422 were being counted as actual protocol errors, and logged if the log
5423 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5424 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5425 and DATA following a set of rejected RCPTs do not count as protocol errors.
5426 In other words, Exim assumes they were pipelined, though this may not
5427 actually be the case. Of course, in all cases the client gets an
5428 appropriate error code.
5430 15. If a lookup fails in an ACL condition, a message about the failure may
5431 be available; it is used if testing the ACL cannot continue, because most
5432 such messages specify what the cause of the deferral is. However, some
5433 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5434 that caused an old message to be retained and used if a later statement
5435 caused a defer, replacing the real cause of the deferral.
5437 16. If an IP address had so many PTR records that the DNS lookup buffer
5438 was not large enough to hold them, Exim could crash while trying to process
5439 the truncated data. It now detects and logs this case.
5441 17. Further to 4.21/58, another change has been made: if (and only if) the
5442 first line of a message (the first header line) ends with CRLF, a bare LF
5443 in a subsequent header line has a space inserted after it, so as not to
5444 terminate the header.
5446 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5447 unnecessarily, used atoi() instead of strtol(), and didn't check the
5448 termination when getting file sizes from file names by regex.
5450 19. Completely re-implemented the support for maildirsize files, in the light
5451 of a number of problems with the previous contributed implementation
5452 (4.30/29). In particular:
5454 . If the quota is zero, the maildirsize file is maintained, but no quota is
5457 . If the maildir directory does not exist, it is created before any attempt
5458 to write a maildirsize file.
5460 . The quota value in the file is just a cache; if the quota is changed in
5461 the transport, the new value overrides.
5463 . A regular expression is available for excluding directories from the
5466 20. The autoreply transport checks the characters in options that define the
5467 message's headers; it allows continued headers, but it was checking with
5468 isspace() after an embedded newline instead of explicitly looking for a
5471 21. If all the "regular" hosts to which an address was routed had passed their
5472 expiry times, and had not reached their retry times, the address was
5473 bounced, even if fallback hosts were defined. Now Exim should go on to try
5476 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5477 equivalent code in the SMTP transport. Some hosts send humungous responses
5478 to HELO/EHLO, more than 1024 it seems.
5480 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5481 gives compiler warnings in some environments. I've now done it "properly",
5484 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5485 (because of problems with the built-in one) was declared to return uschar *
5486 instead of char *, causing compiler failure.
5488 25. Fixed a file descriptor leak when processing alias/forward files.
5490 26. Fixed a minor format string issue in dbfn.c.
5492 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5494 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5495 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5498 29. When a sender address is verified, it is cached, to save repeating the test
5499 when there is more than one recipient in a message. However, when the
5500 verification involves a callout, it is possible for different callout
5501 options to be set for different recipients. It is too complicated to keep
5502 track of this in the cache, so now Exim always runs a verification when a
5503 callout is required, relying on the callout cache for the optimization.
5504 The overhead is duplication of the address routing, but this should not be
5507 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5508 to be verified with callout=postmaster, and the main callout worked but the
5509 postmaster check failed, the verification correctly failed. However, if a
5510 subsequent RCPT command asked for sender verification *without* the
5511 postmaster check, incorrect caching caused this verification also to fail,
5514 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5515 it was not caching the DNS options (qualify_single, search_parents) that
5516 were used when the lookup failed. A subsequent lookup with different
5517 options therefore always gave the same answer, though there were cases
5518 where it should not have. (Example: a "domains = !$mx_any" option on a
5519 dnslookup router: the "domains" option is always processed without any
5520 widening, but the router might have qualify_single set.) Now Exim uses the
5521 cached value only when the same options are set.
5523 32. Added John Jetmore's "exipick" utility to the distribution.
5525 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5526 than a timeout (e.g. a certificate is required, and is not provided), an
5527 Exim server now closes the connection immediately. Previously it waited for
5528 the client to close - but if the client is SSL, it seems that they each
5529 wait for each other, leading to a delay before one of them times out.
5531 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5532 maintained 0.8.x compatibility because I don't think many are using it, and
5533 it is clearly obsolete.
5535 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5538 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5539 client certificate was expired. A simple patch fixes this, though I don't
5540 understand the full logic of why the verify callback is called multiple
5543 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5544 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5545 which causes problems with some clients (such as the Certicom SSL Plus
5546 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5547 disables the coutermeasure allowing Eudora to connect."
5549 38. Exim was not checking that a write() to a log file succeeded. This could
5550 lead to Bad Things if a log got too big, in particular if it hit a file
5551 size limit. Exim now panics and dies if it cannot write to a log file, just
5552 as it does if it cannot open a log file.
5554 39. Modified OS/Makefile-Linux so that it now contains
5556 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5558 The two -D definitions ensure that Exim is compiled with large file
5559 support, which makes it possible to handle log files that are bigger than
5562 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5563 instance) a domain was checked against a named list that involved a lookup,
5564 causing $domain_data to be set, then another domain was checked against the
5565 same list, then the first domain was re-checked, the value of $domain_data
5566 after the final check could be wrong. In particular, if the second check
5567 failed, it could be set empty. This bug probably also applied to
5570 41. The strip_trailing_dot option was not being applied to the address given
5571 with the -f command-line option.
5573 42. The code for reading a message's header from the spool was incrementing
5574 $received_count, but never initializing it. This meant that the value was
5575 incorrect (doubled) while delivering a message in the same process in which
5576 it was received. In the most common configuration of Exim, this never
5577 happens - a fresh exec is done - but it can happen when
5578 deliver_drop_privilege is set.
5580 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5581 it now includes up to 150 characters of the unexpected data in the log
5584 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5585 and building data strings. The size of both of these buffers was 10 000
5586 bytes - far larger than anybody would *ever* want, thought I. Needless to
5587 say, somebody hit the limit. I have increased the maximum line length to
5588 20 000 and the maximum data length of concatenated lines to 100 000. I have
5589 also fixed two bugs, because there was no checking on these buffers. Tsk,
5590 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5591 buffer is too small.
5593 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5594 lsearch lookups. Now it does.
5596 46. When parsing a route_list item in a manualroute router, a fixed-length
5597 buffer was used for the list of hosts. I made this 1024 bytes long,
5598 thinking that nobody would ever have a list of hosts that long. Wrong.
5599 Somebody had a whole pile of complicated expansion conditions, and the
5600 string was silently truncated, leading to an expansion error. It turns out
5601 that it is easier to change to an unlimited length (owing to other changes
5602 that have happened since this code was originally written) than to build
5603 structure for giving a limitation error. The length of the item that
5604 expands into the list of hosts is now unlimited.
5606 47. The lsearch lookup could not handle data where the length of text line was
5607 more than 4095 characters. Such lines were truncated, leading to shortened
5608 data being returned. It should now handle lines of any length.
5610 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5611 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5614 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5615 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5616 no longer happens in comment lines. A list of replacements is now placed
5617 at the head of all of the source files, except those whose only change is
5618 to replace PERL_COMMAND in the very first #! line.
5620 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5621 messages on the queue, with a bottom-up merge sort, using code contributed
5622 by Michael Haardt. This should make operations like -bp somewhat faster on
5623 large queues. It won't affect queue runners, except when queue_run_in_order
5626 51. Installed eximstats 1.31 in the distribution.
5628 52. Added support for SRV lookups to the dnslookup router.
5630 53. If an ACL referred to $message_body or $message_body_end, the value was not
5631 reset for any messages that followed in the same SMTP session.
5633 54. The store-handling optimization for building very long strings was not
5634 differentiating between the different store pools. I don't think this
5635 actually made any difference in practice, but I've tidied it.
5637 55. While running the routers to verify a sender address, $sender_address
5638 was still set to the sender address. This is wrong, because when routing to
5639 send a bounce to the sender, it would be empty. Therefore, I have changed
5640 it so that, while verifying a sender address, $sender_address is set to <>.
5641 (There is no change to what happens when verifying a recipient address.)
5643 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5644 target A or AAAA records (if not already returned) without resetting the
5645 qualify_single or search_parents options of the DNS resolver. These are
5646 inappropriate in this case because the targets of MX and SRV records must
5647 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5648 target that, when qualified, matched something in the local domain. These
5649 two options are now turned off when doing these lookups.
5651 57. It seems that at least some releases of Reiserfs (which does not have the
5652 concept of a fixed number of inodes) returns zero and not -1 for the
5653 number of available inodes. This interacted badly with check_spool_inodes,
5654 which assumed that -1 was the "no such thing" setting. What I have done is
5655 to check that the total number of inodes is greater than zero before doing
5656 the test of how many are available.
5658 58. When a "warn" ACL statement has a log_message modifier, the message is
5659 remembered, and not repeated. This is to avoid a lot of repetition when a
5660 message has many recipients that cause the same warning to be written.
5661 However, Exim was preserving the list of already written lines for an
5662 entire SMTP session, which doesn't seem right. The memory is now reset if a
5663 new message is started.
5665 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5666 debugging output unless log_rewrite was also set.
5668 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5669 of (int)(handle) when we know that handle contains (void *)(-1).
5671 61. The Exim daemon panic-logs an error return when it closes the incoming
5672 connection. However "connection reset by peer" seems to be common, and
5673 isn't really an error worthy of noting specially, so that particular error
5676 62. When Exim is trying to find all the local interfaces, it used to panic and
5677 die if the ioctl to get the interface flags failed. However, it seems that
5678 on at least one OS (Solaris 9) it is possible to have an interface that is
5679 included in the list of interfaces, but for which you get a failure error
5680 for this call. This happens when the interface is not "plumbed" into a
5681 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5682 failure of the "get flags" call assumes that the interface is down.
5684 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5685 makes life easier for people who are doing arithmetic on fields extracted
5686 from dates, where you often get leading zeros that should not be
5687 interpreted as octal.
5689 64. Added qualify_domain to the redirect router, to override the global
5692 65. If a pathologically long header line contained very many addresses (the
5693 report of this problem mentioned 10 000) and each of them was rewritten,
5694 Exim could use up a very large amount of memory. (It kept on making new
5695 copies of the header line as it rewrote, and never released the old ones.)
5696 At the expense of a bit more processing, the header rewriting function has
5697 been changed so that it no longer eats memory in this way.
5699 66. The generation of the Received: header has been moved from the time that a
5700 message starts to be received, to the time that it finishes. The timestamp
5701 in the Received: header should now be very close to that of the <= log
5702 line. There are two side-effects of this change:
5704 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5705 logged header lines no longer include the local Received: line, because
5706 it has not yet been created. The same applies to a copy of the message
5707 that is returned to a non-SMTP sender when a message is rejected.
5709 (b) When a filter file is tested using -bf, no additional Received: header
5710 is added to the test message. After some thought, I decided that this
5713 This change does not affect the value of $received_for. It is still set
5714 after address rewriting, but before local_scan() is called.
5716 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5718 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5719 gave an unhelpful panic error message, and a defer error. I have managed to
5720 change this behaviour so that it now rejects any supplied certificate,
5721 which seems right, as the list of acceptable certificates is empty.
5723 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5724 gave an unhelpful defer error. I have not managed to make this reject any
5725 supplied certificates, but the error message it gives is "no certificate
5726 supplied", which is not helpful.
5728 70. exigrep's output now also includes lines that are not associated with any
5729 message, but which match the given pattern. Implemented by a patch from
5730 Martin Sluka, which also tidied up the Perl a bit.
5732 71. Recipient callout verification, like sender verification, was using <> in
5733 the MAIL FROM command. This isn't really the right thing, since the actual
5734 sender may affect whether the remote host accepts the recipient or not. I
5735 have changed it to use the actual sender in the callout; this means that
5736 the cache record is now keyed on a recipient/sender pair, not just the
5737 recipient address. There doesn't seem to be a real danger of callout loops,
5738 since a callout by the remote host to check the sender would use <>.
5739 [SEE ABOVE: changed after hitting problems.]
5741 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5742 temporary errors. However, in the case of such a code being given after
5743 the end of a data transmission (i.e. after ".") Exim was failing to write
5744 a retry record for the message. (Yes, there was some broken host that was
5745 actually sending 8xx at this point.)
5747 73. An unknown lookup type in a host list could cause Exim to panic-die when
5748 the list was checked. (An example that provoked this was putting <; in the
5749 middle of a list instead of at the start.) If this happened during a DATA
5750 ACL check, a -D file could be left lying around. This kind of configuration
5751 error no longer causes Exim to die; instead it causes a defer error. The
5752 incident is still logged to the main and panic logs.
5754 74. Buglet left over from Exim 3 conversion. The message "too many messages
5755 in one connection" was written to the rejectlog but not the mainlog, except
5756 when address rewriting (yes!) was being logged.
5758 75. Added write_rejectlog option.
5760 76. When a system filter was run not as root (that is, when system_filter_user
5761 was set), the values of the $n variables were not being returned to the
5762 main process; thus, they were not subsequently available in the $sn
5765 77. Added +return_path_on_delivery log selector.
5767 78. A connection timeout was being treated differently from recipients deferred
5768 when testing hosts_max_try with a message that was older than the host's
5769 retry timeout. (The host should not be counted, thus allowing all hosts to
5770 be tried at least once before bouncing.) This may have been the cause of an
5771 occasionally reported bug whereby a message would remain on the queue
5772 longer than the retry timeout, but would be bounced if a delivery was
5773 forced. I say "may" because I never totally pinned down the problem;
5774 setting up timeout/retry tests is difficult. See also the next item.
5776 79. The ultimate address timeout was not being applied to errors that involved
5777 a combination of host plus message (for example, a timeout on a MAIL
5778 command). When an address resolved to a number of possible hosts, and they
5779 were not all tried for each delivery (e.g. because of hosts_max_try), a
5780 message could remain on the queue longer than the retry timeout.
5782 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5785 81. Fixed an obscure SMTP outgoing bug which required at least the following
5786 conditions: (a) there was another message waiting for the same server;
5787 (b) the server returned 5xx to all RCPT commands in the first message so
5788 that the message was not completed; (c) the server dropped the connection
5789 or gave a negative response to the RSET that Exim sends to abort the
5790 transaction. The observed case was a dropped connection after DATA that had
5791 been sent in pipelining mode. That is, the server had advertised PIPELINING
5792 but was not implementing it correctly. The effect of the bug was incorrect
5793 behaviour, such as trying another host, and this could lead to a crash.
5799 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5800 and daemon.c were passed as pointers to ints; they should have been
5801 pointers to socklen_t variables (which are typically unsigned ints).
5803 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5806 3. Fixed a really odd bug that affected only the testing scheme; patching a
5807 certain fixed string in the binary changed the value of another string that
5808 happened to be identical to the end of the original first string.
5810 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5811 name", it returns that address as the IP address. On some operating
5812 systems (e.g. Solaris), it also passes back the IP address string as the
5813 "host name". However, on others (e.g. Linux), it passes back an empty
5814 string. Exim wasn't checking for this, and was changing the host name to an
5815 empty string, assuming it had been canonicalized.
5817 5. Although rare, it is permitted to have more than one PTR record for a given
5818 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5819 all the names associated with an address, because they do in Solaris.
5820 However, it seems that they do not in Linux for data that comes from the
5821 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5822 I found this out when I moved to a new Linux workstation and tried to run
5823 the Exim test suite.
5825 To get round this problem I have changed the code so that it now does its
5826 own call to the DNS to look up PTR records when searching for a host name.
5827 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5828 addresses that are only in /etc/hosts are still found.
5830 This behaviour is, however, controlled by an option called host_lookup_
5831 order, which defaults to "bydns:byaddr". If people want to use the other
5832 order, or indeed, just use one or the other means of lookup, they can
5833 specify it in this variable.
5835 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5836 some operating systems, this comes back from gethostbyaddr() as an empty
5837 string, and this is what Exim used to test for. However, it seems that in
5838 other systems, "." is yielded. Exim now tests for this case too.
5840 7. The values of check_spool_space and check_log_space are now held internally
5841 as a number of kilobytes instead of an absolute number of bytes. If a
5842 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5843 kilobyte. This means that much larger values can be stored.
5845 8. Exim monitor: an attempt to get the action menu when not actually pointing
5846 at a message produces an empty menu entitled "No message selected". This
5847 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5848 no entries in it ("Shell widget menu has zero width and/or height"). So I
5849 have added a single, blank menu entry in this case.
5851 9. Added ${quote_local_part.
5853 10. MIME decoding is now applied to the contents of Subject: header lines when
5856 11. Now that a reference to $sender_host_address automatically causes a reverse
5857 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5858 host lookup before query-style lookups in lists that might use this
5859 variable. This has therefore been abolished, and the "net-" prefix is no
5860 longer necessary for query-style lookups.
5862 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5863 have been a typo for LFLAGS, so it has been changed.
5865 13. The install script calls Exim with "-C /dev/null" in order to find the
5866 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5867 to be output. However, since Exim outputs its version number before the
5868 error, it didn't break the script. It just looked ugly. I fixed this by
5869 always allowing "-C /dev/null" if the caller is root.
5871 14. Ignore overlarge ACL variable number when reading spool file - insurance
5872 against a later release with more variables having written the file.
5874 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5875 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5876 this, as well as the form without the "IPv6" on the front.
5878 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5879 OS/Makefile-Darwin file.
5881 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5882 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5884 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5885 variable. However, some OS use ints for inodes. I've added cast to long int
5886 to get rid of the compiler warning.
5888 19. I had forgotten to lock out "/../" in configuration file names when
5889 ALT_CONFIG_PREFIX was set.
5891 20. Routers used for verification do not need to specify transports. However,
5892 if such a router generated a host list, and callout was configured, Exim
5893 crashed, because it could not find a port number from the (non-existent)
5894 transport. It now assumes port 25 in this circumstance.
5896 21. Added the -t option to exigrep.
5898 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5899 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5900 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5901 them is set, LOOKUP_LSEARCH is forced.
5903 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5905 24. Added sender and host information to the "rejected by local_scan()" log
5906 line; previously there was no indication of these.
5908 25. Added .include_if_exists.
5910 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5911 Linux. It turns out that not all file systems support this. Apparently some
5912 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5913 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5914 not supported on the file descriptor, is now ignored when Exim is trying to
5915 sync a directory. This applies only to Linux.
5917 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5919 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5920 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5923 29. Added support for maildirsize files from supplied patch (modified a bit).
5925 30. The use of :fail: followed by an empty string could lead Exim to respond to
5926 sender verification failures with (e.g.):
5928 550 Verification failed for <xxx>
5929 550 Sender verify failed
5931 where the first response line was missing the '-' that indicates it is not
5932 the final line of the response.
5934 31. The loop for finding the name of the user that called Exim had a hardwired
5935 limit of 10; it now uses the value of finduser_retries, which is used for
5936 all other user lookups.
5938 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5941 33. Exim was neglecting to zero errno before one call of strtol() when
5942 expanding a string and expecting an integer value. On some systems this
5943 resulted in spurious "integer overflow" errors. Also, it was casting the
5944 result into an int without checking.
5946 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5947 did not work. The code looks as if it has *never* worked, though it appears
5948 to have been documented since at least release 1.62. I have made it work.
5950 35. The "timeout_DNS" error in retry rules, also documented since at least
5951 1.62, also never worked. As it isn't clear exactly what this means, and
5952 clearly it isn't a major issue, I have abolished the feature by treating it
5953 as "timeout", and writing a warning to the main and panic logs.
5955 36. The display of retry rules for -brt wasn't always showing the error code
5958 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5959 timeout_connect_A, timeout_connect_MX.
5961 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5962 to the empty sender.
5964 39. The daemon was not analysing the content of -oX till after it had closed
5965 stderr and disconnected from the controlling terminal. This meant that any
5966 syntax errors were only noted on the panic log, and the return code from
5967 the command was 0. By re-arranging the code a little, I've made the
5968 decoding happen first, so such errors now appear on stderr, and the return
5969 code is 1. However, the actual setting up of the sockets still happens in
5970 the disconnected process, so errors there are still only recorded on the
5973 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5974 connections (as happens on some IP stacks) was logged at start up time as
5975 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5976 it from "IPv6 and IPv4", which means that two separate sockets are being
5979 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5980 says whether AF_INET or AF_INET6 was passed as an argument.
5982 42. Exiwhat output was messed up when time zones were included in log
5985 43. Exiwhat now gives more information about the daemon's listening ports,
5986 and whether -tls-on-connect was used.
5988 44. The "port" option of the smtp transport is now expanded.
5990 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5991 silently ignored. Now an error message is written to the main and panic
5994 46. There's a new ACL modifier called "logwrite" which writes to a log file
5995 as soon as it is encountered.
5997 47. Added $local_user_uid and $local_user_gid at routing time.
5999 48. Exim crashed when trying to verify a sender address that was being
6002 49. Exim was recognizing only a space character after ".include". It now also
6003 recognizes a tab character.
6005 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6006 extracting the relevant information from the specification. The man page no
6007 longer contains scrambled data for the -d option, and I've added a section
6008 at the front about calling Exim under different names.
6010 51. Added "extra_headers" argument to the "mail" command in filter files.
6012 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6015 53. Installed eximstats 1.29.
6017 54. Added transport_filter_timeout as a generic transport option.
6019 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6020 Cc: header lines. This was required by RFC 822, but it not required by RFC
6023 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6024 incoming messages that did not have them. Now it does so only if the
6025 message originates locally, that is, if there is no associated remote host
6026 address. When Resent- header lines are present, this applies to the Resent-
6027 lines rather than the non-Resent- lines.
6029 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6030 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6032 58. Messages for configuration errors now include the name of the main
6033 configuration file - useful now that there may be more than one file in a
6034 list (.included file names were always shown).
6036 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6037 for those rare installations that do not start the daemon as root or run it
6038 setuid root. I've cut out the call to initgroups() if the daemon is not
6041 60. The Exim user and group can now be bound into the binary as text strings
6042 that are looked up at the start of Exim's processing.
6044 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6046 62. Added $mailstore_basename variable.
6048 63. Installed patch to sieve.c from Michael Haardt.
6050 64. When Exim failed to open the panic log after failing to open the main log,
6051 the original message it was trying to log was written to stderr and debug
6052 output, but if they were not available (the usual case in production), it
6053 was lost. Now it is written to syslog before the two lines that record the
6054 failures to open the logs.
6056 65. Users' Exim filters run in subprocesses under the user's uid. It is
6057 possible for a "deliver" command or an alias in a "personal" command to
6058 provoke an address rewrite. If logging of address rewriting is configured,
6059 this fails because the process is not running as root or exim. There may be
6060 a better way of dealing with this, but for the moment (because 4.30 needs
6061 to be released), I have disabled address rewrite logging when running a
6062 filter in a non-root, non-exim process.
6068 1. The buildconfig auxiliary program wasn't quoting the value set for
6069 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6070 not defined. This bug was masked in 4.22 by the effect that was fixed in
6073 2. Some messages that were rejected after a message id was allocated were
6074 shown as "incomplete" by exigrep. It no longer does this for messages that
6075 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6077 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6078 have allow_domain_literals set, the ID did not get logged in the <= line.
6079 Domain literals are now always recognized in Message-ID: header lines.
6081 4. The first argument for a ${extract expansion item is the key name or field
6082 number. Leading and trailing spaces in this item were not being ignored,
6083 causing some misleading effects.
6085 5. When deliver_drop_privilege was set, single queue runner processes started
6086 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6087 same command in the process it spins off) were not dropping privilege.
6089 6. When the daemon running as "exim" started a queue runner, it always
6090 re-executed Exim in the spun-off process. This is a waste of effort when
6091 deliver_drop_privilege is set. The new process now just calls the
6092 queue-runner function directly.
6098 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6101 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6102 ignored. Though the use of -oP was forcing the writing of a pid file, it
6103 was always written to the default place.
6105 3. If the message "no IP address found for host xxxx" is generated during
6106 incoming verification, it is now followed by identification of the incoming
6107 connection (so you can more easily find what provoked it).
6109 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6111 5. Added some features to "harden" Exim a bit more against certain attacks:
6113 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6114 be put in Local/Makefile. This is like the never_users runtime option,
6115 but it cannot be overridden. The default setting is "root".
6117 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6118 prefix string with which any file named in a -C command line option
6121 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6122 is retained for -C and -D only if the caller of Exim is root. Without
6123 it, the exim user may also use -C and -D and retain privilege.
6125 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6126 command line option is disabled.
6128 6. Macro names set by the -D option must start with an upper case letter, just
6129 like macro names defined in the configuration file.
6131 7. Added "dereference=" facility to LDAP.
6133 8. Two instances of the typo "uknown" in the source files are fixed.
6135 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6136 the Configure-Makefile script screwed up while processing it.
6138 10. Incorporated PCRE 4.4.
6140 11. The SMTP synchronization check was not operating right at the start of an
6141 SMTP session. For example, it could not catch a HELO sent before the client
6142 waited for the greeting. There is now a check for outstanding input at the
6143 point when the greeting is written. Because of the duplex, asynchronous
6144 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6145 way, but not yet received, when the check is performed.
6147 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6148 on TCP/IP sockets, because this apparently causes some broken clients to
6151 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6152 unchanged) from the Cygwin maintainer.
6154 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6155 was supported instead of testing for mbx. Effectively a typo.
6157 15. The spa authenticator server code was not checking that the input it
6158 received was valid base64.
6160 16. The debug output line for the "set" modifier in ACLs was not showing the
6161 name of the variable that was being set.
6163 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6165 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6166 lookup on its own. Something else was needed to trigger the lookup. For
6167 example, a match in host_lookup or the need for a host name in a host list.
6168 Now, if $sender_host_name is referenced and the host name has not yet been
6169 looked up, a lookup is performed. If the lookup fails, the variable remains
6170 empty, and $host_lookup_failed is set to "1".
6172 19. Added "eqi" as a case-independent comparison operator.
6174 20. The saslauthd authentication condition could segfault if neither service
6175 nor realm was specified.
6177 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6178 error message that was logged was misleading, and incoming SMTP
6179 connections were dropped. The message is now more accurate, and temporary
6180 errors are given to SMTP connections.
6182 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6183 (except RSET) with a 421 error, until QUIT is received. However, it was
6184 failing to send a response to QUIT.
6186 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6187 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6188 ACL is now run after the helo_try_verify_hosts code.
6190 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6191 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6192 case-independent, so other case variants are also recognized). Apparently
6193 some systems use these upper case variants.
6195 25. If more than two messages were waiting for the same host, and a transport
6196 filter was specified for the transport, Exim sent two messages over the
6197 same TCP/IP connection, and then failed with "socket operation on non-
6198 socket" when it tried to send the third.
6200 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6202 27. The extern definition of crypt16() in expand.c was not being excluded when
6203 the OS had its own crypt16() function.
6205 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6206 as a preferred synonym for return_size_limit, both as an option and as an
6209 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6211 30. Changed the default configuration ACL to relax the local part checking rule
6212 for addresses that are not in any local domains. For these addresses,
6213 slashes and pipe symbols are allowed within local parts, but the sequence
6214 /../ is explicitly forbidden.
6216 31. SPA server authentication was not clearing the challenge buffer before
6219 32. log_message in a "warn" ACL statement was writing to the reject log as
6220 well as to the main log, which contradicts the documentation and doesn't
6221 seem right (because no rejection is happening). So I have stopped it.
6223 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6224 However, I am unable to do any testing of this.
6226 34. Fixed an infelicity in the appendfile transport. When checking directories
6227 for a mailbox, to see if any needed to be created, it was accidentally
6228 using path names with one or more superfluous leading slashes; tracing
6229 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6231 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6232 contents are added to the log line that is written for every discarded
6233 recipient. (Previously a log_message setting was ignored.)
6235 36. The ${quote: operator now quotes the string if it is empty.
6237 37. The install script runs exim in order to find its version number. If for
6238 some reason other than non-existence or emptiness, which it checks, it
6239 could not run './exim', it was installing it with an empty version number,
6240 i.e. as "exim-". This error state is now caught, and the installation is
6243 38. An argument was missing from the function that creates an error message
6244 when Exim fails to connect to the socket for saslauthd authentication.
6245 This could cause Exim to crash, or give a corrupted message.
6247 39. Added isip, isip4, and isip6 to ${if conditions.
6249 40. The ACL variables $acl_xx are now saved with the message, and can be
6250 accessed later in routers, transports, and filters.
6252 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6253 strings in the file are not string-expanded.
6255 42. If a MAIL command specified a SIZE value that was too large to fit into an
6256 int variable, the check against message_size_limit failed. Such values are
6257 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6258 day this will have to be increased, but I don't think I want to be around
6259 when emails are that large.
6266 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6267 iconv() is not standard in FreeBSD.
6269 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6270 IPv6 enabled. The observed symptom was a segmentation fault on return from
6271 the function os_common_find_running_interfaces() in src/os.c.
6273 3. In the check_special_case() function in daemon.c I had used "errno" as an
6274 argument name, which causes warnings on some systems. This was basically a
6275 typo, since it was named "eno" in the comments!
6277 4. The code that waits for the clock to tick (at a resolution of some fraction
6278 of a second) so as to ensure message-id uniqueness was always waiting for
6279 at least one whole tick, when it could have waited for less. [This is
6280 almost certainly not relevant at current processor speeds, where it is
6281 unlikely to ever wait at all. But we try to future-proof.]
6283 5. The function that sleeps for a time interval that includes fractions of a
6284 second contained a race. It did not block SIGALRM between setting the
6285 timer, and suspending (a couple of lines later). If the interval was short
6286 and the sigsuspend() was delayed until after it had expired, the suspension
6287 never ended. On busy systems this could lead to processes getting stuck for
6290 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6291 process, before it forks any delivery processes. The open lookup caching
6292 mechanism meant that the open file or database connection was passed into
6293 the delivery process. The problem was that delivery processes always tidy
6294 up cached lookup data. This could cause a problem for the next delivery
6295 process started by the queue runner, because the external queue runner
6296 process does not know about the closure. So the next delivery process
6297 still has data in the lookup cache. In the case of a file lookup, there was
6298 no problem because closing a file descriptor in a subprocess doesn't affect
6299 the parent. However, if the lookup was caching a connection to a database,
6300 the connection was closed, and the second delivery process was likely to
6301 see errors such as "PGSQL: query failed: server closed the connection
6302 unexpectedly". The problem has been fixed by closing all cached lookups
6303 in a queue runner before running a delivery process.
6305 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6306 seem to have the "const" qualifier which it has on other OS. I've
6309 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6310 *of the same type* (client or server) with the same public name that an
6311 error should be diagnosed.
6313 9. When Exim looked up a host name for an IP address, but failed to find the
6314 original IP address when looking up the host name (a safety check), it
6315 output the message "<ip address> does not match any IP for NULL", which was
6316 confusing, to say the least. The bug was that the host name should have
6317 appeared instead of "NULL".
6319 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6320 user that is built into the binary, and the -C or -D options is used, root
6321 privilege is dropped before the configuration file is read. In addition,
6322 logging is switched to stderr instead of the normal log files. If the
6323 configuration then re-defines the Exim user, the unprivileged environment
6324 is probably not what is expected, so Exim logs a panic warning message (but
6327 However, if deliver_drop_privilege is set, the unprivileged state may well
6328 be exactly what is intended, so the warning has been cut out in that case,
6329 and Exim is allowed to try to write to its normal log files.
6335 1. smtp_return_error_details was not giving details for temporary sender
6336 or receiver verification errors.
6338 2. Diagnose a configuration error if two authenticators have the same public
6341 3. Exim used not to create the message log file for a message until the first
6342 delivery attempt. This could be confusing when incoming messages were held
6343 for policy or load reasons. The message log file is now created at the time
6344 the message is received, and an initial "Received" line is written to it.
6346 4. The automatically generated man page for command line options had a minor
6347 bug that caused no ill effects; however, a more serious problem was that
6348 the procedure for building the man page automatically didn't always
6349 operate. Consequently, release 4.20 contains an out-of-date version. This
6350 shouldn't happen again.
6352 5. When building Exim with embedded Perl support, the script that builds the
6353 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6354 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6356 6. The freeze_tell option was not being used for messages that were frozen on
6357 arrival, either by an ACL or by local_scan().
6359 7. Added the smtp_incomplete_transaction log selector.
6361 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6362 was accepting AUTH without a new EHLO.
6364 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6365 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6368 10. Logging of TCP/IP connections (when configured) now happens in the main
6369 daemon process instead of the child process, so that the TCP/IP connection
6370 count is more accurate (but it can never be perfect).
6372 11. The use of "drop" in a nested ACL was not being handled correctly in the
6373 outer ACL. Now, if condition failure induced by the nested "drop" causes
6374 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6375 or "require"), the connection is dropped.
6377 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6378 that yield "discard" can now be used with an "accept" or a "discard" verb,
6379 but an error is generated for any others (because I can't see a useful way
6380 to define what should happen).
6382 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6383 are now processed in the same way as lines in the Exim configuration file.
6384 In particular, continuation lines are supported.
6386 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6388 15. Added -ti meaning -t -i.
6390 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6391 domains, and warn by logging if others are found.
6393 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6394 returned from GIFCONF when Exim is trying to find the list of interfaces on
6395 a host. The code in os.c has been modified to copy each ifreq to an aligned
6396 structure in all cases.
6398 Also, in some cases, the returned ifreq's were being copied to a 'struct
6399 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6400 means the last couple of bytes of an IPv6 address could be chopped if the
6401 ifreq contained only a normal sockaddr (14 bytes storage).
6403 18. Named domain lists were not supported in the hosts_treat_as_local option.
6404 An entry such as +xxxx was not recognized, and was treated as a literal
6407 19. Ensure that header lines added by a DATA ACL are included in the reject log
6408 if the ACL subsequently rejects the message.
6410 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6411 MD5 (which is deprecated).
6413 21. When testing a filter file using -bf, Exim was writing a message when it
6414 took the sender from a "From " line in the message, but it was not doing so
6415 when it took $return_path from a Return-Path: header line. It now does.
6417 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6418 with a valid header line field name (a series of printing characters
6419 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6421 23. Changed "disc" in the source to "disk" to conform to the documentation and
6422 the book and for uniformity.
6424 24. Ignore Sendmail's -Ooption=value command line item.
6426 25. When execve() failed while trying to run a command in a pipe transport,
6427 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6428 could be confused with a return value of 69 from the command itself. This
6429 has been changed to 127, the value the shell returns if it is asked to run
6430 a non-existent command. The wording for the related log line suggests a
6431 non-existent command as the problem.
6433 26. If received_header_text expands to an empty string, do not add a Received:
6434 header line to the message. (Well, it adds a token one on the spool, but
6435 marks it "old" so that it doesn't get used or transmitted.)
6437 27. Installed eximstats 1.28 (addition of -nt option).
6439 28. There was no check for failure on the call to getsockname() in the daemon
6440 code. This can fail if there is a shortage of resources on the system, with
6441 ENOMEM, for example. A temporary error is now given on failure.
6443 29. Contrary to the C standard, it seems that in some environments, the
6444 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6445 program. Exim now does this explicitly; it affects the formatting of
6446 timestamps using strftime().
6448 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6449 complaints. I've now initialized all the variables, to avoid this.
6451 32. Header lines added by a system filter were not being "seen" during
6452 transport-time rewrites.
6454 33. The info_callback() function passed to OpenSSL is set up with type void
6455 (*)(SSL *, int, int), as described somewhere. However, when calling the
6456 function (actually a macro) that sets it up, the type void(*)() is
6457 expected. I've put in a cast to prevent warnings from picky compilers.
6459 34. If a DNS black list lookup found a CNAME record, but there were no A
6460 records associated with the domain it pointed at, Exim crashed.
6462 35. If a DNS black list lookup returned more than one A record, Exim ignored
6463 all but the first. It now scans all returned addresses if a particular IP
6464 value is being sought. In this situation, the contents of the
6465 $dnslist_value variable are a list of all the addresses, separated by a
6468 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6469 to accept a host name and all its aliases if the forward lookup for any of
6470 them yielded the IP address of the incoming connection. Now it accepts only
6471 those names whose forward lookup yields the correct IP address. Any other
6472 names are discarded. This closes a loophole whereby a rogue DNS
6473 administrator could create reverse DNS records to break through a
6474 wildcarded host restriction in an ACL.
6476 37. If a user filter or a system filter that ran in a subprocess used any of
6477 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6478 the wrong values were passed to the pipe command ($thisaddress had the
6479 value of $0, $0 had the value of $1, etc). This bug was introduced by
6480 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6482 38. Improved the line breaking for long SMTP error messages from ACLs.
6483 Previously, if there was no break point between 40 and 75 characters, Exim
6484 left the rest of the message alone. Two changes have been made: (a) I've
6485 reduced the minimum length to 35 characters; (b) if it can't find a break
6486 point between 35 and 75 characters, it looks ahead and uses the first one
6487 that it finds. This may give the occasional overlong line, but at least the
6488 remaining text gets split now.
6490 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6491 file descriptors might be low, and that setting 1000 would always raise it.
6492 It turns out that in some environments, the limit is already over 1000 and
6493 that lowering it causes trouble. So now Exim takes care not to decrease it.
6495 40. When delivering a message, the value of $return_path is set to $sender_
6496 address at the start of routing (routers may change the value). By an
6497 oversight, this default was not being set up when an address was tested by
6498 -bt or -bv, which affected the outcome if any router or filter referred to
6501 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6502 log only when "message" or "log_message" are set. However, if one of the
6503 conditions was an address verification, or a call to a nested ACL, the
6504 messages generated by the underlying test were being passed through. This
6505 no longer happens. The underlying message is available in $acl_verify_
6506 message for both "message" and "log_message" expansions, so it can be
6507 passed through if needed.
6509 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6510 new expansion $bh_ to give the encoded byte string without charset
6511 translation. Translation happens only if iconv() is available; HAVE_ICONV
6512 indicates this at build time. HEADERS_CHARSET gives the charset to
6513 translate to; headers_charset can change it in the configuration, and
6514 "headers charset" can change it in an individual filter file.
6516 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6517 that creates RFC 2047 encoded "words" labels them as that charset instead
6518 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6519 expansion operator; (ii) when Exim creates a From: line for a local
6520 message; (iii) when a header line is rewritten to include a "phrase" part.
6522 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6523 buggy, causing it to skip the first lines of messages whose message ID
6524 ended in 'D'. This would not have bitten before Exim release 4.14, because
6525 message IDs were unlikely to end in 'D' before then. The effect was to have
6526 incorrect size information for certain domains.
6528 45. #include "config.h" was missing at the start of the crypt16.c module. This
6529 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6532 46. If there was a timeout during a "random" callout check, Exim treated it as
6533 a failure of the random address, and carried on sending RSET and the real
6534 address. If the delay was just some slowness somewhere, the response to the
6535 original RCPT would be taken as a response to RSET and so on, causing
6536 mayhem of various kinds.
6538 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6539 when I implemented it. It didn't allow for the fact that some option values
6540 may legitimately be negative (e.g. size_addition), and it didn't even do
6541 the right test for positive values.
6543 48. Domain names in DNS records are case-independent. Exim always looks them up
6544 in lower case. Some resolvers return domain names in exactly the case they
6545 appear in the zone file, that is, they may contain uppercase letters. Not
6546 all resolvers do this - some return always lower case. Exim was treating a
6547 change of case by a resolver as a change of domain, similar to a widening
6548 of a domain abbreviation. This triggered its re-routing code and so it was
6549 trying to route what was effectively the same domain again. This normally
6550 caused routing to fail (because the router wouldn't handle the domain
6551 twice). Now Exim checks for this case specially, and just changes the
6552 casing of the domain that it ultimately uses when it transmits the message
6555 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6558 50. If a filter generated a file delivery with a non-absolute name (possible if
6559 no home directory exists for the router), the forbid_file option was not
6562 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6563 the existing equality matching.
6565 52. Exim was using ints instead of ino_t variables in some places where it was
6566 dealing with inode numbers.
6568 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6569 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6570 called TMPDIR, and if it finds it is different, it changes its value.
6572 54. The smtp_printf() function is now made available to local_scan() so
6573 additional output lines can be written before returning. There is also an
6574 smtp_fflush() function to enable the detection of a dropped connection.
6575 The variables smtp_input and smtp_batched_input are exported to
6578 55. Changed the default runtime configuration: the message "Unknown user"
6579 has been removed from the ACL, and instead placed on the localuser router,
6580 using the cannot_route_message feature. This means that any verification
6581 failures that generate their own messages won't get overridden. Similarly,
6582 the "Unrouteable address" message that was in the ACL for unverifiable
6583 relay addresses has also been removed.
6585 56. Added hosts_avoid_esmtp to the smtp transport.
6587 57. The exicyclog script was not checking for the esoteric option
6588 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6589 will work only if exicyclog is run under the appropriate euid.
6591 58. Following a discussion on the list, the rules by which Exim recognises line
6592 endings on incoming messages have been changed. The -dropcr and drop_cr
6593 options are now no-ops, retained only for backwards compatibility. The
6594 following line terminators are recognized: LF CRLF CR. However, special
6595 processing applies to CR:
6597 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6598 nor a local message in the state where . is a terminator.
6600 (ii) If a bare CR is encountered in a header line, an extra space is added
6601 after the line terminator so as not to end the header. The reasoning
6602 behind this is that bare CRs in header lines are most likely either
6603 to be mistakes, or people trying to play silly games.
6605 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6606 was being incorrectly given as 18 bytes larger than it should have been.
6607 This is a VOB (very old bug).
6609 60. This may never have affected anything current, but just in case it has:
6610 When the local host is found other than at the start of a list of hosts,
6611 the local host, those with the same MX, and any that follow, are discarded.
6612 When the list in question was part of a longer list of hosts, the following
6613 hosts (not currently being processed) were also being discarded. This no
6614 longer happens. I'm not sure if this situation could ever has previously
6617 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6620 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6621 header. This is something that is recommended in a new Internet Draft, and
6622 is something that is documented as being done by Sendmail. There are two
6623 possible values. For messages generated by the autoreply transport, Exim
6626 Auto-Submitted: auto-replied
6628 whereas for all other generated messages (e.g. bounces) it adds
6630 Auto-Submitted: auto-generated
6632 63. The "personal" condition in filters now includes a test for the
6633 Auto-Submitted: header. If it contains the string "auto-" the message it
6634 not considered personal.
6636 64. Added rcpt_include_affixes as a generic transport option.
6638 65. Added queue_only_override (default true).
6640 66. Added the syslog_duplication option.
6642 67. If what should have been the first header line of a message consisted of
6643 a space followed by a colon, Exim was mis-interpreting it as a header line.
6644 It isn't of course - it is syntactically invalid and should therefore be
6645 treated as the start of the message body. The misbehaviour could have
6646 caused a number of strange effects, including loss of data in subsequent
6647 header lines, and spool format errors.
6649 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6650 client host had authenticated. This control can now be exercised by an ACL
6651 for more flexibility.
6653 69. By default, callouts do not happen when testing with -bh. There is now a
6654 variant, -bhc, which does actually run the callout code, including
6655 consulting and updating the callout cache.
6657 70. Added support for saslauthd authentication, courtesy of Alexander
6660 71. If statvfs() failed on the spool or log directories while checking their
6661 size for availability, Exim confusingly gave the error "space shortage".
6662 Furthermore, in debugging mode it crashed with a floating point exception.
6663 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6664 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6665 serious problem, Exim now writes to the main and panic logs when this
6666 happens, with details of the failure. It then refuses to accept the
6667 incoming message, giving the message "spool directory problem" or "log
6668 directory problem" with a 421 code for SMTP messages.
6670 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6671 0, 1, and 2 exist, because some OS complain for execs without them (see
6672 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6673 descriptors. However, the code omitted to check that the open succeeded,
6674 causing mysterious errors if for some reason the permissions on /dev/null
6675 got screwed. Now Exim writes a message to the main and panic logs, and
6676 bombs out if it can't open /dev/null.
6678 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6679 interact so that it is all more flexible. It is supposed to remain
6680 backwards compatible. Also added extra_local_interfaces.
6682 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6683 to bomb out with an assertion failure - to the client this appears as a
6684 connection drop. This problem occurs in the part of the code that was taken
6685 from the Samba project. Fortunately, the assertion is in a very simple
6686 function, so I have fixed this by reproducing the function inline in the
6687 one place where it is called, and arranging for authentication to fail
6688 instead of killing the process with assert().
6690 75. The SPA client code was not working when the server requested OEM rather
6691 than Unicode encoding.
6693 76. Added code to make require_files with a specific uid setting more usable in
6694 the case where statting the file as root fails - usually a non-root-mounted
6695 NFS file system. When this happens and the failure is EACCES, Exim now
6696 forks a subprocess and does the per-uid checking as the relevant uid.
6698 77. Added process_log_path.
6700 78. If log_file_path was not explicitly set, a setting of check_log_space or
6701 check_log_inodes was ignored.
6703 79. If a space check for the spool or log partitions fails, the incident is now
6704 logged. Of course, in the latter case the data may get lost...
6706 80. Added the %p formatting code to string_format() so that it can be used to
6707 print addresses in debug_print(). Adjusted all the address printing in the
6708 debugging in store.c to use %p rather than %d.
6710 81. There was a concern that a line of code in smtp_in.c could overflow a
6711 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6712 initially expressed, the concern was not well-founded, because trailing
6713 spaces are removed early. However, if the trailing spaces were followed by
6714 a NULL, they did not get removed, so the overflow was possible. Two fixes
6717 (a) I re-wrote the offending code in a cleaner fashion.
6718 (b) If an incoming SMTP command contains a NULL character, it is rejected
6721 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6722 runs initgroups(), so that if the Exim user is in any additional groups,
6723 they will be used during message reception.
6729 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob