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.
76 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
77 supports it and a size is available (ie. the sending peer gave us one).
79 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
80 years of being deprecated). The replacements are the ACLs
81 acl_smtp_mime and acl_not_smtp_mime.
83 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
84 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
85 or even an in-clear connection were permitted. Now, if the host lookup
86 was dnssec and dane was requested then the host is only used if the
87 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
88 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
89 if one fails this test.
90 This means that a poorly-configured remote DNS will make it incommunicado;
91 but it protects against a DNS-interception attack on it.
93 JH/04 Bug 1810: make continued-use of an open smtp transport connection
94 non-noisy when a race steals the message being considered.
96 JH/05 If main configuration option tls_certificate is unset, generate a
97 self-signed certificate for inbound TLS connections.
99 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
100 in rewrites and routers.
102 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
103 and logged a warning sing 4.83; now they are a configuration file error.
105 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
106 (lacking @domain). Apply the same qualification processing as RCPT.
108 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
110 JH/10 Support ${sha256:} applied to a string (as well as the previous
113 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
114 a cutthrough deliver is pending, as we always want to make a connection.
115 This also avoids re-routing the message when later placing the cutthrough
116 connection after a verify cache hit.
117 Do not update it with the verify result either.
119 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
120 when routing results in more than one destination address.
122 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
123 signing (which inhibits the cutthrough capability). Previously only
124 the presence of an option was tested; now an expansion evaluating as
125 empty is permissible (obviously it should depend only on data available
126 when the cutthrough connection is made).
128 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
129 the relevant preceding SMTP command did not note the pipelining mode.
131 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
132 Previously they were not counted.
134 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
135 as one having no matching records. Previously we deferred the message
136 that needed the lookup.
138 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
139 distinguished as "(=".
141 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
142 for missing MX records. Previously it only worked for missing A records.
144 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
146 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
147 after the data-go-ahead and data-ack. Patch from Jason Betts.
149 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
150 even for a "none" policy. Patch from Tony Meyer.
152 JH/22 Fix continued use of a connection for further deliveries. If a port was
153 specified by a router, it must also match for the delivery to be
156 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
157 When one of the recipients of a message was unsuitable for the connection
158 (has no matching addresses), we lost track of needing to mark it
159 deferred. As a result mail would be lost.
161 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
163 JH/25 Decoding ACL controls is now done using a binary search; the source code
164 takes up less space and should be simpler to maintain. Merge the ACL
165 condition decode tables also, with similar effect.
167 JH/26 Fix problem with one_time used on a redirect router which returned the
168 parent address unchanged. A retry would see the parent address marked as
169 delivered, so not attempt the (identical) child. As a result mail would
172 JH/27 Fix a possible security hole, wherein a process operating with the Exim
173 UID can gain a root shell. Credit to http://www.halfdog.net/ for
174 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
177 JH/28 Enable {spool,log} filesystem space and inode checks as default.
178 Main config options check_{log,spool}_{inodes,space} are now
179 100 inodes, 10MB unless set otherwise in the configuration.
181 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
182 Previously it only applied to the main-section connection policy
185 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
187 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
188 by me. Added RFC7919 DH primes as an alternative.
190 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
191 are not in the system include path.
193 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
194 GnuTLS, when a session startup failed (eg because the client disconnected)
195 Exim did stdio operations after fclose. This was exposed by a recent
196 change which nulled out the file handle after the fclose.
198 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
199 signed directly by the cert-signing cert, rather than an intermediate
200 OCSP-signing cert. This is the model used by LetsEncrypt.
202 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
204 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
205 an incoming connection.
207 HS/02 Bug 1802: Do not half-close the connection after sending a request
210 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
211 fallback to "prime256v1".
213 JH/34 SECURITY: Use proper copy of DATA command in error message.
214 Could leak key material. Remotely exploitable. CVE-2016-9963.
220 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
221 and 3.4.4 - once the server is enabled to respond to an OCSP request
222 it does even when not requested, resulting in a stapling non-aware
223 client dropping the TLS connection.
225 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
226 support variable-length bit vectors. No functional change.
228 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
229 The I= interface field on outgoing lines is now after the H= remote
230 host field, same as incoming lines. There is a separate
231 outgoing_interface log selector which allows you to disable the
234 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
235 If not running log_selector +smtp_connection the mainlog would be held
236 open indefinitely after a "too many connections" event, including to a
237 deleted file after a log rotate. Leave the per net connection logging
238 leaving it open for efficiency as that will be quickly detected by the
239 check on the next write.
241 HS/01 Bug 1671: Fix post transport crash.
242 Processing the wait-<transport> messages could crash the delivery
243 process if the message IDs didn't exist for some reason. When
244 using 'split_spool_directory=yes' the construction of the spool
245 file name failed already, exposing the same netto behaviour.
247 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
248 mime_regex ACL conditions.
250 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
251 to DSN fail messages (bounces): remote IP, remote greeting, remote response
252 to HELO, local diagnostic string.
254 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
255 log line to debug. Even when configured with a tls authenticator many
256 client connections are expected to not authenticate in this way, so
257 an authenticate fail is not an error.
259 HS/02 Add the Exim version string to the process info. This way exiwhat
260 gives some more detail about the running daemon.
262 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
263 matter for fast-change records such as DNSBLs.
265 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
266 retry record, even if constant. There may be multiple transports with
267 different interface settings and the retry behaviour needs to be kept
270 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
272 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
274 JH/10 Bug 840: fix log_defer_output option of pipe transport
276 JH/11 Bug 830: use same host for all RCPTS of a message, even under
277 hosts_randomize. This matters a lot when combined with mua_wrapper.
279 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
280 ${quote_pgsql:<string>} operator.
282 JH/13 Bug 1708: avoid misaligned access in cached lookup.
284 JH/14 Change header file name for freeradius-client. Relevant if compiling
285 with Radius support; from the Gentoo tree and checked under Fedora.
287 JH/15 Bug 1712: Introduce $prdr_requested flag variable
289 JH/16 Bug 1714: Permit an empty string as expansion result for transport
290 option transport_filter, meaning no filtering.
292 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
294 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
295 defaults to "*" (all hosts). The variable is now available when not built
296 with TLS, default unset, mainly to enable keeping the testsuite sane.
297 If a server certificate is not supplied (via tls_certificate) an error is
298 logged, and clients will find TLS connections fail on startup. Presumably
299 they will retry in-clear.
300 Packagers of Exim are strongly encouraged to create a server certificate
301 at installation time.
303 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
304 with the $config_file variable.
306 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
307 in transport context, after the attempt, and per-recipient. The latter type
308 is per host attempted. The event data is the error message, and the errno
309 information encodes the lookup type (A vs. MX) used for the (first) host,
310 and the trailing two digits of the smtp 4xx response.
312 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
313 to write to mainlog (or rejectlog, paniclog) in the window between file
314 creation and permissions/ownership being changed. Particularly affects
315 installations where exicyclog is run as root, rather than exim user;
316 result is that the running daemon panics and dies.
318 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
320 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
321 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
322 "pri" and "weight". Note that the previous implicit priority given by the
323 list order is no longer honoured.
325 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
328 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
329 by defining SUPPORT_SOCKS.
331 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
332 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
333 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
334 variables are renamed to proxy_{local,external}_{address,port}.
336 JH/27 Move Internationalisation support from Experimental to mainline, enabled
337 for a build by defining SUPPORT_I18N
339 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
340 of the query string, and make ${quote_redis:} do that quoting.
342 JH/29 Move Events support from Experimental to mainline, enabled by default
343 and removable for a build by defining DISABLE_EVENT.
345 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
347 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
348 cached by the daemon.
350 JH/32 Move Redis support from Experimental to mainline, enabled for a build
351 by defining LOOKUP_REDIS. The libhiredis library is required.
353 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
354 keys are given for lookup.
356 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
357 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
358 only supported when built with TLS support. The PolarSSL SHA routines
359 are still used when the TLS library is too old for convenient support.
361 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
362 openssl_options), for security. OpenSSL forces this from version 1.1.0
363 server-side so match that on older versions.
365 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
366 allocation for $value could be released as the expansion processing
367 concluded, but leaving the global pointer active for it.
369 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
370 and to use the domains and local_parts ACL conditions.
372 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
373 incorrectly not doubled on cutthrough transmission, hence seen as a
374 body-termination at the receiving system - resulting in truncated mails.
375 Commonly the sender saw a TCP-level error, and retransmitted the message
376 via the normal store-and-forward channel. This could result in duplicates
377 received - but deduplicating mailstores were liable to retain only the
378 initial truncated version.
380 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
382 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
384 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
385 we're in there, support oversigning also; bug 1309.
387 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
389 HS/04 Add support for keep_environment and add_environment options.
391 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
392 either intentional arithmetic overflow during PRNG, or testing config-
395 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
396 delivery resulted in actual delivery. Cancel cutthrough before DATA
399 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
400 rejects a recipient: pass the reject to the originator.
402 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
403 Many were false-positives and ignorable, but it's worth fixing the
406 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
407 for the new environment-manipulation done at startup. Move the routines
408 from being local to tls.c to being global via the os.c file.
410 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
411 an extract embedded as result-arg for a map, the first arg for extract
412 is unavailable so we cannot tell if this is a numbered or keyed
413 extraction. Accept either.
419 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
422 JH/02 The smtp transport option "multi_domain" is now expanded.
424 JH/03 The smtp transport now requests PRDR by default, if the server offers
427 JH/04 Certificate name checking on server certificates, when exim is a client,
428 is now done by default. The transport option tls_verify_cert_hostnames
429 can be used to disable this per-host. The build option
430 EXPERIMENTAL_CERTNAMES is withdrawn.
432 JH/05 The value of the tls_verify_certificates smtp transport and main options
433 default to the word "system" to access the system default CA bundle.
434 For GnuTLS, only version 3.0.20 or later.
436 JH/06 Verification of the server certificate for a TLS connection is now tried
437 (but not required) by default. The verification status is now logged by
438 default, for both outbound TLS and client-certificate supplying inbound
441 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
444 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
445 Status Notification (bounce) messages are now MIME format per RFC 3464.
446 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
447 under the control of the dsn_advertise_hosts option, and routers may
448 have a dsn_lasthop option.
450 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
451 default, modifiable by a malware= option. The list separator for
452 the options can now be changed in the usual way. Bug 68.
454 JH/10 The smtp_receive_timeout main option is now expanded before use.
456 JH/11 The incoming_interface log option now also enables logging of the
457 local interface on delivery outgoing connections.
459 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
460 if the interface and destination host and port all match.
462 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
465 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
466 Patch from Andrew Lewis.
468 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
469 now supports optional time-restrictions, weighting, and priority
470 modifiers per server. Patch originally by <rommer@active.by>.
472 JH/16 The spamd_address main option now supports a mixed list of local
473 and remote servers. Remote servers can be IPv6 addresses, and
474 specify a port-range.
476 JH/17 Bug 68: The spamd_address main option now supports an optional
477 timeout value per server.
479 JH/18 Bug 1581: Router and transport options headers_add/remove can
480 now have the list separator specified.
482 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
485 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
488 JH/21 Support for the A6 type of dns record is withdrawn.
490 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
491 rather than the verbs used.
493 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
494 from 255 to 1024 chars.
496 JH/24 Verification callouts now attempt to use TLS by default.
498 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
499 are generic router options now. The defaults didn't change.
501 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
502 Original patch from Alexander Shikoff, worked over by JH.
504 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
505 files if ZCAT_COMMAND is not executable.
507 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
509 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
511 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
512 Normally benign, it bites when the pair was led to by a CNAME;
513 modern usage is to not canonicalize the domain to a CNAME target
514 (and we were inconsistent anyway for A-only vs AAAA+A).
516 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
518 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
519 when evaluating $sender_host_dnssec.
521 JH/31 Check the HELO verification lookup for DNSSEC, adding new
522 $sender_helo_dnssec variable.
524 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
526 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
528 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
530 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
531 documented as working, but never had. Support all but $spam_report.
533 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
534 added for tls authenticator.
536 HS/03 Add perl_taintmode main config option
542 TL/01 When running the test suite, the README says that variables such as
543 no_msglog_check are global and can be placed anywhere in a specific
544 test's script, however it was observed that placement needed to be near
545 the beginning for it to behave that way. Changed the runtest perl
546 script to read through the entire script once to detect and set these
547 variables, reset to the beginning of the script, and then run through
548 the script parsing/test process like normal.
550 TL/02 The BSD's have an arc4random API. One of the functions to induce
551 adding randomness was arc4random_stir(), but it has been removed in
552 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
553 function when detected.
555 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
556 cause callback expansion.
558 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
559 syntax errors in an expansion can be treated as a string instead of
560 logging or causing an error, due to the internal use of bool_lax
561 instead of bool when processing it.
563 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
564 server certificates when making smtp deliveries.
566 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
568 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
570 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
572 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
573 Merged patch from Sebastian Wiedenroth.
575 JH/05 Fix results-pipe from transport process. Several recipients, combined
576 with certificate use, exposed issues where response data items split
577 over buffer boundaries were not parsed properly. This eventually
578 resulted in duplicates being sent. This issue only became common enough
579 to notice due to the introduction of connection certificate information,
580 the item size being so much larger. Found and fixed by Wolfgang Breyha.
582 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
583 size buffer was used, resulting in syntax errors when an expansion
586 JH/07 Add support for directories of certificates when compiled with a GnuTLS
587 version 3.3.6 or later.
589 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
590 is EXPERIMENTAL_EVENT, the main-configuration and transport options
591 both become "event_action", the variables become $event_name, $event_data
592 and $event_defer_errno. There is a new variable $verify_mode, usable in
593 routers, transports and related events. The tls:cert event is now also
594 raised for inbound connections, if the main configuration event_action
597 TL/06 In test suite, disable OCSP for old versions of openssl which contained
598 early OCSP support, but no stapling (appears to be less than 1.0.0).
600 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
601 server certificate names available under the smtp transport option
602 "tls_verify_cert_hostname" now do not permit multi-component wildcard
605 JH/10 Time-related extraction expansions from certificates now use the main
606 option "timezone" setting for output formatting, and are consistent
607 between OpenSSL and GnuTLS compilations. Bug 1541.
609 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
610 encoded parameter in the incoming message. Bug 1558.
612 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
613 include certificate info, eximon was claiming there were spoolfile
616 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
618 JH/14 Log delivery-related information more consistently, using the sequence
619 "H=<name> [<ip>]" wherever possible.
621 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
622 are problematic for Debian distribution, omit them from the release
625 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
627 JH/16 Fix string representation of time values on 64bit time_t architectures.
630 JH/17 Fix a null-indirection in certextract expansions when a nondefault
631 output list separator was used.
636 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
637 checkers that were complaining about end of non-void function with no
640 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
641 This was a regression introduced in 4.83 by another bugfix.
643 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
645 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
646 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
652 TF/01 Correctly close the server side of TLS when forking for delivery.
654 When a message was received over SMTP with TLS, Exim failed to clear up
655 the incoming connection properly after forking off the child process to
656 deliver the message. In some situations the subsequent outgoing
657 delivery connection happened to have the same fd number as the incoming
658 connection previously had. Exim would try to use TLS and fail, logging
659 a "Bad file descriptor" error.
661 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
662 utilities have not been installed.
664 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
665 temporary space as the ACL may create new global variables.
667 TL/01 LDAP support uses per connection or global context settings, depending
668 upon the detected version of the libraries at build time.
670 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
671 to extract and use the src ip:port in logging and expansions as if it
672 were a direct connection from the outside internet. PPv2 support was
673 updated based on HAProxy spec change in May 2014.
675 JH/02 Add ${listextract {number}{list}{success}{fail}}.
677 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
678 Properly escape header and check for NULL return.
680 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
683 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
685 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
686 characters in header names, implemented as a verify condition.
687 Contributed by Michael Fischer v. Mollard.
689 TL/05 Rename SPF condition results err_perm and err_temp to standardized
690 results permerror and temperror. Previous values are deprecated but
691 still accepted. In a future release, err_perm and err_temp will be
692 completely removed, which will be a backward incompatibility if the
693 ACL tests for either of these two old results. Patch contributed by
694 user bes-internal on the mailing list.
696 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
698 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
699 selectors, in both main and reject logs.
701 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
704 JH/07 Add malware type "sock" for talking to simple daemon.
706 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
708 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
709 routers/transports under cutthrough routing.
711 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
712 numbers. Touch up "bool" conditional to keep the same definition.
714 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
716 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
718 JH/12 Expand items in router/transport headers_add or headers_remove lists
719 individually rather than the list as a whole. Bug 1452.
721 Required for reasonable handling of multiple headers_ options when
722 they may be empty; requires that headers_remove items with embedded
723 colons must have them doubled (or the list-separator changed).
725 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
726 view the policy declared in the DMARC record. Currently, $dmarc_status
727 is a combined value of both the record presence and the result of the
730 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
732 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
733 dnslookup router and the smtp transport (applying to the forward
736 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
737 of ldap servers used for a specific lookup. Patch provided by Heiko
740 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
741 New variable $lookup_dnssec_authenticated for observability.
743 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
744 Patch submitted by Lars Timman.
746 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
748 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
749 Requires trusted mode and valid format message id, aborts otherwise.
750 Patch contributed by Heiko Schlichting.
752 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
753 certextract with support for various fields. Bug 1358.
755 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
756 is requested by default, modifiable by smtp transport option
759 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
760 operate on certificate variables to give certificate fingerprints
761 Also new ${sha256:cert_variable}.
763 JH/23 The PRDR feature is moved from being Experimental into the mainline.
765 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
766 Christian Aistleitner.
768 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
770 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
771 file. Patch from Wolfgang Breyha.
773 JH/25 Expand the coverage of the delivery $host and $host_address to
774 client authenticators run in verify callout. Bug 1476.
776 JH/26 Port service names are now accepted for tls_on_connect_ports, to
777 align with daemon_smtp_ports. Bug 72.
779 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
780 support and error reporting did not work properly.
782 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
783 and is readable. Patch from Andrew Colin Kissa.
785 TL/14 Enhance documentation of ${run expansion and how it parses the
786 commandline after expansion, particularly in the case when an
787 unquoted variable expansion results in an empty value.
789 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
791 PP/02 Fix internal collision of T_APL on systems which support RFC3123
792 by renaming away from it. Addresses GH issue 15, reported by
795 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
797 TL/15 SECURITY: prevent double expansion in math comparison functions
798 (can expand unsanitized data). Not remotely exploitable.
805 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
807 PP/02 Make -n do something, by making it not do something.
808 When combined with -bP, the name of an option is not output.
810 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
813 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
814 $sender_host_name and config options to manage this, and basic check
817 PP/05 DSCP support for outbound connections and control modifier for inbound.
819 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
820 (Only plugin which currently uses this is kerberos4, which nobody should
821 be using, but we should make it available and other future plugins might
822 conceivably use it, even though it would break NAT; stuff *should* be
823 using channel bindings instead).
825 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
826 name; added for Sendmail compatibility; requires admin caller.
827 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
828 just ignore it); requires trusted caller.
829 Also parse but ignore: -Ac -Am -X<logfile>
832 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
834 TL/02 Add +smtp_confirmation as a default logging option.
836 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
837 Patch by Magnus Holmgren from 2007-02-20.
839 TL/04 Bugzilla 1281 - Spec typo.
840 Bugzilla 1283 - Spec typo.
841 Bugzilla 1290 - Spec grammar fixes.
843 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
845 TL/06 Add Experimental DMARC support using libopendmarc libraries.
847 TL/07 Fix an out of order global option causing a segfault. Reported to dev
848 mailing list by by Dmitry Isaikin.
850 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
852 JH/02 Support "G" suffix to numbers in ${if comparisons.
854 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
856 NM/01 Bugzilla 1197 - Spec typo
857 Bugzilla 1196 - Spec examples corrections
859 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
861 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
862 gnutls_enable_pkcs11, but renamed to more accurately indicate its
865 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
866 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
868 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
869 "acl {{name}{arg}...}", and optional args on acl condition
872 JH/05 Permit multiple router/transport headers_add/remove lines.
874 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
876 JH/07 Avoid using a waiting database for a single-message-only transport.
877 Performance patch from Paul Fisher. Bugzilla 1262.
879 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
882 JH/09 Add $headers_added variable, with content from use of ACL modifier
883 add_header (but not yet added to the message). Bugzilla 199.
885 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
886 Pulled from Bugzilla 817 by Wolfgang Breyha.
888 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
890 (nb: this is the same fix as in Exim 4.80.1)
892 JH/11 Add A= logging on delivery lines, and a client_set_id option on
895 JH/12 Add optional authenticated_sender logging to A= and a log_selector
898 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
900 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
901 advertise SMTP AUTH mechanism to us, instead of a generic
902 protocol violation error. Also, make Exim more robust to bad
903 data from the Dovecot auth socket.
905 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
907 When a queue runner is handling a message, Exim first routes the
908 recipient addresses, during which it prunes them based on the retry
909 hints database. After that it attempts to deliver the message to
910 any remaining recipients. It then updates the hints database using
913 So if a recipient address works intermittently, it can get repeatedly
914 deferred at routing time. The retry hints record remains fresh so the
915 address never reaches the final cutoff time.
917 This is a fairly common occurrence when a user is bumping up against
918 their storage quota. Exim had some logic in its local delivery code
919 to deal with this. However it did not apply to per-recipient defers
920 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
922 This change adds a proper retry rule check during routing so that the
923 final cutoff time is checked against the message's age. We only do
924 this check if there is an address retry record and there is not a
925 domain retry record; this implies that previous attempts to handle
926 the address had the retry_use_local_parts option turned on. We use
927 this as an approximation for the destination being like a local
928 delivery, as in LMTP.
930 I suspect this new check makes the old local delivery cutoff check
931 redundant, but I have not verified this so I left the code in place.
933 TF/02 Correct gecos expansion when From: is a prefix of the username.
935 Test 0254 submits a message to Exim with the header
939 When I ran the test suite under the user fanf2, Exim expanded
940 the header to contain my full name, whereas it should have added
941 a Resent-Sender: header. It erroneously treats any prefix of the
942 username as equal to the username.
944 This change corrects that bug.
946 GF/01 DCC debug and logging tidyup
947 Error conditions log to paniclog rather than rejectlog.
948 Debug lines prefixed by "DCC: " to remove any ambiguity.
950 TF/03 Avoid unnecessary rebuilds of lookup-related code.
952 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
953 Bug spotted by Jeremy Harris; was flawed since initial commit.
954 Would have resulted in OCSP responses post-SNI triggering an Exim
955 NULL dereference and crash.
957 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
959 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
960 Bug detection, analysis and fix by Samuel Thibault.
961 Bugzilla 1331, Debian bug #698092.
963 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
965 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
966 Server implementation by Todd Lyons, client by JH.
967 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
968 config variable "prdr_enable" controls whether the server
969 advertises the facility. If the client requests PRDR a new
970 acl_data_smtp_prdr ACL is called once for each recipient, after
971 the body content is received and before the acl_smtp_data ACL.
972 The client is controlled by both of: a hosts_try_prdr option
973 on the smtp transport, and the server advertisement.
974 Default client logging of deliveries and rejections involving
975 PRDR are flagged with the string "PRDR".
977 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
978 fclose(). Diagnosis by Todd Lyons.
980 PP/17 Update configure.default to handle IPv6 localhost better.
981 Patch by Alain Williams (plus minor tweaks).
984 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
985 This is now consistent with GnuTLS, and is now documented: the
986 previous undocumented portable approach to treating the option as
987 unset was to force an expansion failure. That still works, and
988 an empty string is now equivalent.
990 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
991 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
992 not performing validation itself.
994 PP/20 Added force_command boolean option to pipe transport.
995 Patch from Nick Koston, of cPanel Inc.
997 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1000 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1002 PP/21 Fix eximon continuous updating with timestamped log-files.
1003 Broken in a format-string cleanup in 4.80, missed when I repaired the
1004 other false fix of the same issue.
1005 Report and fix from Heiko Schlichting.
1008 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1009 Report from Prashanth Katuri.
1011 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1012 It's SecureTransport, so affects any MacOS clients which use the
1013 system-integrated TLS libraries, including email clients.
1015 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1016 using a MIME ACL for non-SMTP local injection.
1017 Report and assistance in diagnosis by Warren Baker.
1019 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1021 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1023 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1024 last id that failed so it may be referenced in subsequent ACL's.
1026 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1029 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1030 ldap library initialization, allowing self-signed CA's to be
1031 used. Also properly sets require_cert option later in code by
1032 using NULL (global ldap config) instead of ldap handle (per
1033 session). Bug diagnosis and testing by alxgomz.
1035 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1036 the src/util/ subdirectory.
1038 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1039 renamed to Transport Post Delivery Action by Jeremy Harris, as
1042 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1043 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1044 redis_servers = needs to be configured which will be used by the redis
1045 lookup. Patch from Warren Baker, of The Packet Hub.
1047 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1049 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1050 hostname or reverse DNS when processing a host list. Used suggestions
1051 from multiple comments on this bug.
1053 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1055 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1056 Missed a few lines, added it to make the runtest require no keyboard
1059 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1060 contains upper case chars. Make router use caseful_local_part.
1062 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1063 support when GnuTLS has been built with p11-kit.
1069 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1071 This, or similar/improved, will also be change PP/11 of 4.82.
1077 PP/01 Handle short writes when writing local log-files.
1078 In practice, only affects FreeBSD (8 onwards).
1079 Bugzilla 1053, with thanks to Dmitry Isaikin.
1081 NM/01 Bugzilla 949 - Documentation tweak
1083 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1086 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1088 PP/02 Implemented gsasl authenticator.
1090 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1092 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1093 `pkg-config foo` for cflags/libs.
1095 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1096 with rest of GSASL and with heimdal_gssapi.
1098 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1099 `pkg-config foo` for cflags/libs for the TLS implementation.
1101 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1102 properties get this fed in as external SSF. A number of robustness
1103 and debugging improvements to the cyrus_sasl authenticator.
1105 PP/08 cyrus_sasl server now expands the server_realm option.
1107 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1108 Patch by Jeremy Harris.
1110 PP/10 Added dbmjz lookup type.
1112 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1114 PP/12 MAIL args handles TAB as well as SP, for better interop with
1115 non-compliant senders.
1116 Analysis and variant patch by Todd Lyons.
1118 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1119 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1120 Patch from Dirk Mueller <dmueller@suse.com>
1122 PP/13 tls_peerdn now print-escaped for spool files.
1123 Observed some $tls_peerdn in wild which contained \n, which resulted
1124 in spool file corruption.
1126 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1127 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1128 or write after TLS renegotiation, which otherwise led to messages
1131 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1132 as a tracking header (ie: a signed header comes before the signature).
1133 Patch from Wolfgang Breyha.
1135 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1136 comma-sep list; embedded commas doubled.
1138 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1140 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1142 Report and patch from Dmitry Banschikov.
1144 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1145 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1146 failures. We appear to now support TLS1.1+ with Exim.
1148 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1149 lets Exim select keys and certificates based upon TLS SNI from client.
1150 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1151 before an outbound SMTP session. New log_selector, +tls_sni.
1153 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1154 NULL dereference. Report and patch from Alun Jones.
1156 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1157 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1158 Not seeing resolver debug output on NetBSD, but suspect this is a
1159 resolver implementation change.
1161 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1162 Left warnings. Added "eximon gdb" invocation mode.
1164 PP/21 Defaulting "accept_8bitmime" to true, not false.
1166 PP/22 Added -bw for inetd wait mode support.
1168 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1169 locate the relevant includes and libraries. Made this the default.
1171 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1172 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1174 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1175 This may cause build issues on older platforms.
1177 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1178 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1179 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1180 Added SNI support via GnuTLS too.
1181 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1183 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1185 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1186 Applied second patch from Janne, implementing suggestion to default
1187 multiple-strings-in-record handling to match SPF spec.
1189 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1191 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1192 read-only, out of scope).
1193 Patch from Wolfgang Breyha, report from Stuart Northfield.
1195 PP/29 Fix three issues highlighted by clang analyser static analysis.
1196 Only crash-plausible issue would require the Cambridge-specific
1197 iplookup router and a misconfiguration.
1198 Report from Marcin Mirosław.
1200 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1202 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1203 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1204 As part of this, removing so much warning spew let me fix some minor
1205 real issues in debug logging.
1207 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1208 assignment on my part. Fixed.
1210 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1211 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1212 Janne Snabb (who went above and beyond: thank you).
1214 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1215 string otherwise requires a connection and a bunch more work and it's
1216 relatively easy to get wrong. Should also expose TLS library linkage
1219 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1220 64-bit ${eval} (JH/03).
1222 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1223 GNU libc to support some of the 64-bit stuff, should not lead to
1224 conflicts. Defined before os.h is pulled in, so if a given platform
1225 needs to override this, it can.
1227 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1228 protection layer was required, which is not implemented.
1229 Bugzilla 1254, patch from Wolfgang Breyha.
1231 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1232 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1233 tls_dhparam take prime identifiers. Also unbreak combination of
1234 OpenSSL+DH_params+TLSSNI.
1236 PP/39 Disable SSLv2 by default in OpenSSL support.
1242 PP/01 Solaris build fix for Oracle's LDAP libraries.
1243 Bugzilla 1109, patch from Stephen Usher.
1245 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1247 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1250 TF/02 Fix a couple more cases where we did not log the error message
1251 when unlink() failed. See also change 4.74-TF/03.
1253 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1254 lock up or crash if it happened to be inside a call to libc when it
1255 got a SIGUSR1 from exiwhat.
1257 The SIGUSR1 handler appends the current process status to the process
1258 log which is later printed by exiwhat. It used to use the general
1259 purpose logging code to do this, but several functions it calls are
1260 not safe for signals.
1262 The new output code in the SIGUSR1 handler is specific to the process
1263 log, and simple enough that it's easy to inspect for signal safety.
1264 Removing some special cases also simplifies the general logging code.
1265 Removing the spurious timestamps from the process log simplifies
1268 TF/04 Improved ratelimit ACL condition.
1270 The /noupdate option has been deprecated in favour of /readonly which
1271 has clearer semantics. The /leaky, /strict, and /readonly update modes
1272 are mutually exclusive. The update mode is no longer included in the
1273 database key; it just determines when the database is updated. (This
1274 means that when you upgrade Exim will forget old rate measurements.)
1276 Exim now checks that the per_* options are used with an update mode that
1277 makes sense for the current ACL. For example, when Exim is processing a
1278 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1279 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1280 must specify per_mail/readonly. If you omit the update mode it defaults to
1281 /leaky where that makes sense (as before) or /readonly where required.
1283 The /noupdate option is now undocumented but still supported for
1284 backwards compatibility. It is equivalent to /readonly except that in
1285 ACLs where /readonly is required you may specify /leaky/noupdate or
1286 /strict/noupdate which are treated the same as /readonly.
1288 A useful new feature is the /count= option. This is a generalization
1289 of the per_byte option, so that you can measure the throughput of other
1290 aggregate values. For example, the per_byte option is now equivalent
1291 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1293 The per_rcpt option has been generalized using the /count= mechanism
1294 (though it's more complicated than the per_byte equivalence). When it is
1295 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1296 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1297 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1298 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1299 Note that using per_rcpt with a non-readonly update mode in more than
1300 one ACL will cause the recipients to be double-counted. (The per_mail
1301 and per_byte options don't have this problem.)
1303 The handling of very low rates has changed slightly. If the computed rate
1304 is less than the event's count (usually one) then this event is the first
1305 after a long gap. In this case the rate is set to the same as this event's
1306 count, so that the first message of a spam run is counted properly.
1308 The major new feature is a mechanism for counting the rate of unique
1309 events. The new per_addr option counts the number of different
1310 recipients that someone has sent messages to in the last time period. It
1311 behaves like per_rcpt if all the recipient addresses are different, but
1312 duplicate recipient addresses do not increase the measured rate. Like
1313 the /count= option this is a general mechanism, so the per_addr option
1314 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1315 example, measure the rate that a client uses different sender addresses
1316 with the options per_mail/unique=$sender_address. There are further
1317 details in the main documentation.
1319 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1321 TF/06 Removed a few PCRE remnants.
1323 TF/07 Automatically extract Exim's version number from tags in the git
1324 repository when doing development or release builds.
1326 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1327 Bugzilla 879. Patch from Paul Fisher.
1329 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1330 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1333 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1335 PP/05 Variable $av_failed, true if the AV scanner deferred.
1336 Bugzilla 1078. Patch from John Horne.
1338 PP/06 Stop make process more reliably on build failure.
1339 Bugzilla 1087. Patch from Heiko Schlittermann.
1341 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1342 Bugzilla 1089. Patch from Heiko Schlittermann.
1344 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1345 Bugzilla 1131. Patch from Holger Weiß.
1347 PP/09 Handle IPv6 addresses with SPF.
1348 Bugzilla 860. Patch from Wolfgang Breyha.
1350 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1352 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1355 PP/11 match_* no longer expand right-hand-side by default.
1356 New compile-time build option, EXPAND_LISTMATCH_RHS.
1357 New expansion conditions, "inlist", "inlisti".
1359 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1361 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1363 PP/14 fix log_write() format string regression from TF/03.
1364 Bugzilla 1152. Patch from Dmitry Isaikin.
1370 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1372 PP/02 Harmonised TLS library version reporting; only show if debugging.
1373 Layout now matches that introduced for other libraries in 4.74 PP/03.
1375 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1377 PP/04 New "dns_use_edns0" global option.
1379 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1382 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1383 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1385 TK/01 Updated PolarSSL code to 0.14.2.
1386 Bugzilla 1097. Patch from Andreas Metzler.
1388 PP/07 Catch divide-by-zero in ${eval:...}.
1389 Fixes bugzilla 1102.
1391 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1394 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1395 format-string attack -- SECURITY: remote arbitrary code execution.
1397 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1398 time unintentionally subject to list matching rules, letting the header
1399 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1400 arbitrary string expansion). This allowed for information disclosure.
1402 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1403 INT_MIN/-1 -- value coerced to INT_MAX.
1409 NM/01 Workaround for PCRE version dependency in version reporting
1412 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1413 This fixes portability to compilers other than gcc, notably
1414 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1416 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1417 makefiles for portability to HP-UX and POSIX correctness.
1419 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1420 Also via indented variable definition in the Makefile.
1421 (Debugging by Oliver Heesakkers).
1423 PP/02 Restore caching of spamd results with expanded spamd_address.
1424 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1426 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1427 Improves build reliability. Fix from: Frank Elsner
1429 NM/02 Fix wide character breakage in the rfc2047 coding
1430 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1432 NM/03 Allow underscore in dnslist lookups
1433 Fixes bug 1026. Patch from Graeme Fowler
1435 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1436 Code patches from Adam Ciarcinski of NetBSD.
1438 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1441 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1442 is logged, to avoid truncation. Patch from John Horne.
1444 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1445 Patch from Jakob Hirsch.
1447 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1448 SQL string expansion failure details.
1449 Patch from Andrey Oktyabrski.
1451 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1452 Patch from Simon Arlott.
1454 PP/09 New lookups functionality failed to compile on old gcc which rejects
1455 extern declarations in function scope.
1456 Patch from Oliver Fleischmann
1458 PP/10 Use sig_atomic_t for flags set from signal handlers.
1459 Check getgroups() return and improve debugging.
1460 Fixed developed for diagnosis in bug 927 (which turned out to be
1463 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1464 Patch from Mark Zealey.
1466 PP/12 Bugzilla 1056: Improved spamd server selection.
1467 Patch from Mark Zealey.
1469 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1470 Based on patch from Heiko Schlittermann.
1472 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1473 Patch from Uwe Doering, sign-off by Michael Haardt.
1475 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1476 variable declaration deep within a block. Bug and patch from
1479 PP/15 lookups-Makefile IRIX compatibility coercion.
1481 PP/16 Make DISABLE_DKIM build knob functional.
1483 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1484 Patch by Simon Arlott
1486 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1487 variable argument macros. Our copy now differs from upstream.
1493 TF/01 Failure to get a lock on a hints database can have serious
1494 consequences so log it to the panic log.
1496 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1497 controlled using the smtp_confirmation log selector.
1499 TF/03 Include the error message when we fail to unlink a spool file.
1501 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1502 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1503 for maintaining out-of-tree patches for some time.
1505 PP/01 Bugzilla 139: Documentation and portability issues.
1506 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1507 Handle per-OS dynamic-module compilation flags.
1509 PP/02 Let /dev/null have normal permissions.
1510 The 4.73 fixes were a little too stringent and complained about the
1511 permissions on /dev/null. Exempt it from some checks.
1512 Reported by Andreas M. Kirchwitz.
1514 PP/03 Report version information for many libraries, including
1515 Exim version information for dynamically loaded libraries. Created
1516 version.h, now support a version extension string for distributors
1517 who patch heavily. Dynamic module ABI change.
1519 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1520 privilege escalation vulnerability whereby the Exim run-time user
1521 can cause root to append content of the attacker's choosing to
1524 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1527 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1528 If dropping privileges for untrusted macros, we disabled normal logging
1529 on the basis that it would fail; for the Exim run-time user, this is not
1530 the case, and it resulted in successful deliveries going unlogged.
1531 Fixed. Reported by Andreas Metzler.
1537 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1538 only prepend for the Resent-* case. Fixes regression introduced in
1539 Exim 4.70 by NM/22 for Bugzilla 607.
1541 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1542 increasing numbers of administrators be bitten by this.
1544 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1546 PP/03 Bugzilla 994: added openssl_options main configuration option.
1548 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1550 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1552 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1554 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1555 without a peer certificate, leading to a segfault because of an
1556 assumption that peers always have certificates. Be a little more
1557 paranoid. Problem reported by Martin Tscholak.
1559 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1560 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1561 NB: ClamAV planning to remove STREAM in "middle of 2010".
1562 CL also introduces -bmalware, various -d+acl logging additions and
1563 more caution in buffer sizes.
1565 PP/09 Implemented reverse_ip expansion operator.
1567 PP/10 Bugzilla 937: provide a "debug" ACL control.
1569 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1571 PP/12 Bugzilla 973: Implement --version.
1573 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1575 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1577 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1579 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1580 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1581 ignore trailing whitespace.
1583 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1585 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1588 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1589 Notification from Dr Andrew Aitchison.
1591 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1592 ExtendedDetectionInfo response format.
1593 Notification from John Horne.
1595 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1598 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1599 XSL and documented dependency on system catalogs, with examples of how
1602 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1605 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1606 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1607 configuration file which is writeable by the Exim user or group.
1609 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1610 of configuration files to cover files specified with the -C option if
1611 they are going to be used with root privileges, not just the default
1614 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1615 option (effectively making it always true).
1617 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1618 files to be used while preserving root privileges.
1620 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1621 that rogue child processes cannot use them.
1623 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1624 run-time user, instead of root.
1626 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1627 Exim run-time user without dropping privileges.
1629 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1630 result string, instead of calling string_vformat() twice with the same
1633 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1634 for other users. Others should always drop root privileges if they use
1635 -C on the command line, even for a whitelisted configure file.
1637 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1639 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1645 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1646 $data_path, and $header_path variables; fixed documentation bugs and
1649 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1650 exipick to access non-standard spools, including the "frozen" queue
1653 NM/01 Bugzilla 965: Support mysql stored procedures.
1654 Patch from Alain Williams
1656 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1658 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1659 Patch from Andreas Metzler
1661 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1662 Patch from Kirill Miazine
1664 NM/05 Bugzilla 671: Added umask to procmail example.
1666 JJ/03 installed exipick 20100323.0, fixing doc bug
1668 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1669 directory. Notification and patch from Dan Rosenberg.
1671 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1673 TK/02 Improve log output when DKIM signing operation fails.
1675 MH/01 Treat the transport option dkim_domain as a colon separated
1676 list, not as a single string, and sign the message with each element,
1677 omitting multiple occurences of the same signer.
1679 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1680 Bugzilla 985, 986. Patch by Simon Arlott
1682 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1683 Patch by Simon Arlott
1685 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1686 MBX locking. Notification from Dan Rosenberg.
1692 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1694 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1696 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1698 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1700 NM/04 Bugzilla 915: Flag broken perl installation during build.
1706 TK/01 Added patch by Johannes Berg that expands the main option
1707 "spamd_address" if it starts with a dollar sign.
1709 TK/02 Write list of recipients to X-Envelope-Sender header when building
1710 the mbox-format spool file for content scanning (suggested by Jakob
1713 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1714 (http://www.dcc-servers.net/) support via dccifd. Activated by
1715 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1717 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1718 by Mark Daniel Reidel <mr@df.eu>.
1720 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1721 When building exim an external PCRE library is now needed -
1722 PCRE is a system library on the majority of modern systems.
1723 See entry on PCRE_LIBS in EDITME file.
1725 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1726 conversation. Added nologin parameter to request.
1727 Patch contributed by Kirill Miazine.
1729 TF/01 Do not log submission mode rewrites if they do not change the address.
1731 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1733 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1734 log files in place. Contributed by Roberto Lima.
1736 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1738 TF/03 Bugzilla 615: When checking the local_parts router precondition
1739 after a local_part_suffix or local_part_prefix option, Exim now
1740 does not use the address's named list lookup cache, since this
1741 contains cached lookups for the whole local part.
1743 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1744 Robert Millan. Documentation is in experimental-spec.txt.
1746 TF/04 Bugzilla 668: Fix parallel build (make -j).
1748 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1750 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1751 Patch provided by Jan Srzednicki.
1753 TF/05 Leading white space used to be stripped from $spam_report which
1754 wrecked the formatting. Now it is preserved.
1756 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1757 that they are available at delivery time.
1759 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1761 TF/08 TLS error reporting now respects the incoming_interface and
1762 incoming_port log selectors.
1764 TF/09 Produce a more useful error message if an SMTP transport's hosts
1765 setting expands to an empty string.
1767 NM/06 Bugzilla 744: EXPN did not work under TLS.
1768 Patch provided by Phil Pennock.
1770 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1771 Patch provided by Richard Godbee.
1773 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1774 acl_smtp_notquit, added index entry.
1776 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1777 Patch provided by Eugene Bujak.
1779 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1780 accept(). Patch provided by Maxim Dounin.
1782 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1783 Patch provided by Phil Pennock.
1785 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1787 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1788 Patch provided by Brad "anomie" Jorsch.
1790 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1791 Patch provided by Dean Brooks.
1793 TK/05 Add native DKIM support (does not depend on external libraries).
1795 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1796 Patch provided by Graeme Fowler.
1798 NM/16 Bugzilla 851: Documentation example syntax fix.
1800 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1802 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1805 NM/19 Bugzilla 745: TLS version reporting.
1806 Patch provided by Phil Pennock.
1808 NM/20 Bugzilla 167: bool: condition support.
1809 Patch provided by Phil Pennock.
1811 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1812 clients. Patch provided by Phil Pennock.
1814 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1815 Patch provided by Brad "anomie" Jorsch.
1817 NM/23 Bugzilla 687: Fix misparses in eximstats.
1818 Patch provided by Heiko Schlittermann.
1820 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1821 Patch provided by Heiko Schlittermann.
1823 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1824 plus update to original patch.
1826 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1828 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1829 Patch provided by David Brownlee.
1831 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1833 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1835 NM/30 Bugzilla 888: TLS documentation bugfixes.
1837 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1839 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1840 Unlike the original bugzilla I have changed all shell scripts in src tree.
1842 NM/33 Bugzilla 898: Transport filter timeout fix.
1843 Patch by Todd Rinaldo.
1845 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1846 Patch by Serge Demonchaux.
1848 NM/35 Bugzilla 39: Base64 decode bug fixes.
1849 Patch by Jakob Hirsch.
1851 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1853 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1855 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1857 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1863 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1864 ALT-N's libdkim that I have put here:
1865 http://duncanthrax.net/exim-experimental/
1867 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1868 were called 'true' and it seems that C99 defines that as a reserved
1869 keyword to be used with 'bool' variable types. That means you could
1870 not include C99-style headers which use bools without triggering
1871 build errors in sieve.c.
1873 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1874 as mailq or other aliases. Changed the --help handling significantly
1875 to do whats expected. exim_usage() emits usage/help information.
1877 SC/01 Added the -bylocaldomain option to eximstats.
1879 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1881 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1883 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1889 PH/01 Another patch from the Sieve maintainer.
1891 PH/02 When an IPv6 address is converted to a string for single-key lookup
1892 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1893 dots are used instead of colons so that keys in lsearch files need not
1894 contain colons. This was done some time before quoting was made available
1895 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1896 (notated using the quote facility) so as to distinguish them from IPv4
1897 keys. This meant that lookups for IP addresses in host lists did not work
1898 for iplsearch lookups.
1900 This has been fixed by arranging for IPv6 addresses to be expressed with
1901 colons if the lookup type is iplsearch. This is not incompatible, because
1902 previously such lookups could never work.
1904 The situation is now rather anomalous, since one *can* have colons in
1905 ordinary lsearch keys. However, making the change in all cases is
1906 incompatible and would probably break a number of configurations.
1908 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1911 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1912 conversion specification without a maximum field width, thereby enabling
1913 a rogue spamd server to cause a buffer overflow. While nobody in their
1914 right mind would setup Exim to query an untrusted spamd server, an
1915 attacker that gains access to a server running spamd could potentially
1916 exploit this vulnerability to run arbitrary code as the Exim user.
1918 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1919 $primary_hostname instead of what libspf2 thinks the hosts name is.
1921 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1922 a directory entry by the name of the lookup key. Previously, if a
1923 symlink pointed to a non-existing file or a file in a directory that
1924 Exim lacked permissions to read, a lookup for a key matching that
1925 symlink would fail. Now it is enough that a matching directory entry
1926 exists, symlink or not. (Bugzilla 503.)
1928 PH/03 The body_linecount and body_zerocount variables are now exported in the
1931 PH/04 Added the $dnslist_matched variable.
1933 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1934 This means they are set thereafter only if the connection becomes
1937 PH/06 Added the client_condition to authenticators so that some can be skipped
1938 by clients under certain conditions.
1940 PH/07 The error message for a badly-placed control=no_multiline_responses left
1941 "_responses" off the end of the name.
1943 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1945 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1946 (without spaces) instead of just copying the configuration text.
1948 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1950 PH/11 Added $max_received_linelength.
1952 PH/12 Added +ignore_defer and +include_defer to host lists.
1954 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1955 way in which PCRE > 7.0 is built.
1957 PH/14 Implemented queue_only_load_latch.
1959 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1960 MAIL command. The effect was to mangle the value on 64-bit systems.
1962 PH/16 Another patch from the Sieve maintainer.
1964 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1966 PH/18 If a system quota error occurred while trying to create the file for
1967 a maildir delivery, the message "Mailbox is full" was not appended to the
1968 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1969 only to a quota excession during the actual writing of the file.
1971 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1972 characters?) which causes problems in log lines. The DN values are now
1973 passed through string_printing() before being added to log lines.
1975 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1976 and InterBase are left for another time.)
1978 PH/21 Added message_body_newlines option.
1980 PH/22 Guard against possible overflow in moan_check_errorcopy().
1982 PH/23 POSIX allows open() to be a macro; guard against that.
1984 PH/24 If the recipient of an error message contained an @ in the local part
1985 (suitably quoted, of course), incorrect values were put in $domain and
1986 $local_part during the evaluation of errors_copy.
1992 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1993 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1996 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1997 issue a MAIL command.
1999 PH/02 In an ACL statement such as
2001 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2003 if a client was not listed at all, or was listed with a value other than
2004 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2005 the condition was not true (as it should be), so access was not denied.
2006 The bug was that the ! inversion was incorrectly passed on to the second
2007 item. This has been fixed.
2009 PH/03 Added additional dnslists conditions == and =& which are different from
2010 = and & when the dns lookup returns more than one IP address.
2012 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2013 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2015 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2016 FSYNC, which compiles an option called disable_fsync that allows for
2017 bypassing fsync(). The documentation is heavily laced with warnings.
2019 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2021 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2022 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2023 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2024 including adding "make clean"; (3) Added -fPIC when compiling the test
2025 dynamically loaded module, to get rid of a warning.
2027 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2028 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2029 = 0s. The bug is otherwise harmless.
2031 PH/07 There was a bug in the dovecot authenticator such that the value of
2032 $auth1 could be overwritten, and so not correctly preserved, after a
2033 successful authentication. This usually meant that the value preserved by
2034 the server_setid option was incorrect.
2036 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2038 PH/09 Installed PCRE release 7.0.
2040 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2041 run for batched SMTP input. It is now run at the start of every message
2042 in the batch. While fixing this I discovered that the process information
2043 (output by running exiwhat) was not always getting set for -bs and -bS
2044 input. This is fixed, and it now also says "batched" for BSMTP.
2046 PH/11 Added control=no_pipelining.
2048 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2049 patch, slightly modified), and move the expansion of helo_data till after
2050 the connection is made in the smtp transport (so it can use these
2053 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2055 PH/14 Added log_selector = +pid.
2057 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2059 PH/16 Add ${if forany and ${if forall.
2061 PH/17 Added dsn_from option to vary the From: line in DSNs.
2063 PH/18 Flush SMTP output before performing a callout, unless control =
2064 no_callout_flush is set.
2066 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2067 was true (the default) a successful delivery failed to delete the retry
2068 item, thus causing premature timeout of the address. The bug is now
2071 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2073 PH/21 Long custom messages for fakedefer and fakereject are now split up
2074 into multiline responses in the same way that messages for "deny" and
2075 other ACL rejections are.
2077 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2078 with slight modification.
2080 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2081 draft, changing the syntax and factoring some duplicate code".
2083 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2084 for deliveries of the second and subsequent messages over the same SMTP
2087 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2088 ${reduce, with only minor "tidies".
2090 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2092 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2093 expansion side effects.
2095 PH/27 When a message times out after an over-quota error from an Exim-imposed
2096 quota, the bounce message says "mailbox is full". This message was not
2097 being given when it was a system quota that was exceeded. It now should
2100 MH/03 Made $recipients available in local_scan(). local_scan() already has
2101 better access to the recipient list through recipients_list[], but
2102 $recipients can be useful in postmaster-provided expansion strings.
2104 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2105 in the case of a MAIL command with additional options following the
2106 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2107 were accidentally chopped off.
2109 PH/29 SMTP synchronization checks are implemented when a command is read -
2110 there is a check that no more input is waiting when there shouldn't be
2111 any. However, for some commands, a delay in an ACL can mean that it is
2112 some time before the response is written. In this time, more input might
2113 arrive, invalidly. So now there are extra checks after an ACL has run for
2114 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2115 pipelining has not been advertised.
2117 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2119 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2120 returned in the SMTP response when smtp_return_error_details was false.
2121 This has been fixed.
2123 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2124 instead of the C I/O that was originally supplied, because problems were
2125 reported on Solaris.
2127 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2128 Exim which did not show up earlier: it was assuming that a call to
2129 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2130 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2131 was a macro that became an assignment, so it seemed to work. This has
2132 changed to a proper function call with a void return, hence the compile
2133 error. Exim's code has been fixed.
2135 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2138 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2140 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2142 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2143 because it turns out to be too big or there is a timeout) there is no
2144 "Completed" line in the log. When some messages of this type were
2145 selected by exigrep, they were listed as "not completed". Others were
2146 picked up by some special patterns. I have improved the selection
2147 criteria to be more general.
2149 PH/38 The host_find_failed option in the manualroute router can now be set
2150 to "ignore", to completely ignore a host whose IP address cannot be
2151 found. If all hosts are ignored, the behaviour is controlled by the new
2152 host_all_ignored option.
2154 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2155 homing or because of multiple MX records with /mx) generated more than
2156 one IP address, and the following item turned out to be the local host,
2157 all the secondary addresses of the first item were incorrectly removed
2158 from the list, along with the local host and any following hosts (which
2159 is what is supposed to happen).
2161 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2162 whoever called Exim into the -H file. In the case of the daemon it was
2163 behaving confusingly. When first started, it used values for whoever
2164 started the daemon, but after a SIGHUP it used the Exim user (because it
2165 calls itself on a restart). I have changed the code so that it now always
2168 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2169 message are rejected with the same error (e.g. no authentication or bad
2170 sender address), and a DATA command is nevertheless sent (as can happen
2171 with PIPELINING or a stupid MUA), the error message that was given to the
2172 RCPT commands is included in the rejection of the DATA command. This is
2173 intended to be helpful for MUAs that show only the final error to their
2176 PH/42 Another patch from the Sieve maintainer.
2178 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2179 Eximstats - Fixed some broken HTML links and added missing column headers
2181 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2182 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2184 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2186 PH/43 Yet another patch from the Sieve maintainer.
2188 PH/44 I found a way to check for a TCP/IP connection going away before sending
2189 the response to the final '.' that terminates a message, but only in the
2190 case where the client has not sent further data following the '.'
2191 (unfortunately, this is allowed). However, in many cases there won't be
2192 any further data because there won't be any more messages to send. A call
2193 to select() can be used: if it shows that the input is "ready", there is
2194 either input waiting, or the socket has been closed. An attempt to read
2195 the next input character can distinguish the two cases. Previously, Exim
2196 would have sent an OK response which the client would never have see.
2197 This could lead to message repetition. This fix should cure that, at
2198 least in a lot of common cases.
2200 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2201 advertised in response to EHLO.
2207 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2208 fixed by 4.65/MH/01 (is this a record?) are fixed:
2210 (i) An empty string was always treated as zero by the numeric comparison
2211 operators. This behaviour has been restored.
2213 (ii) It is documented that the numeric comparison operators always treat
2214 their arguments as decimal numbers. This was broken in that numbers
2215 starting with 0 were being interpreted as octal.
2217 While fixing these problems I realized that there was another issue that
2218 hadn't been noticed. Values of message_size_limit (both the global option
2219 and the transport option) were treated as octal if they started with 0.
2220 The documentation was vague. These values are now always treated as
2221 decimal, and I will make that clear in the documentation.
2227 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2228 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2231 MH/01 Don't check that the operands of numeric comparison operators are
2232 integers when their expansion is in "skipping" mode (fixes bug
2233 introduced by 4.64-PH/07).
2235 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2236 child addresses, Exim now panics and dies. Previously, because the count
2237 is held in a short int, deliveries were likely to be lost. As such a
2238 large number of recipients for a single message is ridiculous
2239 (performance will be very, very poor), I have chosen to impose a limit
2240 rather than extend the field.
2246 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2247 leftover -K file (the existence of which was triggered by #402).
2248 While we were at it, introduced process PID as part of the -K
2249 filename. This should rule out race conditions when creating
2252 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2253 processing considerably. Previous code took too long for large mails,
2254 triggering a timeout which in turn triggers #401.
2256 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2257 in the DK code in transports.c. sendfile() is not really portable,
2258 hence the _LINUX specificness.
2260 TF/01 In the add_headers option to the mail command in an Exim filter,
2261 there was a bug that Exim would claim a syntax error in any
2262 header after the first one which had an odd number of characters
2265 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2266 callout verification, Exim cached a "reject" for the entire domain. This
2267 is correct for most verifications, but it is not correct for a recipient
2268 verification with use_sender or use_postmaster set, because in that case
2269 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2270 case of MAIL FROM:<> rejection from other early rejections (e.g.
2271 rejection of HELO). When verifying a recipient using a non-null MAIL
2272 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2273 Whatever the result of the callout, the value of the domain cache is
2274 left unchanged (for any other kind of callout, getting as far as trying
2275 RCPT means that the domain itself is ok).
2277 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2280 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2281 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2282 session, the daemon ignores ECONNECT errors and logs others; it now
2283 ignores EPIPE as well.
2285 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2286 (quoted-printable decoding).
2288 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2289 later the small subsequent patch to fix an introduced bug.
2291 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2293 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2295 PH/08 An error is now given if message_size_limit is specified negative.
2297 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2298 to be given (somewhat) arbitrary names.
2300 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2303 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2304 miscellaneous code fixes
2306 PH/10 Added the log_reject_target ACL modifier to specify where to log
2309 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2310 hostname. This is wrong, because it relates to the incoming message (and
2311 probably the interface on which it is arriving) and not to the outgoing
2312 callout (which could be using a different interface). This has been
2313 changed to use the value of the helo_data option from the smtp transport
2314 instead - this is what is used when a message is actually being sent. If
2315 there is no remote transport (possible with a router that sets up host
2316 addresses), $smtp_active_hostname is used.
2318 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2319 tweaks were necessary in order to get it to work (see also 21 below):
2320 (a) The code assumed that strncpy() returns a negative number on buffer
2321 overflow, which isn't the case. Replaced with Exim's string_format()
2323 (b) There were several signed/unsigned issues. I just did the minimum
2324 hacking in of casts. There is scope for a larger refactoring.
2325 (c) The code used strcasecmp() which is not a standard C function.
2326 Replaced with Exim's strcmpic() function.
2327 (d) The code set only $1; it now sets $auth1 as well.
2328 (e) A simple test gave the error "authentication client didn't specify
2329 service in request". It would seem that Dovecot has changed its
2330 interface. Fortunately there's a specification; I followed it and
2331 changed what the client sends and it appears to be working now.
2333 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2336 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2337 address A is aliased to B and C, where B exists and C does not. Without
2338 -v the output is "A verified" because verification stops after a
2339 successful redirection if more than one address is generated. However,
2340 with -v the child addresses are also verified. Exim was outputting "A
2341 failed to verify" and then showing the successful verification for C,
2342 with its parentage. It now outputs "B failed to verify", showing B's
2343 parentage before showing the successful verification of C.
2345 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2346 look up a TXT record in a specific list after matching in a combined
2349 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2350 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2351 they consult the DNS. I had assumed they would set it the way they
2352 wanted; and indeed my experiments on Linux seem to show that in some
2353 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2354 To be on the safe side, however, I have now made the interface to
2355 host_find_byname() similar to host_find_bydns(), with an argument
2356 containing the DNS resolver options. The host_find_byname() function now
2357 sets these options at its start, just as host_find_bydns() does. The smtp
2358 transport options dns_qualify_single and dns_search_parents are passed to
2359 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2360 of host_find_byname() use the default settings of RES_DEFNAMES
2361 (qualify_single) but not RES_DNSRCH (search_parents).
2363 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2364 spool_read_header() do less string testing, by means of a preliminary
2365 switch on the second character of optional "-foo" lines. (This is
2366 overdue, caused by the large number of possibilities that now exist.
2367 Originally there were few.) While I was there, I also converted the
2368 str(n)cmp tests so they don't re-test the leading "-" and the first
2369 character, in the hope this might squeeze out yet more improvement.
2371 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2372 flag allowing group syntax was set by the header_syntax check but not
2373 turned off, possible causing trouble later; (2) The flag was not being
2374 set at all for the header_verify test, causing "group"-style headers to
2375 be rejected. I have now set it in this case, and also caused header_
2376 verify to ignore an empty address taken from a group. While doing this, I
2377 came across some other cases where the code for allowing group syntax
2378 while scanning a header line wasn't quite right (mostly, not resetting
2379 the flag correctly in the right place). These bugs could have caused
2380 trouble for malformed header lines. I hope it is now all correct.
2382 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2383 with the "reply" argument non-NULL. The code, however (which originally
2384 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2385 but it didn't always do it. This confused somebody who was copying the
2386 code for some other use. I have removed all the tests.
2388 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2389 feature that was used to support insecure browsers during the U.S. crypto
2390 embargo. It requires special client support, and Exim is probably the
2391 only MTA that supported it -- and would never use it because real RSA is
2392 always available. This code has been removed, because it had the bad
2393 effect of slowing Exim down by computing (never used) parameters for the
2394 RSA_EXPORT functionality.
2396 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2397 authenticator to fail if there's a tab character in the incoming data
2398 (there should never be unless someone is messing about, as it's supposed
2399 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2400 if the connection is using TLS or if the remote IP is the same as the
2401 local IP, and the "valid-client-cert option" if a client certificate has
2404 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2405 authenticators. This can be used for authorization after authentication
2406 succeeds. (In the case of plaintext, it servers for both authentication
2409 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2410 if any retry times were supplied.
2412 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2413 connection, where there is no client IP address to check. In this
2414 situation, the verify now always succeeds.
2416 PH/25 Applied John Jetmore's -Mset patch.
2418 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2420 PH/27 In a string expansion for a processed (not raw) header when multiple
2421 headers of the same name were present, leading whitespace was being
2422 removed from all of them, but trailing whitespace was being removed only
2423 from the last one. Now trailing whitespace is removed from each header
2424 before concatenation. Completely empty headers in a concatenation (as
2425 before) are ignored.
2427 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2428 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2430 PH/29 [Removed. This was a change that I later backed out, and forgot to
2431 correct the ChangeLog entry (that I had efficiently created) before
2432 committing the later change.]
2434 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2435 address errors (4xx response to RCPT) over the same connection as other
2436 messages routed to the same hosts. Such deliveries are always "forced",
2437 so retry times are not inspected. This resulted in far too many retries
2438 for the affected addresses. The effect occurred only when there were more
2439 hosts than the hosts_max_try setting in the smtp transport when it had
2440 the 4xx errors. Those hosts that it had tried were not added to the list
2441 of hosts for which the message was waiting, so if all were tried, there
2442 was no problem. Two fixes have been applied:
2444 (i) If there are any address or message errors in an SMTP delivery, none
2445 of the hosts (tried or untried) are now added to the list of hosts
2446 for which the message is waiting, so the message should not be a
2447 candidate for sending over the same connection that was used for a
2448 successful delivery of some other message. This seems entirely
2449 reasonable: after all the message is NOT "waiting for some host".
2450 This is so "obvious" that I'm not sure why it wasn't done
2451 previously. Hope I haven't missed anything, but it can't do any
2452 harm, as the worst effect is to miss an optimization.
2454 (ii) If, despite (i), such a delivery is accidentally attempted, the
2455 routing retry time is respected, so at least it doesn't keep
2456 hammering the server.
2458 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2459 in ${readsocket because some servers need this prod.
2461 PH/32 Added some extra debug output when updating a wait-xxx database.
2463 PH/33 The hint "could be header name not terminated by colon", which has been
2464 given for certain expansion errors for a long time, was not being given
2465 for the ${if def:h_colon_omitted{... case.
2467 PH/34 The spec says: "With one important exception, whenever a domain list is
2468 being scanned, $domain contains the subject domain." There was at least
2469 one case where this was not true.
2471 PH/35 The error "getsockname() failed: connection reset by peer" was being
2472 written to the panic log as well as the main log, but it isn't really
2473 panic-worthy as it just means the connection died rather early on. I have
2474 removed the panic log writing for the ECONNRESET error when getsockname()
2477 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2478 runs only) independently of the message's sender address. This meant
2479 that, if the 4xx error was in fact related to the sender, a different
2480 message to the same recipient with a different sender could confuse
2481 things. In particular, this can happen when sending to a greylisting
2482 server, but other circumstances could also provoke similar problems.
2483 I have changed the default so that the retry time for these errors is now
2484 based a combination of the sender and recipient addresses. This change
2485 can be overridden by setting address_retry_include_sender=false in the
2488 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2489 remote server are returned as part of bounce messages. This was not
2490 happening for LMTP over a pipe (the lmtp transport), but now it is the
2491 same for both kinds of LMTP.
2493 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2494 in header lines that were in fact CNAMEs. This is no longer the case.
2496 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2497 and queue runs started by the daemon processed all messages. This has
2498 been fixed so that -R and -S can now usefully be given with -q<time>.
2500 PH/40 Import PCRE release 6.7 (fixes some bugs).
2502 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2504 PH/42 Give an error if -q is specified more than once.
2506 PH/43 Renamed the variables $interface_address and $interface_port as
2507 $received_ip_address and $received_port, to make it clear that these
2508 values apply to message reception, and not to the outgoing interface when
2509 a message is delivered. (The old names remain recognized, of course.)
2511 PH/44 There was no timeout on the connect() call when using a Unix domain
2512 socket in the ${readsocket expansion. There now is.
2514 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2515 be meaningful with "accept".
2517 SC/01 Eximstats V1.43
2518 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2520 SC/02 Eximstats V1.44
2521 Use a glob alias rather than an array ref in the generated
2522 parser. This improves both readability and performance.
2524 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2525 Collect SpamAssassin and rejection statistics.
2526 Don't display local sender or destination tables unless
2527 there is data to show.
2528 Added average volumes into the top table text output.
2530 SC/04 Eximstats V1.46
2531 Collect data on the number of addresses (recipients)
2532 as well as the number of messages.
2534 SC/05 Eximstats V1.47
2535 Added 'Message too big' to the list of mail rejection
2536 reasons (thanks to Marco Gaiarin).
2538 SC/06 Eximstats V1.48
2539 Mainlog lines which have GMT offsets and are too short to
2540 have a flag are now skipped.
2542 SC/07 Eximstats V1.49 (Alain Williams)
2543 Added the -emptyok flag.
2545 SC/08 Eximstats V1.50
2546 Fixes for obtaining the IP address from reject messages.
2548 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2549 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2550 whitespace changes from 4.64-PH/27
2552 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2555 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2556 are found, allow negative numbers in numeric criteria)
2558 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2560 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2561 to match changes made in 4.64-PH/43
2563 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2565 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2566 "domain" is actually an IP address, and force a failure. This locks out
2567 those revolvers/nameservers that support "A-for-A" lookups, in
2568 contravention of the specifications.
2570 PH/48 When a host name was looked up from an IP address, and the subsequent
2571 forward lookup of the name timed out, the host name was left in
2572 $sender_host_name, contrary to the specification.
2574 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2575 restricted to single-key lookups, Exim was not diagnosing an error if
2576 * or *@ was used with a query-style lookup.
2578 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2580 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2581 long ago, but noone interested enough thought of it. Let's just say that
2582 the "1.1" means that there are some new functions that weren't there at
2583 some point in the past.
2585 PH/51 Error processing for expansion failure of helo_data from an smtp
2586 transport during callout processing was broken.
2588 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2589 tested/used via the -bh/-bhc/-bs options.
2591 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2592 bug, fixed in subsequent PCRE releases).
2594 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2595 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2597 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2603 SC/01 Use a glob alias rather than an array ref in eximstats generated
2604 parser. This improves both readability and performance.
2606 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2607 Don't display local sender or destination tables in eximstats unless
2608 there is data to show.
2609 Added average volumes into the eximstats top table text output.
2611 SC/03 Collect data on the number of addresses (recipients) as well
2612 as the number of messages in eximstats.
2614 TF/01 Correct an error in the documentation for the redirect router. Exim
2615 does (usually) call initgroups() when daemonizing.
2617 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2618 with consistent privilege compared to when running as a daemon.
2620 TF/03 Note in the spec that $authenticated_id is not set for local
2621 submissions from trusted users.
2623 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2624 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2626 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2627 by adding some example configuration directives to the default
2628 configuration file. A little bit of work is required to uncomment the
2629 directives and define how usernames and passwords are checked, but
2630 there is now a framework to start from.
2632 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2633 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2634 without this. I don't know how relevant this is to other LDAP libraries.
2636 PH/02 Add the verb name to the "unknown ACL verb" error.
2638 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2640 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2642 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2643 directory not expanded when it should be if an expanded home directory
2644 was set for the address (which is overridden by the transport).
2646 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2649 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2650 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2651 because it is too late at that time, and has no effect.
2653 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2654 security issue with \' (bugzilla #107). I could not use the
2655 PQescapeStringConn() function, because it needs a PGconn value as one of
2658 PH/08 When testing addresses using -bt, indicate those final addresses that
2659 are duplicates that would not cause an additional delivery. At least one
2660 person was confused, thinking that -bt output corresponded to deliveries.
2661 (Suppressing duplicates isn't a good idea as you lose the information
2662 about possibly different redirections that led to the duplicates.)
2664 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2665 systems where poll() doesn't work, in particular OS X.
2667 PH/10 Added more information to debugging output for retry time not reached.
2669 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2670 operations in malware.c.
2672 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2675 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2676 syslog_duplication set false, log lines that would normally be written
2677 both the the main log and to the reject log were not written to syslog at
2680 PH/14 In the default configuration, change the use of "message" in ACL warn
2681 statements to "add_header".
2683 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2684 not followed by a command (e.g. "seen endif").
2686 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2687 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2690 PH/17 Added extra conditions to the default value of delay_warning_condition
2694 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2695 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2696 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2699 The Auto-Submitted: and various List- headers are standardised, whereas I
2700 don't think Precedence: ever was.
2702 PH/18 Refactored debugging code in route_finduser() to show more information,
2703 in particular, the error code if getpwnam() issues one.
2705 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2706 This is apparently needed in addition to the PH/07 change above to avoid
2707 any possible encoding problems.
2709 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2710 but not after initializing Perl.
2712 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2713 output them only if debugging. By default they are written stderr,
2714 apparently, which is not desirable.
2716 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2719 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2722 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2724 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2725 authenticated or an ident call has been made. Suppress the default
2726 values for $authenticated_id and $authenticated_sender (but permit -oMai
2727 and -oMas) when testing with -bh.
2729 PH/24 Re-jigged the order of the tests in the default configuration so that the
2730 tests for valid domains and recipients precede the DNS black list and CSA
2731 tests, on the grounds that those ones are more expensive.
2733 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2734 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2735 This bug exists in every version of Exim that I still have, right back to
2738 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2739 However, an attempt to turn on case-sensitivity in a regex key by
2740 including (?-i) didn't work because the subject string was already
2741 lowercased, and the effects were non-intuitive. It turns out that a
2742 one-line patch can be used to allow (?-i) to work as expected.
2748 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2749 other effects) broke the use of negated acl sub-conditions.
2751 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2754 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2755 "Deny" causes Exim to reject the incoming connection with a 554 error.
2756 Unfortunately, if there is a major crisis, such as a disk failure,
2757 tcp-wrappers gives "deny", whereas what one would like would be some
2758 kind of temporary error. A kludge has been added to help with this.
2759 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2760 554 error is used if errno is still zero or contains ENOENT (which occurs
2761 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2764 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2766 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2767 errors. Otherwise a message that provokes a temporary error (when other
2768 messages do not) can cause a whole host to time out.
2770 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2771 addresses were routed directly to files or pipes from a redirect router.
2772 File deliveries just didn't batch; pipe deliveries might have suffered
2775 PH/06 A failure to get a lock for a hints database would erroneously always say
2776 "Failed to get write lock", even when it was really a read lock.
2778 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2779 of 0600. This has been changed to use the value of the lockfile_mode
2780 option (which defaults to 0600).
2782 PH/08 Applied small patch from the Sieve maintainer.
2784 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2785 folder from quota calculations, a direct delivery into this folder messed
2786 up the contents of the maildirsize file. This was because the regex was
2787 used only to exclude .Trash (or whatever) when the size of the mailbox
2788 was calculated. There was no check that a delivery was happening into an
2789 excluded directory. This bug has been fixed by ignoring all quota
2790 processing for deliveries into excluded directories.
2792 PH/10 Added the maildirfolder_create_regex option to appendfile.
2798 PH/01 The code for finding all the local interface addresses on a FreeBSD
2799 system running IPv6 was broken. This may well have applied to all BSD
2800 systems, as well as to others that have similar system calls. The broken
2801 code found IPv4 interfaces correctly, but gave incorrect values for the
2802 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2803 that it would not match correctly against @[] and not recognize the IPv6
2806 PH/02 The ipliteral router was not recognizing addresses of the form user@
2807 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2809 PH/03 Added disable_ipv6.
2811 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2812 decoded form, because it is most often used to construct To: headers
2813 lines in autoreplies, and the decoded form may well be syntactically
2814 invalid. However, $reply_address has leading white space removed, and all
2815 newlines turned into spaces so that the autoreply transport does not
2818 PH/05 If group was specified without a user on a router, and no group or user
2819 was specified on a transport, the group from the router was ignored.
2821 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2822 for visible compile-time settings that can be used to change these
2823 numbers, for those that want even more. Backwards compatibility with old
2824 spool files has been maintained. However, going back to a previous Exim
2825 release will lost any variables that are in spool files.
2827 PH/07 Two small changes when running in the test harness: increase delay when
2828 passing a TCP/IP connection to a new process, in case the original
2829 process has to generate a bounce, and remove special handling of
2830 127.0.0.2 (sic), which is no longer necessary.
2832 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2833 be the same on different OS.
2835 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2838 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2839 whether --show-vars was specified or not
2841 JJ/02 exipick: Added support for new ACL variable spool format introduced
2844 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2845 syntactically invalid From: or Reply-to: line, and a filter used this to
2846 generate an autoreply, and therefore failed to obtain an address for the
2847 autoreply, Exim could try to deliver to a non-existent relative file
2848 name, causing unrelated and misleading errors. What now happens is that
2849 it logs this as a hard delivery error, but does not attempt to create a
2852 PH/11 The exinext utility has a -C option for testing purposes, but although
2853 the given file was scanned by exinext itself; it wasn't being passed on
2854 when Exim was called.
2856 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2857 an end-of-file indication when reading a command response.
2859 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2860 compiled. In many other places in Exim, IPv6 addresses are always
2861 recognized, so I have changed this. It also means that IPv4 domain
2862 literals of the form [IPV4:n.n.n.n] are now always recognized.
2864 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2865 used if the router is not running as root, for example, when verifying at
2866 ACL time, or when using -bh. The debugging output from this situation was
2867 non-existent - all you got was a failure to exec. I have made two
2870 (a) Failures to set uid/gid, the current directory, or a process leader
2871 in a subprocess such as that created by queryprogram now generate
2872 suitable debugging output when -d is set.
2874 (b) The queryprogram router detects when it is not running as root,
2875 outputs suitable debugging information if -d is set, and then runs
2876 the subprocess without attempting to change uid/gid.
2878 PH/15 Minor change to Makefile for building test_host (undocumented testing
2881 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2882 additional section of a DNS packet that returns MX or SRV records.
2883 Instead, it always explicitly searches for A/AAAA records. This avoids
2884 major problems that occur when a DNS server includes only records of one
2885 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2886 fixed another bug: if SRV records were looked up and the corresponding
2887 address records were *not* found in the additional section, the port
2888 values from the SRV records were lost.
2890 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2891 using the correct key (the original address) when searching the retry
2892 rules in order to find which one to use for generating the retry hint.
2894 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2895 adding the default one. Similarly, if it contains a Reply-To: header, the
2896 errors_reply_to option, if set, is not used.
2898 PH/19 When calculating a retry time, Exim used to measure the "time since
2899 failure" by looking at the "first failed" field in the retry record. Now
2900 it does not use this if it is later than than the arrival time of the
2901 message. Instead it uses the arrival time. This makes for better
2902 behaviour in cases where some deliveries succeed, thus re-setting the
2903 "first failed" field. An example is a quota failure for a huge message
2904 when small messages continue to be delivered. Without this change, the
2905 "time since failure" will always be short, possible causing more frequent
2906 delivery attempts for the huge message than are intended.
2907 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2909 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2910 $1, $2, $3) because the numerical variables can be reset during some
2911 expansion items (e.g. "match"), thereby losing the authentication data.
2913 PH/21 Make -bV show the size of off_t variables so that the test suite can
2914 decide whether to run tests for quotas > 2G.
2916 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2917 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2918 filecount value is greater than 2G or if a quota value is greater than 2G
2919 on a system where the size of off_t is not greater than 4, a panic error
2922 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2923 never match. The debug and -bh output now contains an explicit error
2924 message indicating a malformed IPv4 address or mask.
2926 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2927 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2928 PH/23 above applies.
2930 PH/25 Do not write to syslog when running in the test harness. The only
2931 occasion when this arises is a failure to open the main or panic logs
2932 (for which there is an explicit test).
2934 PH/26 Added the /no_tell option to "control=freeze".
2936 PH/27 If a host name lookup failed very early in a connection, for example, if
2937 the IP address matched host_lookup and the reverse lookup yielded a name
2938 that did not have a forward lookup, an error message of the form "no IP
2939 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2940 could be logged. Now it outputs the IP address instead of "NULL".
2942 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2943 allows the sender and the authenticated sender to be set when
2944 submitting a message from within Exim. Since child_open_exim() is
2945 documented for local_scan(), the new function should be too.
2947 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2948 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2949 results in an empty string is now treated as unset.
2951 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2953 PH/31 Added sender_verify_fail logging option.
2955 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2956 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2957 tidied the source and removed it altogether.
2959 PH/33 When a queue run was abandoned because the load average was too high, a
2960 log line was always written; now it is written only if the queue_run log
2961 selector is set. In addition, the log line for abandonment now contains
2962 information about the queue run such as the pid. This is always present
2963 in "start" and "stop" lines but was omitted from the "abandon" line.
2965 PH/34 Omit spaces between a header name and the colon in the error message that
2966 is given when verify = headers_syntax fails (if there are lots of them,
2967 the message gets confusing).
2969 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2970 names, as there are now some PTR records that contain slashes. This check
2971 is only to protect against broken name servers that fall over on strange
2972 characters, so the fact that it applies to all lookups doesn't matter.
2974 PH/36 Now that the new test suite is complete, we can remove some of the
2975 special code in Exim that was needed for the old test suite. For example,
2976 sorting DNS records because real resolvers return them in an arbitrary
2977 order. The new test suite's fake resolver always returns records in the
2980 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2981 bounces) except when queue_only is set, to avoid logging races between
2982 the different processes.
2984 PH/38 Panic-die if .include specifies a non-absolute path.
2986 PH/39 A tweak to the "H" retry rule from its user.
2988 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2989 a label. They prevented compilation on older perls.
2991 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2992 a warning to be raised on newish perls.
2994 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2995 on queue. Changes to match documented behaviour of showing count of
2996 messages matching specified criteria.
2998 PH/40 Changed the default ident timeout from 30s to 5s.
3000 PH/41 Added support for the use of login_cap features, on those BSD systems
3001 that have them, for controlling the resources used by pipe deliveries.
3003 PH/42 The content-scanning code uses fopen() to create files in which to put
3004 message data. Previously it was not paying any attention to the mode of
3005 the files. Exim runs with umask(0) because the rest of the code creates
3006 files with open(), and sets the required mode explicitly. Thus, these
3007 files were ending up world-writeable. This was not a big issue, because,
3008 being within the spool directory, they were not world-accessible. I have
3009 created a function called modefopen, which takes an additional mode
3010 argument. It sets umask(777), creates the file, chmods it to the required
3011 mode, then resets the umask. All the relevant calls to fopen() in the
3012 content scanning code have been changed to use this function.
3014 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3015 to 24 hours. This avoids potential overflow problems when processing G
3016 and H retry rules. I suspect nobody ever tinkers with this value.
3018 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3020 PH/45 When the plaintext authenticator is running as a client, the server's
3021 challenges are checked to ensure they are valid base64 strings. By
3022 default, the authentication attempt is cancelled if an invalid string is
3023 received. Setting client_ignore_invalid_base64 true ignores these errors.
3024 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3025 they are received. Thus, the responses can be made to depend on the
3026 challenges. If an invalid string is ignored, an empty string is placed in
3029 PH/46 Messages that are created by the autoreply transport now contains a
3030 References: header, in accordance with RFCs 2822 and 3834.
3032 PH/47 Added authenticated_sender_force to the smtp transport.
3034 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3036 PH/49 Installed latest patch from the Sieve maintainer.
3038 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3039 was also set, the appendfile transport was unnecessarily scanning a
3040 directory of message files (e.g. for maildir delivery) to find the count
3041 of files (along with the size), even though it did not need this
3042 information. It now does the scan only if it needs to find either the
3043 size of the count of files.
3045 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3047 PH/52 Two bugs concerned with error handling when the smtp transport is
3050 (i) Exim was not creating retry information for temporary errors given
3051 for individual recipients after the DATA command when the smtp transport
3052 was used in LMTP mode. This meant that they could be retried too
3053 frequently, and not timed out correctly.
3055 (ii) Exim was setting the flag that allows error details to be returned
3056 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3057 recipients that were returned after the DATA command.
3059 PH/53 This is related to PH/52, but is more general: for any failing address,
3060 when detailed error information was permitted to be returned to the
3061 sender, but the error was temporary, then after the final timeout, only
3062 "retry timeout exceeded" was returned. Now it returns the full error as
3063 well as "retry timeout exceeded".
3065 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3066 do this, and (what is worse) MTAs that accept it.
3068 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3069 will now be deprecated.
3071 PH/56 New os.c-cygwin from the Cygwin maintainer.
3073 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3074 formats (previously only available in exim formats via -bpr, -bpru,
3075 and -bpra. Now also available in native and exiqgrep formats)
3077 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3078 with very large, slow to parse queues
3080 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3082 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3084 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3085 responses to authentication challenges, though it was showing the
3086 challenges; (ii) I've removed the CR characters from the debug output for
3089 PH/58 Allow for the insertion of a newline as well as a space when a string
3090 is turned into more than one encoded-word during RFC 2047 encoding. The
3091 Sieve code now uses this.
3093 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3094 data_4xx, lost_connection, tls_required.
3096 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3097 message was being sent as an SMTP response.
3099 PH/61 Add -l and -k options to exicyclog.
3101 PH/62 When verifying, if an address was redirected to one new address, so that
3102 verification continued, and the new address failed or deferred after
3103 having set something in $address_data, the value of $address_data was not
3104 passed back to the ACL. This was different to the case when no
3105 redirection occurred. The value is now passed back in both cases.
3107 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3108 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3109 use login_cap.h, so on its own it isn't the distinguishing feature. The
3110 new name refers directly to the setclassresources() function.
3112 PH/65 Added configuration files for NetBSD3.
3114 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3116 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3117 is preferred over IPv4.
3119 PH/68 The bounce_return_message and bounce_return_body options were not being
3120 honoured for bounces generated during the reception of non-SMTP messages.
3121 In particular, this applied to messages rejected by the ACL. This bug has
3122 been fixed. However, if bounce_return_message is true and bounce_return_
3123 body is false, the headers that are returned for a non-SMTP message
3124 include only those that have been read before the error was detected.
3125 (In the case of an ACL rejection, they have all been read.)
3127 PH/69 The HTML version of the specification is now built in a directory called
3128 spec_html instead of spec.html, because the latter looks like a path with
3129 a MIME-type, and this confuses some software.
3131 PH/70 Catch two compiler warnings in sieve.c.
3133 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3134 function verify_get_ident() calls ip_connect() to connect a socket, but
3135 if the "connect()" function timed out, ip_connect() used to close the
3136 socket. However, verify_get_ident() also closes the socket later, and in
3137 between Exim writes to the log, which may get opened at this point. When
3138 the socket was closed in ip_connect(), the log could get the same file
3139 descriptor number as the socket. This naturally causes chaos. The fix is
3140 not to close the socket in ip_connect(); the socket should be closed by
3141 the function that creates it. There was only one place in the code where
3142 this was missing, in the iplookup router, which I don't think anybody now
3143 uses, but I've fixed it anyway.
3145 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3146 well as to direct DNS lookups. Otherwise the handling of names in host
3147 lists is inconsistent and therefore confusing.
3153 PH/01 Two changes to the default runtime configuration:
3155 (1) Move the checks for relay_from_hosts and authenticated clients from
3156 after to before the (commented out) DNS black list checks.
3158 (2) Add control=submission to the relay_from_hosts and authenticated
3159 clients checks, on the grounds that messages accepted by these
3160 statements are most likely to be submissions.
3162 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3164 (1) Generate an error if the third argument for the ${prvs expansion is
3167 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3170 (3) Reset the variables that are obtained from the first argument of
3171 ${prvscheck and used in the second argument before leaving the code,
3172 because their memory is reclaimed, so using them afterwards may do
3175 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3176 one (it's much easier than Tom thought :-).
3178 (5) Because of (4), we can now allow for the use of $prvscheck_result
3179 inside the third argument.
3181 PH/03 For some reason, the default setting of PATH when running a command from
3182 a pipe transport was just "/usr/bin". I have changed it to
3185 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3186 anything to be listed in the output from -bV.
3188 PH/05 When a filter generated an autoreply, the entire To: header line was
3189 quoted in the delivery log line, like this:
3191 => >A.N.Other <ano@some.domain> <original@ddress> ...
3193 This has been changed so that it extracts the operative address. There
3194 may be more than one such address. If so, they are comma-separated, like
3197 => >ano@some.domain,ona@other.domain <original@ddress> ...
3199 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3200 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3201 not being looked up in the rDNS to get a host name, Exim was showing the
3202 IP address twice in Received: lines, even though the IP addresses were
3203 identical. For example:
3205 Received: from [1.2.3.4] (helo=[1.2.3.4])
3207 However, if the real host name was known, it was omitting the HELO data
3208 if it matched the actual IP address. This has been tidied up so that it
3209 doesn't show the same IP address twice.
3211 PH/07 When both +timestamp and +memory debugging was on, the value given by
3212 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3213 called by the debug printing, thereby overwriting the timestamp buffer.
3214 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3216 PH/08 When the original message was included in an autoreply transport, it
3217 always said "this is a copy of the message, including all the headers",
3218 even if body_only or headers_only was set. It now gives an appropriate
3221 PH/09 Applied a patch from the Sieve maintainer which:
3223 o fixes some comments
3224 o adds the (disabled) notify extension core
3225 o adds some debug output for the result of if/elsif tests
3226 o points to the current vacation draft in the documentation
3227 and documents the missing references header update
3231 o fixes a bug in processing the envelope test (when testing
3232 multiple envelope elements, the last element determined the
3235 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3236 Electronic Mail") by including:
3238 Auto-submitted: auto-generated
3240 in the messages that it generates (bounce messages and others, such as
3241 warnings). In the case of bounce messages for non-SMTP messages, there was
3242 also a typo: it was using "Auto_submitted" (underscore instead of
3243 hyphen). Since every message generated by Exim is necessarily in response
3244 to another message, thes have all been changed to:
3246 Auto-Submitted: auto-replied
3248 in accordance with these statements in the RFC:
3250 The auto-replied keyword:
3252 - SHOULD be used on messages sent in direct response to another
3253 message by an automatic process,
3255 - MUST NOT be used on manually-generated messages,
3257 - MAY be used on Delivery Status Notifications (DSNs) and Message
3258 Disposition Notifications (MDNs),
3260 - MUST NOT be used on messages generated by automatic or periodic
3261 processes, except for messages which are automatic responses to
3264 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3265 to the default Received: header definition.
3267 PH/12 Added log selector acl_warn_skipped (default on).
3269 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3270 variables because (a) they are in the wrong storage pool and (b) even if
3271 they were copied, it wouldn't work properly because of the caching.
3273 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3274 checking when decoding. Apparently there are clients that generate
3275 overlong encoded strings. Why am I not surprised?
3277 PH/15 If the first argument of "${if match_address" was not empty, but did not
3278 contain an "@" character, Exim crashed. Now it writes a panic log message
3279 and treats the condition as false.
3281 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3283 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3284 extension "envelope-auth". The code is finished and in agreement with
3285 other implementations, but there is no documentation so far and in fact,
3286 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3287 not changing the active code.
3289 Print executed "if" and "elsif" statements when debugging is used. This
3290 helps a great deal to understand what a filter does.
3292 Document more things not specified clearly in RFC3028. I had all this
3293 sorted out, when out of a sudden new issues came to my mind. Oops."
3295 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3298 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3299 canonical form (as documented). However, after a host name lookup from
3300 the IP address, check_host() was doing a simple string comparison with
3301 addresses acquired from the DNS when checking that the found name did
3302 have the original IP as one of its addresses. Since any found IPv6
3303 addresses are likely to be in abbreviated form, the comparison could
3304 fail. Luckily, there already exists a function for doing the comparison
3305 by converting both addresses to binary, so now that is used instead of
3306 the text comparison.
3308 PH/20 There was another similar case to PH/19, when a complete host name was
3309 given in a host list; looking up its IP address could give an abbreviated
3310 form, whereas the current host's name might or might not be abbreviated.
3311 The same fix has been applied.
3317 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3318 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3321 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3322 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3324 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3326 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3327 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3328 Also, for "Auto-Submitted", treat anything other than "no" as
3329 non-personal, in accordance with RFC 3834. (Previously it treated
3330 anything starting "auto-" as non-personal.)
3332 TF/01 The control=submission/name=... option had a problem with syntax
3333 errors if the name included a slash character. The /name= option
3334 now slurps the rest of the string, so it can include any characters
3335 but it must come last in the list of options (after /sender_retain
3338 PH/05 Some modifications to the interface to the fake nameserver for the new
3346 TK/01 Added the "success_on_redirect" address verification option. See
3347 NewStuff for rationale and an example.
3349 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3351 PH/02 Patch to exigrep to allow it to work on syslog lines.
3353 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3354 fread() to skip over the body file's header line, because in Cygwin the
3355 header line is locked and is inaccessible.
3357 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3358 co-exist for some time) to make it clear that it is the Exim ID that is
3359 referenced, not the Message-ID: header line.
3361 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3362 string_format() function, because snprintf() does not exist on all
3365 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3366 ${stat: expansion item.
3368 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3369 protocol synchronization error", to keep the pedants happy.
3371 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3372 well as for IRIX systems, when gcc is being used. See the host.c source
3375 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3377 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3380 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3381 transport and to the smtp transport in LMTP mode.
3383 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3385 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3386 run a filter in a subprocess. This could lead to confusion in subsequent
3387 lookups in the parent process. There should also be a search_tidyup() at
3388 the end of the subprocess.
3390 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3391 only if the host matched helo_try_verify_hosts, which caused the
3392 verification to occur when the EHLO/HELO command was issued. The ACL just
3393 tested the remembered result. Now, if a previous verification attempt has
3394 not happened, "verify = helo" does it there and then.
3396 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3398 TK/03 Fix log output including CR from clamd.
3400 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3401 exist provoked a memory error which could cause a segfault.
3403 PH/15 Installed PCRE 6.2
3405 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3407 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3408 of the problem. Specifically, suggested +O2 rather than +O1 for the
3411 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3413 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3414 record, the port was not correctly logged when the outgoing_port log
3415 selector was set (it logged the transort's default port).
3417 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3418 fallback_hosts, and "hosts" in the smtp transport.
3420 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3421 host errors such as "Connection refused".
3423 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3424 authentication with radiusclient 0.4.9:
3426 - Error returned from rc_read_config was caught wrongly
3427 - Username/password not passed on to radius server due to wrong length.
3429 The presumption is that some radiusclient API changes for 4.51/PH/17
3430 were not taken care of correctly. The code is still untested by me (my
3431 Linux distribution still has 0.3.2 of radiusclient), but it was
3432 contributed by a Radius user.
3434 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3435 expanding the "port" option of the smtp transport.
3437 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3438 while reading a MIME header. Thanks to Tom Hughes for a patch.
3440 PH/24 Include config.h inside local_scan.h so that configuration settings are
3443 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3444 that in an ACL for RCPT (for example), you can examine exactly what was
3447 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3448 commands, but it was not correctly comparing the address with the actual
3449 client host address. Thus, it would show the EHLO address in Received:
3450 header lines when this was not necessary.
3452 PH/27 Added the % operator to ${eval:}.
3454 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3455 it should be ignoring failures (because with -C, for example, it has lost
3456 privilege). It wasn't ignoring creation failures other than "already
3459 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3462 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3463 because an input line was too long, either on its own, or by virtue of
3464 too many continuations, the temporary file was not being removed, and the
3465 return code was incorrect.
3467 PH/31 Missing "BOOL" in function definition in filtertest.c.
3469 PH/32 Applied Sieve patches from the maintainer.
3471 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3473 PH/33 Added "verify = not_blind".
3475 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3476 Local/Makefile (with some defaults set). These are used in built scripts
3477 such as exicyclog, but they have never been used in the exim_install
3478 script (though there are many overriding facilities there). I have
3479 arranged that the exim_install script now takes note of these two
3482 PH/35 Installed configuration files for Dragonfly.
3484 PH/36 When a locally submitted message by a trusted user did not contain a
3485 From: header, and the sender address was obtained from -f or from an SMTP
3486 MAIL command, and the trusted user did not use -F to supply a sender
3487 name, $originator_name was incorrectly used when constructing a From:
3488 header. Furthermore, $originator_name was used for submission mode
3489 messages from external hosts without From: headers in a similar way,
3490 which is clearly wrong.
3492 PH/37 Added control=suppress_local_fixups.
3494 PH/38 When log_selector = +received_sender was set, and the addition of the
3495 sender made the log line's construction buffer exactly full, or one byte
3496 less than full, an overflow happened when the terminating "\n" was
3499 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3500 when the result of a list match is failure because a DNS lookup failed.
3502 PH/40 RM_COMMAND is now used in the building process.
3504 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3505 the "build-* directories that it finds.
3507 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3508 address in a domain literal was a prefix of an interface address.
3510 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3511 when verifying a sender address, unless rewrite_headers is false.
3513 PH/44 Wrote a long comment about why errors_to addresses are verified as
3514 recipients, not senders.
3516 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3517 the ratelimit ACL was added.
3519 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3521 PH/46 Added extra information about PostgreSQL errors to the error string.
3523 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3524 avoiding the need to install special zones in a real server. This is
3525 backwards compatible; if it can't find the fake resolver, it drops back.
3526 Thus, both old and new test suites can be run.
3528 TF/02 Added util/ratelimit.pl
3530 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3531 clock is set back in time.
3533 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3534 Candler <B.Candler@pobox.com>.
3536 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3537 OK for addresses that are the result of redirections.
3539 PH/48 A number of further additions for the benefit of the new test suite,
3540 including a fake gethostbyname() that interfaces to the fake DNS resolver
3543 TF/06 The fix for widen_domains has also been applied to qualify_single and
3544 search_parents which are the other dnslookup options that can cause
3547 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3550 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3552 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3553 service" if it encountered an MX record with an empty target hostname.
3554 The message is now "an MX or SRV record indicated no SMTP service".
3556 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3557 if the DNS of the sending site is misconfigured. This is quite a
3558 common situation. This change restores the behaviour of treating a
3559 helo verification defer as a failure.
3561 PH/51 If self=fail was set on a router, the bounce message did not include the
3562 actual error message.
3568 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3570 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3571 command itself ended in error, the underlying message about the transport
3572 filter timeout was being overwritten with the pipe command error. Now the
3573 underlying error message should be appended to the second error message.
3575 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3577 PH/02 Reduce the amount of output that "make" produces by default. Full output
3578 can still be requested.
3580 PH/03 The warning log line about a condition test deferring for a "warn" verb
3581 was being output only once per connection, rather than after each
3582 occurrence (because it was using the same function as for successful
3583 "warn" verbs). This seems wrong, so I have changed it.
3585 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3586 it should not have, which might have caused a crash in the right
3587 circumstances, but probably never did.
3589 PH/04 Installed a modified version of Tony Finch's patch to make submission
3590 mode fix the return path as well as the Sender: header line, and to
3591 add a /name= option so that you can make the user's friendly name appear
3594 TF/03 Added the control = fakedefer ACL modifier.
3596 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3597 Mark Lowes for thorough testing.
3599 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3601 TK/03 Merged latest SRS patch from Miles Wilton.
3603 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3604 with the definition in sysexits.h (which is #included earlier).
3605 Fortunately, Exim does not actually use EX_OK. The code used to try to
3606 preserve the sysexits.h value, by assuming that macro definitions were
3607 scanned for macro replacements. I have been disabused of this notion,
3608 so now the code just undefines EX_OK before #including unistd.h.
3610 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3611 in the smtp transport. When a block could not be written in a single
3612 write() function, the timeout was being re-applied to each part-write.
3613 This seems wrong - if the receiver was accepting one byte at a time it
3614 would take for ever. The timeout is now adjusted when this happens. It
3615 doesn't have to be particularly precise.
3617 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3618 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3620 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3621 without a domain if the check to <postmaster@domain> fails.
3623 SC/01 Eximstats: added -xls and the ability to specify output files
3624 (patch written by Frank Heydlauf).
3626 SC/02 Eximstats: use FileHandles for outputting results.
3628 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3630 SC/04 Eximstats: fixed display of large numbers with -nvr option
3632 SC/05 Eximstats: fixed merging of reports with empty tables.
3634 SC/06 Eximstats: added the -include_original_destination flag
3636 SC/07 Eximstats: removed tabs and trailing whitespace.
3638 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3640 TK/06 MBOX spool code: Add real "From " MBOX separator line
3641 so the .eml file is really in mbox format (even though
3642 most programs do not really care). Patch from Alex Miller.
3644 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3645 The latter is generated from $received_to and is only set if the
3646 message has one envelope recipient. SA can use these headers,
3647 obviously out-of-the-box. Patch from Alex Miller.
3649 PH/08 The ${def test on a variable was returning false if the variable's
3650 value was "0", contrary to what the specification has always said!
3651 The result should be true unless the variable is empty.
3653 PH/09 The syntax error of a character other than { following "${if
3654 def:variable_name" (after optional whitespace) was not being diagnosed.
3655 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3656 accidental colon was present, for example, could give incorrect results.
3658 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3659 result is used (not including appendfile, where other changes are about
3662 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3663 This involved changing a lot of size variables from int to off_t. It
3664 should work with maildirs and everything.
3666 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3667 spamd dying while we are connected to it.
3669 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3672 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3673 function for BDB 4.3.
3675 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3677 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3678 the sources for PCRE have been re-arranged and more files are now
3681 PH/15 The code I had for printing potentially long long variables in PH/11
3682 above was not the best (it lost precision). The length of off_t variables
3683 is now inspected at build time, and an appropriate printing format (%ld
3684 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3685 to be "long long int" or "long int". This is needed for the internal
3686 formatting function string_vformat().
3688 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3689 the configuration file to be ":syslog", then the script "guesses" where
3690 the logs files are, rather than using the compiled in default. In our
3691 case the guess is not the same as the compiled default, so the script
3692 suddenly stopped working when I started to use syslog. The patch checks
3693 to see if log_file_path is "". If so, it attempts to read it from exim
3694 with no configuration file to get the compiled in version, before it
3695 falls back to the previous guessing code."
3697 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3698 implementing BATV in an Exim configuration. See NewStuff for the gory
3701 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3702 Makefile that are specific to HP-UX.
3704 PH/18 If the "use_postmaster" option was set for a recipient callout together
3705 with the "random" option, the postmaster address was used as the MAIL
3706 FROM address for the random test, but not for the subsequent recipient
3707 test. It is now used for both.
3709 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3710 patch removes a few documentation additions to RFC 3028, because the
3711 latest draft now contains them. It adds the new en;ascii-case comparator
3712 and a new error check for 8bit text in MIME parts. Comparator and
3713 require names are now matched exactly. I enabled the subaddress
3714 extension, but it is not well tested yet (read: it works for me)."
3716 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3717 rework some of the code of TK/09 above to avoid the hardwired use of
3718 "%lld" and "long long". Replaced the call to snprintf() with a call to
3721 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3722 records point to non-existent hosts", "retry timeout exceeded", and
3723 "retry time not reached for any host after a long failure period".
3725 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3726 experimental DomainKeys support:
3728 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3729 (2) On an error such as an illegally used "control", the wrong name for
3730 the control was given.
3732 These problems did NOT occur unless DomainKeys support was compiled.
3734 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3736 PH/24 Added ${if match_ip condition.
3738 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3739 databases so that it will be absolutely obvious if a crash occurs in the
3740 DB library. This is a regular occurrence (often caused by mis-matched
3743 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3744 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3745 picked up on a user's system that detects such things. There doesn't seem
3746 to be a gcc warning option for this - only an attribute that has to be
3747 put on the function's prototype. It seems that in Fedora Core 4 they have
3748 set this on a number of new functions. No doubt there will be more in due
3751 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3752 specify a transport. However, if an address that was verified by such a
3753 router was the subject of a callout, Exim crashed because it tried to
3754 read the rcpt_include_affixes from the non-existent transport. Now it
3755 just assumes that the setting of that option is false. This bug was
3756 introduced by 4.51/PH/31.
3758 PH/28 Changed -d+all to exclude +memory, because that information is very
3759 rarely of interest, but it makes the output a lot bigger. People tend to
3760 do -d+all out of habit.
3762 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3763 code in os-type was giving problems when libc.so lives in lib64, like on
3766 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3767 aren't the modern standard, and it seems that some systems' include files
3768 don't always have them. Exim was already checking for some of the newer
3769 ones like T_AAAA, and defining it itself. I've added checks for all the
3770 record types that Exim uses.
3772 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3773 not automatically generating a new one, as it is supposed to. This
3774 prevented TLS from working. If the file did exist, but contained invalid
3775 data, a new version was generated, as expected. It was only the case of a
3776 non-existent file that was broken.
3778 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3779 with a change in libdomainkeys > 0.64.
3781 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3782 from DNS. If the selector record carries the flag, it now has
3783 precedence over the domain-wide flag.
3785 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3787 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3788 the use of a port name that isn't defined in /etc/services) occurred, the
3789 message was deferred as in a normal delivery, and thus remained on the
3790 spool, instead of being failed because of the mua_wrapper setting. This
3791 is now fixed, and I tidied up some of the mua_wrapper messages at the
3794 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3795 the messages in a hash of arrays rather than using individual hashes.
3796 This is a bit cleaner and results in dramatic memory savings, albeit
3797 at a slight CPU cost.
3799 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3800 as requested by Marc Sherman.
3802 SC/10 Eximstats: added histograms for user specified patterns as requested
3805 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3807 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3808 fopen() in the content-scanning modules that did not already have it.
3814 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3815 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3817 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3819 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3821 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3822 to test Sieve filters that use "vacation".
3824 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3825 that changes the way the GnuTLS parameters are stored in the cache file.
3826 The new format can be generated externally. For backward compatibility,
3827 if the data in the cache doesn't make sense, Exim assumes it has read an
3828 old-format file, and it generates new data and writes a new file. This
3829 means that you can't go back to an older release without removing the
3832 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3833 work if there are any delivery delays because "one_time" forces the
3834 parent to be marked "delivered", so its unseen clone is never tried
3835 again. For this reason, Exim now forbids the simultaneous setting of
3838 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3839 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3840 entry if you want to know the details. The fix, however, neglected to
3841 consider the case when local delivery batching is involved. The test for
3842 "previously delivered" was not happening when checking to see if an
3843 address could be batched with a previous (undelivered) one; under
3844 certain circumstances this could lead to multiple deliveries to the same
3847 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3848 in its include files, and this causes problems building Exim.
3850 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3851 header_syntax) but Exim was just ignoring anything given after a slash.
3852 In particular, this caused confusion with an attempt to use "verify =
3853 reverse_host_lookup/defer_ok". An error is now given when options are
3854 supplied for verify items that do not have them. (Maybe reverse_host_
3855 lookup should have a defer_ok option, but that's a different point.)
3857 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3858 defined by RFC 821) to 2048, because there were problems with some AUTH
3859 commands, and RFC 1869 says the size should be increased for extended
3860 SMTP commands that take arguments.
3862 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3865 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3866 "unknown" error; now it says that the functionality isn't in the binary.
3868 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3869 an address' error message when a string expansion fails (syntax or
3870 whatever). Otherwise the password may appear in the log. Following change
3871 PH/42 below, there is no longer a chance of it appearing in a bounce
3874 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3876 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3877 of its addresses was ever tried. (Bugzilla bug #2.)
3879 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3880 the result incorrectly in the debug output. (It correctly added a newline
3881 to what was transported.)
3883 TF/01 Added $received_time.
3885 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3886 commented out examples of how to interface to a virus scanner and to
3887 SpamAssassin. Also added commented examples of av_scanner and
3888 spamd_address settings.
3890 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3891 and controls are allowed in which ACLs. There were a couple of minor
3892 errors. Some of the entries in the conditions table (which is a table of
3893 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3894 negation of where the condition IS allowed.
3896 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3898 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3899 header file does not have a version number, so I've had to invent a new
3900 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3901 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3902 radiusclient), but it was contributed by a Radius user.
3904 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3905 files or directories, for OpenSSL.
3907 PH/19 When an Exim process that is running as root has to create an Exim log
3908 file, it does so in a subprocess that runs as exim:exim so as to get the
3909 ownership right at creation (otherwise, other Exim processes might see
3910 the file with the wrong ownership). There was no test for failure of this
3911 fork() call, which would lead to the process getting stuck as it waited
3912 for a non-existent subprocess. Forks do occasionally fail when resources
3913 run out. I reviewed all the other calls to fork(); they all seem to check
3916 PH/20 When checking for unexpected SMTP input at connect time (before writing
3917 the banner), Exim was not dealing correctly with a non-positive return
3918 from the read() function. If the client had disconnected by this time,
3919 the result was a log entry for a synchronization error with an empty
3920 string after "input=" when read() returned zero. If read() returned -1
3921 (an event I could not check), uninitialized data bytes were printed.
3922 There were reports of junk text (parts of files, etc) appearing after
3925 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3927 PH/22 Added support for macro redefinition, and (re)definition in between
3928 driver and ACL definitions.
3930 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3931 forgetting to use the resulting value; it was using the unexpanded value.
3933 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3934 hadn't been configured. The fix is from Juergen Kreileder, who
3935 understands it better than I do:
3937 "Here's what I see happening with three configured cyrus_sasl
3938 authenticators configured (plain, login, cram-md5):
3940 On startup auth_cyrus_sasl_init() gets called for each of these.
3941 This means three calls to sasl_listmech() without a specified mech_list.
3942 => SASL tests which mechs of all available mechs actually work
3943 => three warnings about OTP not working
3944 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3946 With the patch, sasl_listmech() also gets called three times. But now
3947 SASL's mech_list option is set to the server_mech specified in the the
3948 authenticator. Or in other words, the answer from sasl_listmech()
3949 gets limited to just the mech you're testing for (which is different
3951 => the return list contains just 'plain' or 'login', 'cram-md5' or
3952 nothing depending on the value of ob->server_mech.
3954 I've just tested the patch: Authentication still works fine,
3955 unavailable mechs specified in the exim configuration are still
3956 caught, and the auth.log warnings about OTP are gone."
3958 PH/25 When debugging is enabled, the contents of the command line are added
3959 to the debugging output, even when log_selector=+arguments is not
3962 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3963 answer is "GNU", and only if the return is "GNU/something" is the answer
3966 PH/27 $acl_verify_message is now set immediately after the failure of a
3967 verification in an ACL, and so is available in subsequent modifiers. In
3968 particular, the message can be preserved by coding like this:
3970 warn !verify = sender
3971 set acl_m0 = $acl_verify_message
3973 Previously, $acl_verify_message was set only while expanding "message"
3974 and "log_message" when a very denied access.
3976 PH/28 Modified OS/os.c-Linux with
3978 -#ifndef OS_LOAD_AVERAGE
3979 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3981 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3982 nomenclature these days.)
3984 PH/29 Installed patch from the Sieve maintainer that adds the options
3985 sieve_useraddress and sieve_subaddress to the redirect router.
3987 PH/30 In these circumstances:
3988 . Two addresses routed to the same list of hosts;
3989 . First host does not offer TLS;
3990 . First host accepts first address;
3991 . First host gives temporary error to second address;
3992 . Second host offers TLS and a TLS session is established;
3993 . Second host accepts second address.
3994 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3995 and peerdn, if requested) that were in fact used only for the second
3998 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3999 attention to any local part prefix or suffix that was matched by the
4000 router that accepted the address. It now behaves in the same way as it
4001 does for delivery: the affixes are removed from the local part unless
4002 rcpt_include_affixes is set on the transport.
4004 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4005 timeout during the DATA phase of an incoming message.
4007 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4008 applied a patch sanctioned by the Sieve maintainer.
4010 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4011 the uid or gid is negative. A case of a negative gid caused this to be
4012 noticed. The fix allows for either to be negative.
4014 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4015 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4018 PH/36 The change PH/12 above was broken. Fixed it.
4020 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4021 the grounds that routing the same address twice would always produce the
4022 same answer. This might have been true once, but it is certainly no
4023 longer true now. Routing a child address may depend on the previous
4024 routing that produced that child. Some complicated redirection strategies
4025 went wrong when messages had multiple recipients, and made Exim's
4026 behaviour dependent on the order in which the addresses were given.
4028 I have moved the duplicate checking until after the routing is complete.
4029 Exim scans the addresses that are assigned to local and remote
4030 transports, and removes any duplicates. This means that more work will be
4031 done, as duplicates will always all be routed, but duplicates are
4032 presumably rare, so I don't expect this is of any significance.
4034 For deliveries to pipes, files, and autoreplies, the duplicate checking
4035 still happens during the routing process, since they are not going to be
4038 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4039 It corrects a timeout issue with spamd. This is Ian's comment: "The
4040 background is that sometimes spamd either never reads data from a
4041 connection it has accepted, or it never writes response data. The exiscan
4042 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4043 blindly assumes that writes won't block so it may never time out."
4045 PH/39 Allow G after quota size as well as K and M.
4047 PH/40 The value set for $authenticated_id in an authenticator may not contain
4048 binary zeroes or newlines because the value is written to log lines and
4049 to spool files. There was no check on this. Now the value is run through
4050 the string_printing() function so that such characters are converted to
4051 printable escape sequences.
4053 PH/41 $message_linecount is a new variable that contains the total number of
4054 lines in the message. Compare $body_linecount, which is the count for the
4057 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4058 bounce and delay warning messages, except in certain special cases, which
4061 (a) An SMTP error message from a remote host;
4062 (b) A message specified in a :fail: redirection;
4063 (c) A message specified in a "fail" command in a system filter;
4064 (d) A message specified in a FAIL return from the queryprogram router;
4065 (e) A message specified by the cannot_route_message router option.
4067 In these cases only, Exim does include the error details in bounce and
4068 warning messages. There are also a few cases where bland messages such
4069 as "unrouteable address" or "local delivery error" are given.
4071 PH/43 $value is now also set for the "else" part of a ${run expansion.
4073 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4074 being worked on, but at least Exim now implements the latest version to
4077 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4078 process to complete was treated as a delivery failure, a timeout while
4079 writing the message to the pipe was logged, but erroneously treated as a
4080 successful delivery. Such timeouts include transport filter timeouts. For
4081 consistency with the overall process timeout, these timeouts are now
4082 treated as errors, giving rise to delivery failures by default. However,
4083 there is now a new Boolean option for the pipe transport called
4084 timeout_defer, which, if set TRUE, converts the failures into defers for
4085 both kinds of timeout. A transport filter timeout is now identified in
4088 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4089 systems where "make" and "gmake" are different, calling "gmake" at top
4090 level broke things. I've arranged for the value of $(MAKE) to be passed
4091 from the Makefile to this script so that it can call the same version of
4095 A note about Exim versions 4.44 and 4.50
4096 ----------------------------------------
4098 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4099 changes of various kinds. As a consequence, a big documentation update was
4100 needed. This delayed the release for rather longer than seemed good, especially
4101 in the light of a couple of (minor) security issues. Therefore, the changes
4102 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4103 release. So 4.44 and 4.50 are in effect two different branches that both start
4106 I have left the 4.50 change log unchanged; it contains all the changes since
4107 4.43. The change log for 4.44 is below; many of its items are identical to
4108 those for 4.50. This seems to be the most sensible way to preserve the
4109 historical information.
4115 1. Minor wording change to the doc/README.SIEVE file.
4117 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4118 computation of the current number of files was incorrect.
4120 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4121 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4124 4. Give more explanation in the error message when the command for a transport
4125 filter fails to execute.
4127 5. There are several places where Exim runs a non-Exim command in a
4128 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4129 was being done only for the command run by the queryprogram router. It is
4130 now done for all such subprocesses. The other cases are: ${run, transport
4131 filters, and the commands run by the lmtp and pipe transports.
4133 6. Added CONFIGURE_GROUP build-time option.
4135 7. Some older OS have a limit of 256 on the maximum number of file
4136 descriptors. Exim was using setrlimit() to set 1000 as a large value
4137 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4138 systems. I've change it so that if it can't get 1000, it tries for 256.
4140 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4141 was an oversight, and furthermore, ever since the addition of extra
4142 controls (e.g. 4.43/32), the checks on when to allow different forms of
4143 "control" were broken. There should now be diagnostics for all cases when a
4144 control that does not make sense is encountered.
4146 9. Added the /retain_sender option to "control=submission".
4148 10. $recipients is now available in the predata ACL (oversight).
4150 11. Tidy the search cache before the fork to do a delivery from a message
4151 received from the command line. Otherwise the child will trigger a lookup
4152 failure and thereby defer the delivery if it tries to use (for example) a
4153 cached ldap connection that the parent has called unbind on.
4155 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4156 of $address_data from the recipient verification was clobbered by the
4157 sender verification.
4159 13. The value of address_data from a sender verification is now available in
4160 $sender_address_data in subsequent conditions in the ACL statement.
4162 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4164 15. Added a new option "connect=<time>" to callout options, to set a different
4167 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4168 was its contents. (It was OK if the option was not defined at all.)
4170 17. A "Completed" log line is now written for messages that are removed from
4171 the spool by the -Mrm option.
4173 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4174 information about exactly what failed.
4176 19. Added -dd to debug only the daemon process.
4178 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4179 handles timeouts, both on the server side and network timeouts. Renamed the
4180 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4182 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4183 It is now set to "smtps".
4185 22. $host_address is now set to the target address during the checking of
4186 ignore_target_hosts.
4188 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4189 being passed; this would have caused $sender_host_name to have been used if
4190 matching the list had actually called for a host name (not very likely,
4191 since this list is usually IP addresses). A host name is now passed as
4194 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4195 code that specifies a non-restarting handler (typically sigaction() in
4196 modern systems) in an attempt to fix a rare and obscure crash bug.
4198 25. Narrowed the window for a race in the daemon that could cause it to ignore
4199 SIGCHLD signals. This is not a major problem, because they are used only to
4200 wake it up if nothing else does.
4202 26. A malformed maildirsize file could cause Exim to calculate negative values
4203 for the mailbox size or file count. Odd effects could occur as a result.
4204 The maildirsize information is now recalculated if the size or filecount
4207 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4208 support for a long time. Removed HAVE_SYS_VFS_H.
4210 28. Installed the latest version of exipick from John Jetmore.
4212 29. In an address list, if the pattern was not a regular expression, an empty
4213 subject address (from a bounce message) matched only if the pattern was an
4214 empty string. Non-empty patterns were not even tested. This was the wrong
4215 because it is perfectly reasonable to use an empty address as part of a
4216 database query. An empty address is now tested by patterns that are
4217 lookups. However, all the other forms of pattern expect the subject to
4218 contain a local part and a domain, and therefore, for them, an empty
4219 address still always fails if the pattern is not itself empty.
4221 30. Exim went into a mad DNS loop when attempting to do a callout where the
4222 host was specified on an smtp transport, and looking it up yielded more
4223 than one IP address.
4225 31. Re-factored the code for checking spool and log partition space into a
4226 function that finds that data and another that does the check. The former
4227 is then used to implement four new variables: $spool_space, $log_space,
4228 $spool_inodes, and $log_inodes.
4230 32. The RFC2047 encoding function was originally intended for short strings
4231 such as real names; it was not keeping to the 75-character limit for
4232 encoded words that the RFC imposes. It now respects the limit, and
4233 generates multiple encoded words if necessary. To be on the safe side, I
4234 have increased the buffer size for the ${rfc2047: expansion operator from
4237 33. It is now permitted to omit both strings after an "if" condition; if the
4238 condition is true, the result is "true". As before, when the second string
4239 is omitted, a false condition yields an empty string. This makes it less
4240 cumbersome to write custom ACL and router conditions.
4242 34. Failure to deliver a bounce message always caused it to be frozen, even if
4243 there was an errors_to setting on the router. The errors_to setting is now
4246 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4247 canonical form (fully expanded) before being placed in
4248 $sender_host_address.
4250 36. The table in the code that translates DNS record types into text (T_A to
4251 "A" for instance) was missing entries for NS and CNAME. It is just possible
4252 that this could have caused confusion if both these types were looked up
4253 for the same domain, because the text type is used as part of Exim's
4254 per-process caching. But the chance of anyone hitting this buglet seems
4257 37. The dnsdb lookup has been extended in a number of ways.
4259 (1) There is a new type, "zns", which walks up the domain tree until it
4260 finds some nameserver records. It should be used with care.
4262 (2) There is a new type, "mxh", which is like "mx" except that it returns
4263 just the host names, not the priorities.
4265 (3) It is now possible to give a list of domains (or IP addresses) to be
4266 looked up. The behaviour when one of the lookups defers can be
4267 controlled by a keyword.
4269 (4) It is now possible to specify the separator character for use when
4270 multiple records are returned.
4272 38. The dnslists ACL condition has been extended: it is now possible to supply
4273 a list of IP addresses and/or domains to be looked up in a particular DNS
4276 39. Added log_selector=+queue_time_overall.
4278 40. When running the queue in the test harness, wait just a tad after forking a
4279 delivery process, to get repeatability of debugging output.
4281 41. Include certificate and key file names in error message when GnuTLS fails
4282 to set them up, because the GnuTLS error message doesn't include the name
4283 of the failing file when there is a problem reading it.
4285 42. Allow both -bf and -bF in the same test run.
4287 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4289 44. The "Exiscan patch" is now merged into the mainline Exim source.
4291 45. Sometimes the final signoff response after QUIT could fail to get
4292 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4293 < 0 before doing a fflush(). This bug looks as though it goes back to the
4294 introduction of TLS in release 3.20, but "sometimes" must have been rare
4295 because the tests only now provoked it.
4297 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4298 (this can affect the format of dates).
4300 47. exim_tidydb, when checking for the continued existence of a message for
4301 which it has found a message-specific retry record, was not finding
4302 messages that were in split spool directories. Consequently, it was
4303 deleting retry records that should have stayed in existence.
4305 48. Steve fixed some bugs in eximstats.
4307 49. The SPA authentication driver was not abandoning authentication and moving
4308 on to the next authenticator when an expansion was forced to fail,
4309 contradicting the general specification for all authenticators. Instead it
4310 was generating a temporary error. It now behaves as specified.
4312 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4313 (the order specifies the preference for clients). The order is now AES256,
4314 AES128, 3DES, ARCFOUR128.
4316 51. Small patch to Sieve code - explicitly set From: when generating an
4319 52. Exim crashed if a remote delivery caused a very long error message to be
4320 recorded - for instance if somebody sent an entire SpamAssassin report back
4321 as a large number of 550 error lines. This bug was coincidentally fixed by
4322 increasing the size of one of Exim's internal buffers (big_buffer) that
4323 happened as part of the Exiscan merge. However, to be on the safe side, I
4324 have made the code more robust (and fixed the comments that describe what
4327 53. Now that there can be additional text after "Completed" in log lines (if
4328 the queue_time_overall log selector is set), a one-byte patch to exigrep
4329 was needed to allow it to recognize "Completed" as not the last thing in
4332 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4333 patch that reportedly fixes this has been added. I am not expert enough to
4334 create a test for it. This is what the patch creator wrote:
4336 "I found a little strange behaviour of ldap code when working with
4337 Windows 2003 AD Domain, where users was placed in more than one
4338 Organization Units. When I tried to give exim partial DN, the exit code
4339 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4340 But simultaneously result of request was absolutely normal ldap result,
4341 so I produce this patch..."
4343 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4344 REFERENCE, so I have modified the code to exclude the patch when that macro
4347 55. Some experimental protocols are using DNS PTR records for new purposes. The
4348 keys for these records are domain names, not reversed IP addresses. The
4349 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4350 leaves it alone. Component reversal etc. now happens only for IP addresses.
4353 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4355 57. Double the size of the debug message buffer (to 2048) so that more of very
4356 long debug lines gets shown.
4358 58. The exicyclog utility now does better if the number of log files to keep
4359 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4361 59. Two changes related to the smtp_active_hostname option:
4363 (1) $smtp_active_hostname is now available as a variable.
4364 (2) The default for smtp_banner uses $smtp_active_hostname instead
4365 of $primary_hostname.
4367 60. The host_aton() function is supposed to be passed a string that is known
4368 to be a valid IP address. However, in the case of IPv6 addresses, it was
4369 not checking this. This is a hostage to fortune. Exim now panics and dies
4370 if the condition is not met. A case was found where this could be provoked
4371 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4372 components; fortuitously, this particular loophole had already been fixed
4373 by change 4.50/55 above.
4375 If there are any other similar loopholes, the new check in host_aton()
4376 itself should stop them being exploited. The report I received stated that
4377 data on the command line could provoke the exploit when Exim was running as
4378 exim, but did not say which command line option was involved. All I could
4379 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4380 running as the user.
4383 61. There was a buffer overflow vulnerability in the SPA authentication code
4384 (which came originally from the Samba project). I have added a test to the
4385 spa_base64_to_bits() function which I hope fixes it.
4388 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4389 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4391 63. The daemon start-up calls getloadavg() while still root for those OS that
4392 need the first call to be done as root, but it missed one case: when
4393 deliver_queue_load_max is set with deliver_drop_privilege. This is
4394 necessary for the benefit of the queue runner, because there is no re-exec
4395 when deliver_drop_privilege is set.
4397 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4398 This has been fixed.
4400 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4401 was in use, was not putting the data itself into the right store pool;
4402 consequently, it could be overwritten for a subsequent message in the same
4403 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4406 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4408 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4409 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4410 the function were treating the return as a boolean value, which happened to
4411 work because 0=false and not-0=true, but is not correct code.
4413 68. The host_aton() function was not handling scoped IPv6 addresses (those
4414 with, for example, "%eth0" on the end) correctly.
4416 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4417 negated items (that is, ~something) in unsigned ints. Some compilers
4418 apparently mutter when there is no cast.
4420 70. If an address verification called from an ACL failed, and did not produce a
4421 user-specific message (i.e. there was only a "system" message), nothing was
4422 put in $acl_verify_message. In this situation, it now puts the system
4425 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4426 session; change 4.31/43 added the unwanted input to the log line - except
4427 that it did not do this in the start of session case. It now does.
4429 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4430 This is wrong and can cause the other end to generate a synchronization
4431 error if it is another Exim or anything else that does the synchronization
4432 check. A QUIT command is no longer sent after a timeout.
4434 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4435 during host lookups.
4437 74. The defer_ok option of callout verification was not working if it was used
4438 when verifying addresses in header lines, that is, for this case:
4440 verify = header_sender/callout=defer_ok
4442 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4443 those file descriptors could be used for SMTP connections. If anything
4444 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4445 could be sent to the SMTP client, causing chaos. The daemon now opens
4446 stdin, stdout, and stderr to /dev/null when it puts itself into the
4449 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4450 log by default. The user can override this with suitable Perl magic.
4452 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4453 the log message when discard triggers, was not working for the DATA ACL or
4454 for the non-SMTP ACL.
4456 78. Error message wording change in sieve.c.
4458 79. If smtp_accept_max_per_host was set, the number of connections could be
4459 restricted to fewer than expected, because the daemon was trying to set up
4460 a new connection before checking whether the processes handling previous
4461 connections had finished. The check for completed processes is now done
4462 earlier. On busy systems, this bug wouldn't be noticed because something
4463 else would have woken the daemon, and it would have reaped the completed
4466 80. If a message was submitted locally by a user whose login name contained one
4467 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4468 It caused a spool format error. I have fixed the spool reading code. A
4469 related problem was that the "from" clause in the Received: line became
4470 illegal because of the space(s). It is now covered by ${quote_local_part.
4472 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4475 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4477 83. Patch from Sieve maintainer to fix unterminated string problem in
4478 "vacation" handling.
4480 84. Some minor changes to the Linux configuration files to help with other
4481 OS variants using glibc.
4483 85. One more patch for Sieve to update vacation handling to latest spec.
4486 ----------------------------------------------------
4487 See the note above about the 4.44 and 4.50 releases.
4488 ----------------------------------------------------
4494 1. Change 4.43/35 introduced a bug that caused file counts to be
4495 incorrectly computed when quota_filecount was set in an appendfile
4498 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4499 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4502 3. Give more explanation in the error message when the command for a transport
4503 filter fails to execute.
4505 4. There are several places where Exim runs a non-Exim command in a
4506 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4507 was being done only for the command run by the queryprogram router. It is
4508 now done for all such subprocesses. The other cases are: ${run, transport
4509 filters, and the commands run by the lmtp and pipe transports.
4511 5. Some older OS have a limit of 256 on the maximum number of file
4512 descriptors. Exim was using setrlimit() to set 1000 as a large value
4513 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4514 systems. I've change it so that if it can't get 1000, it tries for 256.
4516 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4517 was an oversight, and furthermore, ever since the addition of extra
4518 controls (e.g. 4.43/32), the checks on when to allow different forms of
4519 "control" were broken. There should now be diagnostics for all cases when a
4520 control that does not make sense is encountered.
4522 7. $recipients is now available in the predata ACL (oversight).
4524 8. Tidy the search cache before the fork to do a delivery from a message
4525 received from the command line. Otherwise the child will trigger a lookup
4526 failure and thereby defer the delivery if it tries to use (for example) a
4527 cached ldap connection that the parent has called unbind on.
4529 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4530 of $address_data from the recipient verification was clobbered by the
4531 sender verification.
4533 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4534 was its contents. (It was OK if the option was not defined at all.)
4536 11. A "Completed" log line is now written for messages that are removed from
4537 the spool by the -Mrm option.
4539 12. $host_address is now set to the target address during the checking of
4540 ignore_target_hosts.
4542 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4543 being passed; this would have caused $sender_host_name to have been used if
4544 matching the list had actually called for a host name (not very likely,
4545 since this list is usually IP addresses). A host name is now passed as
4548 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4549 code that specifies a non-restarting handler (typically sigaction() in
4550 modern systems) in an attempt to fix a rare and obscure crash bug.
4552 15. Narrowed the window for a race in the daemon that could cause it to ignore
4553 SIGCHLD signals. This is not a major problem, because they are used only to
4554 wake it up if nothing else does.
4556 16. A malformed maildirsize file could cause Exim to calculate negative values
4557 for the mailbox size or file count. Odd effects could occur as a result.
4558 The maildirsize information is now recalculated if the size or filecount
4561 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4562 support for a long time. Removed HAVE_SYS_VFS_H.
4564 18. Updated exipick to current release from John Jetmore.
4566 19. Allow an empty sender to be matched against a lookup in an address list.
4567 Previously the only cases considered were a regular expression, or an
4570 20. Exim went into a mad DNS lookup loop when doing a callout where the
4571 host was specified on the transport, if the DNS lookup yielded more than
4574 21. The RFC2047 encoding function was originally intended for short strings
4575 such as real names; it was not keeping to the 75-character limit for
4576 encoded words that the RFC imposes. It now respects the limit, and
4577 generates multiple encoded words if necessary. To be on the safe side, I
4578 have increased the buffer size for the ${rfc2047: expansion operator from
4581 22. Failure to deliver a bounce message always caused it to be frozen, even if
4582 there was an errors_to setting on the router. The errors_to setting is now
4585 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4586 canonical form (fully expanded) before being placed in
4587 $sender_host_address.
4589 24. Updated eximstats to version 1.33
4591 25. Include certificate and key file names in error message when GnuTLS fails
4592 to set them up, because the GnuTLS error message doesn't include the name
4593 of the failing file when there is a problem reading it.
4595 26. Expand error message when OpenSSL has problems setting up cert/key files.
4598 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4599 (this can affect the format of dates).
4601 28. exim_tidydb, when checking for the continued existence of a message for
4602 which it has found a message-specific retry record, was not finding
4603 messages that were in split spool directories. Consequently, it was
4604 deleting retry records that should have stayed in existence.
4606 29. eximstats updated to version 1.35
4607 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4608 1.35 - bugfix such that pie charts by volume are generated correctly
4610 30. The SPA authentication driver was not abandoning authentication and moving
4611 on to the next authenticator when an expansion was forced to fail,
4612 contradicting the general specification for all authenticators. Instead it
4613 was generating a temporary error. It now behaves as specified.
4615 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4616 (the order specifies the preference for clients). The order is now AES256,
4617 AES128, 3DES, ARCFOUR128.
4619 31. Small patch to Sieve code - explicitly set From: when generating an
4622 32. Exim crashed if a remote delivery caused a very long error message to be
4623 recorded - for instance if somebody sent an entire SpamAssassin report back
4624 as a large number of 550 error lines. This bug was coincidentally fixed by
4625 increasing the size of one of Exim's internal buffers (big_buffer) that
4626 happened as part of the Exiscan merge. However, to be on the safe side, I
4627 have made the code more robust (and fixed the comments that describe what
4630 33. Some experimental protocols are using DNS PTR records for new purposes. The
4631 keys for these records are domain names, not reversed IP addresses. The
4632 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4633 leaves it alone. Component reversal etc. now happens only for IP addresses.
4636 34. The host_aton() function is supposed to be passed a string that is known
4637 to be a valid IP address. However, in the case of IPv6 addresses, it was
4638 not checking this. This is a hostage to fortune. Exim now panics and dies
4639 if the condition is not met. A case was found where this could be provoked
4640 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4641 components; fortuitously, this particular loophole had already been fixed
4642 by change 4.50/55 or 4.44/33 above.
4644 If there are any other similar loopholes, the new check in host_aton()
4645 itself should stop them being exploited. The report I received stated that
4646 data on the command line could provoke the exploit when Exim was running as
4647 exim, but did not say which command line option was involved. All I could
4648 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4649 running as the user.
4652 35. There was a buffer overflow vulnerability in the SPA authentication code
4653 (which came originally from the Samba project). I have added a test to the
4654 spa_base64_to_bits() function which I hope fixes it.
4657 36. The daemon start-up calls getloadavg() while still root for those OS that
4658 need the first call to be done as root, but it missed one case: when
4659 deliver_queue_load_max is set with deliver_drop_privilege. This is
4660 necessary for the benefit of the queue runner, because there is no re-exec
4661 when deliver_drop_privilege is set.
4663 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4664 was in use, was not putting the data itself into the right store pool;
4665 consequently, it could be overwritten for a subsequent message in the same
4666 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4669 38. Sometimes the final signoff response after QUIT could fail to get
4670 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4671 < 0 before doing a fflush(). This bug looks as though it goes back to the
4672 introduction of TLS in release 3.20, but "sometimes" must have been rare
4673 because the tests only now provoked it.
4679 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4680 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4681 Now it returns an integer. A number of calls were still expecting a T/F
4682 return. Fortuitously, in all cases, the tests worked in OK situations,
4683 which is the norm. However, things would have gone wrong on any write
4684 failures on the smtp file descriptor. This function is used when sending
4685 messages over SMTP and also when doing verify callouts.
4687 2. When Exim is called to do synchronous delivery of a locally submitted
4688 message (the -odf or -odi options), it no longer closes stderr before doing
4691 3. Implemented the mua_wrapper option.
4693 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4695 5. Implemented the functions header_remove(), header_testname(),
4696 header_add_at_position(), and receive_remove_recipient(), and exported them
4699 6. If an ACL "warn" statement specified the addition of headers, Exim already
4700 inserted X-ACL-Warn: at the start if there was no header name. However, it
4701 was not making this test for the second and subsequent header lines if
4702 there were newlines in the string. This meant that an invalid header could
4703 be inserted if Exim was badly configured.
4705 7. Allow an ACL "warn" statement to add header lines at the start or after all
4706 the Received: headers, as well as at the end.
4708 8. Added the rcpt_4xx retry error code.
4710 9. Added postmaster_mailfrom=xxx to callout verification option.
4712 10. Added mailfrom=xxxx to the callout verification option, for verify=
4715 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4716 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4718 12. Inserted some casts to stop certain compilers warning when using pointer
4719 differences as field lengths or precisions in printf-type calls (mostly
4720 affecting debugging statements).
4722 13. Added optional readline() support for -be (dynamically loaded).
4724 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4725 same clock tick as a message's arrival, so that its received time was the
4726 same as the "first fail" time on the retry record, and that message
4727 remained on the queue past the ultimate address timeout, every queue runner
4728 would try a delivery (because it was past the ultimate address timeout) but
4729 after another failure, the ultimate address timeout, which should have then
4730 bounced the address, did not kick in. This was a "< instead of <=" error;
4731 in most cases the first failure would have been in the next clock tick
4732 after the received time, and all would be well.
4734 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4735 being recognized when the domain list was tested by the match_domain
4736 condition in an expansion string.
4738 16. Added the ${str2b64: operator.
4740 17. Exim was always calling setrlimit() to set a large limit for the number of
4741 processes, without checking whether the existing limit was already
4742 adequate. (It did check for the limit on file descriptors.) Furthermore,
4743 errors from getrlimit() and setrlimit() were being ignored. Now they are
4744 logged to the main and panic logs, but Exim does carry on, to try to do its
4745 job under whatever limits there are.
4747 18. Imported PCRE 5.0.
4749 19. Trivial typo in log message " temporarily refused connection" (the leading
4752 20. If the log selector return_path_on_delivery was set and an address was
4753 redirected to /dev/null, the delivery process crashed because it assumed
4754 that a return path would always be set for a "successful" delivery. In this
4755 case, the whole delivery is bypassed as an optimization, and therefore no
4758 21. Internal re-arrangement: the function for sending a challenge and reading
4759 a response while authentication was assuming a zero-terminated challenge
4760 string. It's now changed to take a pointer and a length, to allow for
4761 binary data in such strings.
4763 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4765 23. Exim was not respecting finduser_retries when seeking the login of the
4766 uid under which it was called; it was always trying 10 times. (The default
4767 setting of finduser_retries is zero.) Also, it was sleeping after the final
4768 failure, which is pointless.
4770 24. Implemented tls_on_connect_ports.
4772 25. Implemented acl_smtp_predata.
4774 26. If the domain in control=submission is set empty, Exim assumes that the
4775 authenticated id is a complete email address when it generates From: or
4776 Sender: header lines.
4778 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4779 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4780 chown and chgrp in /bin and hostname in /usr/bin.
4782 28. Exim was keeping the "process log" file open after each use, just as it
4783 does for the main log. This opens the possibility of it remaining open for
4784 long periods when the USR1 signal hits a daemon. Occasional processlog
4785 errors were reported, that could have been caused by this. Anyway, it seems
4786 much more sensible not to leave this file open at all, so that is what now
4789 29. The long-running daemon process does not normally write to the log once it
4790 has entered its main loop, and it closes the log before doing so. This is
4791 so that log files can straightforwardly be renamed and moved. However,
4792 there are a couple of unusual error situations where the daemon does write
4793 log entries, and I had neglected to close the log afterwards.
4795 30. The text of an SMTP error response that was received during a remote
4796 delivery was being truncated at 512 bytes. This is too short for some of
4797 the long messages that one sometimes sees. I've increased the limit to
4800 31. It is now possible to make retry rules that apply only when a message has a
4801 specific sender, in particular, an empty sender.
4803 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4804 it possible to be selective about when SMTP synchronization is enforced.
4806 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4808 32. Implemented hosts_connection_nolog.
4810 33. Added an ACL for QUIT.
4812 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4815 35. Added mailbox_size and mailbox_filecount to appendfile.
4817 36. Added control = no_multiline_responses to ACLs.
4819 37. There was a bug in the logic of the code that waits for the clock to tick
4820 in the case where the clock went backwards by a substantial amount such
4821 that the microsecond fraction of "now" was more than the microsecond
4822 fraction of "then" (but the whole seconds number was less).
4824 38. Added support for the libradius Radius client library this is found on
4825 FreeBSD (previously only the radiusclient library was supported).
4831 1. When certain lookups returned multiple values in the form name=value, the
4832 quoting of the values was not always being done properly. Specifically:
4833 (a) If the value started with a double quote, but contained no whitespace,
4835 (b) If the value contained whitespace other than a space character (i.e.
4836 tabs or newlines or carriage returns) it was not quoted.
4837 This fix has been applied to the mysql and pgsql lookups by writing a
4838 separate quoting function and calling it from the lookup code. The fix
4839 should probably also be applied to nisplus, ibase and oracle lookups, but
4840 since I cannot test any of those, I have not disturbed their existing code.
4842 2. A hit in the callout cache for a specific address caused a log line with no
4843 reason for rejecting RCPT. Now it says "Previous (cached) callout
4844 verification failure".
4846 3. There was an off-by-one bug in the queryprogram router. An over-long
4847 return line was truncated at 256 instead of 255 characters, thereby
4848 overflowing its buffer with the terminating zero. As well as fixing this, I
4849 have increased the buffer size to 1024 (and made a note to document this).
4851 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4852 when Exim is waiting for an SMTP response from a remote server, Exim
4853 restarts its select() call on the socket, thereby resetting its timeout.
4854 This is not a problem when such interrupts are rare. Somebody set up a cron
4855 job to run exiwhat every 2 minutes, which is less than the normal select()
4856 timeout (5 or 10 minutes). This meant that the select() timeout never
4857 kicked in because it was always reset. I have fixed this by comparing the
4858 time when an interrupt arrives with the time at the start of the first call
4859 to select(). If more time than the timeout has elapsed, the interrupt is
4860 treated as a timeout.
4862 5. Some internal re-factoring in preparation for the addition of Sieve
4863 extensions (by MH). In particular, the "personal" test is moved to a
4864 separate function, and given an option for scanning Cc: and Bcc: (which is
4865 not set for Exim filters).
4867 6. When Exim created an email address using the login of the caller as the
4868 local part (e.g. when creating a From: or Sender: header line), it was not
4869 quoting the local part when it contained special characters such as @.
4871 7. Installed new OpenBSD configuration files.
4873 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4874 try to make them clearer.
4876 9. Callout options, other than the timeout value, were being ignored when
4877 verifying sender addresses in header lines. For example, when using
4879 verify = header_sender/callout=no_cache
4881 the cache was (incorrectly) being used.
4883 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4884 only the Cygwin environment.
4886 11. When return_path_on_delivery was set as a log selector, if different remote
4887 addresses in the same message used different return paths and parallel
4888 remote delivery occurred, the wrong values would sometimes be logged.
4889 (Whenever a remote delivery process finished, the return path value from
4890 the most recently started remote delivery process was logged.)
4892 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4893 lines when AUTH and/or TLS are in use. This is the "received protocol"
4894 field. Exim used to use "asmtp" for authenticated SMTP, without any
4895 indication (in the protocol name) for TLS use. Now it follows the RFC and
4896 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4897 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4898 names appear in log lines as well as in Received: header lines.
4900 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4901 extensions, and comparison tests, and to fix some bugs.
4903 14. Changes to the "personal" filter test:
4905 (1) The test was buggy in that it was just doing the equivalent of
4906 "contains" tests on header lines. For example, if a user's address was
4907 anne@some.where, the "personal" test would incorrectly be true for
4909 To: susanne@some.where
4911 This test is now done by extracting each address from the header in turn,
4912 and checking the entire address. Other tests that are part of "personal"
4913 are now done using regular expressions (for example, to check local parts
4914 of addresses in From: header lines).
4916 (2) The list of non-personal local parts in From: addresses has been
4917 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4918 taken from the Sieve specification recommendations.
4920 (3) If the message contains any header line starting with "List-" it is
4921 treated as non-personal.
4923 (4) The test for "circular" in the Subject: header line has been removed
4924 because it now seems ill-conceived.
4926 15. Minor typos in src/EDITME comments corrected.
4928 16. Installed latest exipick from John Jetmore.
4930 17. If headers_add on a router specified a text string that was too long for
4931 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4932 of string_sprintf() is now avoided.
4934 18. $message_body_size was not set (it was always zero) when running the DATA
4935 ACL and the local_scan() function.
4937 19. For the "mail" command in an Exim filter, no default was being set for
4938 the once_repeat time, causing a random time value to be used if "once" was
4939 specified. (If the value happened to be <= 0, no repeat happened.) The
4940 default is now 0s, meaning "never repeat". The "vacation" command was OK
4941 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4942 (I found it when inspecting the code).
4944 20. There is now an overall timeout for performing a callout verification. It
4945 defaults to 4 times the callout timeout, which applies to individual SMTP
4946 commands during the callout. The overall timeout applies when there is more
4947 than one host that can be tried. The timeout is checked before trying the
4948 next host. This prevents very long delays if there are a large number of
4949 hosts and all are timing out (e.g. when the network connections are timing
4950 out). The value of the overall timeout can be changed by specifying an
4951 additional sub-option for "callout", called "maxwait". For example:
4953 verify = sender/callout=5s,maxwait=20s
4955 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4956 to the end before writing, but this should make it even safer).
4958 22. Exim was forgetting that it had advertised PIPELINING for the second and
4959 subsequent messages on an SMTP connection. It was also not resetting its
4960 memory on STARTTLS and an internal HELO.
4962 23. When Exim logs an SMTP synchronization error within a session, it now
4963 records whether PIPELINING has been advertised or not.
4965 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4966 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4967 rather than long int.
4969 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4971 26. Added the never_mail option to autoreply.
4977 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4978 crash if the getsockname() call failed; this can happen if a connection is
4979 closed very soon after it is established. The problem was simply in the
4980 order in which certain operations were done, causing Exim to try to write
4981 to the SMTP stream before it had set up the file descriptor. The bug has
4982 been fixed by making things happen in the correct order.
4988 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4989 before the connection was closed, thus losing the rejection response.
4991 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4992 some early Solaris releases, but causes trouble in current releases where
4993 socklen_t is defined.
4995 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4998 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5001 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5002 that releasing the top of it at the end releases what was used for sub-
5003 expansions (unless the block got too big). However, discard this block if
5004 the first thing is a variable or header, so that we can use its block when
5005 it is dynamic (useful for very large $message_headers, for example).
5007 6. Lookups now cache *every* query, not just the most recent. A new, separate
5008 store pool is used for this. It can be recovered when all lookup caches are
5009 flushed. Lookups now release memory at the end of their result strings.
5010 This has involved some general refactoring of the lookup sources.
5012 7. Some code has been added to the store_xxx() functions to reduce the amount
5013 of flapping under certain conditions.
5015 8. log_incoming_interface used to affect only the <= reception log lines. Now
5016 it causes the local interface and port to be added to several more SMTP log
5017 lines, for example "SMTP connection from", and rejection lines.
5019 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5021 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5023 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5024 could be overwritten at the end of the current message (or the start of a
5025 new message if it was set in a HELO ACL). The value is now preserved for
5026 the duration of the SMTP connection.
5028 12. If a transport had a headers_rewrite setting, and a matching header line
5029 contained an unqualified address, that address was qualified, even if it
5030 did not match any rewriting rules. The underlying bug was that the values
5031 of the flags that permit the existence of unqualified sender and recipient
5032 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5033 non-local messages, and by -bnq for local messages) were not being
5034 preserved with the message after it was received.
5036 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5037 "next input=" as part of the text comprising the host identity instead of
5038 the correct text. The code was using the same buffer for two different
5039 strings. However, depending on which order the printing function evaluated
5040 its arguments, the bug did not always show up. Under Linux, for example, my
5041 test suite worked just fine.
5043 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5044 doesn't work with some older versions of Perl. It has been changed to "my",
5045 which in any case is probably the better facility to use.
5047 15. A really picky compiler found some instances of statements for creating
5048 error messages that either had too many or two few arguments for the format
5051 16. The size of the buffer for calls to the DNS resolver has been increased
5052 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5053 for addresses that have a lot of PTR records. This alleviates a problem; it
5054 does not fully solve it.
5056 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5057 buffer now truncates the list and logs the incident, which is the same
5058 action as happens when Exim is looking up a host name and its aliases.
5059 Previously in this situation something unpredictable would happen;
5060 sometimes it was "internal error: store_reset failed".
5062 18. If a server dropped the connection unexpectedly when an Exim client was
5063 using GnuTLS and trying to read a response, the client delivery process
5064 crashed while trying to generate an error log message.
5066 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5067 string, for example:
5069 warn message = H1: something\nH2: something
5071 the text was added as a single header line from Exim's point of view
5072 though it ended up OK in the delivered message. However, searching for the
5073 second and subsequent header lines using $h_h2: did not work. This has been
5074 fixed. Similarly, if a system filter added multiple headers in this way,
5075 the routers could not see them.
5077 20. Expanded the error message when iplsearch is called with an invalid key to
5078 suggest using net-iplsearch in a host list.
5080 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5081 ACLs are no longer actually imposed (and a message to that effect is
5084 22. If a "gecos" field in a passwd entry contained escaped characters, in
5085 particular, if it contained a \" sequence, Exim got it wrong when building
5086 a From: or a Sender: header from that name. A second bug also caused
5087 incorrect handling when an unquoted " was present following a character
5088 that needed quoting.
5090 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5091 was not being matched caselessly.
5093 24. Arranged for all hyphens in the exim.8 source to be escaped with
5096 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5097 properly. Recipient callout cache records were still being keyed to include
5098 the sender, even when use_sender was set false. This led to far more
5099 callouts that were necessary. The sender is no longer included in the key
5100 when use_sender is false.
5102 26. Added "control = submission" modifier to ACLs.
5104 27. Added the ${base62d: operator to decode base 62 numbers.
5106 28. dnsdb lookups can now access SRV records.
5108 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5109 the configuration file.
5111 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5112 (-v) mode. This makes the output for a verbose queue run more intelligible.
5114 31. Added a use_postmaster feature to recipient callouts.
5116 32. Added the $body_zerocount variable, containing the number of binary zero
5117 bytes in the message body.
5119 33. The time of last modification of the "new" subdirectory is now used as the
5120 "mailbox time last read" when there is a quota error for a maildir
5123 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5125 35. Added +ignore_unknown as a special item in host lists.
5127 36. Code for decoding IPv6 addresses in host lists is now included, even if
5128 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5129 address was recognized as an IP address, but was then not correctly decoded
5130 into binary, causing unexpected and incorrect effects when compared with
5137 1. Very minor rewording of debugging text in manualroute to say "list of
5138 hosts" instead of "hostlist".
5140 2. If verify=header_syntax was set, and a header line with an unqualified
5141 address (no domain) and a large number of spaces between the end of the
5142 name and the colon was received, the reception process suffered a buffer
5143 overflow, and (when I tested it) crashed. This was caused by some obsolete
5144 code that should have been removed. The fix is to remove it!
5146 3. When running in the test harness, delay a bit after writing a bounce
5147 message to get a bit more predictability in the log output.
5149 4. Added a call to search_tidyup() just before forking a reception process. In
5150 theory, someone could use a lookup in the expansion of smtp_accept_max_
5151 per_host which, without the tidyup, could leave open a database connection.
5153 5. Added the variables $recipient_data and $sender_data which get set from a
5154 lookup success in an ACL "recipients" or "senders" condition, or a router
5155 "senders" option, similar to $domain_data and $local_part_data.
5157 6. Moved the writing of debug_print from before to after the "senders" test
5160 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5161 problems for message scanning, either using a data ACL, or using
5162 local_scan() because the Received: header was not generated till after they
5163 were called (in order to set the time as the time of reception completion).
5164 I have revised the way this works. The header is now generated after the
5165 body is received, but before the ACL or local_scan() are called. After they
5166 are run, the timestamp in the header is updated.
5172 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5173 before starting a queue runner without re-exec. This happened only when
5174 deliver_drop_privilege was set or when the Exim user was set to root. The
5175 effect of the bug was that timeouts during subsequent deliveries caused
5176 crashes instead of being properly handled. The handler is now left at its
5177 default (and expected) setting.
5179 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5180 message, again when deliver_drop_privilege is set or Exim is run as root.
5181 The bug described in (1) was not present in this case, but the tidying up
5182 of the other signals was missing. I have made the two cases consistent.
5184 3. The ignore_target_hosts setting on a manualroute router was being ignored
5185 for hosts that were looked up using the /MX notation.
5187 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5190 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5191 operated on the sender address. After changing the $sender_address to <>
5192 for the sender address verify, Exim was re-instated it as the original
5193 (before rewriting) address, but remembering that it had rewritten it, so it
5194 wasn't rewriting it again. This bug also had the effect of breaking the
5195 sender address verification caching when the sender address was rewritten.
5197 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5198 This has been changed so that if the ip literal address matches
5199 ignore_target_hosts, the router declines.
5201 7. Added expansion conditions match_domain, match_address, and match_local_
5202 part (NOT match_host).
5204 8. The placeholder for the Received: header didn't have a length field set.
5206 9. Added code to Exim itself and to exim_lock to test for a specific race
5207 condition that could lead to file corruption when using MBX delivery. The
5208 issue is with the lockfile that is created in /tmp. If this file is removed
5209 after a process has opened it but before that process has acquired a lock,
5210 there is the potential for a second process to recreate the file and also
5211 acquire a lock. This could lead to two Exim processes writing to the file
5212 at the same time. The added code performs the same test as UW imapd; it
5213 checks after acquiring the lock that its file descriptor still refers to
5214 the same named file.
5216 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5217 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5220 11. Added the smtp_active_hostname option. If used, this will typically be made
5221 to depend on the incoming interface address. Because $interface_address is
5222 not set up until the daemon has forked a reception process, error responses
5223 that can happen earlier (such as "too many connections") no longer contain
5226 12. If an expansion in a condition on a "warn" statement fails because a lookup
5227 defers, the "warn" statement is abandoned, and the next ACL statement is
5228 processed. Previously this caused the whole ACL to be aborted.
5230 13. Added the iplsearch lookup type.
5232 14. Added ident_timeout as a log selector.
5234 15. Added tls_certificate_verified as a log selector.
5236 16. Added a global option tls_require_ciphers (compare the smtp transport
5237 option of the same name). This controls incoming TLS connections.
5239 17. I finally figured out how to make tls_require_ciphers do a similar thing
5240 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5241 before starting the TLS session.
5243 18. Tabs are now shown as \t in -bP output.
5245 19. If the log selector return_path_on_delivery was set, Exim crashed when
5246 bouncing a message because it had too many Received: header lines.
5248 20. If two routers both had headers_remove settings, and the first one included
5249 a superfluous trailing colon, the final name in the first list and the
5250 first name in the second list were incorrectly joined into one item (with a
5251 colon in the middle).
5257 1. Added -C and -D options to the exinext utility, mainly to make it easier
5258 to include in the automated testing, but these could be helpful when
5259 multiple configurations are in use.
5261 2. The exinext utility was not formatting the output nicely when there was
5262 an alternate port involved in the retry record key, nor when there was a
5263 message id as well (for retries that were specific to a specific message
5264 and a specific host). It was also confused by IPv6 addresses, because of
5265 the additional colons they contain. I have fixed the IPv4 problem, and
5266 patched it up to do a reasonable job for IPv6.
5268 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5269 delivery, the log line now contains "pipelined" if PIPELINING was used.
5271 4. An SMTP transport process used to panic and die if the bind() call to set
5272 an explicit outgoing interface failed. This has been changed; it is now
5273 treated in the same way as a connect() failure.
5275 5. A reference to $sender_host_name in the part of a conditional expansion
5276 that was being skipped was still causing a DNS lookup. This no longer
5279 6. The def: expansion condition was not recognizing references to header lines
5280 that used bh_ and bheader_.
5282 7. Added the _cache feature to named lists.
5284 8. The code for checking quota_filecount in the appendfile transport was
5285 allowing one more file than it should have been.
5287 9. For compatibility with Sendmail, the command line option
5295 and sets the incoming protocol and host name (for trusted callers). The
5296 host name and its colon can be omitted when only the protocol is to be set.
5297 Note the Exim already has two private options, -pd and -ps, that refer to
5298 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5299 "s", but I don't think that's a major issue.
5301 10. A number of refactoring changes to the code, none of which should affect
5304 (a) The number of logging options was getting close to filling up the
5305 32-bit word that was used as a bit map. I have split them into two classes:
5306 those that are passed in the argument to log_write(), and those that are
5307 only ever tested independently outside of that function. These are now in
5308 separate 32-bit words, so there is plenty of room for expansion again.
5309 There is no change in the user interface or the logging behaviour.
5311 (b) When building, for example, log lines, the code previously used a
5312 macro that called string_cat() twice, in order to add two strings. This is
5313 not really sufficiently general. Furthermore, there was one instance where
5314 it was actually wrong because one of the argument was used twice, and in
5315 one call a function was used. (As it happened, calling the function twice
5316 did not affect the overall behaviour.) The macro has been replaced by a
5317 function that can join an arbitrary number of extra strings onto a growing
5320 (c) The code for expansion conditions now uses a table and a binary chop
5321 instead of a serial search (which was left over from when there were very
5322 few conditions). Also, it now recognizes conditions like "pam" even when
5323 the relevant support is not compiled in: a suitably worded error message is
5324 given if an attempt is made to use such a condition.
5326 11. Added ${time_interval:xxxxx}.
5328 12. A bug was causing one of the ddress fields not to be passed back correctly
5329 from remote delivery subprocesses. The field in question was not being
5330 subsequently used, so this caused to problems in practice.
5332 13. Added new log selectors queue_time and deliver_time.
5334 14. Might have fixed a bug in maildirsizefile handling that threw up
5335 "unexpected character" debug warnings, and recalculated the data
5336 unnecessarily. In any case, I expanded the warning message to give more
5339 15. Added the message "Restricted characters in address" to the statements in
5340 the default ACL that block characters like @ and % in local parts.
5342 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5343 Three changes have been made:
5345 (a) There was a serious bug; a negative response to MAIL caused the whole
5346 recipient domain to be cached as invalid, thereby blocking all messages
5347 to all local parts at the same domain, from all senders. This bug has
5348 been fixed. The domain is no longer cached after a negative response to
5349 MAIL if the sender used is not empty.
5351 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5354 (c) A new callout option, "use_sender" has been added for people who want
5355 the modified behaviour.
5361 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5364 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5365 indeed breaks things for older releases.
5367 3. Added additional logging to the case where there is a problem reading data
5368 from a filter that is running in a subprocess using a pipe, in order to
5369 try to track down a specific problem.
5371 4. Testing facility fudge: when running in the test harness and attempting
5372 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5373 getting "No route to host". Convert this to a timeout.
5375 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5378 6. Some OS don't have socklen_t but use size_t instead. This affects the
5379 fifth argument of getsockopt() amongst other things. This is now
5380 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5381 can be set for individual OS. I have set it for SunOS5, OSF1, and
5382 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5383 some earlier ones do not.
5385 7. Change 4.30/15 was not doing the test caselessly.
5387 8. The standard form for an IPv6 address literal was being rejected by address
5388 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5389 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5390 this, as well as the form without the "IPv6" on the front (but only when
5391 address literals are enabled, of course).
5393 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5395 10. Exim crashed if a message with an empty sender address specified by -f
5396 encountered a router with an errors_to setting. This could be provoked only
5397 by a command such as
5401 where an empty string was supplied; "<>" did not hit this bug.
5403 11. Installed PCRE release 4.5.
5405 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5406 remained set. It is now erased.
5408 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5409 times from message ids (which are base 36 rather than the normal 62).
5411 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5412 were being counted as actual protocol errors, and logged if the log
5413 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5414 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5415 and DATA following a set of rejected RCPTs do not count as protocol errors.
5416 In other words, Exim assumes they were pipelined, though this may not
5417 actually be the case. Of course, in all cases the client gets an
5418 appropriate error code.
5420 15. If a lookup fails in an ACL condition, a message about the failure may
5421 be available; it is used if testing the ACL cannot continue, because most
5422 such messages specify what the cause of the deferral is. However, some
5423 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5424 that caused an old message to be retained and used if a later statement
5425 caused a defer, replacing the real cause of the deferral.
5427 16. If an IP address had so many PTR records that the DNS lookup buffer
5428 was not large enough to hold them, Exim could crash while trying to process
5429 the truncated data. It now detects and logs this case.
5431 17. Further to 4.21/58, another change has been made: if (and only if) the
5432 first line of a message (the first header line) ends with CRLF, a bare LF
5433 in a subsequent header line has a space inserted after it, so as not to
5434 terminate the header.
5436 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5437 unnecessarily, used atoi() instead of strtol(), and didn't check the
5438 termination when getting file sizes from file names by regex.
5440 19. Completely re-implemented the support for maildirsize files, in the light
5441 of a number of problems with the previous contributed implementation
5442 (4.30/29). In particular:
5444 . If the quota is zero, the maildirsize file is maintained, but no quota is
5447 . If the maildir directory does not exist, it is created before any attempt
5448 to write a maildirsize file.
5450 . The quota value in the file is just a cache; if the quota is changed in
5451 the transport, the new value overrides.
5453 . A regular expression is available for excluding directories from the
5456 20. The autoreply transport checks the characters in options that define the
5457 message's headers; it allows continued headers, but it was checking with
5458 isspace() after an embedded newline instead of explicitly looking for a
5461 21. If all the "regular" hosts to which an address was routed had passed their
5462 expiry times, and had not reached their retry times, the address was
5463 bounced, even if fallback hosts were defined. Now Exim should go on to try
5466 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5467 equivalent code in the SMTP transport. Some hosts send humungous responses
5468 to HELO/EHLO, more than 1024 it seems.
5470 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5471 gives compiler warnings in some environments. I've now done it "properly",
5474 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5475 (because of problems with the built-in one) was declared to return uschar *
5476 instead of char *, causing compiler failure.
5478 25. Fixed a file descriptor leak when processing alias/forward files.
5480 26. Fixed a minor format string issue in dbfn.c.
5482 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5484 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5485 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5488 29. When a sender address is verified, it is cached, to save repeating the test
5489 when there is more than one recipient in a message. However, when the
5490 verification involves a callout, it is possible for different callout
5491 options to be set for different recipients. It is too complicated to keep
5492 track of this in the cache, so now Exim always runs a verification when a
5493 callout is required, relying on the callout cache for the optimization.
5494 The overhead is duplication of the address routing, but this should not be
5497 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5498 to be verified with callout=postmaster, and the main callout worked but the
5499 postmaster check failed, the verification correctly failed. However, if a
5500 subsequent RCPT command asked for sender verification *without* the
5501 postmaster check, incorrect caching caused this verification also to fail,
5504 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5505 it was not caching the DNS options (qualify_single, search_parents) that
5506 were used when the lookup failed. A subsequent lookup with different
5507 options therefore always gave the same answer, though there were cases
5508 where it should not have. (Example: a "domains = !$mx_any" option on a
5509 dnslookup router: the "domains" option is always processed without any
5510 widening, but the router might have qualify_single set.) Now Exim uses the
5511 cached value only when the same options are set.
5513 32. Added John Jetmore's "exipick" utility to the distribution.
5515 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5516 than a timeout (e.g. a certificate is required, and is not provided), an
5517 Exim server now closes the connection immediately. Previously it waited for
5518 the client to close - but if the client is SSL, it seems that they each
5519 wait for each other, leading to a delay before one of them times out.
5521 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5522 maintained 0.8.x compatibility because I don't think many are using it, and
5523 it is clearly obsolete.
5525 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5528 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5529 client certificate was expired. A simple patch fixes this, though I don't
5530 understand the full logic of why the verify callback is called multiple
5533 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5534 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5535 which causes problems with some clients (such as the Certicom SSL Plus
5536 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5537 disables the coutermeasure allowing Eudora to connect."
5539 38. Exim was not checking that a write() to a log file succeeded. This could
5540 lead to Bad Things if a log got too big, in particular if it hit a file
5541 size limit. Exim now panics and dies if it cannot write to a log file, just
5542 as it does if it cannot open a log file.
5544 39. Modified OS/Makefile-Linux so that it now contains
5546 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5548 The two -D definitions ensure that Exim is compiled with large file
5549 support, which makes it possible to handle log files that are bigger than
5552 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5553 instance) a domain was checked against a named list that involved a lookup,
5554 causing $domain_data to be set, then another domain was checked against the
5555 same list, then the first domain was re-checked, the value of $domain_data
5556 after the final check could be wrong. In particular, if the second check
5557 failed, it could be set empty. This bug probably also applied to
5560 41. The strip_trailing_dot option was not being applied to the address given
5561 with the -f command-line option.
5563 42. The code for reading a message's header from the spool was incrementing
5564 $received_count, but never initializing it. This meant that the value was
5565 incorrect (doubled) while delivering a message in the same process in which
5566 it was received. In the most common configuration of Exim, this never
5567 happens - a fresh exec is done - but it can happen when
5568 deliver_drop_privilege is set.
5570 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5571 it now includes up to 150 characters of the unexpected data in the log
5574 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5575 and building data strings. The size of both of these buffers was 10 000
5576 bytes - far larger than anybody would *ever* want, thought I. Needless to
5577 say, somebody hit the limit. I have increased the maximum line length to
5578 20 000 and the maximum data length of concatenated lines to 100 000. I have
5579 also fixed two bugs, because there was no checking on these buffers. Tsk,
5580 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5581 buffer is too small.
5583 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5584 lsearch lookups. Now it does.
5586 46. When parsing a route_list item in a manualroute router, a fixed-length
5587 buffer was used for the list of hosts. I made this 1024 bytes long,
5588 thinking that nobody would ever have a list of hosts that long. Wrong.
5589 Somebody had a whole pile of complicated expansion conditions, and the
5590 string was silently truncated, leading to an expansion error. It turns out
5591 that it is easier to change to an unlimited length (owing to other changes
5592 that have happened since this code was originally written) than to build
5593 structure for giving a limitation error. The length of the item that
5594 expands into the list of hosts is now unlimited.
5596 47. The lsearch lookup could not handle data where the length of text line was
5597 more than 4095 characters. Such lines were truncated, leading to shortened
5598 data being returned. It should now handle lines of any length.
5600 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5601 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5604 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5605 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5606 no longer happens in comment lines. A list of replacements is now placed
5607 at the head of all of the source files, except those whose only change is
5608 to replace PERL_COMMAND in the very first #! line.
5610 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5611 messages on the queue, with a bottom-up merge sort, using code contributed
5612 by Michael Haardt. This should make operations like -bp somewhat faster on
5613 large queues. It won't affect queue runners, except when queue_run_in_order
5616 51. Installed eximstats 1.31 in the distribution.
5618 52. Added support for SRV lookups to the dnslookup router.
5620 53. If an ACL referred to $message_body or $message_body_end, the value was not
5621 reset for any messages that followed in the same SMTP session.
5623 54. The store-handling optimization for building very long strings was not
5624 differentiating between the different store pools. I don't think this
5625 actually made any difference in practice, but I've tidied it.
5627 55. While running the routers to verify a sender address, $sender_address
5628 was still set to the sender address. This is wrong, because when routing to
5629 send a bounce to the sender, it would be empty. Therefore, I have changed
5630 it so that, while verifying a sender address, $sender_address is set to <>.
5631 (There is no change to what happens when verifying a recipient address.)
5633 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5634 target A or AAAA records (if not already returned) without resetting the
5635 qualify_single or search_parents options of the DNS resolver. These are
5636 inappropriate in this case because the targets of MX and SRV records must
5637 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5638 target that, when qualified, matched something in the local domain. These
5639 two options are now turned off when doing these lookups.
5641 57. It seems that at least some releases of Reiserfs (which does not have the
5642 concept of a fixed number of inodes) returns zero and not -1 for the
5643 number of available inodes. This interacted badly with check_spool_inodes,
5644 which assumed that -1 was the "no such thing" setting. What I have done is
5645 to check that the total number of inodes is greater than zero before doing
5646 the test of how many are available.
5648 58. When a "warn" ACL statement has a log_message modifier, the message is
5649 remembered, and not repeated. This is to avoid a lot of repetition when a
5650 message has many recipients that cause the same warning to be written.
5651 However, Exim was preserving the list of already written lines for an
5652 entire SMTP session, which doesn't seem right. The memory is now reset if a
5653 new message is started.
5655 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5656 debugging output unless log_rewrite was also set.
5658 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5659 of (int)(handle) when we know that handle contains (void *)(-1).
5661 61. The Exim daemon panic-logs an error return when it closes the incoming
5662 connection. However "connection reset by peer" seems to be common, and
5663 isn't really an error worthy of noting specially, so that particular error
5666 62. When Exim is trying to find all the local interfaces, it used to panic and
5667 die if the ioctl to get the interface flags failed. However, it seems that
5668 on at least one OS (Solaris 9) it is possible to have an interface that is
5669 included in the list of interfaces, but for which you get a failure error
5670 for this call. This happens when the interface is not "plumbed" into a
5671 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5672 failure of the "get flags" call assumes that the interface is down.
5674 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5675 makes life easier for people who are doing arithmetic on fields extracted
5676 from dates, where you often get leading zeros that should not be
5677 interpreted as octal.
5679 64. Added qualify_domain to the redirect router, to override the global
5682 65. If a pathologically long header line contained very many addresses (the
5683 report of this problem mentioned 10 000) and each of them was rewritten,
5684 Exim could use up a very large amount of memory. (It kept on making new
5685 copies of the header line as it rewrote, and never released the old ones.)
5686 At the expense of a bit more processing, the header rewriting function has
5687 been changed so that it no longer eats memory in this way.
5689 66. The generation of the Received: header has been moved from the time that a
5690 message starts to be received, to the time that it finishes. The timestamp
5691 in the Received: header should now be very close to that of the <= log
5692 line. There are two side-effects of this change:
5694 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5695 logged header lines no longer include the local Received: line, because
5696 it has not yet been created. The same applies to a copy of the message
5697 that is returned to a non-SMTP sender when a message is rejected.
5699 (b) When a filter file is tested using -bf, no additional Received: header
5700 is added to the test message. After some thought, I decided that this
5703 This change does not affect the value of $received_for. It is still set
5704 after address rewriting, but before local_scan() is called.
5706 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5708 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5709 gave an unhelpful panic error message, and a defer error. I have managed to
5710 change this behaviour so that it now rejects any supplied certificate,
5711 which seems right, as the list of acceptable certificates is empty.
5713 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5714 gave an unhelpful defer error. I have not managed to make this reject any
5715 supplied certificates, but the error message it gives is "no certificate
5716 supplied", which is not helpful.
5718 70. exigrep's output now also includes lines that are not associated with any
5719 message, but which match the given pattern. Implemented by a patch from
5720 Martin Sluka, which also tidied up the Perl a bit.
5722 71. Recipient callout verification, like sender verification, was using <> in
5723 the MAIL FROM command. This isn't really the right thing, since the actual
5724 sender may affect whether the remote host accepts the recipient or not. I
5725 have changed it to use the actual sender in the callout; this means that
5726 the cache record is now keyed on a recipient/sender pair, not just the
5727 recipient address. There doesn't seem to be a real danger of callout loops,
5728 since a callout by the remote host to check the sender would use <>.
5729 [SEE ABOVE: changed after hitting problems.]
5731 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5732 temporary errors. However, in the case of such a code being given after
5733 the end of a data transmission (i.e. after ".") Exim was failing to write
5734 a retry record for the message. (Yes, there was some broken host that was
5735 actually sending 8xx at this point.)
5737 73. An unknown lookup type in a host list could cause Exim to panic-die when
5738 the list was checked. (An example that provoked this was putting <; in the
5739 middle of a list instead of at the start.) If this happened during a DATA
5740 ACL check, a -D file could be left lying around. This kind of configuration
5741 error no longer causes Exim to die; instead it causes a defer error. The
5742 incident is still logged to the main and panic logs.
5744 74. Buglet left over from Exim 3 conversion. The message "too many messages
5745 in one connection" was written to the rejectlog but not the mainlog, except
5746 when address rewriting (yes!) was being logged.
5748 75. Added write_rejectlog option.
5750 76. When a system filter was run not as root (that is, when system_filter_user
5751 was set), the values of the $n variables were not being returned to the
5752 main process; thus, they were not subsequently available in the $sn
5755 77. Added +return_path_on_delivery log selector.
5757 78. A connection timeout was being treated differently from recipients deferred
5758 when testing hosts_max_try with a message that was older than the host's
5759 retry timeout. (The host should not be counted, thus allowing all hosts to
5760 be tried at least once before bouncing.) This may have been the cause of an
5761 occasionally reported bug whereby a message would remain on the queue
5762 longer than the retry timeout, but would be bounced if a delivery was
5763 forced. I say "may" because I never totally pinned down the problem;
5764 setting up timeout/retry tests is difficult. See also the next item.
5766 79. The ultimate address timeout was not being applied to errors that involved
5767 a combination of host plus message (for example, a timeout on a MAIL
5768 command). When an address resolved to a number of possible hosts, and they
5769 were not all tried for each delivery (e.g. because of hosts_max_try), a
5770 message could remain on the queue longer than the retry timeout.
5772 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5775 81. Fixed an obscure SMTP outgoing bug which required at least the following
5776 conditions: (a) there was another message waiting for the same server;
5777 (b) the server returned 5xx to all RCPT commands in the first message so
5778 that the message was not completed; (c) the server dropped the connection
5779 or gave a negative response to the RSET that Exim sends to abort the
5780 transaction. The observed case was a dropped connection after DATA that had
5781 been sent in pipelining mode. That is, the server had advertised PIPELINING
5782 but was not implementing it correctly. The effect of the bug was incorrect
5783 behaviour, such as trying another host, and this could lead to a crash.
5789 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5790 and daemon.c were passed as pointers to ints; they should have been
5791 pointers to socklen_t variables (which are typically unsigned ints).
5793 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5796 3. Fixed a really odd bug that affected only the testing scheme; patching a
5797 certain fixed string in the binary changed the value of another string that
5798 happened to be identical to the end of the original first string.
5800 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5801 name", it returns that address as the IP address. On some operating
5802 systems (e.g. Solaris), it also passes back the IP address string as the
5803 "host name". However, on others (e.g. Linux), it passes back an empty
5804 string. Exim wasn't checking for this, and was changing the host name to an
5805 empty string, assuming it had been canonicalized.
5807 5. Although rare, it is permitted to have more than one PTR record for a given
5808 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5809 all the names associated with an address, because they do in Solaris.
5810 However, it seems that they do not in Linux for data that comes from the
5811 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5812 I found this out when I moved to a new Linux workstation and tried to run
5813 the Exim test suite.
5815 To get round this problem I have changed the code so that it now does its
5816 own call to the DNS to look up PTR records when searching for a host name.
5817 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5818 addresses that are only in /etc/hosts are still found.
5820 This behaviour is, however, controlled by an option called host_lookup_
5821 order, which defaults to "bydns:byaddr". If people want to use the other
5822 order, or indeed, just use one or the other means of lookup, they can
5823 specify it in this variable.
5825 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5826 some operating systems, this comes back from gethostbyaddr() as an empty
5827 string, and this is what Exim used to test for. However, it seems that in
5828 other systems, "." is yielded. Exim now tests for this case too.
5830 7. The values of check_spool_space and check_log_space are now held internally
5831 as a number of kilobytes instead of an absolute number of bytes. If a
5832 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5833 kilobyte. This means that much larger values can be stored.
5835 8. Exim monitor: an attempt to get the action menu when not actually pointing
5836 at a message produces an empty menu entitled "No message selected". This
5837 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5838 no entries in it ("Shell widget menu has zero width and/or height"). So I
5839 have added a single, blank menu entry in this case.
5841 9. Added ${quote_local_part.
5843 10. MIME decoding is now applied to the contents of Subject: header lines when
5846 11. Now that a reference to $sender_host_address automatically causes a reverse
5847 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5848 host lookup before query-style lookups in lists that might use this
5849 variable. This has therefore been abolished, and the "net-" prefix is no
5850 longer necessary for query-style lookups.
5852 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5853 have been a typo for LFLAGS, so it has been changed.
5855 13. The install script calls Exim with "-C /dev/null" in order to find the
5856 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5857 to be output. However, since Exim outputs its version number before the
5858 error, it didn't break the script. It just looked ugly. I fixed this by
5859 always allowing "-C /dev/null" if the caller is root.
5861 14. Ignore overlarge ACL variable number when reading spool file - insurance
5862 against a later release with more variables having written the file.
5864 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5865 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5866 this, as well as the form without the "IPv6" on the front.
5868 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5869 OS/Makefile-Darwin file.
5871 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5872 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5874 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5875 variable. However, some OS use ints for inodes. I've added cast to long int
5876 to get rid of the compiler warning.
5878 19. I had forgotten to lock out "/../" in configuration file names when
5879 ALT_CONFIG_PREFIX was set.
5881 20. Routers used for verification do not need to specify transports. However,
5882 if such a router generated a host list, and callout was configured, Exim
5883 crashed, because it could not find a port number from the (non-existent)
5884 transport. It now assumes port 25 in this circumstance.
5886 21. Added the -t option to exigrep.
5888 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5889 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5890 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5891 them is set, LOOKUP_LSEARCH is forced.
5893 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5895 24. Added sender and host information to the "rejected by local_scan()" log
5896 line; previously there was no indication of these.
5898 25. Added .include_if_exists.
5900 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5901 Linux. It turns out that not all file systems support this. Apparently some
5902 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5903 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5904 not supported on the file descriptor, is now ignored when Exim is trying to
5905 sync a directory. This applies only to Linux.
5907 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5909 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5910 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5913 29. Added support for maildirsize files from supplied patch (modified a bit).
5915 30. The use of :fail: followed by an empty string could lead Exim to respond to
5916 sender verification failures with (e.g.):
5918 550 Verification failed for <xxx>
5919 550 Sender verify failed
5921 where the first response line was missing the '-' that indicates it is not
5922 the final line of the response.
5924 31. The loop for finding the name of the user that called Exim had a hardwired
5925 limit of 10; it now uses the value of finduser_retries, which is used for
5926 all other user lookups.
5928 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5931 33. Exim was neglecting to zero errno before one call of strtol() when
5932 expanding a string and expecting an integer value. On some systems this
5933 resulted in spurious "integer overflow" errors. Also, it was casting the
5934 result into an int without checking.
5936 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5937 did not work. The code looks as if it has *never* worked, though it appears
5938 to have been documented since at least release 1.62. I have made it work.
5940 35. The "timeout_DNS" error in retry rules, also documented since at least
5941 1.62, also never worked. As it isn't clear exactly what this means, and
5942 clearly it isn't a major issue, I have abolished the feature by treating it
5943 as "timeout", and writing a warning to the main and panic logs.
5945 36. The display of retry rules for -brt wasn't always showing the error code
5948 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5949 timeout_connect_A, timeout_connect_MX.
5951 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5952 to the empty sender.
5954 39. The daemon was not analysing the content of -oX till after it had closed
5955 stderr and disconnected from the controlling terminal. This meant that any
5956 syntax errors were only noted on the panic log, and the return code from
5957 the command was 0. By re-arranging the code a little, I've made the
5958 decoding happen first, so such errors now appear on stderr, and the return
5959 code is 1. However, the actual setting up of the sockets still happens in
5960 the disconnected process, so errors there are still only recorded on the
5963 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5964 connections (as happens on some IP stacks) was logged at start up time as
5965 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5966 it from "IPv6 and IPv4", which means that two separate sockets are being
5969 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5970 says whether AF_INET or AF_INET6 was passed as an argument.
5972 42. Exiwhat output was messed up when time zones were included in log
5975 43. Exiwhat now gives more information about the daemon's listening ports,
5976 and whether -tls-on-connect was used.
5978 44. The "port" option of the smtp transport is now expanded.
5980 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5981 silently ignored. Now an error message is written to the main and panic
5984 46. There's a new ACL modifier called "logwrite" which writes to a log file
5985 as soon as it is encountered.
5987 47. Added $local_user_uid and $local_user_gid at routing time.
5989 48. Exim crashed when trying to verify a sender address that was being
5992 49. Exim was recognizing only a space character after ".include". It now also
5993 recognizes a tab character.
5995 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5996 extracting the relevant information from the specification. The man page no
5997 longer contains scrambled data for the -d option, and I've added a section
5998 at the front about calling Exim under different names.
6000 51. Added "extra_headers" argument to the "mail" command in filter files.
6002 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6005 53. Installed eximstats 1.29.
6007 54. Added transport_filter_timeout as a generic transport option.
6009 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6010 Cc: header lines. This was required by RFC 822, but it not required by RFC
6013 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6014 incoming messages that did not have them. Now it does so only if the
6015 message originates locally, that is, if there is no associated remote host
6016 address. When Resent- header lines are present, this applies to the Resent-
6017 lines rather than the non-Resent- lines.
6019 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6020 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6022 58. Messages for configuration errors now include the name of the main
6023 configuration file - useful now that there may be more than one file in a
6024 list (.included file names were always shown).
6026 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6027 for those rare installations that do not start the daemon as root or run it
6028 setuid root. I've cut out the call to initgroups() if the daemon is not
6031 60. The Exim user and group can now be bound into the binary as text strings
6032 that are looked up at the start of Exim's processing.
6034 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6036 62. Added $mailstore_basename variable.
6038 63. Installed patch to sieve.c from Michael Haardt.
6040 64. When Exim failed to open the panic log after failing to open the main log,
6041 the original message it was trying to log was written to stderr and debug
6042 output, but if they were not available (the usual case in production), it
6043 was lost. Now it is written to syslog before the two lines that record the
6044 failures to open the logs.
6046 65. Users' Exim filters run in subprocesses under the user's uid. It is
6047 possible for a "deliver" command or an alias in a "personal" command to
6048 provoke an address rewrite. If logging of address rewriting is configured,
6049 this fails because the process is not running as root or exim. There may be
6050 a better way of dealing with this, but for the moment (because 4.30 needs
6051 to be released), I have disabled address rewrite logging when running a
6052 filter in a non-root, non-exim process.
6058 1. The buildconfig auxiliary program wasn't quoting the value set for
6059 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6060 not defined. This bug was masked in 4.22 by the effect that was fixed in
6063 2. Some messages that were rejected after a message id was allocated were
6064 shown as "incomplete" by exigrep. It no longer does this for messages that
6065 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6067 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6068 have allow_domain_literals set, the ID did not get logged in the <= line.
6069 Domain literals are now always recognized in Message-ID: header lines.
6071 4. The first argument for a ${extract expansion item is the key name or field
6072 number. Leading and trailing spaces in this item were not being ignored,
6073 causing some misleading effects.
6075 5. When deliver_drop_privilege was set, single queue runner processes started
6076 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6077 same command in the process it spins off) were not dropping privilege.
6079 6. When the daemon running as "exim" started a queue runner, it always
6080 re-executed Exim in the spun-off process. This is a waste of effort when
6081 deliver_drop_privilege is set. The new process now just calls the
6082 queue-runner function directly.
6088 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6091 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6092 ignored. Though the use of -oP was forcing the writing of a pid file, it
6093 was always written to the default place.
6095 3. If the message "no IP address found for host xxxx" is generated during
6096 incoming verification, it is now followed by identification of the incoming
6097 connection (so you can more easily find what provoked it).
6099 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6101 5. Added some features to "harden" Exim a bit more against certain attacks:
6103 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6104 be put in Local/Makefile. This is like the never_users runtime option,
6105 but it cannot be overridden. The default setting is "root".
6107 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6108 prefix string with which any file named in a -C command line option
6111 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6112 is retained for -C and -D only if the caller of Exim is root. Without
6113 it, the exim user may also use -C and -D and retain privilege.
6115 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6116 command line option is disabled.
6118 6. Macro names set by the -D option must start with an upper case letter, just
6119 like macro names defined in the configuration file.
6121 7. Added "dereference=" facility to LDAP.
6123 8. Two instances of the typo "uknown" in the source files are fixed.
6125 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6126 the Configure-Makefile script screwed up while processing it.
6128 10. Incorporated PCRE 4.4.
6130 11. The SMTP synchronization check was not operating right at the start of an
6131 SMTP session. For example, it could not catch a HELO sent before the client
6132 waited for the greeting. There is now a check for outstanding input at the
6133 point when the greeting is written. Because of the duplex, asynchronous
6134 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6135 way, but not yet received, when the check is performed.
6137 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6138 on TCP/IP sockets, because this apparently causes some broken clients to
6141 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6142 unchanged) from the Cygwin maintainer.
6144 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6145 was supported instead of testing for mbx. Effectively a typo.
6147 15. The spa authenticator server code was not checking that the input it
6148 received was valid base64.
6150 16. The debug output line for the "set" modifier in ACLs was not showing the
6151 name of the variable that was being set.
6153 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6155 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6156 lookup on its own. Something else was needed to trigger the lookup. For
6157 example, a match in host_lookup or the need for a host name in a host list.
6158 Now, if $sender_host_name is referenced and the host name has not yet been
6159 looked up, a lookup is performed. If the lookup fails, the variable remains
6160 empty, and $host_lookup_failed is set to "1".
6162 19. Added "eqi" as a case-independent comparison operator.
6164 20. The saslauthd authentication condition could segfault if neither service
6165 nor realm was specified.
6167 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6168 error message that was logged was misleading, and incoming SMTP
6169 connections were dropped. The message is now more accurate, and temporary
6170 errors are given to SMTP connections.
6172 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6173 (except RSET) with a 421 error, until QUIT is received. However, it was
6174 failing to send a response to QUIT.
6176 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6177 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6178 ACL is now run after the helo_try_verify_hosts code.
6180 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6181 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6182 case-independent, so other case variants are also recognized). Apparently
6183 some systems use these upper case variants.
6185 25. If more than two messages were waiting for the same host, and a transport
6186 filter was specified for the transport, Exim sent two messages over the
6187 same TCP/IP connection, and then failed with "socket operation on non-
6188 socket" when it tried to send the third.
6190 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6192 27. The extern definition of crypt16() in expand.c was not being excluded when
6193 the OS had its own crypt16() function.
6195 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6196 as a preferred synonym for return_size_limit, both as an option and as an
6199 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6201 30. Changed the default configuration ACL to relax the local part checking rule
6202 for addresses that are not in any local domains. For these addresses,
6203 slashes and pipe symbols are allowed within local parts, but the sequence
6204 /../ is explicitly forbidden.
6206 31. SPA server authentication was not clearing the challenge buffer before
6209 32. log_message in a "warn" ACL statement was writing to the reject log as
6210 well as to the main log, which contradicts the documentation and doesn't
6211 seem right (because no rejection is happening). So I have stopped it.
6213 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6214 However, I am unable to do any testing of this.
6216 34. Fixed an infelicity in the appendfile transport. When checking directories
6217 for a mailbox, to see if any needed to be created, it was accidentally
6218 using path names with one or more superfluous leading slashes; tracing
6219 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6221 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6222 contents are added to the log line that is written for every discarded
6223 recipient. (Previously a log_message setting was ignored.)
6225 36. The ${quote: operator now quotes the string if it is empty.
6227 37. The install script runs exim in order to find its version number. If for
6228 some reason other than non-existence or emptiness, which it checks, it
6229 could not run './exim', it was installing it with an empty version number,
6230 i.e. as "exim-". This error state is now caught, and the installation is
6233 38. An argument was missing from the function that creates an error message
6234 when Exim fails to connect to the socket for saslauthd authentication.
6235 This could cause Exim to crash, or give a corrupted message.
6237 39. Added isip, isip4, and isip6 to ${if conditions.
6239 40. The ACL variables $acl_xx are now saved with the message, and can be
6240 accessed later in routers, transports, and filters.
6242 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6243 strings in the file are not string-expanded.
6245 42. If a MAIL command specified a SIZE value that was too large to fit into an
6246 int variable, the check against message_size_limit failed. Such values are
6247 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6248 day this will have to be increased, but I don't think I want to be around
6249 when emails are that large.
6256 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6257 iconv() is not standard in FreeBSD.
6259 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6260 IPv6 enabled. The observed symptom was a segmentation fault on return from
6261 the function os_common_find_running_interfaces() in src/os.c.
6263 3. In the check_special_case() function in daemon.c I had used "errno" as an
6264 argument name, which causes warnings on some systems. This was basically a
6265 typo, since it was named "eno" in the comments!
6267 4. The code that waits for the clock to tick (at a resolution of some fraction
6268 of a second) so as to ensure message-id uniqueness was always waiting for
6269 at least one whole tick, when it could have waited for less. [This is
6270 almost certainly not relevant at current processor speeds, where it is
6271 unlikely to ever wait at all. But we try to future-proof.]
6273 5. The function that sleeps for a time interval that includes fractions of a
6274 second contained a race. It did not block SIGALRM between setting the
6275 timer, and suspending (a couple of lines later). If the interval was short
6276 and the sigsuspend() was delayed until after it had expired, the suspension
6277 never ended. On busy systems this could lead to processes getting stuck for
6280 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6281 process, before it forks any delivery processes. The open lookup caching
6282 mechanism meant that the open file or database connection was passed into
6283 the delivery process. The problem was that delivery processes always tidy
6284 up cached lookup data. This could cause a problem for the next delivery
6285 process started by the queue runner, because the external queue runner
6286 process does not know about the closure. So the next delivery process
6287 still has data in the lookup cache. In the case of a file lookup, there was
6288 no problem because closing a file descriptor in a subprocess doesn't affect
6289 the parent. However, if the lookup was caching a connection to a database,
6290 the connection was closed, and the second delivery process was likely to
6291 see errors such as "PGSQL: query failed: server closed the connection
6292 unexpectedly". The problem has been fixed by closing all cached lookups
6293 in a queue runner before running a delivery process.
6295 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6296 seem to have the "const" qualifier which it has on other OS. I've
6299 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6300 *of the same type* (client or server) with the same public name that an
6301 error should be diagnosed.
6303 9. When Exim looked up a host name for an IP address, but failed to find the
6304 original IP address when looking up the host name (a safety check), it
6305 output the message "<ip address> does not match any IP for NULL", which was
6306 confusing, to say the least. The bug was that the host name should have
6307 appeared instead of "NULL".
6309 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6310 user that is built into the binary, and the -C or -D options is used, root
6311 privilege is dropped before the configuration file is read. In addition,
6312 logging is switched to stderr instead of the normal log files. If the
6313 configuration then re-defines the Exim user, the unprivileged environment
6314 is probably not what is expected, so Exim logs a panic warning message (but
6317 However, if deliver_drop_privilege is set, the unprivileged state may well
6318 be exactly what is intended, so the warning has been cut out in that case,
6319 and Exim is allowed to try to write to its normal log files.
6325 1. smtp_return_error_details was not giving details for temporary sender
6326 or receiver verification errors.
6328 2. Diagnose a configuration error if two authenticators have the same public
6331 3. Exim used not to create the message log file for a message until the first
6332 delivery attempt. This could be confusing when incoming messages were held
6333 for policy or load reasons. The message log file is now created at the time
6334 the message is received, and an initial "Received" line is written to it.
6336 4. The automatically generated man page for command line options had a minor
6337 bug that caused no ill effects; however, a more serious problem was that
6338 the procedure for building the man page automatically didn't always
6339 operate. Consequently, release 4.20 contains an out-of-date version. This
6340 shouldn't happen again.
6342 5. When building Exim with embedded Perl support, the script that builds the
6343 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6344 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6346 6. The freeze_tell option was not being used for messages that were frozen on
6347 arrival, either by an ACL or by local_scan().
6349 7. Added the smtp_incomplete_transaction log selector.
6351 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6352 was accepting AUTH without a new EHLO.
6354 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6355 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6358 10. Logging of TCP/IP connections (when configured) now happens in the main
6359 daemon process instead of the child process, so that the TCP/IP connection
6360 count is more accurate (but it can never be perfect).
6362 11. The use of "drop" in a nested ACL was not being handled correctly in the
6363 outer ACL. Now, if condition failure induced by the nested "drop" causes
6364 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6365 or "require"), the connection is dropped.
6367 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6368 that yield "discard" can now be used with an "accept" or a "discard" verb,
6369 but an error is generated for any others (because I can't see a useful way
6370 to define what should happen).
6372 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6373 are now processed in the same way as lines in the Exim configuration file.
6374 In particular, continuation lines are supported.
6376 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6378 15. Added -ti meaning -t -i.
6380 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6381 domains, and warn by logging if others are found.
6383 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6384 returned from GIFCONF when Exim is trying to find the list of interfaces on
6385 a host. The code in os.c has been modified to copy each ifreq to an aligned
6386 structure in all cases.
6388 Also, in some cases, the returned ifreq's were being copied to a 'struct
6389 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6390 means the last couple of bytes of an IPv6 address could be chopped if the
6391 ifreq contained only a normal sockaddr (14 bytes storage).
6393 18. Named domain lists were not supported in the hosts_treat_as_local option.
6394 An entry such as +xxxx was not recognized, and was treated as a literal
6397 19. Ensure that header lines added by a DATA ACL are included in the reject log
6398 if the ACL subsequently rejects the message.
6400 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6401 MD5 (which is deprecated).
6403 21. When testing a filter file using -bf, Exim was writing a message when it
6404 took the sender from a "From " line in the message, but it was not doing so
6405 when it took $return_path from a Return-Path: header line. It now does.
6407 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6408 with a valid header line field name (a series of printing characters
6409 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6411 23. Changed "disc" in the source to "disk" to conform to the documentation and
6412 the book and for uniformity.
6414 24. Ignore Sendmail's -Ooption=value command line item.
6416 25. When execve() failed while trying to run a command in a pipe transport,
6417 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6418 could be confused with a return value of 69 from the command itself. This
6419 has been changed to 127, the value the shell returns if it is asked to run
6420 a non-existent command. The wording for the related log line suggests a
6421 non-existent command as the problem.
6423 26. If received_header_text expands to an empty string, do not add a Received:
6424 header line to the message. (Well, it adds a token one on the spool, but
6425 marks it "old" so that it doesn't get used or transmitted.)
6427 27. Installed eximstats 1.28 (addition of -nt option).
6429 28. There was no check for failure on the call to getsockname() in the daemon
6430 code. This can fail if there is a shortage of resources on the system, with
6431 ENOMEM, for example. A temporary error is now given on failure.
6433 29. Contrary to the C standard, it seems that in some environments, the
6434 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6435 program. Exim now does this explicitly; it affects the formatting of
6436 timestamps using strftime().
6438 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6439 complaints. I've now initialized all the variables, to avoid this.
6441 32. Header lines added by a system filter were not being "seen" during
6442 transport-time rewrites.
6444 33. The info_callback() function passed to OpenSSL is set up with type void
6445 (*)(SSL *, int, int), as described somewhere. However, when calling the
6446 function (actually a macro) that sets it up, the type void(*)() is
6447 expected. I've put in a cast to prevent warnings from picky compilers.
6449 34. If a DNS black list lookup found a CNAME record, but there were no A
6450 records associated with the domain it pointed at, Exim crashed.
6452 35. If a DNS black list lookup returned more than one A record, Exim ignored
6453 all but the first. It now scans all returned addresses if a particular IP
6454 value is being sought. In this situation, the contents of the
6455 $dnslist_value variable are a list of all the addresses, separated by a
6458 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6459 to accept a host name and all its aliases if the forward lookup for any of
6460 them yielded the IP address of the incoming connection. Now it accepts only
6461 those names whose forward lookup yields the correct IP address. Any other
6462 names are discarded. This closes a loophole whereby a rogue DNS
6463 administrator could create reverse DNS records to break through a
6464 wildcarded host restriction in an ACL.
6466 37. If a user filter or a system filter that ran in a subprocess used any of
6467 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6468 the wrong values were passed to the pipe command ($thisaddress had the
6469 value of $0, $0 had the value of $1, etc). This bug was introduced by
6470 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6472 38. Improved the line breaking for long SMTP error messages from ACLs.
6473 Previously, if there was no break point between 40 and 75 characters, Exim
6474 left the rest of the message alone. Two changes have been made: (a) I've
6475 reduced the minimum length to 35 characters; (b) if it can't find a break
6476 point between 35 and 75 characters, it looks ahead and uses the first one
6477 that it finds. This may give the occasional overlong line, but at least the
6478 remaining text gets split now.
6480 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6481 file descriptors might be low, and that setting 1000 would always raise it.
6482 It turns out that in some environments, the limit is already over 1000 and
6483 that lowering it causes trouble. So now Exim takes care not to decrease it.
6485 40. When delivering a message, the value of $return_path is set to $sender_
6486 address at the start of routing (routers may change the value). By an
6487 oversight, this default was not being set up when an address was tested by
6488 -bt or -bv, which affected the outcome if any router or filter referred to
6491 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6492 log only when "message" or "log_message" are set. However, if one of the
6493 conditions was an address verification, or a call to a nested ACL, the
6494 messages generated by the underlying test were being passed through. This
6495 no longer happens. The underlying message is available in $acl_verify_
6496 message for both "message" and "log_message" expansions, so it can be
6497 passed through if needed.
6499 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6500 new expansion $bh_ to give the encoded byte string without charset
6501 translation. Translation happens only if iconv() is available; HAVE_ICONV
6502 indicates this at build time. HEADERS_CHARSET gives the charset to
6503 translate to; headers_charset can change it in the configuration, and
6504 "headers charset" can change it in an individual filter file.
6506 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6507 that creates RFC 2047 encoded "words" labels them as that charset instead
6508 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6509 expansion operator; (ii) when Exim creates a From: line for a local
6510 message; (iii) when a header line is rewritten to include a "phrase" part.
6512 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6513 buggy, causing it to skip the first lines of messages whose message ID
6514 ended in 'D'. This would not have bitten before Exim release 4.14, because
6515 message IDs were unlikely to end in 'D' before then. The effect was to have
6516 incorrect size information for certain domains.
6518 45. #include "config.h" was missing at the start of the crypt16.c module. This
6519 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6522 46. If there was a timeout during a "random" callout check, Exim treated it as
6523 a failure of the random address, and carried on sending RSET and the real
6524 address. If the delay was just some slowness somewhere, the response to the
6525 original RCPT would be taken as a response to RSET and so on, causing
6526 mayhem of various kinds.
6528 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6529 when I implemented it. It didn't allow for the fact that some option values
6530 may legitimately be negative (e.g. size_addition), and it didn't even do
6531 the right test for positive values.
6533 48. Domain names in DNS records are case-independent. Exim always looks them up
6534 in lower case. Some resolvers return domain names in exactly the case they
6535 appear in the zone file, that is, they may contain uppercase letters. Not
6536 all resolvers do this - some return always lower case. Exim was treating a
6537 change of case by a resolver as a change of domain, similar to a widening
6538 of a domain abbreviation. This triggered its re-routing code and so it was
6539 trying to route what was effectively the same domain again. This normally
6540 caused routing to fail (because the router wouldn't handle the domain
6541 twice). Now Exim checks for this case specially, and just changes the
6542 casing of the domain that it ultimately uses when it transmits the message
6545 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6548 50. If a filter generated a file delivery with a non-absolute name (possible if
6549 no home directory exists for the router), the forbid_file option was not
6552 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6553 the existing equality matching.
6555 52. Exim was using ints instead of ino_t variables in some places where it was
6556 dealing with inode numbers.
6558 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6559 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6560 called TMPDIR, and if it finds it is different, it changes its value.
6562 54. The smtp_printf() function is now made available to local_scan() so
6563 additional output lines can be written before returning. There is also an
6564 smtp_fflush() function to enable the detection of a dropped connection.
6565 The variables smtp_input and smtp_batched_input are exported to
6568 55. Changed the default runtime configuration: the message "Unknown user"
6569 has been removed from the ACL, and instead placed on the localuser router,
6570 using the cannot_route_message feature. This means that any verification
6571 failures that generate their own messages won't get overridden. Similarly,
6572 the "Unrouteable address" message that was in the ACL for unverifiable
6573 relay addresses has also been removed.
6575 56. Added hosts_avoid_esmtp to the smtp transport.
6577 57. The exicyclog script was not checking for the esoteric option
6578 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6579 will work only if exicyclog is run under the appropriate euid.
6581 58. Following a discussion on the list, the rules by which Exim recognises line
6582 endings on incoming messages have been changed. The -dropcr and drop_cr
6583 options are now no-ops, retained only for backwards compatibility. The
6584 following line terminators are recognized: LF CRLF CR. However, special
6585 processing applies to CR:
6587 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6588 nor a local message in the state where . is a terminator.
6590 (ii) If a bare CR is encountered in a header line, an extra space is added
6591 after the line terminator so as not to end the header. The reasoning
6592 behind this is that bare CRs in header lines are most likely either
6593 to be mistakes, or people trying to play silly games.
6595 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6596 was being incorrectly given as 18 bytes larger than it should have been.
6597 This is a VOB (very old bug).
6599 60. This may never have affected anything current, but just in case it has:
6600 When the local host is found other than at the start of a list of hosts,
6601 the local host, those with the same MX, and any that follow, are discarded.
6602 When the list in question was part of a longer list of hosts, the following
6603 hosts (not currently being processed) were also being discarded. This no
6604 longer happens. I'm not sure if this situation could ever has previously
6607 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6610 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6611 header. This is something that is recommended in a new Internet Draft, and
6612 is something that is documented as being done by Sendmail. There are two
6613 possible values. For messages generated by the autoreply transport, Exim
6616 Auto-Submitted: auto-replied
6618 whereas for all other generated messages (e.g. bounces) it adds
6620 Auto-Submitted: auto-generated
6622 63. The "personal" condition in filters now includes a test for the
6623 Auto-Submitted: header. If it contains the string "auto-" the message it
6624 not considered personal.
6626 64. Added rcpt_include_affixes as a generic transport option.
6628 65. Added queue_only_override (default true).
6630 66. Added the syslog_duplication option.
6632 67. If what should have been the first header line of a message consisted of
6633 a space followed by a colon, Exim was mis-interpreting it as a header line.
6634 It isn't of course - it is syntactically invalid and should therefore be
6635 treated as the start of the message body. The misbehaviour could have
6636 caused a number of strange effects, including loss of data in subsequent
6637 header lines, and spool format errors.
6639 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6640 client host had authenticated. This control can now be exercised by an ACL
6641 for more flexibility.
6643 69. By default, callouts do not happen when testing with -bh. There is now a
6644 variant, -bhc, which does actually run the callout code, including
6645 consulting and updating the callout cache.
6647 70. Added support for saslauthd authentication, courtesy of Alexander
6650 71. If statvfs() failed on the spool or log directories while checking their
6651 size for availability, Exim confusingly gave the error "space shortage".
6652 Furthermore, in debugging mode it crashed with a floating point exception.
6653 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6654 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6655 serious problem, Exim now writes to the main and panic logs when this
6656 happens, with details of the failure. It then refuses to accept the
6657 incoming message, giving the message "spool directory problem" or "log
6658 directory problem" with a 421 code for SMTP messages.
6660 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6661 0, 1, and 2 exist, because some OS complain for execs without them (see
6662 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6663 descriptors. However, the code omitted to check that the open succeeded,
6664 causing mysterious errors if for some reason the permissions on /dev/null
6665 got screwed. Now Exim writes a message to the main and panic logs, and
6666 bombs out if it can't open /dev/null.
6668 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6669 interact so that it is all more flexible. It is supposed to remain
6670 backwards compatible. Also added extra_local_interfaces.
6672 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6673 to bomb out with an assertion failure - to the client this appears as a
6674 connection drop. This problem occurs in the part of the code that was taken
6675 from the Samba project. Fortunately, the assertion is in a very simple
6676 function, so I have fixed this by reproducing the function inline in the
6677 one place where it is called, and arranging for authentication to fail
6678 instead of killing the process with assert().
6680 75. The SPA client code was not working when the server requested OEM rather
6681 than Unicode encoding.
6683 76. Added code to make require_files with a specific uid setting more usable in
6684 the case where statting the file as root fails - usually a non-root-mounted
6685 NFS file system. When this happens and the failure is EACCES, Exim now
6686 forks a subprocess and does the per-uid checking as the relevant uid.
6688 77. Added process_log_path.
6690 78. If log_file_path was not explicitly set, a setting of check_log_space or
6691 check_log_inodes was ignored.
6693 79. If a space check for the spool or log partitions fails, the incident is now
6694 logged. Of course, in the latter case the data may get lost...
6696 80. Added the %p formatting code to string_format() so that it can be used to
6697 print addresses in debug_print(). Adjusted all the address printing in the
6698 debugging in store.c to use %p rather than %d.
6700 81. There was a concern that a line of code in smtp_in.c could overflow a
6701 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6702 initially expressed, the concern was not well-founded, because trailing
6703 spaces are removed early. However, if the trailing spaces were followed by
6704 a NULL, they did not get removed, so the overflow was possible. Two fixes
6707 (a) I re-wrote the offending code in a cleaner fashion.
6708 (b) If an incoming SMTP command contains a NULL character, it is rejected
6711 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6712 runs initgroups(), so that if the Exim user is in any additional groups,
6713 they will be used during message reception.
6719 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob