1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
12 than -2003 did; needs libidn2 in addition to libidn.
14 JH/02 The path option on a pipe transport is now expanded before use.
16 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
17 Patch provided by "Björn", documentation fix added too.
19 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
20 missing a wire-to-host endian conversion.
22 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
23 close after a BDAT command line could be taken as a following command,
24 giving a synch failure. Fix by only checking for synch immediately
25 before acknowledging the chunk.
27 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
28 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
29 macro. Patches provided by Josh Soref.
31 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
32 Previously we did not; the RFC seems ambiguous and VRFY is not listed
33 by IANA as a service extension. However, John Klensin suggests that we
36 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
37 the dkim code may be unix-mode line endings rather than smtp wire-format
38 CRLF, so prepend a CR to any bare LF.
40 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
41 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
43 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
44 the first were themselves being wrongly included in the feed into dkim
45 processing; with most chunk sizes in use this resulted in an incorrect
46 body hash calculated value.
48 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
49 DKIM signature block, for verification. Although advised against by
50 standards it is specifically not ruled illegal.
52 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
54 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
55 missing a body hash (the bh= tag).
57 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
58 It seems that HAProxy sends the Proxy Protocol information in clear and
59 only then does a TLS startup, so do the same.
61 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
62 TCP connections (such as for Spamd) unless the daemon successfully set
63 Fast Open mode on its listening sockets. This fixes breakage seen on
64 too-old kernels or those not configured for Fast Open, at the cost of
65 requiring both directions being enabled for TFO, and TFO never being used
66 by non-daemon-related Exim processes.
68 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
69 endings, at least on the first header line. Try to canonify any that get
70 past that check, despite the cost.
72 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
73 now limited to an arbitrary five deep, while parsing addresses with the
74 strip_excess_angle_brackets option enabled.
76 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
77 instead leave the unprompted TLS handshake in socket buffer for the
78 TLS library to consume.
80 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
82 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
84 JH/16 Drop variables when they go out of scope. Memory management drops a whole
85 region in one operation, for speed, and this leaves assigned pointers
86 dangling. Add checks run only under the testsuite which checks all
87 variables at a store-reset and panics on a dangling pointer; add code
88 explicitly nulling out all the variables discovered. Fixes one known
89 bug: a transport crash, where a dangling pointer for $sending_ip_address
90 originally assigned in a verify callout, is re-used.
96 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
97 supports it and a size is available (ie. the sending peer gave us one).
99 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
100 years of being deprecated). The replacements are the ACLs
101 acl_smtp_mime and acl_not_smtp_mime.
103 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
104 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
105 or even an in-clear connection were permitted. Now, if the host lookup
106 was dnssec and dane was requested then the host is only used if the
107 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
108 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
109 if one fails this test.
110 This means that a poorly-configured remote DNS will make it incommunicado;
111 but it protects against a DNS-interception attack on it.
113 JH/04 Bug 1810: make continued-use of an open smtp transport connection
114 non-noisy when a race steals the message being considered.
116 JH/05 If main configuration option tls_certificate is unset, generate a
117 self-signed certificate for inbound TLS connections.
119 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
120 in rewrites and routers.
122 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
123 and logged a warning sing 4.83; now they are a configuration file error.
125 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
126 (lacking @domain). Apply the same qualification processing as RCPT.
128 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
130 JH/10 Support ${sha256:} applied to a string (as well as the previous
133 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
134 a cutthrough deliver is pending, as we always want to make a connection.
135 This also avoids re-routing the message when later placing the cutthrough
136 connection after a verify cache hit.
137 Do not update it with the verify result either.
139 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
140 when routing results in more than one destination address.
142 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
143 signing (which inhibits the cutthrough capability). Previously only
144 the presence of an option was tested; now an expansion evaluating as
145 empty is permissible (obviously it should depend only on data available
146 when the cutthrough connection is made).
148 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
149 the relevant preceding SMTP command did not note the pipelining mode.
151 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
152 Previously they were not counted.
154 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
155 as one having no matching records. Previously we deferred the message
156 that needed the lookup.
158 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
159 distinguished as "(=".
161 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
162 for missing MX records. Previously it only worked for missing A records.
164 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
166 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
167 after the data-go-ahead and data-ack. Patch from Jason Betts.
169 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
170 even for a "none" policy. Patch from Tony Meyer.
172 JH/22 Fix continued use of a connection for further deliveries. If a port was
173 specified by a router, it must also match for the delivery to be
176 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
177 When one of the recipients of a message was unsuitable for the connection
178 (has no matching addresses), we lost track of needing to mark it
179 deferred. As a result mail would be lost.
181 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
183 JH/25 Decoding ACL controls is now done using a binary search; the source code
184 takes up less space and should be simpler to maintain. Merge the ACL
185 condition decode tables also, with similar effect.
187 JH/26 Fix problem with one_time used on a redirect router which returned the
188 parent address unchanged. A retry would see the parent address marked as
189 delivered, so not attempt the (identical) child. As a result mail would
192 JH/27 Fix a possible security hole, wherein a process operating with the Exim
193 UID can gain a root shell. Credit to http://www.halfdog.net/ for
194 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
197 JH/28 Enable {spool,log} filesystem space and inode checks as default.
198 Main config options check_{log,spool}_{inodes,space} are now
199 100 inodes, 10MB unless set otherwise in the configuration.
201 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
202 Previously it only applied to the main-section connection policy
205 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
207 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
208 by me. Added RFC7919 DH primes as an alternative.
210 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
211 are not in the system include path.
213 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
214 GnuTLS, when a session startup failed (eg because the client disconnected)
215 Exim did stdio operations after fclose. This was exposed by a recent
216 change which nulled out the file handle after the fclose.
218 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
219 signed directly by the cert-signing cert, rather than an intermediate
220 OCSP-signing cert. This is the model used by LetsEncrypt.
222 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
224 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
225 an incoming connection.
227 HS/02 Bug 1802: Do not half-close the connection after sending a request
230 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
231 fallback to "prime256v1".
233 JH/34 SECURITY: Use proper copy of DATA command in error message.
234 Could leak key material. Remotely exploitable. CVE-2016-9963.
240 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
241 and 3.4.4 - once the server is enabled to respond to an OCSP request
242 it does even when not requested, resulting in a stapling non-aware
243 client dropping the TLS connection.
245 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
246 support variable-length bit vectors. No functional change.
248 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
249 The I= interface field on outgoing lines is now after the H= remote
250 host field, same as incoming lines. There is a separate
251 outgoing_interface log selector which allows you to disable the
254 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
255 If not running log_selector +smtp_connection the mainlog would be held
256 open indefinitely after a "too many connections" event, including to a
257 deleted file after a log rotate. Leave the per net connection logging
258 leaving it open for efficiency as that will be quickly detected by the
259 check on the next write.
261 HS/01 Bug 1671: Fix post transport crash.
262 Processing the wait-<transport> messages could crash the delivery
263 process if the message IDs didn't exist for some reason. When
264 using 'split_spool_directory=yes' the construction of the spool
265 file name failed already, exposing the same netto behaviour.
267 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
268 mime_regex ACL conditions.
270 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
271 to DSN fail messages (bounces): remote IP, remote greeting, remote response
272 to HELO, local diagnostic string.
274 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
275 log line to debug. Even when configured with a tls authenticator many
276 client connections are expected to not authenticate in this way, so
277 an authenticate fail is not an error.
279 HS/02 Add the Exim version string to the process info. This way exiwhat
280 gives some more detail about the running daemon.
282 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
283 matter for fast-change records such as DNSBLs.
285 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
286 retry record, even if constant. There may be multiple transports with
287 different interface settings and the retry behaviour needs to be kept
290 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
292 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
294 JH/10 Bug 840: fix log_defer_output option of pipe transport
296 JH/11 Bug 830: use same host for all RCPTS of a message, even under
297 hosts_randomize. This matters a lot when combined with mua_wrapper.
299 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
300 ${quote_pgsql:<string>} operator.
302 JH/13 Bug 1708: avoid misaligned access in cached lookup.
304 JH/14 Change header file name for freeradius-client. Relevant if compiling
305 with Radius support; from the Gentoo tree and checked under Fedora.
307 JH/15 Bug 1712: Introduce $prdr_requested flag variable
309 JH/16 Bug 1714: Permit an empty string as expansion result for transport
310 option transport_filter, meaning no filtering.
312 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
314 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
315 defaults to "*" (all hosts). The variable is now available when not built
316 with TLS, default unset, mainly to enable keeping the testsuite sane.
317 If a server certificate is not supplied (via tls_certificate) an error is
318 logged, and clients will find TLS connections fail on startup. Presumably
319 they will retry in-clear.
320 Packagers of Exim are strongly encouraged to create a server certificate
321 at installation time.
323 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
324 with the $config_file variable.
326 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
327 in transport context, after the attempt, and per-recipient. The latter type
328 is per host attempted. The event data is the error message, and the errno
329 information encodes the lookup type (A vs. MX) used for the (first) host,
330 and the trailing two digits of the smtp 4xx response.
332 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
333 to write to mainlog (or rejectlog, paniclog) in the window between file
334 creation and permissions/ownership being changed. Particularly affects
335 installations where exicyclog is run as root, rather than exim user;
336 result is that the running daemon panics and dies.
338 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
340 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
341 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
342 "pri" and "weight". Note that the previous implicit priority given by the
343 list order is no longer honoured.
345 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
348 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
349 by defining SUPPORT_SOCKS.
351 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
352 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
353 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
354 variables are renamed to proxy_{local,external}_{address,port}.
356 JH/27 Move Internationalisation support from Experimental to mainline, enabled
357 for a build by defining SUPPORT_I18N
359 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
360 of the query string, and make ${quote_redis:} do that quoting.
362 JH/29 Move Events support from Experimental to mainline, enabled by default
363 and removable for a build by defining DISABLE_EVENT.
365 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
367 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
368 cached by the daemon.
370 JH/32 Move Redis support from Experimental to mainline, enabled for a build
371 by defining LOOKUP_REDIS. The libhiredis library is required.
373 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
374 keys are given for lookup.
376 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
377 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
378 only supported when built with TLS support. The PolarSSL SHA routines
379 are still used when the TLS library is too old for convenient support.
381 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
382 openssl_options), for security. OpenSSL forces this from version 1.1.0
383 server-side so match that on older versions.
385 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
386 allocation for $value could be released as the expansion processing
387 concluded, but leaving the global pointer active for it.
389 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
390 and to use the domains and local_parts ACL conditions.
392 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
393 incorrectly not doubled on cutthrough transmission, hence seen as a
394 body-termination at the receiving system - resulting in truncated mails.
395 Commonly the sender saw a TCP-level error, and retransmitted the message
396 via the normal store-and-forward channel. This could result in duplicates
397 received - but deduplicating mailstores were liable to retain only the
398 initial truncated version.
400 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
402 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
404 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
405 we're in there, support oversigning also; bug 1309.
407 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
409 HS/04 Add support for keep_environment and add_environment options.
411 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
412 either intentional arithmetic overflow during PRNG, or testing config-
415 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
416 delivery resulted in actual delivery. Cancel cutthrough before DATA
419 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
420 rejects a recipient: pass the reject to the originator.
422 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
423 Many were false-positives and ignorable, but it's worth fixing the
426 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
427 for the new environment-manipulation done at startup. Move the routines
428 from being local to tls.c to being global via the os.c file.
430 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
431 an extract embedded as result-arg for a map, the first arg for extract
432 is unavailable so we cannot tell if this is a numbered or keyed
433 extraction. Accept either.
439 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
442 JH/02 The smtp transport option "multi_domain" is now expanded.
444 JH/03 The smtp transport now requests PRDR by default, if the server offers
447 JH/04 Certificate name checking on server certificates, when exim is a client,
448 is now done by default. The transport option tls_verify_cert_hostnames
449 can be used to disable this per-host. The build option
450 EXPERIMENTAL_CERTNAMES is withdrawn.
452 JH/05 The value of the tls_verify_certificates smtp transport and main options
453 default to the word "system" to access the system default CA bundle.
454 For GnuTLS, only version 3.0.20 or later.
456 JH/06 Verification of the server certificate for a TLS connection is now tried
457 (but not required) by default. The verification status is now logged by
458 default, for both outbound TLS and client-certificate supplying inbound
461 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
464 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
465 Status Notification (bounce) messages are now MIME format per RFC 3464.
466 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
467 under the control of the dsn_advertise_hosts option, and routers may
468 have a dsn_lasthop option.
470 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
471 default, modifiable by a malware= option. The list separator for
472 the options can now be changed in the usual way. Bug 68.
474 JH/10 The smtp_receive_timeout main option is now expanded before use.
476 JH/11 The incoming_interface log option now also enables logging of the
477 local interface on delivery outgoing connections.
479 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
480 if the interface and destination host and port all match.
482 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
485 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
486 Patch from Andrew Lewis.
488 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
489 now supports optional time-restrictions, weighting, and priority
490 modifiers per server. Patch originally by <rommer@active.by>.
492 JH/16 The spamd_address main option now supports a mixed list of local
493 and remote servers. Remote servers can be IPv6 addresses, and
494 specify a port-range.
496 JH/17 Bug 68: The spamd_address main option now supports an optional
497 timeout value per server.
499 JH/18 Bug 1581: Router and transport options headers_add/remove can
500 now have the list separator specified.
502 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
505 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
508 JH/21 Support for the A6 type of dns record is withdrawn.
510 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
511 rather than the verbs used.
513 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
514 from 255 to 1024 chars.
516 JH/24 Verification callouts now attempt to use TLS by default.
518 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
519 are generic router options now. The defaults didn't change.
521 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
522 Original patch from Alexander Shikoff, worked over by JH.
524 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
525 files if ZCAT_COMMAND is not executable.
527 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
529 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
531 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
532 Normally benign, it bites when the pair was led to by a CNAME;
533 modern usage is to not canonicalize the domain to a CNAME target
534 (and we were inconsistent anyway for A-only vs AAAA+A).
536 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
538 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
539 when evaluating $sender_host_dnssec.
541 JH/31 Check the HELO verification lookup for DNSSEC, adding new
542 $sender_helo_dnssec variable.
544 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
546 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
548 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
550 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
551 documented as working, but never had. Support all but $spam_report.
553 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
554 added for tls authenticator.
556 HS/03 Add perl_taintmode main config option
562 TL/01 When running the test suite, the README says that variables such as
563 no_msglog_check are global and can be placed anywhere in a specific
564 test's script, however it was observed that placement needed to be near
565 the beginning for it to behave that way. Changed the runtest perl
566 script to read through the entire script once to detect and set these
567 variables, reset to the beginning of the script, and then run through
568 the script parsing/test process like normal.
570 TL/02 The BSD's have an arc4random API. One of the functions to induce
571 adding randomness was arc4random_stir(), but it has been removed in
572 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
573 function when detected.
575 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
576 cause callback expansion.
578 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
579 syntax errors in an expansion can be treated as a string instead of
580 logging or causing an error, due to the internal use of bool_lax
581 instead of bool when processing it.
583 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
584 server certificates when making smtp deliveries.
586 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
588 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
590 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
592 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
593 Merged patch from Sebastian Wiedenroth.
595 JH/05 Fix results-pipe from transport process. Several recipients, combined
596 with certificate use, exposed issues where response data items split
597 over buffer boundaries were not parsed properly. This eventually
598 resulted in duplicates being sent. This issue only became common enough
599 to notice due to the introduction of connection certificate information,
600 the item size being so much larger. Found and fixed by Wolfgang Breyha.
602 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
603 size buffer was used, resulting in syntax errors when an expansion
606 JH/07 Add support for directories of certificates when compiled with a GnuTLS
607 version 3.3.6 or later.
609 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
610 is EXPERIMENTAL_EVENT, the main-configuration and transport options
611 both become "event_action", the variables become $event_name, $event_data
612 and $event_defer_errno. There is a new variable $verify_mode, usable in
613 routers, transports and related events. The tls:cert event is now also
614 raised for inbound connections, if the main configuration event_action
617 TL/06 In test suite, disable OCSP for old versions of openssl which contained
618 early OCSP support, but no stapling (appears to be less than 1.0.0).
620 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
621 server certificate names available under the smtp transport option
622 "tls_verify_cert_hostname" now do not permit multi-component wildcard
625 JH/10 Time-related extraction expansions from certificates now use the main
626 option "timezone" setting for output formatting, and are consistent
627 between OpenSSL and GnuTLS compilations. Bug 1541.
629 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
630 encoded parameter in the incoming message. Bug 1558.
632 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
633 include certificate info, eximon was claiming there were spoolfile
636 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
638 JH/14 Log delivery-related information more consistently, using the sequence
639 "H=<name> [<ip>]" wherever possible.
641 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
642 are problematic for Debian distribution, omit them from the release
645 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
647 JH/16 Fix string representation of time values on 64bit time_t architectures.
650 JH/17 Fix a null-indirection in certextract expansions when a nondefault
651 output list separator was used.
656 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
657 checkers that were complaining about end of non-void function with no
660 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
661 This was a regression introduced in 4.83 by another bugfix.
663 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
665 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
666 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
672 TF/01 Correctly close the server side of TLS when forking for delivery.
674 When a message was received over SMTP with TLS, Exim failed to clear up
675 the incoming connection properly after forking off the child process to
676 deliver the message. In some situations the subsequent outgoing
677 delivery connection happened to have the same fd number as the incoming
678 connection previously had. Exim would try to use TLS and fail, logging
679 a "Bad file descriptor" error.
681 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
682 utilities have not been installed.
684 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
685 temporary space as the ACL may create new global variables.
687 TL/01 LDAP support uses per connection or global context settings, depending
688 upon the detected version of the libraries at build time.
690 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
691 to extract and use the src ip:port in logging and expansions as if it
692 were a direct connection from the outside internet. PPv2 support was
693 updated based on HAProxy spec change in May 2014.
695 JH/02 Add ${listextract {number}{list}{success}{fail}}.
697 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
698 Properly escape header and check for NULL return.
700 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
703 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
705 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
706 characters in header names, implemented as a verify condition.
707 Contributed by Michael Fischer v. Mollard.
709 TL/05 Rename SPF condition results err_perm and err_temp to standardized
710 results permerror and temperror. Previous values are deprecated but
711 still accepted. In a future release, err_perm and err_temp will be
712 completely removed, which will be a backward incompatibility if the
713 ACL tests for either of these two old results. Patch contributed by
714 user bes-internal on the mailing list.
716 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
718 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
719 selectors, in both main and reject logs.
721 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
724 JH/07 Add malware type "sock" for talking to simple daemon.
726 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
728 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
729 routers/transports under cutthrough routing.
731 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
732 numbers. Touch up "bool" conditional to keep the same definition.
734 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
736 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
738 JH/12 Expand items in router/transport headers_add or headers_remove lists
739 individually rather than the list as a whole. Bug 1452.
741 Required for reasonable handling of multiple headers_ options when
742 they may be empty; requires that headers_remove items with embedded
743 colons must have them doubled (or the list-separator changed).
745 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
746 view the policy declared in the DMARC record. Currently, $dmarc_status
747 is a combined value of both the record presence and the result of the
750 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
752 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
753 dnslookup router and the smtp transport (applying to the forward
756 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
757 of ldap servers used for a specific lookup. Patch provided by Heiko
760 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
761 New variable $lookup_dnssec_authenticated for observability.
763 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
764 Patch submitted by Lars Timman.
766 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
768 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
769 Requires trusted mode and valid format message id, aborts otherwise.
770 Patch contributed by Heiko Schlichting.
772 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
773 certextract with support for various fields. Bug 1358.
775 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
776 is requested by default, modifiable by smtp transport option
779 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
780 operate on certificate variables to give certificate fingerprints
781 Also new ${sha256:cert_variable}.
783 JH/23 The PRDR feature is moved from being Experimental into the mainline.
785 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
786 Christian Aistleitner.
788 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
790 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
791 file. Patch from Wolfgang Breyha.
793 JH/25 Expand the coverage of the delivery $host and $host_address to
794 client authenticators run in verify callout. Bug 1476.
796 JH/26 Port service names are now accepted for tls_on_connect_ports, to
797 align with daemon_smtp_ports. Bug 72.
799 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
800 support and error reporting did not work properly.
802 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
803 and is readable. Patch from Andrew Colin Kissa.
805 TL/14 Enhance documentation of ${run expansion and how it parses the
806 commandline after expansion, particularly in the case when an
807 unquoted variable expansion results in an empty value.
809 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
811 PP/02 Fix internal collision of T_APL on systems which support RFC3123
812 by renaming away from it. Addresses GH issue 15, reported by
815 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
817 TL/15 SECURITY: prevent double expansion in math comparison functions
818 (can expand unsanitized data). Not remotely exploitable.
825 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
827 PP/02 Make -n do something, by making it not do something.
828 When combined with -bP, the name of an option is not output.
830 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
833 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
834 $sender_host_name and config options to manage this, and basic check
837 PP/05 DSCP support for outbound connections and control modifier for inbound.
839 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
840 (Only plugin which currently uses this is kerberos4, which nobody should
841 be using, but we should make it available and other future plugins might
842 conceivably use it, even though it would break NAT; stuff *should* be
843 using channel bindings instead).
845 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
846 name; added for Sendmail compatibility; requires admin caller.
847 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
848 just ignore it); requires trusted caller.
849 Also parse but ignore: -Ac -Am -X<logfile>
852 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
854 TL/02 Add +smtp_confirmation as a default logging option.
856 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
857 Patch by Magnus Holmgren from 2007-02-20.
859 TL/04 Bugzilla 1281 - Spec typo.
860 Bugzilla 1283 - Spec typo.
861 Bugzilla 1290 - Spec grammar fixes.
863 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
865 TL/06 Add Experimental DMARC support using libopendmarc libraries.
867 TL/07 Fix an out of order global option causing a segfault. Reported to dev
868 mailing list by by Dmitry Isaikin.
870 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
872 JH/02 Support "G" suffix to numbers in ${if comparisons.
874 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
876 NM/01 Bugzilla 1197 - Spec typo
877 Bugzilla 1196 - Spec examples corrections
879 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
881 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
882 gnutls_enable_pkcs11, but renamed to more accurately indicate its
885 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
886 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
888 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
889 "acl {{name}{arg}...}", and optional args on acl condition
892 JH/05 Permit multiple router/transport headers_add/remove lines.
894 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
896 JH/07 Avoid using a waiting database for a single-message-only transport.
897 Performance patch from Paul Fisher. Bugzilla 1262.
899 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
902 JH/09 Add $headers_added variable, with content from use of ACL modifier
903 add_header (but not yet added to the message). Bugzilla 199.
905 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
906 Pulled from Bugzilla 817 by Wolfgang Breyha.
908 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
910 (nb: this is the same fix as in Exim 4.80.1)
912 JH/11 Add A= logging on delivery lines, and a client_set_id option on
915 JH/12 Add optional authenticated_sender logging to A= and a log_selector
918 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
920 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
921 advertise SMTP AUTH mechanism to us, instead of a generic
922 protocol violation error. Also, make Exim more robust to bad
923 data from the Dovecot auth socket.
925 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
927 When a queue runner is handling a message, Exim first routes the
928 recipient addresses, during which it prunes them based on the retry
929 hints database. After that it attempts to deliver the message to
930 any remaining recipients. It then updates the hints database using
933 So if a recipient address works intermittently, it can get repeatedly
934 deferred at routing time. The retry hints record remains fresh so the
935 address never reaches the final cutoff time.
937 This is a fairly common occurrence when a user is bumping up against
938 their storage quota. Exim had some logic in its local delivery code
939 to deal with this. However it did not apply to per-recipient defers
940 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
942 This change adds a proper retry rule check during routing so that the
943 final cutoff time is checked against the message's age. We only do
944 this check if there is an address retry record and there is not a
945 domain retry record; this implies that previous attempts to handle
946 the address had the retry_use_local_parts option turned on. We use
947 this as an approximation for the destination being like a local
948 delivery, as in LMTP.
950 I suspect this new check makes the old local delivery cutoff check
951 redundant, but I have not verified this so I left the code in place.
953 TF/02 Correct gecos expansion when From: is a prefix of the username.
955 Test 0254 submits a message to Exim with the header
959 When I ran the test suite under the user fanf2, Exim expanded
960 the header to contain my full name, whereas it should have added
961 a Resent-Sender: header. It erroneously treats any prefix of the
962 username as equal to the username.
964 This change corrects that bug.
966 GF/01 DCC debug and logging tidyup
967 Error conditions log to paniclog rather than rejectlog.
968 Debug lines prefixed by "DCC: " to remove any ambiguity.
970 TF/03 Avoid unnecessary rebuilds of lookup-related code.
972 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
973 Bug spotted by Jeremy Harris; was flawed since initial commit.
974 Would have resulted in OCSP responses post-SNI triggering an Exim
975 NULL dereference and crash.
977 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
979 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
980 Bug detection, analysis and fix by Samuel Thibault.
981 Bugzilla 1331, Debian bug #698092.
983 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
985 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
986 Server implementation by Todd Lyons, client by JH.
987 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
988 config variable "prdr_enable" controls whether the server
989 advertises the facility. If the client requests PRDR a new
990 acl_data_smtp_prdr ACL is called once for each recipient, after
991 the body content is received and before the acl_smtp_data ACL.
992 The client is controlled by both of: a hosts_try_prdr option
993 on the smtp transport, and the server advertisement.
994 Default client logging of deliveries and rejections involving
995 PRDR are flagged with the string "PRDR".
997 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
998 fclose(). Diagnosis by Todd Lyons.
1000 PP/17 Update configure.default to handle IPv6 localhost better.
1001 Patch by Alain Williams (plus minor tweaks).
1004 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1005 This is now consistent with GnuTLS, and is now documented: the
1006 previous undocumented portable approach to treating the option as
1007 unset was to force an expansion failure. That still works, and
1008 an empty string is now equivalent.
1010 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1011 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1012 not performing validation itself.
1014 PP/20 Added force_command boolean option to pipe transport.
1015 Patch from Nick Koston, of cPanel Inc.
1017 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1020 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1022 PP/21 Fix eximon continuous updating with timestamped log-files.
1023 Broken in a format-string cleanup in 4.80, missed when I repaired the
1024 other false fix of the same issue.
1025 Report and fix from Heiko Schlichting.
1028 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1029 Report from Prashanth Katuri.
1031 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1032 It's SecureTransport, so affects any MacOS clients which use the
1033 system-integrated TLS libraries, including email clients.
1035 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1036 using a MIME ACL for non-SMTP local injection.
1037 Report and assistance in diagnosis by Warren Baker.
1039 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1041 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1043 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1044 last id that failed so it may be referenced in subsequent ACL's.
1046 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1049 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1050 ldap library initialization, allowing self-signed CA's to be
1051 used. Also properly sets require_cert option later in code by
1052 using NULL (global ldap config) instead of ldap handle (per
1053 session). Bug diagnosis and testing by alxgomz.
1055 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1056 the src/util/ subdirectory.
1058 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1059 renamed to Transport Post Delivery Action by Jeremy Harris, as
1062 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1063 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1064 redis_servers = needs to be configured which will be used by the redis
1065 lookup. Patch from Warren Baker, of The Packet Hub.
1067 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1069 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1070 hostname or reverse DNS when processing a host list. Used suggestions
1071 from multiple comments on this bug.
1073 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1075 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1076 Missed a few lines, added it to make the runtest require no keyboard
1079 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1080 contains upper case chars. Make router use caseful_local_part.
1082 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1083 support when GnuTLS has been built with p11-kit.
1089 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1091 This, or similar/improved, will also be change PP/11 of 4.82.
1097 PP/01 Handle short writes when writing local log-files.
1098 In practice, only affects FreeBSD (8 onwards).
1099 Bugzilla 1053, with thanks to Dmitry Isaikin.
1101 NM/01 Bugzilla 949 - Documentation tweak
1103 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1106 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1108 PP/02 Implemented gsasl authenticator.
1110 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1112 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1113 `pkg-config foo` for cflags/libs.
1115 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1116 with rest of GSASL and with heimdal_gssapi.
1118 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1119 `pkg-config foo` for cflags/libs for the TLS implementation.
1121 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1122 properties get this fed in as external SSF. A number of robustness
1123 and debugging improvements to the cyrus_sasl authenticator.
1125 PP/08 cyrus_sasl server now expands the server_realm option.
1127 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1128 Patch by Jeremy Harris.
1130 PP/10 Added dbmjz lookup type.
1132 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1134 PP/12 MAIL args handles TAB as well as SP, for better interop with
1135 non-compliant senders.
1136 Analysis and variant patch by Todd Lyons.
1138 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1139 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1140 Patch from Dirk Mueller <dmueller@suse.com>
1142 PP/13 tls_peerdn now print-escaped for spool files.
1143 Observed some $tls_peerdn in wild which contained \n, which resulted
1144 in spool file corruption.
1146 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1147 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1148 or write after TLS renegotiation, which otherwise led to messages
1151 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1152 as a tracking header (ie: a signed header comes before the signature).
1153 Patch from Wolfgang Breyha.
1155 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1156 comma-sep list; embedded commas doubled.
1158 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1160 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1162 Report and patch from Dmitry Banschikov.
1164 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1165 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1166 failures. We appear to now support TLS1.1+ with Exim.
1168 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1169 lets Exim select keys and certificates based upon TLS SNI from client.
1170 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1171 before an outbound SMTP session. New log_selector, +tls_sni.
1173 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1174 NULL dereference. Report and patch from Alun Jones.
1176 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1177 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1178 Not seeing resolver debug output on NetBSD, but suspect this is a
1179 resolver implementation change.
1181 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1182 Left warnings. Added "eximon gdb" invocation mode.
1184 PP/21 Defaulting "accept_8bitmime" to true, not false.
1186 PP/22 Added -bw for inetd wait mode support.
1188 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1189 locate the relevant includes and libraries. Made this the default.
1191 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1192 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1194 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1195 This may cause build issues on older platforms.
1197 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1198 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1199 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1200 Added SNI support via GnuTLS too.
1201 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1203 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1205 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1206 Applied second patch from Janne, implementing suggestion to default
1207 multiple-strings-in-record handling to match SPF spec.
1209 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1211 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1212 read-only, out of scope).
1213 Patch from Wolfgang Breyha, report from Stuart Northfield.
1215 PP/29 Fix three issues highlighted by clang analyser static analysis.
1216 Only crash-plausible issue would require the Cambridge-specific
1217 iplookup router and a misconfiguration.
1218 Report from Marcin Mirosław.
1220 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1222 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1223 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1224 As part of this, removing so much warning spew let me fix some minor
1225 real issues in debug logging.
1227 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1228 assignment on my part. Fixed.
1230 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1231 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1232 Janne Snabb (who went above and beyond: thank you).
1234 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1235 string otherwise requires a connection and a bunch more work and it's
1236 relatively easy to get wrong. Should also expose TLS library linkage
1239 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1240 64-bit ${eval} (JH/03).
1242 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1243 GNU libc to support some of the 64-bit stuff, should not lead to
1244 conflicts. Defined before os.h is pulled in, so if a given platform
1245 needs to override this, it can.
1247 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1248 protection layer was required, which is not implemented.
1249 Bugzilla 1254, patch from Wolfgang Breyha.
1251 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1252 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1253 tls_dhparam take prime identifiers. Also unbreak combination of
1254 OpenSSL+DH_params+TLSSNI.
1256 PP/39 Disable SSLv2 by default in OpenSSL support.
1262 PP/01 Solaris build fix for Oracle's LDAP libraries.
1263 Bugzilla 1109, patch from Stephen Usher.
1265 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1267 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1270 TF/02 Fix a couple more cases where we did not log the error message
1271 when unlink() failed. See also change 4.74-TF/03.
1273 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1274 lock up or crash if it happened to be inside a call to libc when it
1275 got a SIGUSR1 from exiwhat.
1277 The SIGUSR1 handler appends the current process status to the process
1278 log which is later printed by exiwhat. It used to use the general
1279 purpose logging code to do this, but several functions it calls are
1280 not safe for signals.
1282 The new output code in the SIGUSR1 handler is specific to the process
1283 log, and simple enough that it's easy to inspect for signal safety.
1284 Removing some special cases also simplifies the general logging code.
1285 Removing the spurious timestamps from the process log simplifies
1288 TF/04 Improved ratelimit ACL condition.
1290 The /noupdate option has been deprecated in favour of /readonly which
1291 has clearer semantics. The /leaky, /strict, and /readonly update modes
1292 are mutually exclusive. The update mode is no longer included in the
1293 database key; it just determines when the database is updated. (This
1294 means that when you upgrade Exim will forget old rate measurements.)
1296 Exim now checks that the per_* options are used with an update mode that
1297 makes sense for the current ACL. For example, when Exim is processing a
1298 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1299 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1300 must specify per_mail/readonly. If you omit the update mode it defaults to
1301 /leaky where that makes sense (as before) or /readonly where required.
1303 The /noupdate option is now undocumented but still supported for
1304 backwards compatibility. It is equivalent to /readonly except that in
1305 ACLs where /readonly is required you may specify /leaky/noupdate or
1306 /strict/noupdate which are treated the same as /readonly.
1308 A useful new feature is the /count= option. This is a generalization
1309 of the per_byte option, so that you can measure the throughput of other
1310 aggregate values. For example, the per_byte option is now equivalent
1311 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1313 The per_rcpt option has been generalized using the /count= mechanism
1314 (though it's more complicated than the per_byte equivalence). When it is
1315 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1316 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1317 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1318 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1319 Note that using per_rcpt with a non-readonly update mode in more than
1320 one ACL will cause the recipients to be double-counted. (The per_mail
1321 and per_byte options don't have this problem.)
1323 The handling of very low rates has changed slightly. If the computed rate
1324 is less than the event's count (usually one) then this event is the first
1325 after a long gap. In this case the rate is set to the same as this event's
1326 count, so that the first message of a spam run is counted properly.
1328 The major new feature is a mechanism for counting the rate of unique
1329 events. The new per_addr option counts the number of different
1330 recipients that someone has sent messages to in the last time period. It
1331 behaves like per_rcpt if all the recipient addresses are different, but
1332 duplicate recipient addresses do not increase the measured rate. Like
1333 the /count= option this is a general mechanism, so the per_addr option
1334 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1335 example, measure the rate that a client uses different sender addresses
1336 with the options per_mail/unique=$sender_address. There are further
1337 details in the main documentation.
1339 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1341 TF/06 Removed a few PCRE remnants.
1343 TF/07 Automatically extract Exim's version number from tags in the git
1344 repository when doing development or release builds.
1346 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1347 Bugzilla 879. Patch from Paul Fisher.
1349 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1350 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1353 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1355 PP/05 Variable $av_failed, true if the AV scanner deferred.
1356 Bugzilla 1078. Patch from John Horne.
1358 PP/06 Stop make process more reliably on build failure.
1359 Bugzilla 1087. Patch from Heiko Schlittermann.
1361 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1362 Bugzilla 1089. Patch from Heiko Schlittermann.
1364 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1365 Bugzilla 1131. Patch from Holger Weiß.
1367 PP/09 Handle IPv6 addresses with SPF.
1368 Bugzilla 860. Patch from Wolfgang Breyha.
1370 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1372 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1375 PP/11 match_* no longer expand right-hand-side by default.
1376 New compile-time build option, EXPAND_LISTMATCH_RHS.
1377 New expansion conditions, "inlist", "inlisti".
1379 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1381 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1383 PP/14 fix log_write() format string regression from TF/03.
1384 Bugzilla 1152. Patch from Dmitry Isaikin.
1390 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1392 PP/02 Harmonised TLS library version reporting; only show if debugging.
1393 Layout now matches that introduced for other libraries in 4.74 PP/03.
1395 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1397 PP/04 New "dns_use_edns0" global option.
1399 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1402 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1403 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1405 TK/01 Updated PolarSSL code to 0.14.2.
1406 Bugzilla 1097. Patch from Andreas Metzler.
1408 PP/07 Catch divide-by-zero in ${eval:...}.
1409 Fixes bugzilla 1102.
1411 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1414 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1415 format-string attack -- SECURITY: remote arbitrary code execution.
1417 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1418 time unintentionally subject to list matching rules, letting the header
1419 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1420 arbitrary string expansion). This allowed for information disclosure.
1422 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1423 INT_MIN/-1 -- value coerced to INT_MAX.
1429 NM/01 Workaround for PCRE version dependency in version reporting
1432 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1433 This fixes portability to compilers other than gcc, notably
1434 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1436 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1437 makefiles for portability to HP-UX and POSIX correctness.
1439 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1440 Also via indented variable definition in the Makefile.
1441 (Debugging by Oliver Heesakkers).
1443 PP/02 Restore caching of spamd results with expanded spamd_address.
1444 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1446 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1447 Improves build reliability. Fix from: Frank Elsner
1449 NM/02 Fix wide character breakage in the rfc2047 coding
1450 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1452 NM/03 Allow underscore in dnslist lookups
1453 Fixes bug 1026. Patch from Graeme Fowler
1455 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1456 Code patches from Adam Ciarcinski of NetBSD.
1458 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1461 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1462 is logged, to avoid truncation. Patch from John Horne.
1464 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1465 Patch from Jakob Hirsch.
1467 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1468 SQL string expansion failure details.
1469 Patch from Andrey Oktyabrski.
1471 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1472 Patch from Simon Arlott.
1474 PP/09 New lookups functionality failed to compile on old gcc which rejects
1475 extern declarations in function scope.
1476 Patch from Oliver Fleischmann
1478 PP/10 Use sig_atomic_t for flags set from signal handlers.
1479 Check getgroups() return and improve debugging.
1480 Fixed developed for diagnosis in bug 927 (which turned out to be
1483 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1484 Patch from Mark Zealey.
1486 PP/12 Bugzilla 1056: Improved spamd server selection.
1487 Patch from Mark Zealey.
1489 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1490 Based on patch from Heiko Schlittermann.
1492 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1493 Patch from Uwe Doering, sign-off by Michael Haardt.
1495 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1496 variable declaration deep within a block. Bug and patch from
1499 PP/15 lookups-Makefile IRIX compatibility coercion.
1501 PP/16 Make DISABLE_DKIM build knob functional.
1503 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1504 Patch by Simon Arlott
1506 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1507 variable argument macros. Our copy now differs from upstream.
1513 TF/01 Failure to get a lock on a hints database can have serious
1514 consequences so log it to the panic log.
1516 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1517 controlled using the smtp_confirmation log selector.
1519 TF/03 Include the error message when we fail to unlink a spool file.
1521 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1522 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1523 for maintaining out-of-tree patches for some time.
1525 PP/01 Bugzilla 139: Documentation and portability issues.
1526 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1527 Handle per-OS dynamic-module compilation flags.
1529 PP/02 Let /dev/null have normal permissions.
1530 The 4.73 fixes were a little too stringent and complained about the
1531 permissions on /dev/null. Exempt it from some checks.
1532 Reported by Andreas M. Kirchwitz.
1534 PP/03 Report version information for many libraries, including
1535 Exim version information for dynamically loaded libraries. Created
1536 version.h, now support a version extension string for distributors
1537 who patch heavily. Dynamic module ABI change.
1539 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1540 privilege escalation vulnerability whereby the Exim run-time user
1541 can cause root to append content of the attacker's choosing to
1544 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1547 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1548 If dropping privileges for untrusted macros, we disabled normal logging
1549 on the basis that it would fail; for the Exim run-time user, this is not
1550 the case, and it resulted in successful deliveries going unlogged.
1551 Fixed. Reported by Andreas Metzler.
1557 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1558 only prepend for the Resent-* case. Fixes regression introduced in
1559 Exim 4.70 by NM/22 for Bugzilla 607.
1561 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1562 increasing numbers of administrators be bitten by this.
1564 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1566 PP/03 Bugzilla 994: added openssl_options main configuration option.
1568 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1570 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1572 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1574 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1575 without a peer certificate, leading to a segfault because of an
1576 assumption that peers always have certificates. Be a little more
1577 paranoid. Problem reported by Martin Tscholak.
1579 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1580 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1581 NB: ClamAV planning to remove STREAM in "middle of 2010".
1582 CL also introduces -bmalware, various -d+acl logging additions and
1583 more caution in buffer sizes.
1585 PP/09 Implemented reverse_ip expansion operator.
1587 PP/10 Bugzilla 937: provide a "debug" ACL control.
1589 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1591 PP/12 Bugzilla 973: Implement --version.
1593 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1595 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1597 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1599 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1600 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1601 ignore trailing whitespace.
1603 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1605 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1608 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1609 Notification from Dr Andrew Aitchison.
1611 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1612 ExtendedDetectionInfo response format.
1613 Notification from John Horne.
1615 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1618 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1619 XSL and documented dependency on system catalogs, with examples of how
1622 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1625 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1626 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1627 configuration file which is writeable by the Exim user or group.
1629 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1630 of configuration files to cover files specified with the -C option if
1631 they are going to be used with root privileges, not just the default
1634 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1635 option (effectively making it always true).
1637 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1638 files to be used while preserving root privileges.
1640 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1641 that rogue child processes cannot use them.
1643 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1644 run-time user, instead of root.
1646 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1647 Exim run-time user without dropping privileges.
1649 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1650 result string, instead of calling string_vformat() twice with the same
1653 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1654 for other users. Others should always drop root privileges if they use
1655 -C on the command line, even for a whitelisted configure file.
1657 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1659 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1665 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1666 $data_path, and $header_path variables; fixed documentation bugs and
1669 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1670 exipick to access non-standard spools, including the "frozen" queue
1673 NM/01 Bugzilla 965: Support mysql stored procedures.
1674 Patch from Alain Williams
1676 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1678 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1679 Patch from Andreas Metzler
1681 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1682 Patch from Kirill Miazine
1684 NM/05 Bugzilla 671: Added umask to procmail example.
1686 JJ/03 installed exipick 20100323.0, fixing doc bug
1688 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1689 directory. Notification and patch from Dan Rosenberg.
1691 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1693 TK/02 Improve log output when DKIM signing operation fails.
1695 MH/01 Treat the transport option dkim_domain as a colon separated
1696 list, not as a single string, and sign the message with each element,
1697 omitting multiple occurences of the same signer.
1699 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1700 Bugzilla 985, 986. Patch by Simon Arlott
1702 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1703 Patch by Simon Arlott
1705 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1706 MBX locking. Notification from Dan Rosenberg.
1712 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1714 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1716 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1718 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1720 NM/04 Bugzilla 915: Flag broken perl installation during build.
1726 TK/01 Added patch by Johannes Berg that expands the main option
1727 "spamd_address" if it starts with a dollar sign.
1729 TK/02 Write list of recipients to X-Envelope-Sender header when building
1730 the mbox-format spool file for content scanning (suggested by Jakob
1733 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1734 (http://www.dcc-servers.net/) support via dccifd. Activated by
1735 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1737 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1738 by Mark Daniel Reidel <mr@df.eu>.
1740 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1741 When building exim an external PCRE library is now needed -
1742 PCRE is a system library on the majority of modern systems.
1743 See entry on PCRE_LIBS in EDITME file.
1745 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1746 conversation. Added nologin parameter to request.
1747 Patch contributed by Kirill Miazine.
1749 TF/01 Do not log submission mode rewrites if they do not change the address.
1751 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1753 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1754 log files in place. Contributed by Roberto Lima.
1756 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1758 TF/03 Bugzilla 615: When checking the local_parts router precondition
1759 after a local_part_suffix or local_part_prefix option, Exim now
1760 does not use the address's named list lookup cache, since this
1761 contains cached lookups for the whole local part.
1763 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1764 Robert Millan. Documentation is in experimental-spec.txt.
1766 TF/04 Bugzilla 668: Fix parallel build (make -j).
1768 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1770 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1771 Patch provided by Jan Srzednicki.
1773 TF/05 Leading white space used to be stripped from $spam_report which
1774 wrecked the formatting. Now it is preserved.
1776 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1777 that they are available at delivery time.
1779 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1781 TF/08 TLS error reporting now respects the incoming_interface and
1782 incoming_port log selectors.
1784 TF/09 Produce a more useful error message if an SMTP transport's hosts
1785 setting expands to an empty string.
1787 NM/06 Bugzilla 744: EXPN did not work under TLS.
1788 Patch provided by Phil Pennock.
1790 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1791 Patch provided by Richard Godbee.
1793 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1794 acl_smtp_notquit, added index entry.
1796 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1797 Patch provided by Eugene Bujak.
1799 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1800 accept(). Patch provided by Maxim Dounin.
1802 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1803 Patch provided by Phil Pennock.
1805 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1807 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1808 Patch provided by Brad "anomie" Jorsch.
1810 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1811 Patch provided by Dean Brooks.
1813 TK/05 Add native DKIM support (does not depend on external libraries).
1815 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1816 Patch provided by Graeme Fowler.
1818 NM/16 Bugzilla 851: Documentation example syntax fix.
1820 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1822 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1825 NM/19 Bugzilla 745: TLS version reporting.
1826 Patch provided by Phil Pennock.
1828 NM/20 Bugzilla 167: bool: condition support.
1829 Patch provided by Phil Pennock.
1831 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1832 clients. Patch provided by Phil Pennock.
1834 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1835 Patch provided by Brad "anomie" Jorsch.
1837 NM/23 Bugzilla 687: Fix misparses in eximstats.
1838 Patch provided by Heiko Schlittermann.
1840 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1841 Patch provided by Heiko Schlittermann.
1843 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1844 plus update to original patch.
1846 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1848 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1849 Patch provided by David Brownlee.
1851 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1853 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1855 NM/30 Bugzilla 888: TLS documentation bugfixes.
1857 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1859 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1860 Unlike the original bugzilla I have changed all shell scripts in src tree.
1862 NM/33 Bugzilla 898: Transport filter timeout fix.
1863 Patch by Todd Rinaldo.
1865 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1866 Patch by Serge Demonchaux.
1868 NM/35 Bugzilla 39: Base64 decode bug fixes.
1869 Patch by Jakob Hirsch.
1871 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1873 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1875 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1877 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1883 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1884 ALT-N's libdkim that I have put here:
1885 http://duncanthrax.net/exim-experimental/
1887 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1888 were called 'true' and it seems that C99 defines that as a reserved
1889 keyword to be used with 'bool' variable types. That means you could
1890 not include C99-style headers which use bools without triggering
1891 build errors in sieve.c.
1893 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1894 as mailq or other aliases. Changed the --help handling significantly
1895 to do whats expected. exim_usage() emits usage/help information.
1897 SC/01 Added the -bylocaldomain option to eximstats.
1899 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1901 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1903 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1909 PH/01 Another patch from the Sieve maintainer.
1911 PH/02 When an IPv6 address is converted to a string for single-key lookup
1912 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1913 dots are used instead of colons so that keys in lsearch files need not
1914 contain colons. This was done some time before quoting was made available
1915 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1916 (notated using the quote facility) so as to distinguish them from IPv4
1917 keys. This meant that lookups for IP addresses in host lists did not work
1918 for iplsearch lookups.
1920 This has been fixed by arranging for IPv6 addresses to be expressed with
1921 colons if the lookup type is iplsearch. This is not incompatible, because
1922 previously such lookups could never work.
1924 The situation is now rather anomalous, since one *can* have colons in
1925 ordinary lsearch keys. However, making the change in all cases is
1926 incompatible and would probably break a number of configurations.
1928 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1931 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1932 conversion specification without a maximum field width, thereby enabling
1933 a rogue spamd server to cause a buffer overflow. While nobody in their
1934 right mind would setup Exim to query an untrusted spamd server, an
1935 attacker that gains access to a server running spamd could potentially
1936 exploit this vulnerability to run arbitrary code as the Exim user.
1938 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1939 $primary_hostname instead of what libspf2 thinks the hosts name is.
1941 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1942 a directory entry by the name of the lookup key. Previously, if a
1943 symlink pointed to a non-existing file or a file in a directory that
1944 Exim lacked permissions to read, a lookup for a key matching that
1945 symlink would fail. Now it is enough that a matching directory entry
1946 exists, symlink or not. (Bugzilla 503.)
1948 PH/03 The body_linecount and body_zerocount variables are now exported in the
1951 PH/04 Added the $dnslist_matched variable.
1953 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1954 This means they are set thereafter only if the connection becomes
1957 PH/06 Added the client_condition to authenticators so that some can be skipped
1958 by clients under certain conditions.
1960 PH/07 The error message for a badly-placed control=no_multiline_responses left
1961 "_responses" off the end of the name.
1963 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1965 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1966 (without spaces) instead of just copying the configuration text.
1968 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1970 PH/11 Added $max_received_linelength.
1972 PH/12 Added +ignore_defer and +include_defer to host lists.
1974 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1975 way in which PCRE > 7.0 is built.
1977 PH/14 Implemented queue_only_load_latch.
1979 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1980 MAIL command. The effect was to mangle the value on 64-bit systems.
1982 PH/16 Another patch from the Sieve maintainer.
1984 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1986 PH/18 If a system quota error occurred while trying to create the file for
1987 a maildir delivery, the message "Mailbox is full" was not appended to the
1988 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1989 only to a quota excession during the actual writing of the file.
1991 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1992 characters?) which causes problems in log lines. The DN values are now
1993 passed through string_printing() before being added to log lines.
1995 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1996 and InterBase are left for another time.)
1998 PH/21 Added message_body_newlines option.
2000 PH/22 Guard against possible overflow in moan_check_errorcopy().
2002 PH/23 POSIX allows open() to be a macro; guard against that.
2004 PH/24 If the recipient of an error message contained an @ in the local part
2005 (suitably quoted, of course), incorrect values were put in $domain and
2006 $local_part during the evaluation of errors_copy.
2012 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2013 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2016 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2017 issue a MAIL command.
2019 PH/02 In an ACL statement such as
2021 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2023 if a client was not listed at all, or was listed with a value other than
2024 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2025 the condition was not true (as it should be), so access was not denied.
2026 The bug was that the ! inversion was incorrectly passed on to the second
2027 item. This has been fixed.
2029 PH/03 Added additional dnslists conditions == and =& which are different from
2030 = and & when the dns lookup returns more than one IP address.
2032 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2033 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2035 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2036 FSYNC, which compiles an option called disable_fsync that allows for
2037 bypassing fsync(). The documentation is heavily laced with warnings.
2039 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2041 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2042 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2043 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2044 including adding "make clean"; (3) Added -fPIC when compiling the test
2045 dynamically loaded module, to get rid of a warning.
2047 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2048 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2049 = 0s. The bug is otherwise harmless.
2051 PH/07 There was a bug in the dovecot authenticator such that the value of
2052 $auth1 could be overwritten, and so not correctly preserved, after a
2053 successful authentication. This usually meant that the value preserved by
2054 the server_setid option was incorrect.
2056 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2058 PH/09 Installed PCRE release 7.0.
2060 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2061 run for batched SMTP input. It is now run at the start of every message
2062 in the batch. While fixing this I discovered that the process information
2063 (output by running exiwhat) was not always getting set for -bs and -bS
2064 input. This is fixed, and it now also says "batched" for BSMTP.
2066 PH/11 Added control=no_pipelining.
2068 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2069 patch, slightly modified), and move the expansion of helo_data till after
2070 the connection is made in the smtp transport (so it can use these
2073 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2075 PH/14 Added log_selector = +pid.
2077 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2079 PH/16 Add ${if forany and ${if forall.
2081 PH/17 Added dsn_from option to vary the From: line in DSNs.
2083 PH/18 Flush SMTP output before performing a callout, unless control =
2084 no_callout_flush is set.
2086 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2087 was true (the default) a successful delivery failed to delete the retry
2088 item, thus causing premature timeout of the address. The bug is now
2091 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2093 PH/21 Long custom messages for fakedefer and fakereject are now split up
2094 into multiline responses in the same way that messages for "deny" and
2095 other ACL rejections are.
2097 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2098 with slight modification.
2100 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2101 draft, changing the syntax and factoring some duplicate code".
2103 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2104 for deliveries of the second and subsequent messages over the same SMTP
2107 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2108 ${reduce, with only minor "tidies".
2110 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2112 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2113 expansion side effects.
2115 PH/27 When a message times out after an over-quota error from an Exim-imposed
2116 quota, the bounce message says "mailbox is full". This message was not
2117 being given when it was a system quota that was exceeded. It now should
2120 MH/03 Made $recipients available in local_scan(). local_scan() already has
2121 better access to the recipient list through recipients_list[], but
2122 $recipients can be useful in postmaster-provided expansion strings.
2124 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2125 in the case of a MAIL command with additional options following the
2126 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2127 were accidentally chopped off.
2129 PH/29 SMTP synchronization checks are implemented when a command is read -
2130 there is a check that no more input is waiting when there shouldn't be
2131 any. However, for some commands, a delay in an ACL can mean that it is
2132 some time before the response is written. In this time, more input might
2133 arrive, invalidly. So now there are extra checks after an ACL has run for
2134 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2135 pipelining has not been advertised.
2137 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2139 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2140 returned in the SMTP response when smtp_return_error_details was false.
2141 This has been fixed.
2143 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2144 instead of the C I/O that was originally supplied, because problems were
2145 reported on Solaris.
2147 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2148 Exim which did not show up earlier: it was assuming that a call to
2149 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2150 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2151 was a macro that became an assignment, so it seemed to work. This has
2152 changed to a proper function call with a void return, hence the compile
2153 error. Exim's code has been fixed.
2155 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2158 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2160 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2162 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2163 because it turns out to be too big or there is a timeout) there is no
2164 "Completed" line in the log. When some messages of this type were
2165 selected by exigrep, they were listed as "not completed". Others were
2166 picked up by some special patterns. I have improved the selection
2167 criteria to be more general.
2169 PH/38 The host_find_failed option in the manualroute router can now be set
2170 to "ignore", to completely ignore a host whose IP address cannot be
2171 found. If all hosts are ignored, the behaviour is controlled by the new
2172 host_all_ignored option.
2174 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2175 homing or because of multiple MX records with /mx) generated more than
2176 one IP address, and the following item turned out to be the local host,
2177 all the secondary addresses of the first item were incorrectly removed
2178 from the list, along with the local host and any following hosts (which
2179 is what is supposed to happen).
2181 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2182 whoever called Exim into the -H file. In the case of the daemon it was
2183 behaving confusingly. When first started, it used values for whoever
2184 started the daemon, but after a SIGHUP it used the Exim user (because it
2185 calls itself on a restart). I have changed the code so that it now always
2188 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2189 message are rejected with the same error (e.g. no authentication or bad
2190 sender address), and a DATA command is nevertheless sent (as can happen
2191 with PIPELINING or a stupid MUA), the error message that was given to the
2192 RCPT commands is included in the rejection of the DATA command. This is
2193 intended to be helpful for MUAs that show only the final error to their
2196 PH/42 Another patch from the Sieve maintainer.
2198 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2199 Eximstats - Fixed some broken HTML links and added missing column headers
2201 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2202 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2204 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2206 PH/43 Yet another patch from the Sieve maintainer.
2208 PH/44 I found a way to check for a TCP/IP connection going away before sending
2209 the response to the final '.' that terminates a message, but only in the
2210 case where the client has not sent further data following the '.'
2211 (unfortunately, this is allowed). However, in many cases there won't be
2212 any further data because there won't be any more messages to send. A call
2213 to select() can be used: if it shows that the input is "ready", there is
2214 either input waiting, or the socket has been closed. An attempt to read
2215 the next input character can distinguish the two cases. Previously, Exim
2216 would have sent an OK response which the client would never have see.
2217 This could lead to message repetition. This fix should cure that, at
2218 least in a lot of common cases.
2220 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2221 advertised in response to EHLO.
2227 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2228 fixed by 4.65/MH/01 (is this a record?) are fixed:
2230 (i) An empty string was always treated as zero by the numeric comparison
2231 operators. This behaviour has been restored.
2233 (ii) It is documented that the numeric comparison operators always treat
2234 their arguments as decimal numbers. This was broken in that numbers
2235 starting with 0 were being interpreted as octal.
2237 While fixing these problems I realized that there was another issue that
2238 hadn't been noticed. Values of message_size_limit (both the global option
2239 and the transport option) were treated as octal if they started with 0.
2240 The documentation was vague. These values are now always treated as
2241 decimal, and I will make that clear in the documentation.
2247 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2248 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2251 MH/01 Don't check that the operands of numeric comparison operators are
2252 integers when their expansion is in "skipping" mode (fixes bug
2253 introduced by 4.64-PH/07).
2255 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2256 child addresses, Exim now panics and dies. Previously, because the count
2257 is held in a short int, deliveries were likely to be lost. As such a
2258 large number of recipients for a single message is ridiculous
2259 (performance will be very, very poor), I have chosen to impose a limit
2260 rather than extend the field.
2266 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2267 leftover -K file (the existence of which was triggered by #402).
2268 While we were at it, introduced process PID as part of the -K
2269 filename. This should rule out race conditions when creating
2272 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2273 processing considerably. Previous code took too long for large mails,
2274 triggering a timeout which in turn triggers #401.
2276 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2277 in the DK code in transports.c. sendfile() is not really portable,
2278 hence the _LINUX specificness.
2280 TF/01 In the add_headers option to the mail command in an Exim filter,
2281 there was a bug that Exim would claim a syntax error in any
2282 header after the first one which had an odd number of characters
2285 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2286 callout verification, Exim cached a "reject" for the entire domain. This
2287 is correct for most verifications, but it is not correct for a recipient
2288 verification with use_sender or use_postmaster set, because in that case
2289 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2290 case of MAIL FROM:<> rejection from other early rejections (e.g.
2291 rejection of HELO). When verifying a recipient using a non-null MAIL
2292 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2293 Whatever the result of the callout, the value of the domain cache is
2294 left unchanged (for any other kind of callout, getting as far as trying
2295 RCPT means that the domain itself is ok).
2297 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2300 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2301 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2302 session, the daemon ignores ECONNECT errors and logs others; it now
2303 ignores EPIPE as well.
2305 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2306 (quoted-printable decoding).
2308 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2309 later the small subsequent patch to fix an introduced bug.
2311 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2313 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2315 PH/08 An error is now given if message_size_limit is specified negative.
2317 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2318 to be given (somewhat) arbitrary names.
2320 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2323 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2324 miscellaneous code fixes
2326 PH/10 Added the log_reject_target ACL modifier to specify where to log
2329 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2330 hostname. This is wrong, because it relates to the incoming message (and
2331 probably the interface on which it is arriving) and not to the outgoing
2332 callout (which could be using a different interface). This has been
2333 changed to use the value of the helo_data option from the smtp transport
2334 instead - this is what is used when a message is actually being sent. If
2335 there is no remote transport (possible with a router that sets up host
2336 addresses), $smtp_active_hostname is used.
2338 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2339 tweaks were necessary in order to get it to work (see also 21 below):
2340 (a) The code assumed that strncpy() returns a negative number on buffer
2341 overflow, which isn't the case. Replaced with Exim's string_format()
2343 (b) There were several signed/unsigned issues. I just did the minimum
2344 hacking in of casts. There is scope for a larger refactoring.
2345 (c) The code used strcasecmp() which is not a standard C function.
2346 Replaced with Exim's strcmpic() function.
2347 (d) The code set only $1; it now sets $auth1 as well.
2348 (e) A simple test gave the error "authentication client didn't specify
2349 service in request". It would seem that Dovecot has changed its
2350 interface. Fortunately there's a specification; I followed it and
2351 changed what the client sends and it appears to be working now.
2353 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2356 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2357 address A is aliased to B and C, where B exists and C does not. Without
2358 -v the output is "A verified" because verification stops after a
2359 successful redirection if more than one address is generated. However,
2360 with -v the child addresses are also verified. Exim was outputting "A
2361 failed to verify" and then showing the successful verification for C,
2362 with its parentage. It now outputs "B failed to verify", showing B's
2363 parentage before showing the successful verification of C.
2365 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2366 look up a TXT record in a specific list after matching in a combined
2369 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2370 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2371 they consult the DNS. I had assumed they would set it the way they
2372 wanted; and indeed my experiments on Linux seem to show that in some
2373 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2374 To be on the safe side, however, I have now made the interface to
2375 host_find_byname() similar to host_find_bydns(), with an argument
2376 containing the DNS resolver options. The host_find_byname() function now
2377 sets these options at its start, just as host_find_bydns() does. The smtp
2378 transport options dns_qualify_single and dns_search_parents are passed to
2379 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2380 of host_find_byname() use the default settings of RES_DEFNAMES
2381 (qualify_single) but not RES_DNSRCH (search_parents).
2383 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2384 spool_read_header() do less string testing, by means of a preliminary
2385 switch on the second character of optional "-foo" lines. (This is
2386 overdue, caused by the large number of possibilities that now exist.
2387 Originally there were few.) While I was there, I also converted the
2388 str(n)cmp tests so they don't re-test the leading "-" and the first
2389 character, in the hope this might squeeze out yet more improvement.
2391 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2392 flag allowing group syntax was set by the header_syntax check but not
2393 turned off, possible causing trouble later; (2) The flag was not being
2394 set at all for the header_verify test, causing "group"-style headers to
2395 be rejected. I have now set it in this case, and also caused header_
2396 verify to ignore an empty address taken from a group. While doing this, I
2397 came across some other cases where the code for allowing group syntax
2398 while scanning a header line wasn't quite right (mostly, not resetting
2399 the flag correctly in the right place). These bugs could have caused
2400 trouble for malformed header lines. I hope it is now all correct.
2402 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2403 with the "reply" argument non-NULL. The code, however (which originally
2404 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2405 but it didn't always do it. This confused somebody who was copying the
2406 code for some other use. I have removed all the tests.
2408 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2409 feature that was used to support insecure browsers during the U.S. crypto
2410 embargo. It requires special client support, and Exim is probably the
2411 only MTA that supported it -- and would never use it because real RSA is
2412 always available. This code has been removed, because it had the bad
2413 effect of slowing Exim down by computing (never used) parameters for the
2414 RSA_EXPORT functionality.
2416 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2417 authenticator to fail if there's a tab character in the incoming data
2418 (there should never be unless someone is messing about, as it's supposed
2419 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2420 if the connection is using TLS or if the remote IP is the same as the
2421 local IP, and the "valid-client-cert option" if a client certificate has
2424 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2425 authenticators. This can be used for authorization after authentication
2426 succeeds. (In the case of plaintext, it servers for both authentication
2429 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2430 if any retry times were supplied.
2432 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2433 connection, where there is no client IP address to check. In this
2434 situation, the verify now always succeeds.
2436 PH/25 Applied John Jetmore's -Mset patch.
2438 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2440 PH/27 In a string expansion for a processed (not raw) header when multiple
2441 headers of the same name were present, leading whitespace was being
2442 removed from all of them, but trailing whitespace was being removed only
2443 from the last one. Now trailing whitespace is removed from each header
2444 before concatenation. Completely empty headers in a concatenation (as
2445 before) are ignored.
2447 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2448 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2450 PH/29 [Removed. This was a change that I later backed out, and forgot to
2451 correct the ChangeLog entry (that I had efficiently created) before
2452 committing the later change.]
2454 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2455 address errors (4xx response to RCPT) over the same connection as other
2456 messages routed to the same hosts. Such deliveries are always "forced",
2457 so retry times are not inspected. This resulted in far too many retries
2458 for the affected addresses. The effect occurred only when there were more
2459 hosts than the hosts_max_try setting in the smtp transport when it had
2460 the 4xx errors. Those hosts that it had tried were not added to the list
2461 of hosts for which the message was waiting, so if all were tried, there
2462 was no problem. Two fixes have been applied:
2464 (i) If there are any address or message errors in an SMTP delivery, none
2465 of the hosts (tried or untried) are now added to the list of hosts
2466 for which the message is waiting, so the message should not be a
2467 candidate for sending over the same connection that was used for a
2468 successful delivery of some other message. This seems entirely
2469 reasonable: after all the message is NOT "waiting for some host".
2470 This is so "obvious" that I'm not sure why it wasn't done
2471 previously. Hope I haven't missed anything, but it can't do any
2472 harm, as the worst effect is to miss an optimization.
2474 (ii) If, despite (i), such a delivery is accidentally attempted, the
2475 routing retry time is respected, so at least it doesn't keep
2476 hammering the server.
2478 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2479 in ${readsocket because some servers need this prod.
2481 PH/32 Added some extra debug output when updating a wait-xxx database.
2483 PH/33 The hint "could be header name not terminated by colon", which has been
2484 given for certain expansion errors for a long time, was not being given
2485 for the ${if def:h_colon_omitted{... case.
2487 PH/34 The spec says: "With one important exception, whenever a domain list is
2488 being scanned, $domain contains the subject domain." There was at least
2489 one case where this was not true.
2491 PH/35 The error "getsockname() failed: connection reset by peer" was being
2492 written to the panic log as well as the main log, but it isn't really
2493 panic-worthy as it just means the connection died rather early on. I have
2494 removed the panic log writing for the ECONNRESET error when getsockname()
2497 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2498 runs only) independently of the message's sender address. This meant
2499 that, if the 4xx error was in fact related to the sender, a different
2500 message to the same recipient with a different sender could confuse
2501 things. In particular, this can happen when sending to a greylisting
2502 server, but other circumstances could also provoke similar problems.
2503 I have changed the default so that the retry time for these errors is now
2504 based a combination of the sender and recipient addresses. This change
2505 can be overridden by setting address_retry_include_sender=false in the
2508 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2509 remote server are returned as part of bounce messages. This was not
2510 happening for LMTP over a pipe (the lmtp transport), but now it is the
2511 same for both kinds of LMTP.
2513 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2514 in header lines that were in fact CNAMEs. This is no longer the case.
2516 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2517 and queue runs started by the daemon processed all messages. This has
2518 been fixed so that -R and -S can now usefully be given with -q<time>.
2520 PH/40 Import PCRE release 6.7 (fixes some bugs).
2522 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2524 PH/42 Give an error if -q is specified more than once.
2526 PH/43 Renamed the variables $interface_address and $interface_port as
2527 $received_ip_address and $received_port, to make it clear that these
2528 values apply to message reception, and not to the outgoing interface when
2529 a message is delivered. (The old names remain recognized, of course.)
2531 PH/44 There was no timeout on the connect() call when using a Unix domain
2532 socket in the ${readsocket expansion. There now is.
2534 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2535 be meaningful with "accept".
2537 SC/01 Eximstats V1.43
2538 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2540 SC/02 Eximstats V1.44
2541 Use a glob alias rather than an array ref in the generated
2542 parser. This improves both readability and performance.
2544 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2545 Collect SpamAssassin and rejection statistics.
2546 Don't display local sender or destination tables unless
2547 there is data to show.
2548 Added average volumes into the top table text output.
2550 SC/04 Eximstats V1.46
2551 Collect data on the number of addresses (recipients)
2552 as well as the number of messages.
2554 SC/05 Eximstats V1.47
2555 Added 'Message too big' to the list of mail rejection
2556 reasons (thanks to Marco Gaiarin).
2558 SC/06 Eximstats V1.48
2559 Mainlog lines which have GMT offsets and are too short to
2560 have a flag are now skipped.
2562 SC/07 Eximstats V1.49 (Alain Williams)
2563 Added the -emptyok flag.
2565 SC/08 Eximstats V1.50
2566 Fixes for obtaining the IP address from reject messages.
2568 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2569 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2570 whitespace changes from 4.64-PH/27
2572 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2575 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2576 are found, allow negative numbers in numeric criteria)
2578 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2580 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2581 to match changes made in 4.64-PH/43
2583 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2585 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2586 "domain" is actually an IP address, and force a failure. This locks out
2587 those revolvers/nameservers that support "A-for-A" lookups, in
2588 contravention of the specifications.
2590 PH/48 When a host name was looked up from an IP address, and the subsequent
2591 forward lookup of the name timed out, the host name was left in
2592 $sender_host_name, contrary to the specification.
2594 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2595 restricted to single-key lookups, Exim was not diagnosing an error if
2596 * or *@ was used with a query-style lookup.
2598 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2600 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2601 long ago, but noone interested enough thought of it. Let's just say that
2602 the "1.1" means that there are some new functions that weren't there at
2603 some point in the past.
2605 PH/51 Error processing for expansion failure of helo_data from an smtp
2606 transport during callout processing was broken.
2608 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2609 tested/used via the -bh/-bhc/-bs options.
2611 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2612 bug, fixed in subsequent PCRE releases).
2614 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2615 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2617 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2623 SC/01 Use a glob alias rather than an array ref in eximstats generated
2624 parser. This improves both readability and performance.
2626 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2627 Don't display local sender or destination tables in eximstats unless
2628 there is data to show.
2629 Added average volumes into the eximstats top table text output.
2631 SC/03 Collect data on the number of addresses (recipients) as well
2632 as the number of messages in eximstats.
2634 TF/01 Correct an error in the documentation for the redirect router. Exim
2635 does (usually) call initgroups() when daemonizing.
2637 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2638 with consistent privilege compared to when running as a daemon.
2640 TF/03 Note in the spec that $authenticated_id is not set for local
2641 submissions from trusted users.
2643 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2644 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2646 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2647 by adding some example configuration directives to the default
2648 configuration file. A little bit of work is required to uncomment the
2649 directives and define how usernames and passwords are checked, but
2650 there is now a framework to start from.
2652 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2653 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2654 without this. I don't know how relevant this is to other LDAP libraries.
2656 PH/02 Add the verb name to the "unknown ACL verb" error.
2658 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2660 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2662 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2663 directory not expanded when it should be if an expanded home directory
2664 was set for the address (which is overridden by the transport).
2666 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2669 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2670 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2671 because it is too late at that time, and has no effect.
2673 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2674 security issue with \' (bugzilla #107). I could not use the
2675 PQescapeStringConn() function, because it needs a PGconn value as one of
2678 PH/08 When testing addresses using -bt, indicate those final addresses that
2679 are duplicates that would not cause an additional delivery. At least one
2680 person was confused, thinking that -bt output corresponded to deliveries.
2681 (Suppressing duplicates isn't a good idea as you lose the information
2682 about possibly different redirections that led to the duplicates.)
2684 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2685 systems where poll() doesn't work, in particular OS X.
2687 PH/10 Added more information to debugging output for retry time not reached.
2689 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2690 operations in malware.c.
2692 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2695 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2696 syslog_duplication set false, log lines that would normally be written
2697 both the the main log and to the reject log were not written to syslog at
2700 PH/14 In the default configuration, change the use of "message" in ACL warn
2701 statements to "add_header".
2703 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2704 not followed by a command (e.g. "seen endif").
2706 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2707 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2710 PH/17 Added extra conditions to the default value of delay_warning_condition
2714 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2715 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2716 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2719 The Auto-Submitted: and various List- headers are standardised, whereas I
2720 don't think Precedence: ever was.
2722 PH/18 Refactored debugging code in route_finduser() to show more information,
2723 in particular, the error code if getpwnam() issues one.
2725 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2726 This is apparently needed in addition to the PH/07 change above to avoid
2727 any possible encoding problems.
2729 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2730 but not after initializing Perl.
2732 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2733 output them only if debugging. By default they are written stderr,
2734 apparently, which is not desirable.
2736 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2739 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2742 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2744 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2745 authenticated or an ident call has been made. Suppress the default
2746 values for $authenticated_id and $authenticated_sender (but permit -oMai
2747 and -oMas) when testing with -bh.
2749 PH/24 Re-jigged the order of the tests in the default configuration so that the
2750 tests for valid domains and recipients precede the DNS black list and CSA
2751 tests, on the grounds that those ones are more expensive.
2753 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2754 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2755 This bug exists in every version of Exim that I still have, right back to
2758 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2759 However, an attempt to turn on case-sensitivity in a regex key by
2760 including (?-i) didn't work because the subject string was already
2761 lowercased, and the effects were non-intuitive. It turns out that a
2762 one-line patch can be used to allow (?-i) to work as expected.
2768 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2769 other effects) broke the use of negated acl sub-conditions.
2771 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2774 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2775 "Deny" causes Exim to reject the incoming connection with a 554 error.
2776 Unfortunately, if there is a major crisis, such as a disk failure,
2777 tcp-wrappers gives "deny", whereas what one would like would be some
2778 kind of temporary error. A kludge has been added to help with this.
2779 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2780 554 error is used if errno is still zero or contains ENOENT (which occurs
2781 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2784 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2786 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2787 errors. Otherwise a message that provokes a temporary error (when other
2788 messages do not) can cause a whole host to time out.
2790 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2791 addresses were routed directly to files or pipes from a redirect router.
2792 File deliveries just didn't batch; pipe deliveries might have suffered
2795 PH/06 A failure to get a lock for a hints database would erroneously always say
2796 "Failed to get write lock", even when it was really a read lock.
2798 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2799 of 0600. This has been changed to use the value of the lockfile_mode
2800 option (which defaults to 0600).
2802 PH/08 Applied small patch from the Sieve maintainer.
2804 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2805 folder from quota calculations, a direct delivery into this folder messed
2806 up the contents of the maildirsize file. This was because the regex was
2807 used only to exclude .Trash (or whatever) when the size of the mailbox
2808 was calculated. There was no check that a delivery was happening into an
2809 excluded directory. This bug has been fixed by ignoring all quota
2810 processing for deliveries into excluded directories.
2812 PH/10 Added the maildirfolder_create_regex option to appendfile.
2818 PH/01 The code for finding all the local interface addresses on a FreeBSD
2819 system running IPv6 was broken. This may well have applied to all BSD
2820 systems, as well as to others that have similar system calls. The broken
2821 code found IPv4 interfaces correctly, but gave incorrect values for the
2822 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2823 that it would not match correctly against @[] and not recognize the IPv6
2826 PH/02 The ipliteral router was not recognizing addresses of the form user@
2827 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2829 PH/03 Added disable_ipv6.
2831 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2832 decoded form, because it is most often used to construct To: headers
2833 lines in autoreplies, and the decoded form may well be syntactically
2834 invalid. However, $reply_address has leading white space removed, and all
2835 newlines turned into spaces so that the autoreply transport does not
2838 PH/05 If group was specified without a user on a router, and no group or user
2839 was specified on a transport, the group from the router was ignored.
2841 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2842 for visible compile-time settings that can be used to change these
2843 numbers, for those that want even more. Backwards compatibility with old
2844 spool files has been maintained. However, going back to a previous Exim
2845 release will lost any variables that are in spool files.
2847 PH/07 Two small changes when running in the test harness: increase delay when
2848 passing a TCP/IP connection to a new process, in case the original
2849 process has to generate a bounce, and remove special handling of
2850 127.0.0.2 (sic), which is no longer necessary.
2852 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2853 be the same on different OS.
2855 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2858 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2859 whether --show-vars was specified or not
2861 JJ/02 exipick: Added support for new ACL variable spool format introduced
2864 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2865 syntactically invalid From: or Reply-to: line, and a filter used this to
2866 generate an autoreply, and therefore failed to obtain an address for the
2867 autoreply, Exim could try to deliver to a non-existent relative file
2868 name, causing unrelated and misleading errors. What now happens is that
2869 it logs this as a hard delivery error, but does not attempt to create a
2872 PH/11 The exinext utility has a -C option for testing purposes, but although
2873 the given file was scanned by exinext itself; it wasn't being passed on
2874 when Exim was called.
2876 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2877 an end-of-file indication when reading a command response.
2879 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2880 compiled. In many other places in Exim, IPv6 addresses are always
2881 recognized, so I have changed this. It also means that IPv4 domain
2882 literals of the form [IPV4:n.n.n.n] are now always recognized.
2884 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2885 used if the router is not running as root, for example, when verifying at
2886 ACL time, or when using -bh. The debugging output from this situation was
2887 non-existent - all you got was a failure to exec. I have made two
2890 (a) Failures to set uid/gid, the current directory, or a process leader
2891 in a subprocess such as that created by queryprogram now generate
2892 suitable debugging output when -d is set.
2894 (b) The queryprogram router detects when it is not running as root,
2895 outputs suitable debugging information if -d is set, and then runs
2896 the subprocess without attempting to change uid/gid.
2898 PH/15 Minor change to Makefile for building test_host (undocumented testing
2901 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2902 additional section of a DNS packet that returns MX or SRV records.
2903 Instead, it always explicitly searches for A/AAAA records. This avoids
2904 major problems that occur when a DNS server includes only records of one
2905 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2906 fixed another bug: if SRV records were looked up and the corresponding
2907 address records were *not* found in the additional section, the port
2908 values from the SRV records were lost.
2910 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2911 using the correct key (the original address) when searching the retry
2912 rules in order to find which one to use for generating the retry hint.
2914 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2915 adding the default one. Similarly, if it contains a Reply-To: header, the
2916 errors_reply_to option, if set, is not used.
2918 PH/19 When calculating a retry time, Exim used to measure the "time since
2919 failure" by looking at the "first failed" field in the retry record. Now
2920 it does not use this if it is later than than the arrival time of the
2921 message. Instead it uses the arrival time. This makes for better
2922 behaviour in cases where some deliveries succeed, thus re-setting the
2923 "first failed" field. An example is a quota failure for a huge message
2924 when small messages continue to be delivered. Without this change, the
2925 "time since failure" will always be short, possible causing more frequent
2926 delivery attempts for the huge message than are intended.
2927 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2929 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2930 $1, $2, $3) because the numerical variables can be reset during some
2931 expansion items (e.g. "match"), thereby losing the authentication data.
2933 PH/21 Make -bV show the size of off_t variables so that the test suite can
2934 decide whether to run tests for quotas > 2G.
2936 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2937 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2938 filecount value is greater than 2G or if a quota value is greater than 2G
2939 on a system where the size of off_t is not greater than 4, a panic error
2942 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2943 never match. The debug and -bh output now contains an explicit error
2944 message indicating a malformed IPv4 address or mask.
2946 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2947 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2948 PH/23 above applies.
2950 PH/25 Do not write to syslog when running in the test harness. The only
2951 occasion when this arises is a failure to open the main or panic logs
2952 (for which there is an explicit test).
2954 PH/26 Added the /no_tell option to "control=freeze".
2956 PH/27 If a host name lookup failed very early in a connection, for example, if
2957 the IP address matched host_lookup and the reverse lookup yielded a name
2958 that did not have a forward lookup, an error message of the form "no IP
2959 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2960 could be logged. Now it outputs the IP address instead of "NULL".
2962 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2963 allows the sender and the authenticated sender to be set when
2964 submitting a message from within Exim. Since child_open_exim() is
2965 documented for local_scan(), the new function should be too.
2967 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2968 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2969 results in an empty string is now treated as unset.
2971 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2973 PH/31 Added sender_verify_fail logging option.
2975 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2976 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2977 tidied the source and removed it altogether.
2979 PH/33 When a queue run was abandoned because the load average was too high, a
2980 log line was always written; now it is written only if the queue_run log
2981 selector is set. In addition, the log line for abandonment now contains
2982 information about the queue run such as the pid. This is always present
2983 in "start" and "stop" lines but was omitted from the "abandon" line.
2985 PH/34 Omit spaces between a header name and the colon in the error message that
2986 is given when verify = headers_syntax fails (if there are lots of them,
2987 the message gets confusing).
2989 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2990 names, as there are now some PTR records that contain slashes. This check
2991 is only to protect against broken name servers that fall over on strange
2992 characters, so the fact that it applies to all lookups doesn't matter.
2994 PH/36 Now that the new test suite is complete, we can remove some of the
2995 special code in Exim that was needed for the old test suite. For example,
2996 sorting DNS records because real resolvers return them in an arbitrary
2997 order. The new test suite's fake resolver always returns records in the
3000 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3001 bounces) except when queue_only is set, to avoid logging races between
3002 the different processes.
3004 PH/38 Panic-die if .include specifies a non-absolute path.
3006 PH/39 A tweak to the "H" retry rule from its user.
3008 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3009 a label. They prevented compilation on older perls.
3011 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3012 a warning to be raised on newish perls.
3014 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3015 on queue. Changes to match documented behaviour of showing count of
3016 messages matching specified criteria.
3018 PH/40 Changed the default ident timeout from 30s to 5s.
3020 PH/41 Added support for the use of login_cap features, on those BSD systems
3021 that have them, for controlling the resources used by pipe deliveries.
3023 PH/42 The content-scanning code uses fopen() to create files in which to put
3024 message data. Previously it was not paying any attention to the mode of
3025 the files. Exim runs with umask(0) because the rest of the code creates
3026 files with open(), and sets the required mode explicitly. Thus, these
3027 files were ending up world-writeable. This was not a big issue, because,
3028 being within the spool directory, they were not world-accessible. I have
3029 created a function called modefopen, which takes an additional mode
3030 argument. It sets umask(777), creates the file, chmods it to the required
3031 mode, then resets the umask. All the relevant calls to fopen() in the
3032 content scanning code have been changed to use this function.
3034 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3035 to 24 hours. This avoids potential overflow problems when processing G
3036 and H retry rules. I suspect nobody ever tinkers with this value.
3038 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3040 PH/45 When the plaintext authenticator is running as a client, the server's
3041 challenges are checked to ensure they are valid base64 strings. By
3042 default, the authentication attempt is cancelled if an invalid string is
3043 received. Setting client_ignore_invalid_base64 true ignores these errors.
3044 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3045 they are received. Thus, the responses can be made to depend on the
3046 challenges. If an invalid string is ignored, an empty string is placed in
3049 PH/46 Messages that are created by the autoreply transport now contains a
3050 References: header, in accordance with RFCs 2822 and 3834.
3052 PH/47 Added authenticated_sender_force to the smtp transport.
3054 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3056 PH/49 Installed latest patch from the Sieve maintainer.
3058 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3059 was also set, the appendfile transport was unnecessarily scanning a
3060 directory of message files (e.g. for maildir delivery) to find the count
3061 of files (along with the size), even though it did not need this
3062 information. It now does the scan only if it needs to find either the
3063 size of the count of files.
3065 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3067 PH/52 Two bugs concerned with error handling when the smtp transport is
3070 (i) Exim was not creating retry information for temporary errors given
3071 for individual recipients after the DATA command when the smtp transport
3072 was used in LMTP mode. This meant that they could be retried too
3073 frequently, and not timed out correctly.
3075 (ii) Exim was setting the flag that allows error details to be returned
3076 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3077 recipients that were returned after the DATA command.
3079 PH/53 This is related to PH/52, but is more general: for any failing address,
3080 when detailed error information was permitted to be returned to the
3081 sender, but the error was temporary, then after the final timeout, only
3082 "retry timeout exceeded" was returned. Now it returns the full error as
3083 well as "retry timeout exceeded".
3085 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3086 do this, and (what is worse) MTAs that accept it.
3088 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3089 will now be deprecated.
3091 PH/56 New os.c-cygwin from the Cygwin maintainer.
3093 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3094 formats (previously only available in exim formats via -bpr, -bpru,
3095 and -bpra. Now also available in native and exiqgrep formats)
3097 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3098 with very large, slow to parse queues
3100 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3102 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3104 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3105 responses to authentication challenges, though it was showing the
3106 challenges; (ii) I've removed the CR characters from the debug output for
3109 PH/58 Allow for the insertion of a newline as well as a space when a string
3110 is turned into more than one encoded-word during RFC 2047 encoding. The
3111 Sieve code now uses this.
3113 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3114 data_4xx, lost_connection, tls_required.
3116 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3117 message was being sent as an SMTP response.
3119 PH/61 Add -l and -k options to exicyclog.
3121 PH/62 When verifying, if an address was redirected to one new address, so that
3122 verification continued, and the new address failed or deferred after
3123 having set something in $address_data, the value of $address_data was not
3124 passed back to the ACL. This was different to the case when no
3125 redirection occurred. The value is now passed back in both cases.
3127 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3128 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3129 use login_cap.h, so on its own it isn't the distinguishing feature. The
3130 new name refers directly to the setclassresources() function.
3132 PH/65 Added configuration files for NetBSD3.
3134 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3136 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3137 is preferred over IPv4.
3139 PH/68 The bounce_return_message and bounce_return_body options were not being
3140 honoured for bounces generated during the reception of non-SMTP messages.
3141 In particular, this applied to messages rejected by the ACL. This bug has
3142 been fixed. However, if bounce_return_message is true and bounce_return_
3143 body is false, the headers that are returned for a non-SMTP message
3144 include only those that have been read before the error was detected.
3145 (In the case of an ACL rejection, they have all been read.)
3147 PH/69 The HTML version of the specification is now built in a directory called
3148 spec_html instead of spec.html, because the latter looks like a path with
3149 a MIME-type, and this confuses some software.
3151 PH/70 Catch two compiler warnings in sieve.c.
3153 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3154 function verify_get_ident() calls ip_connect() to connect a socket, but
3155 if the "connect()" function timed out, ip_connect() used to close the
3156 socket. However, verify_get_ident() also closes the socket later, and in
3157 between Exim writes to the log, which may get opened at this point. When
3158 the socket was closed in ip_connect(), the log could get the same file
3159 descriptor number as the socket. This naturally causes chaos. The fix is
3160 not to close the socket in ip_connect(); the socket should be closed by
3161 the function that creates it. There was only one place in the code where
3162 this was missing, in the iplookup router, which I don't think anybody now
3163 uses, but I've fixed it anyway.
3165 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3166 well as to direct DNS lookups. Otherwise the handling of names in host
3167 lists is inconsistent and therefore confusing.
3173 PH/01 Two changes to the default runtime configuration:
3175 (1) Move the checks for relay_from_hosts and authenticated clients from
3176 after to before the (commented out) DNS black list checks.
3178 (2) Add control=submission to the relay_from_hosts and authenticated
3179 clients checks, on the grounds that messages accepted by these
3180 statements are most likely to be submissions.
3182 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3184 (1) Generate an error if the third argument for the ${prvs expansion is
3187 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3190 (3) Reset the variables that are obtained from the first argument of
3191 ${prvscheck and used in the second argument before leaving the code,
3192 because their memory is reclaimed, so using them afterwards may do
3195 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3196 one (it's much easier than Tom thought :-).
3198 (5) Because of (4), we can now allow for the use of $prvscheck_result
3199 inside the third argument.
3201 PH/03 For some reason, the default setting of PATH when running a command from
3202 a pipe transport was just "/usr/bin". I have changed it to
3205 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3206 anything to be listed in the output from -bV.
3208 PH/05 When a filter generated an autoreply, the entire To: header line was
3209 quoted in the delivery log line, like this:
3211 => >A.N.Other <ano@some.domain> <original@ddress> ...
3213 This has been changed so that it extracts the operative address. There
3214 may be more than one such address. If so, they are comma-separated, like
3217 => >ano@some.domain,ona@other.domain <original@ddress> ...
3219 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3220 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3221 not being looked up in the rDNS to get a host name, Exim was showing the
3222 IP address twice in Received: lines, even though the IP addresses were
3223 identical. For example:
3225 Received: from [1.2.3.4] (helo=[1.2.3.4])
3227 However, if the real host name was known, it was omitting the HELO data
3228 if it matched the actual IP address. This has been tidied up so that it
3229 doesn't show the same IP address twice.
3231 PH/07 When both +timestamp and +memory debugging was on, the value given by
3232 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3233 called by the debug printing, thereby overwriting the timestamp buffer.
3234 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3236 PH/08 When the original message was included in an autoreply transport, it
3237 always said "this is a copy of the message, including all the headers",
3238 even if body_only or headers_only was set. It now gives an appropriate
3241 PH/09 Applied a patch from the Sieve maintainer which:
3243 o fixes some comments
3244 o adds the (disabled) notify extension core
3245 o adds some debug output for the result of if/elsif tests
3246 o points to the current vacation draft in the documentation
3247 and documents the missing references header update
3251 o fixes a bug in processing the envelope test (when testing
3252 multiple envelope elements, the last element determined the
3255 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3256 Electronic Mail") by including:
3258 Auto-submitted: auto-generated
3260 in the messages that it generates (bounce messages and others, such as
3261 warnings). In the case of bounce messages for non-SMTP messages, there was
3262 also a typo: it was using "Auto_submitted" (underscore instead of
3263 hyphen). Since every message generated by Exim is necessarily in response
3264 to another message, thes have all been changed to:
3266 Auto-Submitted: auto-replied
3268 in accordance with these statements in the RFC:
3270 The auto-replied keyword:
3272 - SHOULD be used on messages sent in direct response to another
3273 message by an automatic process,
3275 - MUST NOT be used on manually-generated messages,
3277 - MAY be used on Delivery Status Notifications (DSNs) and Message
3278 Disposition Notifications (MDNs),
3280 - MUST NOT be used on messages generated by automatic or periodic
3281 processes, except for messages which are automatic responses to
3284 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3285 to the default Received: header definition.
3287 PH/12 Added log selector acl_warn_skipped (default on).
3289 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3290 variables because (a) they are in the wrong storage pool and (b) even if
3291 they were copied, it wouldn't work properly because of the caching.
3293 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3294 checking when decoding. Apparently there are clients that generate
3295 overlong encoded strings. Why am I not surprised?
3297 PH/15 If the first argument of "${if match_address" was not empty, but did not
3298 contain an "@" character, Exim crashed. Now it writes a panic log message
3299 and treats the condition as false.
3301 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3303 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3304 extension "envelope-auth". The code is finished and in agreement with
3305 other implementations, but there is no documentation so far and in fact,
3306 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3307 not changing the active code.
3309 Print executed "if" and "elsif" statements when debugging is used. This
3310 helps a great deal to understand what a filter does.
3312 Document more things not specified clearly in RFC3028. I had all this
3313 sorted out, when out of a sudden new issues came to my mind. Oops."
3315 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3318 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3319 canonical form (as documented). However, after a host name lookup from
3320 the IP address, check_host() was doing a simple string comparison with
3321 addresses acquired from the DNS when checking that the found name did
3322 have the original IP as one of its addresses. Since any found IPv6
3323 addresses are likely to be in abbreviated form, the comparison could
3324 fail. Luckily, there already exists a function for doing the comparison
3325 by converting both addresses to binary, so now that is used instead of
3326 the text comparison.
3328 PH/20 There was another similar case to PH/19, when a complete host name was
3329 given in a host list; looking up its IP address could give an abbreviated
3330 form, whereas the current host's name might or might not be abbreviated.
3331 The same fix has been applied.
3337 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3338 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3341 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3342 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3344 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3346 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3347 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3348 Also, for "Auto-Submitted", treat anything other than "no" as
3349 non-personal, in accordance with RFC 3834. (Previously it treated
3350 anything starting "auto-" as non-personal.)
3352 TF/01 The control=submission/name=... option had a problem with syntax
3353 errors if the name included a slash character. The /name= option
3354 now slurps the rest of the string, so it can include any characters
3355 but it must come last in the list of options (after /sender_retain
3358 PH/05 Some modifications to the interface to the fake nameserver for the new
3366 TK/01 Added the "success_on_redirect" address verification option. See
3367 NewStuff for rationale and an example.
3369 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3371 PH/02 Patch to exigrep to allow it to work on syslog lines.
3373 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3374 fread() to skip over the body file's header line, because in Cygwin the
3375 header line is locked and is inaccessible.
3377 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3378 co-exist for some time) to make it clear that it is the Exim ID that is
3379 referenced, not the Message-ID: header line.
3381 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3382 string_format() function, because snprintf() does not exist on all
3385 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3386 ${stat: expansion item.
3388 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3389 protocol synchronization error", to keep the pedants happy.
3391 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3392 well as for IRIX systems, when gcc is being used. See the host.c source
3395 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3397 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3400 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3401 transport and to the smtp transport in LMTP mode.
3403 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3405 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3406 run a filter in a subprocess. This could lead to confusion in subsequent
3407 lookups in the parent process. There should also be a search_tidyup() at
3408 the end of the subprocess.
3410 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3411 only if the host matched helo_try_verify_hosts, which caused the
3412 verification to occur when the EHLO/HELO command was issued. The ACL just
3413 tested the remembered result. Now, if a previous verification attempt has
3414 not happened, "verify = helo" does it there and then.
3416 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3418 TK/03 Fix log output including CR from clamd.
3420 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3421 exist provoked a memory error which could cause a segfault.
3423 PH/15 Installed PCRE 6.2
3425 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3427 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3428 of the problem. Specifically, suggested +O2 rather than +O1 for the
3431 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3433 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3434 record, the port was not correctly logged when the outgoing_port log
3435 selector was set (it logged the transort's default port).
3437 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3438 fallback_hosts, and "hosts" in the smtp transport.
3440 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3441 host errors such as "Connection refused".
3443 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3444 authentication with radiusclient 0.4.9:
3446 - Error returned from rc_read_config was caught wrongly
3447 - Username/password not passed on to radius server due to wrong length.
3449 The presumption is that some radiusclient API changes for 4.51/PH/17
3450 were not taken care of correctly. The code is still untested by me (my
3451 Linux distribution still has 0.3.2 of radiusclient), but it was
3452 contributed by a Radius user.
3454 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3455 expanding the "port" option of the smtp transport.
3457 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3458 while reading a MIME header. Thanks to Tom Hughes for a patch.
3460 PH/24 Include config.h inside local_scan.h so that configuration settings are
3463 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3464 that in an ACL for RCPT (for example), you can examine exactly what was
3467 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3468 commands, but it was not correctly comparing the address with the actual
3469 client host address. Thus, it would show the EHLO address in Received:
3470 header lines when this was not necessary.
3472 PH/27 Added the % operator to ${eval:}.
3474 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3475 it should be ignoring failures (because with -C, for example, it has lost
3476 privilege). It wasn't ignoring creation failures other than "already
3479 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3482 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3483 because an input line was too long, either on its own, or by virtue of
3484 too many continuations, the temporary file was not being removed, and the
3485 return code was incorrect.
3487 PH/31 Missing "BOOL" in function definition in filtertest.c.
3489 PH/32 Applied Sieve patches from the maintainer.
3491 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3493 PH/33 Added "verify = not_blind".
3495 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3496 Local/Makefile (with some defaults set). These are used in built scripts
3497 such as exicyclog, but they have never been used in the exim_install
3498 script (though there are many overriding facilities there). I have
3499 arranged that the exim_install script now takes note of these two
3502 PH/35 Installed configuration files for Dragonfly.
3504 PH/36 When a locally submitted message by a trusted user did not contain a
3505 From: header, and the sender address was obtained from -f or from an SMTP
3506 MAIL command, and the trusted user did not use -F to supply a sender
3507 name, $originator_name was incorrectly used when constructing a From:
3508 header. Furthermore, $originator_name was used for submission mode
3509 messages from external hosts without From: headers in a similar way,
3510 which is clearly wrong.
3512 PH/37 Added control=suppress_local_fixups.
3514 PH/38 When log_selector = +received_sender was set, and the addition of the
3515 sender made the log line's construction buffer exactly full, or one byte
3516 less than full, an overflow happened when the terminating "\n" was
3519 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3520 when the result of a list match is failure because a DNS lookup failed.
3522 PH/40 RM_COMMAND is now used in the building process.
3524 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3525 the "build-* directories that it finds.
3527 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3528 address in a domain literal was a prefix of an interface address.
3530 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3531 when verifying a sender address, unless rewrite_headers is false.
3533 PH/44 Wrote a long comment about why errors_to addresses are verified as
3534 recipients, not senders.
3536 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3537 the ratelimit ACL was added.
3539 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3541 PH/46 Added extra information about PostgreSQL errors to the error string.
3543 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3544 avoiding the need to install special zones in a real server. This is
3545 backwards compatible; if it can't find the fake resolver, it drops back.
3546 Thus, both old and new test suites can be run.
3548 TF/02 Added util/ratelimit.pl
3550 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3551 clock is set back in time.
3553 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3554 Candler <B.Candler@pobox.com>.
3556 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3557 OK for addresses that are the result of redirections.
3559 PH/48 A number of further additions for the benefit of the new test suite,
3560 including a fake gethostbyname() that interfaces to the fake DNS resolver
3563 TF/06 The fix for widen_domains has also been applied to qualify_single and
3564 search_parents which are the other dnslookup options that can cause
3567 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3570 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3572 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3573 service" if it encountered an MX record with an empty target hostname.
3574 The message is now "an MX or SRV record indicated no SMTP service".
3576 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3577 if the DNS of the sending site is misconfigured. This is quite a
3578 common situation. This change restores the behaviour of treating a
3579 helo verification defer as a failure.
3581 PH/51 If self=fail was set on a router, the bounce message did not include the
3582 actual error message.
3588 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3590 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3591 command itself ended in error, the underlying message about the transport
3592 filter timeout was being overwritten with the pipe command error. Now the
3593 underlying error message should be appended to the second error message.
3595 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3597 PH/02 Reduce the amount of output that "make" produces by default. Full output
3598 can still be requested.
3600 PH/03 The warning log line about a condition test deferring for a "warn" verb
3601 was being output only once per connection, rather than after each
3602 occurrence (because it was using the same function as for successful
3603 "warn" verbs). This seems wrong, so I have changed it.
3605 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3606 it should not have, which might have caused a crash in the right
3607 circumstances, but probably never did.
3609 PH/04 Installed a modified version of Tony Finch's patch to make submission
3610 mode fix the return path as well as the Sender: header line, and to
3611 add a /name= option so that you can make the user's friendly name appear
3614 TF/03 Added the control = fakedefer ACL modifier.
3616 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3617 Mark Lowes for thorough testing.
3619 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3621 TK/03 Merged latest SRS patch from Miles Wilton.
3623 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3624 with the definition in sysexits.h (which is #included earlier).
3625 Fortunately, Exim does not actually use EX_OK. The code used to try to
3626 preserve the sysexits.h value, by assuming that macro definitions were
3627 scanned for macro replacements. I have been disabused of this notion,
3628 so now the code just undefines EX_OK before #including unistd.h.
3630 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3631 in the smtp transport. When a block could not be written in a single
3632 write() function, the timeout was being re-applied to each part-write.
3633 This seems wrong - if the receiver was accepting one byte at a time it
3634 would take for ever. The timeout is now adjusted when this happens. It
3635 doesn't have to be particularly precise.
3637 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3638 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3640 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3641 without a domain if the check to <postmaster@domain> fails.
3643 SC/01 Eximstats: added -xls and the ability to specify output files
3644 (patch written by Frank Heydlauf).
3646 SC/02 Eximstats: use FileHandles for outputting results.
3648 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3650 SC/04 Eximstats: fixed display of large numbers with -nvr option
3652 SC/05 Eximstats: fixed merging of reports with empty tables.
3654 SC/06 Eximstats: added the -include_original_destination flag
3656 SC/07 Eximstats: removed tabs and trailing whitespace.
3658 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3660 TK/06 MBOX spool code: Add real "From " MBOX separator line
3661 so the .eml file is really in mbox format (even though
3662 most programs do not really care). Patch from Alex Miller.
3664 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3665 The latter is generated from $received_to and is only set if the
3666 message has one envelope recipient. SA can use these headers,
3667 obviously out-of-the-box. Patch from Alex Miller.
3669 PH/08 The ${def test on a variable was returning false if the variable's
3670 value was "0", contrary to what the specification has always said!
3671 The result should be true unless the variable is empty.
3673 PH/09 The syntax error of a character other than { following "${if
3674 def:variable_name" (after optional whitespace) was not being diagnosed.
3675 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3676 accidental colon was present, for example, could give incorrect results.
3678 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3679 result is used (not including appendfile, where other changes are about
3682 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3683 This involved changing a lot of size variables from int to off_t. It
3684 should work with maildirs and everything.
3686 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3687 spamd dying while we are connected to it.
3689 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3692 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3693 function for BDB 4.3.
3695 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3697 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3698 the sources for PCRE have been re-arranged and more files are now
3701 PH/15 The code I had for printing potentially long long variables in PH/11
3702 above was not the best (it lost precision). The length of off_t variables
3703 is now inspected at build time, and an appropriate printing format (%ld
3704 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3705 to be "long long int" or "long int". This is needed for the internal
3706 formatting function string_vformat().
3708 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3709 the configuration file to be ":syslog", then the script "guesses" where
3710 the logs files are, rather than using the compiled in default. In our
3711 case the guess is not the same as the compiled default, so the script
3712 suddenly stopped working when I started to use syslog. The patch checks
3713 to see if log_file_path is "". If so, it attempts to read it from exim
3714 with no configuration file to get the compiled in version, before it
3715 falls back to the previous guessing code."
3717 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3718 implementing BATV in an Exim configuration. See NewStuff for the gory
3721 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3722 Makefile that are specific to HP-UX.
3724 PH/18 If the "use_postmaster" option was set for a recipient callout together
3725 with the "random" option, the postmaster address was used as the MAIL
3726 FROM address for the random test, but not for the subsequent recipient
3727 test. It is now used for both.
3729 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3730 patch removes a few documentation additions to RFC 3028, because the
3731 latest draft now contains them. It adds the new en;ascii-case comparator
3732 and a new error check for 8bit text in MIME parts. Comparator and
3733 require names are now matched exactly. I enabled the subaddress
3734 extension, but it is not well tested yet (read: it works for me)."
3736 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3737 rework some of the code of TK/09 above to avoid the hardwired use of
3738 "%lld" and "long long". Replaced the call to snprintf() with a call to
3741 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3742 records point to non-existent hosts", "retry timeout exceeded", and
3743 "retry time not reached for any host after a long failure period".
3745 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3746 experimental DomainKeys support:
3748 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3749 (2) On an error such as an illegally used "control", the wrong name for
3750 the control was given.
3752 These problems did NOT occur unless DomainKeys support was compiled.
3754 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3756 PH/24 Added ${if match_ip condition.
3758 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3759 databases so that it will be absolutely obvious if a crash occurs in the
3760 DB library. This is a regular occurrence (often caused by mis-matched
3763 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3764 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3765 picked up on a user's system that detects such things. There doesn't seem
3766 to be a gcc warning option for this - only an attribute that has to be
3767 put on the function's prototype. It seems that in Fedora Core 4 they have
3768 set this on a number of new functions. No doubt there will be more in due
3771 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3772 specify a transport. However, if an address that was verified by such a
3773 router was the subject of a callout, Exim crashed because it tried to
3774 read the rcpt_include_affixes from the non-existent transport. Now it
3775 just assumes that the setting of that option is false. This bug was
3776 introduced by 4.51/PH/31.
3778 PH/28 Changed -d+all to exclude +memory, because that information is very
3779 rarely of interest, but it makes the output a lot bigger. People tend to
3780 do -d+all out of habit.
3782 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3783 code in os-type was giving problems when libc.so lives in lib64, like on
3786 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3787 aren't the modern standard, and it seems that some systems' include files
3788 don't always have them. Exim was already checking for some of the newer
3789 ones like T_AAAA, and defining it itself. I've added checks for all the
3790 record types that Exim uses.
3792 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3793 not automatically generating a new one, as it is supposed to. This
3794 prevented TLS from working. If the file did exist, but contained invalid
3795 data, a new version was generated, as expected. It was only the case of a
3796 non-existent file that was broken.
3798 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3799 with a change in libdomainkeys > 0.64.
3801 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3802 from DNS. If the selector record carries the flag, it now has
3803 precedence over the domain-wide flag.
3805 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3807 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3808 the use of a port name that isn't defined in /etc/services) occurred, the
3809 message was deferred as in a normal delivery, and thus remained on the
3810 spool, instead of being failed because of the mua_wrapper setting. This
3811 is now fixed, and I tidied up some of the mua_wrapper messages at the
3814 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3815 the messages in a hash of arrays rather than using individual hashes.
3816 This is a bit cleaner and results in dramatic memory savings, albeit
3817 at a slight CPU cost.
3819 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3820 as requested by Marc Sherman.
3822 SC/10 Eximstats: added histograms for user specified patterns as requested
3825 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3827 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3828 fopen() in the content-scanning modules that did not already have it.
3834 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3835 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3837 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3839 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3841 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3842 to test Sieve filters that use "vacation".
3844 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3845 that changes the way the GnuTLS parameters are stored in the cache file.
3846 The new format can be generated externally. For backward compatibility,
3847 if the data in the cache doesn't make sense, Exim assumes it has read an
3848 old-format file, and it generates new data and writes a new file. This
3849 means that you can't go back to an older release without removing the
3852 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3853 work if there are any delivery delays because "one_time" forces the
3854 parent to be marked "delivered", so its unseen clone is never tried
3855 again. For this reason, Exim now forbids the simultaneous setting of
3858 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3859 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3860 entry if you want to know the details. The fix, however, neglected to
3861 consider the case when local delivery batching is involved. The test for
3862 "previously delivered" was not happening when checking to see if an
3863 address could be batched with a previous (undelivered) one; under
3864 certain circumstances this could lead to multiple deliveries to the same
3867 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3868 in its include files, and this causes problems building Exim.
3870 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3871 header_syntax) but Exim was just ignoring anything given after a slash.
3872 In particular, this caused confusion with an attempt to use "verify =
3873 reverse_host_lookup/defer_ok". An error is now given when options are
3874 supplied for verify items that do not have them. (Maybe reverse_host_
3875 lookup should have a defer_ok option, but that's a different point.)
3877 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3878 defined by RFC 821) to 2048, because there were problems with some AUTH
3879 commands, and RFC 1869 says the size should be increased for extended
3880 SMTP commands that take arguments.
3882 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3885 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3886 "unknown" error; now it says that the functionality isn't in the binary.
3888 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3889 an address' error message when a string expansion fails (syntax or
3890 whatever). Otherwise the password may appear in the log. Following change
3891 PH/42 below, there is no longer a chance of it appearing in a bounce
3894 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3896 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3897 of its addresses was ever tried. (Bugzilla bug #2.)
3899 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3900 the result incorrectly in the debug output. (It correctly added a newline
3901 to what was transported.)
3903 TF/01 Added $received_time.
3905 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3906 commented out examples of how to interface to a virus scanner and to
3907 SpamAssassin. Also added commented examples of av_scanner and
3908 spamd_address settings.
3910 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3911 and controls are allowed in which ACLs. There were a couple of minor
3912 errors. Some of the entries in the conditions table (which is a table of
3913 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3914 negation of where the condition IS allowed.
3916 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3918 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3919 header file does not have a version number, so I've had to invent a new
3920 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3921 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3922 radiusclient), but it was contributed by a Radius user.
3924 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3925 files or directories, for OpenSSL.
3927 PH/19 When an Exim process that is running as root has to create an Exim log
3928 file, it does so in a subprocess that runs as exim:exim so as to get the
3929 ownership right at creation (otherwise, other Exim processes might see
3930 the file with the wrong ownership). There was no test for failure of this
3931 fork() call, which would lead to the process getting stuck as it waited
3932 for a non-existent subprocess. Forks do occasionally fail when resources
3933 run out. I reviewed all the other calls to fork(); they all seem to check
3936 PH/20 When checking for unexpected SMTP input at connect time (before writing
3937 the banner), Exim was not dealing correctly with a non-positive return
3938 from the read() function. If the client had disconnected by this time,
3939 the result was a log entry for a synchronization error with an empty
3940 string after "input=" when read() returned zero. If read() returned -1
3941 (an event I could not check), uninitialized data bytes were printed.
3942 There were reports of junk text (parts of files, etc) appearing after
3945 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3947 PH/22 Added support for macro redefinition, and (re)definition in between
3948 driver and ACL definitions.
3950 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3951 forgetting to use the resulting value; it was using the unexpanded value.
3953 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3954 hadn't been configured. The fix is from Juergen Kreileder, who
3955 understands it better than I do:
3957 "Here's what I see happening with three configured cyrus_sasl
3958 authenticators configured (plain, login, cram-md5):
3960 On startup auth_cyrus_sasl_init() gets called for each of these.
3961 This means three calls to sasl_listmech() without a specified mech_list.
3962 => SASL tests which mechs of all available mechs actually work
3963 => three warnings about OTP not working
3964 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3966 With the patch, sasl_listmech() also gets called three times. But now
3967 SASL's mech_list option is set to the server_mech specified in the the
3968 authenticator. Or in other words, the answer from sasl_listmech()
3969 gets limited to just the mech you're testing for (which is different
3971 => the return list contains just 'plain' or 'login', 'cram-md5' or
3972 nothing depending on the value of ob->server_mech.
3974 I've just tested the patch: Authentication still works fine,
3975 unavailable mechs specified in the exim configuration are still
3976 caught, and the auth.log warnings about OTP are gone."
3978 PH/25 When debugging is enabled, the contents of the command line are added
3979 to the debugging output, even when log_selector=+arguments is not
3982 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3983 answer is "GNU", and only if the return is "GNU/something" is the answer
3986 PH/27 $acl_verify_message is now set immediately after the failure of a
3987 verification in an ACL, and so is available in subsequent modifiers. In
3988 particular, the message can be preserved by coding like this:
3990 warn !verify = sender
3991 set acl_m0 = $acl_verify_message
3993 Previously, $acl_verify_message was set only while expanding "message"
3994 and "log_message" when a very denied access.
3996 PH/28 Modified OS/os.c-Linux with
3998 -#ifndef OS_LOAD_AVERAGE
3999 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4001 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4002 nomenclature these days.)
4004 PH/29 Installed patch from the Sieve maintainer that adds the options
4005 sieve_useraddress and sieve_subaddress to the redirect router.
4007 PH/30 In these circumstances:
4008 . Two addresses routed to the same list of hosts;
4009 . First host does not offer TLS;
4010 . First host accepts first address;
4011 . First host gives temporary error to second address;
4012 . Second host offers TLS and a TLS session is established;
4013 . Second host accepts second address.
4014 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4015 and peerdn, if requested) that were in fact used only for the second
4018 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4019 attention to any local part prefix or suffix that was matched by the
4020 router that accepted the address. It now behaves in the same way as it
4021 does for delivery: the affixes are removed from the local part unless
4022 rcpt_include_affixes is set on the transport.
4024 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4025 timeout during the DATA phase of an incoming message.
4027 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4028 applied a patch sanctioned by the Sieve maintainer.
4030 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4031 the uid or gid is negative. A case of a negative gid caused this to be
4032 noticed. The fix allows for either to be negative.
4034 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4035 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4038 PH/36 The change PH/12 above was broken. Fixed it.
4040 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4041 the grounds that routing the same address twice would always produce the
4042 same answer. This might have been true once, but it is certainly no
4043 longer true now. Routing a child address may depend on the previous
4044 routing that produced that child. Some complicated redirection strategies
4045 went wrong when messages had multiple recipients, and made Exim's
4046 behaviour dependent on the order in which the addresses were given.
4048 I have moved the duplicate checking until after the routing is complete.
4049 Exim scans the addresses that are assigned to local and remote
4050 transports, and removes any duplicates. This means that more work will be
4051 done, as duplicates will always all be routed, but duplicates are
4052 presumably rare, so I don't expect this is of any significance.
4054 For deliveries to pipes, files, and autoreplies, the duplicate checking
4055 still happens during the routing process, since they are not going to be
4058 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4059 It corrects a timeout issue with spamd. This is Ian's comment: "The
4060 background is that sometimes spamd either never reads data from a
4061 connection it has accepted, or it never writes response data. The exiscan
4062 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4063 blindly assumes that writes won't block so it may never time out."
4065 PH/39 Allow G after quota size as well as K and M.
4067 PH/40 The value set for $authenticated_id in an authenticator may not contain
4068 binary zeroes or newlines because the value is written to log lines and
4069 to spool files. There was no check on this. Now the value is run through
4070 the string_printing() function so that such characters are converted to
4071 printable escape sequences.
4073 PH/41 $message_linecount is a new variable that contains the total number of
4074 lines in the message. Compare $body_linecount, which is the count for the
4077 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4078 bounce and delay warning messages, except in certain special cases, which
4081 (a) An SMTP error message from a remote host;
4082 (b) A message specified in a :fail: redirection;
4083 (c) A message specified in a "fail" command in a system filter;
4084 (d) A message specified in a FAIL return from the queryprogram router;
4085 (e) A message specified by the cannot_route_message router option.
4087 In these cases only, Exim does include the error details in bounce and
4088 warning messages. There are also a few cases where bland messages such
4089 as "unrouteable address" or "local delivery error" are given.
4091 PH/43 $value is now also set for the "else" part of a ${run expansion.
4093 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4094 being worked on, but at least Exim now implements the latest version to
4097 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4098 process to complete was treated as a delivery failure, a timeout while
4099 writing the message to the pipe was logged, but erroneously treated as a
4100 successful delivery. Such timeouts include transport filter timeouts. For
4101 consistency with the overall process timeout, these timeouts are now
4102 treated as errors, giving rise to delivery failures by default. However,
4103 there is now a new Boolean option for the pipe transport called
4104 timeout_defer, which, if set TRUE, converts the failures into defers for
4105 both kinds of timeout. A transport filter timeout is now identified in
4108 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4109 systems where "make" and "gmake" are different, calling "gmake" at top
4110 level broke things. I've arranged for the value of $(MAKE) to be passed
4111 from the Makefile to this script so that it can call the same version of
4115 A note about Exim versions 4.44 and 4.50
4116 ----------------------------------------
4118 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4119 changes of various kinds. As a consequence, a big documentation update was
4120 needed. This delayed the release for rather longer than seemed good, especially
4121 in the light of a couple of (minor) security issues. Therefore, the changes
4122 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4123 release. So 4.44 and 4.50 are in effect two different branches that both start
4126 I have left the 4.50 change log unchanged; it contains all the changes since
4127 4.43. The change log for 4.44 is below; many of its items are identical to
4128 those for 4.50. This seems to be the most sensible way to preserve the
4129 historical information.
4135 1. Minor wording change to the doc/README.SIEVE file.
4137 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4138 computation of the current number of files was incorrect.
4140 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4141 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4144 4. Give more explanation in the error message when the command for a transport
4145 filter fails to execute.
4147 5. There are several places where Exim runs a non-Exim command in a
4148 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4149 was being done only for the command run by the queryprogram router. It is
4150 now done for all such subprocesses. The other cases are: ${run, transport
4151 filters, and the commands run by the lmtp and pipe transports.
4153 6. Added CONFIGURE_GROUP build-time option.
4155 7. Some older OS have a limit of 256 on the maximum number of file
4156 descriptors. Exim was using setrlimit() to set 1000 as a large value
4157 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4158 systems. I've change it so that if it can't get 1000, it tries for 256.
4160 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4161 was an oversight, and furthermore, ever since the addition of extra
4162 controls (e.g. 4.43/32), the checks on when to allow different forms of
4163 "control" were broken. There should now be diagnostics for all cases when a
4164 control that does not make sense is encountered.
4166 9. Added the /retain_sender option to "control=submission".
4168 10. $recipients is now available in the predata ACL (oversight).
4170 11. Tidy the search cache before the fork to do a delivery from a message
4171 received from the command line. Otherwise the child will trigger a lookup
4172 failure and thereby defer the delivery if it tries to use (for example) a
4173 cached ldap connection that the parent has called unbind on.
4175 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4176 of $address_data from the recipient verification was clobbered by the
4177 sender verification.
4179 13. The value of address_data from a sender verification is now available in
4180 $sender_address_data in subsequent conditions in the ACL statement.
4182 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4184 15. Added a new option "connect=<time>" to callout options, to set a different
4187 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4188 was its contents. (It was OK if the option was not defined at all.)
4190 17. A "Completed" log line is now written for messages that are removed from
4191 the spool by the -Mrm option.
4193 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4194 information about exactly what failed.
4196 19. Added -dd to debug only the daemon process.
4198 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4199 handles timeouts, both on the server side and network timeouts. Renamed the
4200 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4202 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4203 It is now set to "smtps".
4205 22. $host_address is now set to the target address during the checking of
4206 ignore_target_hosts.
4208 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4209 being passed; this would have caused $sender_host_name to have been used if
4210 matching the list had actually called for a host name (not very likely,
4211 since this list is usually IP addresses). A host name is now passed as
4214 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4215 code that specifies a non-restarting handler (typically sigaction() in
4216 modern systems) in an attempt to fix a rare and obscure crash bug.
4218 25. Narrowed the window for a race in the daemon that could cause it to ignore
4219 SIGCHLD signals. This is not a major problem, because they are used only to
4220 wake it up if nothing else does.
4222 26. A malformed maildirsize file could cause Exim to calculate negative values
4223 for the mailbox size or file count. Odd effects could occur as a result.
4224 The maildirsize information is now recalculated if the size or filecount
4227 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4228 support for a long time. Removed HAVE_SYS_VFS_H.
4230 28. Installed the latest version of exipick from John Jetmore.
4232 29. In an address list, if the pattern was not a regular expression, an empty
4233 subject address (from a bounce message) matched only if the pattern was an
4234 empty string. Non-empty patterns were not even tested. This was the wrong
4235 because it is perfectly reasonable to use an empty address as part of a
4236 database query. An empty address is now tested by patterns that are
4237 lookups. However, all the other forms of pattern expect the subject to
4238 contain a local part and a domain, and therefore, for them, an empty
4239 address still always fails if the pattern is not itself empty.
4241 30. Exim went into a mad DNS loop when attempting to do a callout where the
4242 host was specified on an smtp transport, and looking it up yielded more
4243 than one IP address.
4245 31. Re-factored the code for checking spool and log partition space into a
4246 function that finds that data and another that does the check. The former
4247 is then used to implement four new variables: $spool_space, $log_space,
4248 $spool_inodes, and $log_inodes.
4250 32. The RFC2047 encoding function was originally intended for short strings
4251 such as real names; it was not keeping to the 75-character limit for
4252 encoded words that the RFC imposes. It now respects the limit, and
4253 generates multiple encoded words if necessary. To be on the safe side, I
4254 have increased the buffer size for the ${rfc2047: expansion operator from
4257 33. It is now permitted to omit both strings after an "if" condition; if the
4258 condition is true, the result is "true". As before, when the second string
4259 is omitted, a false condition yields an empty string. This makes it less
4260 cumbersome to write custom ACL and router conditions.
4262 34. Failure to deliver a bounce message always caused it to be frozen, even if
4263 there was an errors_to setting on the router. The errors_to setting is now
4266 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4267 canonical form (fully expanded) before being placed in
4268 $sender_host_address.
4270 36. The table in the code that translates DNS record types into text (T_A to
4271 "A" for instance) was missing entries for NS and CNAME. It is just possible
4272 that this could have caused confusion if both these types were looked up
4273 for the same domain, because the text type is used as part of Exim's
4274 per-process caching. But the chance of anyone hitting this buglet seems
4277 37. The dnsdb lookup has been extended in a number of ways.
4279 (1) There is a new type, "zns", which walks up the domain tree until it
4280 finds some nameserver records. It should be used with care.
4282 (2) There is a new type, "mxh", which is like "mx" except that it returns
4283 just the host names, not the priorities.
4285 (3) It is now possible to give a list of domains (or IP addresses) to be
4286 looked up. The behaviour when one of the lookups defers can be
4287 controlled by a keyword.
4289 (4) It is now possible to specify the separator character for use when
4290 multiple records are returned.
4292 38. The dnslists ACL condition has been extended: it is now possible to supply
4293 a list of IP addresses and/or domains to be looked up in a particular DNS
4296 39. Added log_selector=+queue_time_overall.
4298 40. When running the queue in the test harness, wait just a tad after forking a
4299 delivery process, to get repeatability of debugging output.
4301 41. Include certificate and key file names in error message when GnuTLS fails
4302 to set them up, because the GnuTLS error message doesn't include the name
4303 of the failing file when there is a problem reading it.
4305 42. Allow both -bf and -bF in the same test run.
4307 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4309 44. The "Exiscan patch" is now merged into the mainline Exim source.
4311 45. Sometimes the final signoff response after QUIT could fail to get
4312 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4313 < 0 before doing a fflush(). This bug looks as though it goes back to the
4314 introduction of TLS in release 3.20, but "sometimes" must have been rare
4315 because the tests only now provoked it.
4317 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4318 (this can affect the format of dates).
4320 47. exim_tidydb, when checking for the continued existence of a message for
4321 which it has found a message-specific retry record, was not finding
4322 messages that were in split spool directories. Consequently, it was
4323 deleting retry records that should have stayed in existence.
4325 48. Steve fixed some bugs in eximstats.
4327 49. The SPA authentication driver was not abandoning authentication and moving
4328 on to the next authenticator when an expansion was forced to fail,
4329 contradicting the general specification for all authenticators. Instead it
4330 was generating a temporary error. It now behaves as specified.
4332 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4333 (the order specifies the preference for clients). The order is now AES256,
4334 AES128, 3DES, ARCFOUR128.
4336 51. Small patch to Sieve code - explicitly set From: when generating an
4339 52. Exim crashed if a remote delivery caused a very long error message to be
4340 recorded - for instance if somebody sent an entire SpamAssassin report back
4341 as a large number of 550 error lines. This bug was coincidentally fixed by
4342 increasing the size of one of Exim's internal buffers (big_buffer) that
4343 happened as part of the Exiscan merge. However, to be on the safe side, I
4344 have made the code more robust (and fixed the comments that describe what
4347 53. Now that there can be additional text after "Completed" in log lines (if
4348 the queue_time_overall log selector is set), a one-byte patch to exigrep
4349 was needed to allow it to recognize "Completed" as not the last thing in
4352 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4353 patch that reportedly fixes this has been added. I am not expert enough to
4354 create a test for it. This is what the patch creator wrote:
4356 "I found a little strange behaviour of ldap code when working with
4357 Windows 2003 AD Domain, where users was placed in more than one
4358 Organization Units. When I tried to give exim partial DN, the exit code
4359 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4360 But simultaneously result of request was absolutely normal ldap result,
4361 so I produce this patch..."
4363 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4364 REFERENCE, so I have modified the code to exclude the patch when that macro
4367 55. Some experimental protocols are using DNS PTR records for new purposes. The
4368 keys for these records are domain names, not reversed IP addresses. The
4369 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4370 leaves it alone. Component reversal etc. now happens only for IP addresses.
4373 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4375 57. Double the size of the debug message buffer (to 2048) so that more of very
4376 long debug lines gets shown.
4378 58. The exicyclog utility now does better if the number of log files to keep
4379 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4381 59. Two changes related to the smtp_active_hostname option:
4383 (1) $smtp_active_hostname is now available as a variable.
4384 (2) The default for smtp_banner uses $smtp_active_hostname instead
4385 of $primary_hostname.
4387 60. The host_aton() function is supposed to be passed a string that is known
4388 to be a valid IP address. However, in the case of IPv6 addresses, it was
4389 not checking this. This is a hostage to fortune. Exim now panics and dies
4390 if the condition is not met. A case was found where this could be provoked
4391 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4392 components; fortuitously, this particular loophole had already been fixed
4393 by change 4.50/55 above.
4395 If there are any other similar loopholes, the new check in host_aton()
4396 itself should stop them being exploited. The report I received stated that
4397 data on the command line could provoke the exploit when Exim was running as
4398 exim, but did not say which command line option was involved. All I could
4399 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4400 running as the user.
4403 61. There was a buffer overflow vulnerability in the SPA authentication code
4404 (which came originally from the Samba project). I have added a test to the
4405 spa_base64_to_bits() function which I hope fixes it.
4408 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4409 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4411 63. The daemon start-up calls getloadavg() while still root for those OS that
4412 need the first call to be done as root, but it missed one case: when
4413 deliver_queue_load_max is set with deliver_drop_privilege. This is
4414 necessary for the benefit of the queue runner, because there is no re-exec
4415 when deliver_drop_privilege is set.
4417 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4418 This has been fixed.
4420 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4421 was in use, was not putting the data itself into the right store pool;
4422 consequently, it could be overwritten for a subsequent message in the same
4423 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4426 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4428 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4429 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4430 the function were treating the return as a boolean value, which happened to
4431 work because 0=false and not-0=true, but is not correct code.
4433 68. The host_aton() function was not handling scoped IPv6 addresses (those
4434 with, for example, "%eth0" on the end) correctly.
4436 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4437 negated items (that is, ~something) in unsigned ints. Some compilers
4438 apparently mutter when there is no cast.
4440 70. If an address verification called from an ACL failed, and did not produce a
4441 user-specific message (i.e. there was only a "system" message), nothing was
4442 put in $acl_verify_message. In this situation, it now puts the system
4445 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4446 session; change 4.31/43 added the unwanted input to the log line - except
4447 that it did not do this in the start of session case. It now does.
4449 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4450 This is wrong and can cause the other end to generate a synchronization
4451 error if it is another Exim or anything else that does the synchronization
4452 check. A QUIT command is no longer sent after a timeout.
4454 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4455 during host lookups.
4457 74. The defer_ok option of callout verification was not working if it was used
4458 when verifying addresses in header lines, that is, for this case:
4460 verify = header_sender/callout=defer_ok
4462 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4463 those file descriptors could be used for SMTP connections. If anything
4464 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4465 could be sent to the SMTP client, causing chaos. The daemon now opens
4466 stdin, stdout, and stderr to /dev/null when it puts itself into the
4469 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4470 log by default. The user can override this with suitable Perl magic.
4472 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4473 the log message when discard triggers, was not working for the DATA ACL or
4474 for the non-SMTP ACL.
4476 78. Error message wording change in sieve.c.
4478 79. If smtp_accept_max_per_host was set, the number of connections could be
4479 restricted to fewer than expected, because the daemon was trying to set up
4480 a new connection before checking whether the processes handling previous
4481 connections had finished. The check for completed processes is now done
4482 earlier. On busy systems, this bug wouldn't be noticed because something
4483 else would have woken the daemon, and it would have reaped the completed
4486 80. If a message was submitted locally by a user whose login name contained one
4487 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4488 It caused a spool format error. I have fixed the spool reading code. A
4489 related problem was that the "from" clause in the Received: line became
4490 illegal because of the space(s). It is now covered by ${quote_local_part.
4492 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4495 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4497 83. Patch from Sieve maintainer to fix unterminated string problem in
4498 "vacation" handling.
4500 84. Some minor changes to the Linux configuration files to help with other
4501 OS variants using glibc.
4503 85. One more patch for Sieve to update vacation handling to latest spec.
4506 ----------------------------------------------------
4507 See the note above about the 4.44 and 4.50 releases.
4508 ----------------------------------------------------
4514 1. Change 4.43/35 introduced a bug that caused file counts to be
4515 incorrectly computed when quota_filecount was set in an appendfile
4518 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4519 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4522 3. Give more explanation in the error message when the command for a transport
4523 filter fails to execute.
4525 4. There are several places where Exim runs a non-Exim command in a
4526 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4527 was being done only for the command run by the queryprogram router. It is
4528 now done for all such subprocesses. The other cases are: ${run, transport
4529 filters, and the commands run by the lmtp and pipe transports.
4531 5. Some older OS have a limit of 256 on the maximum number of file
4532 descriptors. Exim was using setrlimit() to set 1000 as a large value
4533 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4534 systems. I've change it so that if it can't get 1000, it tries for 256.
4536 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4537 was an oversight, and furthermore, ever since the addition of extra
4538 controls (e.g. 4.43/32), the checks on when to allow different forms of
4539 "control" were broken. There should now be diagnostics for all cases when a
4540 control that does not make sense is encountered.
4542 7. $recipients is now available in the predata ACL (oversight).
4544 8. Tidy the search cache before the fork to do a delivery from a message
4545 received from the command line. Otherwise the child will trigger a lookup
4546 failure and thereby defer the delivery if it tries to use (for example) a
4547 cached ldap connection that the parent has called unbind on.
4549 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4550 of $address_data from the recipient verification was clobbered by the
4551 sender verification.
4553 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4554 was its contents. (It was OK if the option was not defined at all.)
4556 11. A "Completed" log line is now written for messages that are removed from
4557 the spool by the -Mrm option.
4559 12. $host_address is now set to the target address during the checking of
4560 ignore_target_hosts.
4562 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4563 being passed; this would have caused $sender_host_name to have been used if
4564 matching the list had actually called for a host name (not very likely,
4565 since this list is usually IP addresses). A host name is now passed as
4568 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4569 code that specifies a non-restarting handler (typically sigaction() in
4570 modern systems) in an attempt to fix a rare and obscure crash bug.
4572 15. Narrowed the window for a race in the daemon that could cause it to ignore
4573 SIGCHLD signals. This is not a major problem, because they are used only to
4574 wake it up if nothing else does.
4576 16. A malformed maildirsize file could cause Exim to calculate negative values
4577 for the mailbox size or file count. Odd effects could occur as a result.
4578 The maildirsize information is now recalculated if the size or filecount
4581 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4582 support for a long time. Removed HAVE_SYS_VFS_H.
4584 18. Updated exipick to current release from John Jetmore.
4586 19. Allow an empty sender to be matched against a lookup in an address list.
4587 Previously the only cases considered were a regular expression, or an
4590 20. Exim went into a mad DNS lookup loop when doing a callout where the
4591 host was specified on the transport, if the DNS lookup yielded more than
4594 21. The RFC2047 encoding function was originally intended for short strings
4595 such as real names; it was not keeping to the 75-character limit for
4596 encoded words that the RFC imposes. It now respects the limit, and
4597 generates multiple encoded words if necessary. To be on the safe side, I
4598 have increased the buffer size for the ${rfc2047: expansion operator from
4601 22. Failure to deliver a bounce message always caused it to be frozen, even if
4602 there was an errors_to setting on the router. The errors_to setting is now
4605 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4606 canonical form (fully expanded) before being placed in
4607 $sender_host_address.
4609 24. Updated eximstats to version 1.33
4611 25. Include certificate and key file names in error message when GnuTLS fails
4612 to set them up, because the GnuTLS error message doesn't include the name
4613 of the failing file when there is a problem reading it.
4615 26. Expand error message when OpenSSL has problems setting up cert/key files.
4618 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4619 (this can affect the format of dates).
4621 28. exim_tidydb, when checking for the continued existence of a message for
4622 which it has found a message-specific retry record, was not finding
4623 messages that were in split spool directories. Consequently, it was
4624 deleting retry records that should have stayed in existence.
4626 29. eximstats updated to version 1.35
4627 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4628 1.35 - bugfix such that pie charts by volume are generated correctly
4630 30. The SPA authentication driver was not abandoning authentication and moving
4631 on to the next authenticator when an expansion was forced to fail,
4632 contradicting the general specification for all authenticators. Instead it
4633 was generating a temporary error. It now behaves as specified.
4635 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4636 (the order specifies the preference for clients). The order is now AES256,
4637 AES128, 3DES, ARCFOUR128.
4639 31. Small patch to Sieve code - explicitly set From: when generating an
4642 32. Exim crashed if a remote delivery caused a very long error message to be
4643 recorded - for instance if somebody sent an entire SpamAssassin report back
4644 as a large number of 550 error lines. This bug was coincidentally fixed by
4645 increasing the size of one of Exim's internal buffers (big_buffer) that
4646 happened as part of the Exiscan merge. However, to be on the safe side, I
4647 have made the code more robust (and fixed the comments that describe what
4650 33. Some experimental protocols are using DNS PTR records for new purposes. The
4651 keys for these records are domain names, not reversed IP addresses. The
4652 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4653 leaves it alone. Component reversal etc. now happens only for IP addresses.
4656 34. The host_aton() function is supposed to be passed a string that is known
4657 to be a valid IP address. However, in the case of IPv6 addresses, it was
4658 not checking this. This is a hostage to fortune. Exim now panics and dies
4659 if the condition is not met. A case was found where this could be provoked
4660 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4661 components; fortuitously, this particular loophole had already been fixed
4662 by change 4.50/55 or 4.44/33 above.
4664 If there are any other similar loopholes, the new check in host_aton()
4665 itself should stop them being exploited. The report I received stated that
4666 data on the command line could provoke the exploit when Exim was running as
4667 exim, but did not say which command line option was involved. All I could
4668 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4669 running as the user.
4672 35. There was a buffer overflow vulnerability in the SPA authentication code
4673 (which came originally from the Samba project). I have added a test to the
4674 spa_base64_to_bits() function which I hope fixes it.
4677 36. The daemon start-up calls getloadavg() while still root for those OS that
4678 need the first call to be done as root, but it missed one case: when
4679 deliver_queue_load_max is set with deliver_drop_privilege. This is
4680 necessary for the benefit of the queue runner, because there is no re-exec
4681 when deliver_drop_privilege is set.
4683 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4684 was in use, was not putting the data itself into the right store pool;
4685 consequently, it could be overwritten for a subsequent message in the same
4686 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4689 38. Sometimes the final signoff response after QUIT could fail to get
4690 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4691 < 0 before doing a fflush(). This bug looks as though it goes back to the
4692 introduction of TLS in release 3.20, but "sometimes" must have been rare
4693 because the tests only now provoked it.
4699 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4700 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4701 Now it returns an integer. A number of calls were still expecting a T/F
4702 return. Fortuitously, in all cases, the tests worked in OK situations,
4703 which is the norm. However, things would have gone wrong on any write
4704 failures on the smtp file descriptor. This function is used when sending
4705 messages over SMTP and also when doing verify callouts.
4707 2. When Exim is called to do synchronous delivery of a locally submitted
4708 message (the -odf or -odi options), it no longer closes stderr before doing
4711 3. Implemented the mua_wrapper option.
4713 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4715 5. Implemented the functions header_remove(), header_testname(),
4716 header_add_at_position(), and receive_remove_recipient(), and exported them
4719 6. If an ACL "warn" statement specified the addition of headers, Exim already
4720 inserted X-ACL-Warn: at the start if there was no header name. However, it
4721 was not making this test for the second and subsequent header lines if
4722 there were newlines in the string. This meant that an invalid header could
4723 be inserted if Exim was badly configured.
4725 7. Allow an ACL "warn" statement to add header lines at the start or after all
4726 the Received: headers, as well as at the end.
4728 8. Added the rcpt_4xx retry error code.
4730 9. Added postmaster_mailfrom=xxx to callout verification option.
4732 10. Added mailfrom=xxxx to the callout verification option, for verify=
4735 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4736 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4738 12. Inserted some casts to stop certain compilers warning when using pointer
4739 differences as field lengths or precisions in printf-type calls (mostly
4740 affecting debugging statements).
4742 13. Added optional readline() support for -be (dynamically loaded).
4744 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4745 same clock tick as a message's arrival, so that its received time was the
4746 same as the "first fail" time on the retry record, and that message
4747 remained on the queue past the ultimate address timeout, every queue runner
4748 would try a delivery (because it was past the ultimate address timeout) but
4749 after another failure, the ultimate address timeout, which should have then
4750 bounced the address, did not kick in. This was a "< instead of <=" error;
4751 in most cases the first failure would have been in the next clock tick
4752 after the received time, and all would be well.
4754 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4755 being recognized when the domain list was tested by the match_domain
4756 condition in an expansion string.
4758 16. Added the ${str2b64: operator.
4760 17. Exim was always calling setrlimit() to set a large limit for the number of
4761 processes, without checking whether the existing limit was already
4762 adequate. (It did check for the limit on file descriptors.) Furthermore,
4763 errors from getrlimit() and setrlimit() were being ignored. Now they are
4764 logged to the main and panic logs, but Exim does carry on, to try to do its
4765 job under whatever limits there are.
4767 18. Imported PCRE 5.0.
4769 19. Trivial typo in log message " temporarily refused connection" (the leading
4772 20. If the log selector return_path_on_delivery was set and an address was
4773 redirected to /dev/null, the delivery process crashed because it assumed
4774 that a return path would always be set for a "successful" delivery. In this
4775 case, the whole delivery is bypassed as an optimization, and therefore no
4778 21. Internal re-arrangement: the function for sending a challenge and reading
4779 a response while authentication was assuming a zero-terminated challenge
4780 string. It's now changed to take a pointer and a length, to allow for
4781 binary data in such strings.
4783 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4785 23. Exim was not respecting finduser_retries when seeking the login of the
4786 uid under which it was called; it was always trying 10 times. (The default
4787 setting of finduser_retries is zero.) Also, it was sleeping after the final
4788 failure, which is pointless.
4790 24. Implemented tls_on_connect_ports.
4792 25. Implemented acl_smtp_predata.
4794 26. If the domain in control=submission is set empty, Exim assumes that the
4795 authenticated id is a complete email address when it generates From: or
4796 Sender: header lines.
4798 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4799 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4800 chown and chgrp in /bin and hostname in /usr/bin.
4802 28. Exim was keeping the "process log" file open after each use, just as it
4803 does for the main log. This opens the possibility of it remaining open for
4804 long periods when the USR1 signal hits a daemon. Occasional processlog
4805 errors were reported, that could have been caused by this. Anyway, it seems
4806 much more sensible not to leave this file open at all, so that is what now
4809 29. The long-running daemon process does not normally write to the log once it
4810 has entered its main loop, and it closes the log before doing so. This is
4811 so that log files can straightforwardly be renamed and moved. However,
4812 there are a couple of unusual error situations where the daemon does write
4813 log entries, and I had neglected to close the log afterwards.
4815 30. The text of an SMTP error response that was received during a remote
4816 delivery was being truncated at 512 bytes. This is too short for some of
4817 the long messages that one sometimes sees. I've increased the limit to
4820 31. It is now possible to make retry rules that apply only when a message has a
4821 specific sender, in particular, an empty sender.
4823 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4824 it possible to be selective about when SMTP synchronization is enforced.
4826 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4828 32. Implemented hosts_connection_nolog.
4830 33. Added an ACL for QUIT.
4832 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4835 35. Added mailbox_size and mailbox_filecount to appendfile.
4837 36. Added control = no_multiline_responses to ACLs.
4839 37. There was a bug in the logic of the code that waits for the clock to tick
4840 in the case where the clock went backwards by a substantial amount such
4841 that the microsecond fraction of "now" was more than the microsecond
4842 fraction of "then" (but the whole seconds number was less).
4844 38. Added support for the libradius Radius client library this is found on
4845 FreeBSD (previously only the radiusclient library was supported).
4851 1. When certain lookups returned multiple values in the form name=value, the
4852 quoting of the values was not always being done properly. Specifically:
4853 (a) If the value started with a double quote, but contained no whitespace,
4855 (b) If the value contained whitespace other than a space character (i.e.
4856 tabs or newlines or carriage returns) it was not quoted.
4857 This fix has been applied to the mysql and pgsql lookups by writing a
4858 separate quoting function and calling it from the lookup code. The fix
4859 should probably also be applied to nisplus, ibase and oracle lookups, but
4860 since I cannot test any of those, I have not disturbed their existing code.
4862 2. A hit in the callout cache for a specific address caused a log line with no
4863 reason for rejecting RCPT. Now it says "Previous (cached) callout
4864 verification failure".
4866 3. There was an off-by-one bug in the queryprogram router. An over-long
4867 return line was truncated at 256 instead of 255 characters, thereby
4868 overflowing its buffer with the terminating zero. As well as fixing this, I
4869 have increased the buffer size to 1024 (and made a note to document this).
4871 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4872 when Exim is waiting for an SMTP response from a remote server, Exim
4873 restarts its select() call on the socket, thereby resetting its timeout.
4874 This is not a problem when such interrupts are rare. Somebody set up a cron
4875 job to run exiwhat every 2 minutes, which is less than the normal select()
4876 timeout (5 or 10 minutes). This meant that the select() timeout never
4877 kicked in because it was always reset. I have fixed this by comparing the
4878 time when an interrupt arrives with the time at the start of the first call
4879 to select(). If more time than the timeout has elapsed, the interrupt is
4880 treated as a timeout.
4882 5. Some internal re-factoring in preparation for the addition of Sieve
4883 extensions (by MH). In particular, the "personal" test is moved to a
4884 separate function, and given an option for scanning Cc: and Bcc: (which is
4885 not set for Exim filters).
4887 6. When Exim created an email address using the login of the caller as the
4888 local part (e.g. when creating a From: or Sender: header line), it was not
4889 quoting the local part when it contained special characters such as @.
4891 7. Installed new OpenBSD configuration files.
4893 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4894 try to make them clearer.
4896 9. Callout options, other than the timeout value, were being ignored when
4897 verifying sender addresses in header lines. For example, when using
4899 verify = header_sender/callout=no_cache
4901 the cache was (incorrectly) being used.
4903 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4904 only the Cygwin environment.
4906 11. When return_path_on_delivery was set as a log selector, if different remote
4907 addresses in the same message used different return paths and parallel
4908 remote delivery occurred, the wrong values would sometimes be logged.
4909 (Whenever a remote delivery process finished, the return path value from
4910 the most recently started remote delivery process was logged.)
4912 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4913 lines when AUTH and/or TLS are in use. This is the "received protocol"
4914 field. Exim used to use "asmtp" for authenticated SMTP, without any
4915 indication (in the protocol name) for TLS use. Now it follows the RFC and
4916 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4917 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4918 names appear in log lines as well as in Received: header lines.
4920 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4921 extensions, and comparison tests, and to fix some bugs.
4923 14. Changes to the "personal" filter test:
4925 (1) The test was buggy in that it was just doing the equivalent of
4926 "contains" tests on header lines. For example, if a user's address was
4927 anne@some.where, the "personal" test would incorrectly be true for
4929 To: susanne@some.where
4931 This test is now done by extracting each address from the header in turn,
4932 and checking the entire address. Other tests that are part of "personal"
4933 are now done using regular expressions (for example, to check local parts
4934 of addresses in From: header lines).
4936 (2) The list of non-personal local parts in From: addresses has been
4937 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4938 taken from the Sieve specification recommendations.
4940 (3) If the message contains any header line starting with "List-" it is
4941 treated as non-personal.
4943 (4) The test for "circular" in the Subject: header line has been removed
4944 because it now seems ill-conceived.
4946 15. Minor typos in src/EDITME comments corrected.
4948 16. Installed latest exipick from John Jetmore.
4950 17. If headers_add on a router specified a text string that was too long for
4951 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4952 of string_sprintf() is now avoided.
4954 18. $message_body_size was not set (it was always zero) when running the DATA
4955 ACL and the local_scan() function.
4957 19. For the "mail" command in an Exim filter, no default was being set for
4958 the once_repeat time, causing a random time value to be used if "once" was
4959 specified. (If the value happened to be <= 0, no repeat happened.) The
4960 default is now 0s, meaning "never repeat". The "vacation" command was OK
4961 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4962 (I found it when inspecting the code).
4964 20. There is now an overall timeout for performing a callout verification. It
4965 defaults to 4 times the callout timeout, which applies to individual SMTP
4966 commands during the callout. The overall timeout applies when there is more
4967 than one host that can be tried. The timeout is checked before trying the
4968 next host. This prevents very long delays if there are a large number of
4969 hosts and all are timing out (e.g. when the network connections are timing
4970 out). The value of the overall timeout can be changed by specifying an
4971 additional sub-option for "callout", called "maxwait". For example:
4973 verify = sender/callout=5s,maxwait=20s
4975 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4976 to the end before writing, but this should make it even safer).
4978 22. Exim was forgetting that it had advertised PIPELINING for the second and
4979 subsequent messages on an SMTP connection. It was also not resetting its
4980 memory on STARTTLS and an internal HELO.
4982 23. When Exim logs an SMTP synchronization error within a session, it now
4983 records whether PIPELINING has been advertised or not.
4985 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4986 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4987 rather than long int.
4989 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4991 26. Added the never_mail option to autoreply.
4997 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4998 crash if the getsockname() call failed; this can happen if a connection is
4999 closed very soon after it is established. The problem was simply in the
5000 order in which certain operations were done, causing Exim to try to write
5001 to the SMTP stream before it had set up the file descriptor. The bug has
5002 been fixed by making things happen in the correct order.
5008 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5009 before the connection was closed, thus losing the rejection response.
5011 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5012 some early Solaris releases, but causes trouble in current releases where
5013 socklen_t is defined.
5015 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5018 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5021 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5022 that releasing the top of it at the end releases what was used for sub-
5023 expansions (unless the block got too big). However, discard this block if
5024 the first thing is a variable or header, so that we can use its block when
5025 it is dynamic (useful for very large $message_headers, for example).
5027 6. Lookups now cache *every* query, not just the most recent. A new, separate
5028 store pool is used for this. It can be recovered when all lookup caches are
5029 flushed. Lookups now release memory at the end of their result strings.
5030 This has involved some general refactoring of the lookup sources.
5032 7. Some code has been added to the store_xxx() functions to reduce the amount
5033 of flapping under certain conditions.
5035 8. log_incoming_interface used to affect only the <= reception log lines. Now
5036 it causes the local interface and port to be added to several more SMTP log
5037 lines, for example "SMTP connection from", and rejection lines.
5039 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5041 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5043 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5044 could be overwritten at the end of the current message (or the start of a
5045 new message if it was set in a HELO ACL). The value is now preserved for
5046 the duration of the SMTP connection.
5048 12. If a transport had a headers_rewrite setting, and a matching header line
5049 contained an unqualified address, that address was qualified, even if it
5050 did not match any rewriting rules. The underlying bug was that the values
5051 of the flags that permit the existence of unqualified sender and recipient
5052 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5053 non-local messages, and by -bnq for local messages) were not being
5054 preserved with the message after it was received.
5056 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5057 "next input=" as part of the text comprising the host identity instead of
5058 the correct text. The code was using the same buffer for two different
5059 strings. However, depending on which order the printing function evaluated
5060 its arguments, the bug did not always show up. Under Linux, for example, my
5061 test suite worked just fine.
5063 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5064 doesn't work with some older versions of Perl. It has been changed to "my",
5065 which in any case is probably the better facility to use.
5067 15. A really picky compiler found some instances of statements for creating
5068 error messages that either had too many or two few arguments for the format
5071 16. The size of the buffer for calls to the DNS resolver has been increased
5072 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5073 for addresses that have a lot of PTR records. This alleviates a problem; it
5074 does not fully solve it.
5076 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5077 buffer now truncates the list and logs the incident, which is the same
5078 action as happens when Exim is looking up a host name and its aliases.
5079 Previously in this situation something unpredictable would happen;
5080 sometimes it was "internal error: store_reset failed".
5082 18. If a server dropped the connection unexpectedly when an Exim client was
5083 using GnuTLS and trying to read a response, the client delivery process
5084 crashed while trying to generate an error log message.
5086 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5087 string, for example:
5089 warn message = H1: something\nH2: something
5091 the text was added as a single header line from Exim's point of view
5092 though it ended up OK in the delivered message. However, searching for the
5093 second and subsequent header lines using $h_h2: did not work. This has been
5094 fixed. Similarly, if a system filter added multiple headers in this way,
5095 the routers could not see them.
5097 20. Expanded the error message when iplsearch is called with an invalid key to
5098 suggest using net-iplsearch in a host list.
5100 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5101 ACLs are no longer actually imposed (and a message to that effect is
5104 22. If a "gecos" field in a passwd entry contained escaped characters, in
5105 particular, if it contained a \" sequence, Exim got it wrong when building
5106 a From: or a Sender: header from that name. A second bug also caused
5107 incorrect handling when an unquoted " was present following a character
5108 that needed quoting.
5110 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5111 was not being matched caselessly.
5113 24. Arranged for all hyphens in the exim.8 source to be escaped with
5116 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5117 properly. Recipient callout cache records were still being keyed to include
5118 the sender, even when use_sender was set false. This led to far more
5119 callouts that were necessary. The sender is no longer included in the key
5120 when use_sender is false.
5122 26. Added "control = submission" modifier to ACLs.
5124 27. Added the ${base62d: operator to decode base 62 numbers.
5126 28. dnsdb lookups can now access SRV records.
5128 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5129 the configuration file.
5131 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5132 (-v) mode. This makes the output for a verbose queue run more intelligible.
5134 31. Added a use_postmaster feature to recipient callouts.
5136 32. Added the $body_zerocount variable, containing the number of binary zero
5137 bytes in the message body.
5139 33. The time of last modification of the "new" subdirectory is now used as the
5140 "mailbox time last read" when there is a quota error for a maildir
5143 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5145 35. Added +ignore_unknown as a special item in host lists.
5147 36. Code for decoding IPv6 addresses in host lists is now included, even if
5148 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5149 address was recognized as an IP address, but was then not correctly decoded
5150 into binary, causing unexpected and incorrect effects when compared with
5157 1. Very minor rewording of debugging text in manualroute to say "list of
5158 hosts" instead of "hostlist".
5160 2. If verify=header_syntax was set, and a header line with an unqualified
5161 address (no domain) and a large number of spaces between the end of the
5162 name and the colon was received, the reception process suffered a buffer
5163 overflow, and (when I tested it) crashed. This was caused by some obsolete
5164 code that should have been removed. The fix is to remove it!
5166 3. When running in the test harness, delay a bit after writing a bounce
5167 message to get a bit more predictability in the log output.
5169 4. Added a call to search_tidyup() just before forking a reception process. In
5170 theory, someone could use a lookup in the expansion of smtp_accept_max_
5171 per_host which, without the tidyup, could leave open a database connection.
5173 5. Added the variables $recipient_data and $sender_data which get set from a
5174 lookup success in an ACL "recipients" or "senders" condition, or a router
5175 "senders" option, similar to $domain_data and $local_part_data.
5177 6. Moved the writing of debug_print from before to after the "senders" test
5180 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5181 problems for message scanning, either using a data ACL, or using
5182 local_scan() because the Received: header was not generated till after they
5183 were called (in order to set the time as the time of reception completion).
5184 I have revised the way this works. The header is now generated after the
5185 body is received, but before the ACL or local_scan() are called. After they
5186 are run, the timestamp in the header is updated.
5192 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5193 before starting a queue runner without re-exec. This happened only when
5194 deliver_drop_privilege was set or when the Exim user was set to root. The
5195 effect of the bug was that timeouts during subsequent deliveries caused
5196 crashes instead of being properly handled. The handler is now left at its
5197 default (and expected) setting.
5199 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5200 message, again when deliver_drop_privilege is set or Exim is run as root.
5201 The bug described in (1) was not present in this case, but the tidying up
5202 of the other signals was missing. I have made the two cases consistent.
5204 3. The ignore_target_hosts setting on a manualroute router was being ignored
5205 for hosts that were looked up using the /MX notation.
5207 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5210 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5211 operated on the sender address. After changing the $sender_address to <>
5212 for the sender address verify, Exim was re-instated it as the original
5213 (before rewriting) address, but remembering that it had rewritten it, so it
5214 wasn't rewriting it again. This bug also had the effect of breaking the
5215 sender address verification caching when the sender address was rewritten.
5217 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5218 This has been changed so that if the ip literal address matches
5219 ignore_target_hosts, the router declines.
5221 7. Added expansion conditions match_domain, match_address, and match_local_
5222 part (NOT match_host).
5224 8. The placeholder for the Received: header didn't have a length field set.
5226 9. Added code to Exim itself and to exim_lock to test for a specific race
5227 condition that could lead to file corruption when using MBX delivery. The
5228 issue is with the lockfile that is created in /tmp. If this file is removed
5229 after a process has opened it but before that process has acquired a lock,
5230 there is the potential for a second process to recreate the file and also
5231 acquire a lock. This could lead to two Exim processes writing to the file
5232 at the same time. The added code performs the same test as UW imapd; it
5233 checks after acquiring the lock that its file descriptor still refers to
5234 the same named file.
5236 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5237 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5240 11. Added the smtp_active_hostname option. If used, this will typically be made
5241 to depend on the incoming interface address. Because $interface_address is
5242 not set up until the daemon has forked a reception process, error responses
5243 that can happen earlier (such as "too many connections") no longer contain
5246 12. If an expansion in a condition on a "warn" statement fails because a lookup
5247 defers, the "warn" statement is abandoned, and the next ACL statement is
5248 processed. Previously this caused the whole ACL to be aborted.
5250 13. Added the iplsearch lookup type.
5252 14. Added ident_timeout as a log selector.
5254 15. Added tls_certificate_verified as a log selector.
5256 16. Added a global option tls_require_ciphers (compare the smtp transport
5257 option of the same name). This controls incoming TLS connections.
5259 17. I finally figured out how to make tls_require_ciphers do a similar thing
5260 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5261 before starting the TLS session.
5263 18. Tabs are now shown as \t in -bP output.
5265 19. If the log selector return_path_on_delivery was set, Exim crashed when
5266 bouncing a message because it had too many Received: header lines.
5268 20. If two routers both had headers_remove settings, and the first one included
5269 a superfluous trailing colon, the final name in the first list and the
5270 first name in the second list were incorrectly joined into one item (with a
5271 colon in the middle).
5277 1. Added -C and -D options to the exinext utility, mainly to make it easier
5278 to include in the automated testing, but these could be helpful when
5279 multiple configurations are in use.
5281 2. The exinext utility was not formatting the output nicely when there was
5282 an alternate port involved in the retry record key, nor when there was a
5283 message id as well (for retries that were specific to a specific message
5284 and a specific host). It was also confused by IPv6 addresses, because of
5285 the additional colons they contain. I have fixed the IPv4 problem, and
5286 patched it up to do a reasonable job for IPv6.
5288 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5289 delivery, the log line now contains "pipelined" if PIPELINING was used.
5291 4. An SMTP transport process used to panic and die if the bind() call to set
5292 an explicit outgoing interface failed. This has been changed; it is now
5293 treated in the same way as a connect() failure.
5295 5. A reference to $sender_host_name in the part of a conditional expansion
5296 that was being skipped was still causing a DNS lookup. This no longer
5299 6. The def: expansion condition was not recognizing references to header lines
5300 that used bh_ and bheader_.
5302 7. Added the _cache feature to named lists.
5304 8. The code for checking quota_filecount in the appendfile transport was
5305 allowing one more file than it should have been.
5307 9. For compatibility with Sendmail, the command line option
5315 and sets the incoming protocol and host name (for trusted callers). The
5316 host name and its colon can be omitted when only the protocol is to be set.
5317 Note the Exim already has two private options, -pd and -ps, that refer to
5318 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5319 "s", but I don't think that's a major issue.
5321 10. A number of refactoring changes to the code, none of which should affect
5324 (a) The number of logging options was getting close to filling up the
5325 32-bit word that was used as a bit map. I have split them into two classes:
5326 those that are passed in the argument to log_write(), and those that are
5327 only ever tested independently outside of that function. These are now in
5328 separate 32-bit words, so there is plenty of room for expansion again.
5329 There is no change in the user interface or the logging behaviour.
5331 (b) When building, for example, log lines, the code previously used a
5332 macro that called string_cat() twice, in order to add two strings. This is
5333 not really sufficiently general. Furthermore, there was one instance where
5334 it was actually wrong because one of the argument was used twice, and in
5335 one call a function was used. (As it happened, calling the function twice
5336 did not affect the overall behaviour.) The macro has been replaced by a
5337 function that can join an arbitrary number of extra strings onto a growing
5340 (c) The code for expansion conditions now uses a table and a binary chop
5341 instead of a serial search (which was left over from when there were very
5342 few conditions). Also, it now recognizes conditions like "pam" even when
5343 the relevant support is not compiled in: a suitably worded error message is
5344 given if an attempt is made to use such a condition.
5346 11. Added ${time_interval:xxxxx}.
5348 12. A bug was causing one of the ddress fields not to be passed back correctly
5349 from remote delivery subprocesses. The field in question was not being
5350 subsequently used, so this caused to problems in practice.
5352 13. Added new log selectors queue_time and deliver_time.
5354 14. Might have fixed a bug in maildirsizefile handling that threw up
5355 "unexpected character" debug warnings, and recalculated the data
5356 unnecessarily. In any case, I expanded the warning message to give more
5359 15. Added the message "Restricted characters in address" to the statements in
5360 the default ACL that block characters like @ and % in local parts.
5362 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5363 Three changes have been made:
5365 (a) There was a serious bug; a negative response to MAIL caused the whole
5366 recipient domain to be cached as invalid, thereby blocking all messages
5367 to all local parts at the same domain, from all senders. This bug has
5368 been fixed. The domain is no longer cached after a negative response to
5369 MAIL if the sender used is not empty.
5371 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5374 (c) A new callout option, "use_sender" has been added for people who want
5375 the modified behaviour.
5381 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5384 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5385 indeed breaks things for older releases.
5387 3. Added additional logging to the case where there is a problem reading data
5388 from a filter that is running in a subprocess using a pipe, in order to
5389 try to track down a specific problem.
5391 4. Testing facility fudge: when running in the test harness and attempting
5392 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5393 getting "No route to host". Convert this to a timeout.
5395 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5398 6. Some OS don't have socklen_t but use size_t instead. This affects the
5399 fifth argument of getsockopt() amongst other things. This is now
5400 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5401 can be set for individual OS. I have set it for SunOS5, OSF1, and
5402 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5403 some earlier ones do not.
5405 7. Change 4.30/15 was not doing the test caselessly.
5407 8. The standard form for an IPv6 address literal was being rejected by address
5408 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5409 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5410 this, as well as the form without the "IPv6" on the front (but only when
5411 address literals are enabled, of course).
5413 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5415 10. Exim crashed if a message with an empty sender address specified by -f
5416 encountered a router with an errors_to setting. This could be provoked only
5417 by a command such as
5421 where an empty string was supplied; "<>" did not hit this bug.
5423 11. Installed PCRE release 4.5.
5425 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5426 remained set. It is now erased.
5428 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5429 times from message ids (which are base 36 rather than the normal 62).
5431 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5432 were being counted as actual protocol errors, and logged if the log
5433 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5434 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5435 and DATA following a set of rejected RCPTs do not count as protocol errors.
5436 In other words, Exim assumes they were pipelined, though this may not
5437 actually be the case. Of course, in all cases the client gets an
5438 appropriate error code.
5440 15. If a lookup fails in an ACL condition, a message about the failure may
5441 be available; it is used if testing the ACL cannot continue, because most
5442 such messages specify what the cause of the deferral is. However, some
5443 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5444 that caused an old message to be retained and used if a later statement
5445 caused a defer, replacing the real cause of the deferral.
5447 16. If an IP address had so many PTR records that the DNS lookup buffer
5448 was not large enough to hold them, Exim could crash while trying to process
5449 the truncated data. It now detects and logs this case.
5451 17. Further to 4.21/58, another change has been made: if (and only if) the
5452 first line of a message (the first header line) ends with CRLF, a bare LF
5453 in a subsequent header line has a space inserted after it, so as not to
5454 terminate the header.
5456 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5457 unnecessarily, used atoi() instead of strtol(), and didn't check the
5458 termination when getting file sizes from file names by regex.
5460 19. Completely re-implemented the support for maildirsize files, in the light
5461 of a number of problems with the previous contributed implementation
5462 (4.30/29). In particular:
5464 . If the quota is zero, the maildirsize file is maintained, but no quota is
5467 . If the maildir directory does not exist, it is created before any attempt
5468 to write a maildirsize file.
5470 . The quota value in the file is just a cache; if the quota is changed in
5471 the transport, the new value overrides.
5473 . A regular expression is available for excluding directories from the
5476 20. The autoreply transport checks the characters in options that define the
5477 message's headers; it allows continued headers, but it was checking with
5478 isspace() after an embedded newline instead of explicitly looking for a
5481 21. If all the "regular" hosts to which an address was routed had passed their
5482 expiry times, and had not reached their retry times, the address was
5483 bounced, even if fallback hosts were defined. Now Exim should go on to try
5486 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5487 equivalent code in the SMTP transport. Some hosts send humungous responses
5488 to HELO/EHLO, more than 1024 it seems.
5490 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5491 gives compiler warnings in some environments. I've now done it "properly",
5494 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5495 (because of problems with the built-in one) was declared to return uschar *
5496 instead of char *, causing compiler failure.
5498 25. Fixed a file descriptor leak when processing alias/forward files.
5500 26. Fixed a minor format string issue in dbfn.c.
5502 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5504 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5505 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5508 29. When a sender address is verified, it is cached, to save repeating the test
5509 when there is more than one recipient in a message. However, when the
5510 verification involves a callout, it is possible for different callout
5511 options to be set for different recipients. It is too complicated to keep
5512 track of this in the cache, so now Exim always runs a verification when a
5513 callout is required, relying on the callout cache for the optimization.
5514 The overhead is duplication of the address routing, but this should not be
5517 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5518 to be verified with callout=postmaster, and the main callout worked but the
5519 postmaster check failed, the verification correctly failed. However, if a
5520 subsequent RCPT command asked for sender verification *without* the
5521 postmaster check, incorrect caching caused this verification also to fail,
5524 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5525 it was not caching the DNS options (qualify_single, search_parents) that
5526 were used when the lookup failed. A subsequent lookup with different
5527 options therefore always gave the same answer, though there were cases
5528 where it should not have. (Example: a "domains = !$mx_any" option on a
5529 dnslookup router: the "domains" option is always processed without any
5530 widening, but the router might have qualify_single set.) Now Exim uses the
5531 cached value only when the same options are set.
5533 32. Added John Jetmore's "exipick" utility to the distribution.
5535 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5536 than a timeout (e.g. a certificate is required, and is not provided), an
5537 Exim server now closes the connection immediately. Previously it waited for
5538 the client to close - but if the client is SSL, it seems that they each
5539 wait for each other, leading to a delay before one of them times out.
5541 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5542 maintained 0.8.x compatibility because I don't think many are using it, and
5543 it is clearly obsolete.
5545 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5548 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5549 client certificate was expired. A simple patch fixes this, though I don't
5550 understand the full logic of why the verify callback is called multiple
5553 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5554 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5555 which causes problems with some clients (such as the Certicom SSL Plus
5556 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5557 disables the coutermeasure allowing Eudora to connect."
5559 38. Exim was not checking that a write() to a log file succeeded. This could
5560 lead to Bad Things if a log got too big, in particular if it hit a file
5561 size limit. Exim now panics and dies if it cannot write to a log file, just
5562 as it does if it cannot open a log file.
5564 39. Modified OS/Makefile-Linux so that it now contains
5566 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5568 The two -D definitions ensure that Exim is compiled with large file
5569 support, which makes it possible to handle log files that are bigger than
5572 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5573 instance) a domain was checked against a named list that involved a lookup,
5574 causing $domain_data to be set, then another domain was checked against the
5575 same list, then the first domain was re-checked, the value of $domain_data
5576 after the final check could be wrong. In particular, if the second check
5577 failed, it could be set empty. This bug probably also applied to
5580 41. The strip_trailing_dot option was not being applied to the address given
5581 with the -f command-line option.
5583 42. The code for reading a message's header from the spool was incrementing
5584 $received_count, but never initializing it. This meant that the value was
5585 incorrect (doubled) while delivering a message in the same process in which
5586 it was received. In the most common configuration of Exim, this never
5587 happens - a fresh exec is done - but it can happen when
5588 deliver_drop_privilege is set.
5590 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5591 it now includes up to 150 characters of the unexpected data in the log
5594 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5595 and building data strings. The size of both of these buffers was 10 000
5596 bytes - far larger than anybody would *ever* want, thought I. Needless to
5597 say, somebody hit the limit. I have increased the maximum line length to
5598 20 000 and the maximum data length of concatenated lines to 100 000. I have
5599 also fixed two bugs, because there was no checking on these buffers. Tsk,
5600 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5601 buffer is too small.
5603 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5604 lsearch lookups. Now it does.
5606 46. When parsing a route_list item in a manualroute router, a fixed-length
5607 buffer was used for the list of hosts. I made this 1024 bytes long,
5608 thinking that nobody would ever have a list of hosts that long. Wrong.
5609 Somebody had a whole pile of complicated expansion conditions, and the
5610 string was silently truncated, leading to an expansion error. It turns out
5611 that it is easier to change to an unlimited length (owing to other changes
5612 that have happened since this code was originally written) than to build
5613 structure for giving a limitation error. The length of the item that
5614 expands into the list of hosts is now unlimited.
5616 47. The lsearch lookup could not handle data where the length of text line was
5617 more than 4095 characters. Such lines were truncated, leading to shortened
5618 data being returned. It should now handle lines of any length.
5620 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5621 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5624 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5625 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5626 no longer happens in comment lines. A list of replacements is now placed
5627 at the head of all of the source files, except those whose only change is
5628 to replace PERL_COMMAND in the very first #! line.
5630 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5631 messages on the queue, with a bottom-up merge sort, using code contributed
5632 by Michael Haardt. This should make operations like -bp somewhat faster on
5633 large queues. It won't affect queue runners, except when queue_run_in_order
5636 51. Installed eximstats 1.31 in the distribution.
5638 52. Added support for SRV lookups to the dnslookup router.
5640 53. If an ACL referred to $message_body or $message_body_end, the value was not
5641 reset for any messages that followed in the same SMTP session.
5643 54. The store-handling optimization for building very long strings was not
5644 differentiating between the different store pools. I don't think this
5645 actually made any difference in practice, but I've tidied it.
5647 55. While running the routers to verify a sender address, $sender_address
5648 was still set to the sender address. This is wrong, because when routing to
5649 send a bounce to the sender, it would be empty. Therefore, I have changed
5650 it so that, while verifying a sender address, $sender_address is set to <>.
5651 (There is no change to what happens when verifying a recipient address.)
5653 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5654 target A or AAAA records (if not already returned) without resetting the
5655 qualify_single or search_parents options of the DNS resolver. These are
5656 inappropriate in this case because the targets of MX and SRV records must
5657 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5658 target that, when qualified, matched something in the local domain. These
5659 two options are now turned off when doing these lookups.
5661 57. It seems that at least some releases of Reiserfs (which does not have the
5662 concept of a fixed number of inodes) returns zero and not -1 for the
5663 number of available inodes. This interacted badly with check_spool_inodes,
5664 which assumed that -1 was the "no such thing" setting. What I have done is
5665 to check that the total number of inodes is greater than zero before doing
5666 the test of how many are available.
5668 58. When a "warn" ACL statement has a log_message modifier, the message is
5669 remembered, and not repeated. This is to avoid a lot of repetition when a
5670 message has many recipients that cause the same warning to be written.
5671 However, Exim was preserving the list of already written lines for an
5672 entire SMTP session, which doesn't seem right. The memory is now reset if a
5673 new message is started.
5675 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5676 debugging output unless log_rewrite was also set.
5678 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5679 of (int)(handle) when we know that handle contains (void *)(-1).
5681 61. The Exim daemon panic-logs an error return when it closes the incoming
5682 connection. However "connection reset by peer" seems to be common, and
5683 isn't really an error worthy of noting specially, so that particular error
5686 62. When Exim is trying to find all the local interfaces, it used to panic and
5687 die if the ioctl to get the interface flags failed. However, it seems that
5688 on at least one OS (Solaris 9) it is possible to have an interface that is
5689 included in the list of interfaces, but for which you get a failure error
5690 for this call. This happens when the interface is not "plumbed" into a
5691 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5692 failure of the "get flags" call assumes that the interface is down.
5694 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5695 makes life easier for people who are doing arithmetic on fields extracted
5696 from dates, where you often get leading zeros that should not be
5697 interpreted as octal.
5699 64. Added qualify_domain to the redirect router, to override the global
5702 65. If a pathologically long header line contained very many addresses (the
5703 report of this problem mentioned 10 000) and each of them was rewritten,
5704 Exim could use up a very large amount of memory. (It kept on making new
5705 copies of the header line as it rewrote, and never released the old ones.)
5706 At the expense of a bit more processing, the header rewriting function has
5707 been changed so that it no longer eats memory in this way.
5709 66. The generation of the Received: header has been moved from the time that a
5710 message starts to be received, to the time that it finishes. The timestamp
5711 in the Received: header should now be very close to that of the <= log
5712 line. There are two side-effects of this change:
5714 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5715 logged header lines no longer include the local Received: line, because
5716 it has not yet been created. The same applies to a copy of the message
5717 that is returned to a non-SMTP sender when a message is rejected.
5719 (b) When a filter file is tested using -bf, no additional Received: header
5720 is added to the test message. After some thought, I decided that this
5723 This change does not affect the value of $received_for. It is still set
5724 after address rewriting, but before local_scan() is called.
5726 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5728 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5729 gave an unhelpful panic error message, and a defer error. I have managed to
5730 change this behaviour so that it now rejects any supplied certificate,
5731 which seems right, as the list of acceptable certificates is empty.
5733 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5734 gave an unhelpful defer error. I have not managed to make this reject any
5735 supplied certificates, but the error message it gives is "no certificate
5736 supplied", which is not helpful.
5738 70. exigrep's output now also includes lines that are not associated with any
5739 message, but which match the given pattern. Implemented by a patch from
5740 Martin Sluka, which also tidied up the Perl a bit.
5742 71. Recipient callout verification, like sender verification, was using <> in
5743 the MAIL FROM command. This isn't really the right thing, since the actual
5744 sender may affect whether the remote host accepts the recipient or not. I
5745 have changed it to use the actual sender in the callout; this means that
5746 the cache record is now keyed on a recipient/sender pair, not just the
5747 recipient address. There doesn't seem to be a real danger of callout loops,
5748 since a callout by the remote host to check the sender would use <>.
5749 [SEE ABOVE: changed after hitting problems.]
5751 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5752 temporary errors. However, in the case of such a code being given after
5753 the end of a data transmission (i.e. after ".") Exim was failing to write
5754 a retry record for the message. (Yes, there was some broken host that was
5755 actually sending 8xx at this point.)
5757 73. An unknown lookup type in a host list could cause Exim to panic-die when
5758 the list was checked. (An example that provoked this was putting <; in the
5759 middle of a list instead of at the start.) If this happened during a DATA
5760 ACL check, a -D file could be left lying around. This kind of configuration
5761 error no longer causes Exim to die; instead it causes a defer error. The
5762 incident is still logged to the main and panic logs.
5764 74. Buglet left over from Exim 3 conversion. The message "too many messages
5765 in one connection" was written to the rejectlog but not the mainlog, except
5766 when address rewriting (yes!) was being logged.
5768 75. Added write_rejectlog option.
5770 76. When a system filter was run not as root (that is, when system_filter_user
5771 was set), the values of the $n variables were not being returned to the
5772 main process; thus, they were not subsequently available in the $sn
5775 77. Added +return_path_on_delivery log selector.
5777 78. A connection timeout was being treated differently from recipients deferred
5778 when testing hosts_max_try with a message that was older than the host's
5779 retry timeout. (The host should not be counted, thus allowing all hosts to
5780 be tried at least once before bouncing.) This may have been the cause of an
5781 occasionally reported bug whereby a message would remain on the queue
5782 longer than the retry timeout, but would be bounced if a delivery was
5783 forced. I say "may" because I never totally pinned down the problem;
5784 setting up timeout/retry tests is difficult. See also the next item.
5786 79. The ultimate address timeout was not being applied to errors that involved
5787 a combination of host plus message (for example, a timeout on a MAIL
5788 command). When an address resolved to a number of possible hosts, and they
5789 were not all tried for each delivery (e.g. because of hosts_max_try), a
5790 message could remain on the queue longer than the retry timeout.
5792 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5795 81. Fixed an obscure SMTP outgoing bug which required at least the following
5796 conditions: (a) there was another message waiting for the same server;
5797 (b) the server returned 5xx to all RCPT commands in the first message so
5798 that the message was not completed; (c) the server dropped the connection
5799 or gave a negative response to the RSET that Exim sends to abort the
5800 transaction. The observed case was a dropped connection after DATA that had
5801 been sent in pipelining mode. That is, the server had advertised PIPELINING
5802 but was not implementing it correctly. The effect of the bug was incorrect
5803 behaviour, such as trying another host, and this could lead to a crash.
5809 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5810 and daemon.c were passed as pointers to ints; they should have been
5811 pointers to socklen_t variables (which are typically unsigned ints).
5813 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5816 3. Fixed a really odd bug that affected only the testing scheme; patching a
5817 certain fixed string in the binary changed the value of another string that
5818 happened to be identical to the end of the original first string.
5820 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5821 name", it returns that address as the IP address. On some operating
5822 systems (e.g. Solaris), it also passes back the IP address string as the
5823 "host name". However, on others (e.g. Linux), it passes back an empty
5824 string. Exim wasn't checking for this, and was changing the host name to an
5825 empty string, assuming it had been canonicalized.
5827 5. Although rare, it is permitted to have more than one PTR record for a given
5828 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5829 all the names associated with an address, because they do in Solaris.
5830 However, it seems that they do not in Linux for data that comes from the
5831 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5832 I found this out when I moved to a new Linux workstation and tried to run
5833 the Exim test suite.
5835 To get round this problem I have changed the code so that it now does its
5836 own call to the DNS to look up PTR records when searching for a host name.
5837 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5838 addresses that are only in /etc/hosts are still found.
5840 This behaviour is, however, controlled by an option called host_lookup_
5841 order, which defaults to "bydns:byaddr". If people want to use the other
5842 order, or indeed, just use one or the other means of lookup, they can
5843 specify it in this variable.
5845 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5846 some operating systems, this comes back from gethostbyaddr() as an empty
5847 string, and this is what Exim used to test for. However, it seems that in
5848 other systems, "." is yielded. Exim now tests for this case too.
5850 7. The values of check_spool_space and check_log_space are now held internally
5851 as a number of kilobytes instead of an absolute number of bytes. If a
5852 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5853 kilobyte. This means that much larger values can be stored.
5855 8. Exim monitor: an attempt to get the action menu when not actually pointing
5856 at a message produces an empty menu entitled "No message selected". This
5857 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5858 no entries in it ("Shell widget menu has zero width and/or height"). So I
5859 have added a single, blank menu entry in this case.
5861 9. Added ${quote_local_part.
5863 10. MIME decoding is now applied to the contents of Subject: header lines when
5866 11. Now that a reference to $sender_host_address automatically causes a reverse
5867 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5868 host lookup before query-style lookups in lists that might use this
5869 variable. This has therefore been abolished, and the "net-" prefix is no
5870 longer necessary for query-style lookups.
5872 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5873 have been a typo for LFLAGS, so it has been changed.
5875 13. The install script calls Exim with "-C /dev/null" in order to find the
5876 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5877 to be output. However, since Exim outputs its version number before the
5878 error, it didn't break the script. It just looked ugly. I fixed this by
5879 always allowing "-C /dev/null" if the caller is root.
5881 14. Ignore overlarge ACL variable number when reading spool file - insurance
5882 against a later release with more variables having written the file.
5884 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5885 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5886 this, as well as the form without the "IPv6" on the front.
5888 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5889 OS/Makefile-Darwin file.
5891 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5892 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5894 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5895 variable. However, some OS use ints for inodes. I've added cast to long int
5896 to get rid of the compiler warning.
5898 19. I had forgotten to lock out "/../" in configuration file names when
5899 ALT_CONFIG_PREFIX was set.
5901 20. Routers used for verification do not need to specify transports. However,
5902 if such a router generated a host list, and callout was configured, Exim
5903 crashed, because it could not find a port number from the (non-existent)
5904 transport. It now assumes port 25 in this circumstance.
5906 21. Added the -t option to exigrep.
5908 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5909 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5910 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5911 them is set, LOOKUP_LSEARCH is forced.
5913 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5915 24. Added sender and host information to the "rejected by local_scan()" log
5916 line; previously there was no indication of these.
5918 25. Added .include_if_exists.
5920 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5921 Linux. It turns out that not all file systems support this. Apparently some
5922 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5923 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5924 not supported on the file descriptor, is now ignored when Exim is trying to
5925 sync a directory. This applies only to Linux.
5927 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5929 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5930 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5933 29. Added support for maildirsize files from supplied patch (modified a bit).
5935 30. The use of :fail: followed by an empty string could lead Exim to respond to
5936 sender verification failures with (e.g.):
5938 550 Verification failed for <xxx>
5939 550 Sender verify failed
5941 where the first response line was missing the '-' that indicates it is not
5942 the final line of the response.
5944 31. The loop for finding the name of the user that called Exim had a hardwired
5945 limit of 10; it now uses the value of finduser_retries, which is used for
5946 all other user lookups.
5948 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5951 33. Exim was neglecting to zero errno before one call of strtol() when
5952 expanding a string and expecting an integer value. On some systems this
5953 resulted in spurious "integer overflow" errors. Also, it was casting the
5954 result into an int without checking.
5956 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5957 did not work. The code looks as if it has *never* worked, though it appears
5958 to have been documented since at least release 1.62. I have made it work.
5960 35. The "timeout_DNS" error in retry rules, also documented since at least
5961 1.62, also never worked. As it isn't clear exactly what this means, and
5962 clearly it isn't a major issue, I have abolished the feature by treating it
5963 as "timeout", and writing a warning to the main and panic logs.
5965 36. The display of retry rules for -brt wasn't always showing the error code
5968 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5969 timeout_connect_A, timeout_connect_MX.
5971 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5972 to the empty sender.
5974 39. The daemon was not analysing the content of -oX till after it had closed
5975 stderr and disconnected from the controlling terminal. This meant that any
5976 syntax errors were only noted on the panic log, and the return code from
5977 the command was 0. By re-arranging the code a little, I've made the
5978 decoding happen first, so such errors now appear on stderr, and the return
5979 code is 1. However, the actual setting up of the sockets still happens in
5980 the disconnected process, so errors there are still only recorded on the
5983 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5984 connections (as happens on some IP stacks) was logged at start up time as
5985 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5986 it from "IPv6 and IPv4", which means that two separate sockets are being
5989 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5990 says whether AF_INET or AF_INET6 was passed as an argument.
5992 42. Exiwhat output was messed up when time zones were included in log
5995 43. Exiwhat now gives more information about the daemon's listening ports,
5996 and whether -tls-on-connect was used.
5998 44. The "port" option of the smtp transport is now expanded.
6000 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6001 silently ignored. Now an error message is written to the main and panic
6004 46. There's a new ACL modifier called "logwrite" which writes to a log file
6005 as soon as it is encountered.
6007 47. Added $local_user_uid and $local_user_gid at routing time.
6009 48. Exim crashed when trying to verify a sender address that was being
6012 49. Exim was recognizing only a space character after ".include". It now also
6013 recognizes a tab character.
6015 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6016 extracting the relevant information from the specification. The man page no
6017 longer contains scrambled data for the -d option, and I've added a section
6018 at the front about calling Exim under different names.
6020 51. Added "extra_headers" argument to the "mail" command in filter files.
6022 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6025 53. Installed eximstats 1.29.
6027 54. Added transport_filter_timeout as a generic transport option.
6029 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6030 Cc: header lines. This was required by RFC 822, but it not required by RFC
6033 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6034 incoming messages that did not have them. Now it does so only if the
6035 message originates locally, that is, if there is no associated remote host
6036 address. When Resent- header lines are present, this applies to the Resent-
6037 lines rather than the non-Resent- lines.
6039 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6040 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6042 58. Messages for configuration errors now include the name of the main
6043 configuration file - useful now that there may be more than one file in a
6044 list (.included file names were always shown).
6046 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6047 for those rare installations that do not start the daemon as root or run it
6048 setuid root. I've cut out the call to initgroups() if the daemon is not
6051 60. The Exim user and group can now be bound into the binary as text strings
6052 that are looked up at the start of Exim's processing.
6054 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6056 62. Added $mailstore_basename variable.
6058 63. Installed patch to sieve.c from Michael Haardt.
6060 64. When Exim failed to open the panic log after failing to open the main log,
6061 the original message it was trying to log was written to stderr and debug
6062 output, but if they were not available (the usual case in production), it
6063 was lost. Now it is written to syslog before the two lines that record the
6064 failures to open the logs.
6066 65. Users' Exim filters run in subprocesses under the user's uid. It is
6067 possible for a "deliver" command or an alias in a "personal" command to
6068 provoke an address rewrite. If logging of address rewriting is configured,
6069 this fails because the process is not running as root or exim. There may be
6070 a better way of dealing with this, but for the moment (because 4.30 needs
6071 to be released), I have disabled address rewrite logging when running a
6072 filter in a non-root, non-exim process.
6078 1. The buildconfig auxiliary program wasn't quoting the value set for
6079 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6080 not defined. This bug was masked in 4.22 by the effect that was fixed in
6083 2. Some messages that were rejected after a message id was allocated were
6084 shown as "incomplete" by exigrep. It no longer does this for messages that
6085 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6087 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6088 have allow_domain_literals set, the ID did not get logged in the <= line.
6089 Domain literals are now always recognized in Message-ID: header lines.
6091 4. The first argument for a ${extract expansion item is the key name or field
6092 number. Leading and trailing spaces in this item were not being ignored,
6093 causing some misleading effects.
6095 5. When deliver_drop_privilege was set, single queue runner processes started
6096 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6097 same command in the process it spins off) were not dropping privilege.
6099 6. When the daemon running as "exim" started a queue runner, it always
6100 re-executed Exim in the spun-off process. This is a waste of effort when
6101 deliver_drop_privilege is set. The new process now just calls the
6102 queue-runner function directly.
6108 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6111 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6112 ignored. Though the use of -oP was forcing the writing of a pid file, it
6113 was always written to the default place.
6115 3. If the message "no IP address found for host xxxx" is generated during
6116 incoming verification, it is now followed by identification of the incoming
6117 connection (so you can more easily find what provoked it).
6119 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6121 5. Added some features to "harden" Exim a bit more against certain attacks:
6123 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6124 be put in Local/Makefile. This is like the never_users runtime option,
6125 but it cannot be overridden. The default setting is "root".
6127 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6128 prefix string with which any file named in a -C command line option
6131 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6132 is retained for -C and -D only if the caller of Exim is root. Without
6133 it, the exim user may also use -C and -D and retain privilege.
6135 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6136 command line option is disabled.
6138 6. Macro names set by the -D option must start with an upper case letter, just
6139 like macro names defined in the configuration file.
6141 7. Added "dereference=" facility to LDAP.
6143 8. Two instances of the typo "uknown" in the source files are fixed.
6145 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6146 the Configure-Makefile script screwed up while processing it.
6148 10. Incorporated PCRE 4.4.
6150 11. The SMTP synchronization check was not operating right at the start of an
6151 SMTP session. For example, it could not catch a HELO sent before the client
6152 waited for the greeting. There is now a check for outstanding input at the
6153 point when the greeting is written. Because of the duplex, asynchronous
6154 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6155 way, but not yet received, when the check is performed.
6157 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6158 on TCP/IP sockets, because this apparently causes some broken clients to
6161 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6162 unchanged) from the Cygwin maintainer.
6164 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6165 was supported instead of testing for mbx. Effectively a typo.
6167 15. The spa authenticator server code was not checking that the input it
6168 received was valid base64.
6170 16. The debug output line for the "set" modifier in ACLs was not showing the
6171 name of the variable that was being set.
6173 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6175 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6176 lookup on its own. Something else was needed to trigger the lookup. For
6177 example, a match in host_lookup or the need for a host name in a host list.
6178 Now, if $sender_host_name is referenced and the host name has not yet been
6179 looked up, a lookup is performed. If the lookup fails, the variable remains
6180 empty, and $host_lookup_failed is set to "1".
6182 19. Added "eqi" as a case-independent comparison operator.
6184 20. The saslauthd authentication condition could segfault if neither service
6185 nor realm was specified.
6187 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6188 error message that was logged was misleading, and incoming SMTP
6189 connections were dropped. The message is now more accurate, and temporary
6190 errors are given to SMTP connections.
6192 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6193 (except RSET) with a 421 error, until QUIT is received. However, it was
6194 failing to send a response to QUIT.
6196 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6197 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6198 ACL is now run after the helo_try_verify_hosts code.
6200 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6201 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6202 case-independent, so other case variants are also recognized). Apparently
6203 some systems use these upper case variants.
6205 25. If more than two messages were waiting for the same host, and a transport
6206 filter was specified for the transport, Exim sent two messages over the
6207 same TCP/IP connection, and then failed with "socket operation on non-
6208 socket" when it tried to send the third.
6210 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6212 27. The extern definition of crypt16() in expand.c was not being excluded when
6213 the OS had its own crypt16() function.
6215 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6216 as a preferred synonym for return_size_limit, both as an option and as an
6219 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6221 30. Changed the default configuration ACL to relax the local part checking rule
6222 for addresses that are not in any local domains. For these addresses,
6223 slashes and pipe symbols are allowed within local parts, but the sequence
6224 /../ is explicitly forbidden.
6226 31. SPA server authentication was not clearing the challenge buffer before
6229 32. log_message in a "warn" ACL statement was writing to the reject log as
6230 well as to the main log, which contradicts the documentation and doesn't
6231 seem right (because no rejection is happening). So I have stopped it.
6233 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6234 However, I am unable to do any testing of this.
6236 34. Fixed an infelicity in the appendfile transport. When checking directories
6237 for a mailbox, to see if any needed to be created, it was accidentally
6238 using path names with one or more superfluous leading slashes; tracing
6239 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6241 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6242 contents are added to the log line that is written for every discarded
6243 recipient. (Previously a log_message setting was ignored.)
6245 36. The ${quote: operator now quotes the string if it is empty.
6247 37. The install script runs exim in order to find its version number. If for
6248 some reason other than non-existence or emptiness, which it checks, it
6249 could not run './exim', it was installing it with an empty version number,
6250 i.e. as "exim-". This error state is now caught, and the installation is
6253 38. An argument was missing from the function that creates an error message
6254 when Exim fails to connect to the socket for saslauthd authentication.
6255 This could cause Exim to crash, or give a corrupted message.
6257 39. Added isip, isip4, and isip6 to ${if conditions.
6259 40. The ACL variables $acl_xx are now saved with the message, and can be
6260 accessed later in routers, transports, and filters.
6262 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6263 strings in the file are not string-expanded.
6265 42. If a MAIL command specified a SIZE value that was too large to fit into an
6266 int variable, the check against message_size_limit failed. Such values are
6267 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6268 day this will have to be increased, but I don't think I want to be around
6269 when emails are that large.
6276 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6277 iconv() is not standard in FreeBSD.
6279 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6280 IPv6 enabled. The observed symptom was a segmentation fault on return from
6281 the function os_common_find_running_interfaces() in src/os.c.
6283 3. In the check_special_case() function in daemon.c I had used "errno" as an
6284 argument name, which causes warnings on some systems. This was basically a
6285 typo, since it was named "eno" in the comments!
6287 4. The code that waits for the clock to tick (at a resolution of some fraction
6288 of a second) so as to ensure message-id uniqueness was always waiting for
6289 at least one whole tick, when it could have waited for less. [This is
6290 almost certainly not relevant at current processor speeds, where it is
6291 unlikely to ever wait at all. But we try to future-proof.]
6293 5. The function that sleeps for a time interval that includes fractions of a
6294 second contained a race. It did not block SIGALRM between setting the
6295 timer, and suspending (a couple of lines later). If the interval was short
6296 and the sigsuspend() was delayed until after it had expired, the suspension
6297 never ended. On busy systems this could lead to processes getting stuck for
6300 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6301 process, before it forks any delivery processes. The open lookup caching
6302 mechanism meant that the open file or database connection was passed into
6303 the delivery process. The problem was that delivery processes always tidy
6304 up cached lookup data. This could cause a problem for the next delivery
6305 process started by the queue runner, because the external queue runner
6306 process does not know about the closure. So the next delivery process
6307 still has data in the lookup cache. In the case of a file lookup, there was
6308 no problem because closing a file descriptor in a subprocess doesn't affect
6309 the parent. However, if the lookup was caching a connection to a database,
6310 the connection was closed, and the second delivery process was likely to
6311 see errors such as "PGSQL: query failed: server closed the connection
6312 unexpectedly". The problem has been fixed by closing all cached lookups
6313 in a queue runner before running a delivery process.
6315 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6316 seem to have the "const" qualifier which it has on other OS. I've
6319 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6320 *of the same type* (client or server) with the same public name that an
6321 error should be diagnosed.
6323 9. When Exim looked up a host name for an IP address, but failed to find the
6324 original IP address when looking up the host name (a safety check), it
6325 output the message "<ip address> does not match any IP for NULL", which was
6326 confusing, to say the least. The bug was that the host name should have
6327 appeared instead of "NULL".
6329 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6330 user that is built into the binary, and the -C or -D options is used, root
6331 privilege is dropped before the configuration file is read. In addition,
6332 logging is switched to stderr instead of the normal log files. If the
6333 configuration then re-defines the Exim user, the unprivileged environment
6334 is probably not what is expected, so Exim logs a panic warning message (but
6337 However, if deliver_drop_privilege is set, the unprivileged state may well
6338 be exactly what is intended, so the warning has been cut out in that case,
6339 and Exim is allowed to try to write to its normal log files.
6345 1. smtp_return_error_details was not giving details for temporary sender
6346 or receiver verification errors.
6348 2. Diagnose a configuration error if two authenticators have the same public
6351 3. Exim used not to create the message log file for a message until the first
6352 delivery attempt. This could be confusing when incoming messages were held
6353 for policy or load reasons. The message log file is now created at the time
6354 the message is received, and an initial "Received" line is written to it.
6356 4. The automatically generated man page for command line options had a minor
6357 bug that caused no ill effects; however, a more serious problem was that
6358 the procedure for building the man page automatically didn't always
6359 operate. Consequently, release 4.20 contains an out-of-date version. This
6360 shouldn't happen again.
6362 5. When building Exim with embedded Perl support, the script that builds the
6363 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6364 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6366 6. The freeze_tell option was not being used for messages that were frozen on
6367 arrival, either by an ACL or by local_scan().
6369 7. Added the smtp_incomplete_transaction log selector.
6371 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6372 was accepting AUTH without a new EHLO.
6374 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6375 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6378 10. Logging of TCP/IP connections (when configured) now happens in the main
6379 daemon process instead of the child process, so that the TCP/IP connection
6380 count is more accurate (but it can never be perfect).
6382 11. The use of "drop" in a nested ACL was not being handled correctly in the
6383 outer ACL. Now, if condition failure induced by the nested "drop" causes
6384 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6385 or "require"), the connection is dropped.
6387 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6388 that yield "discard" can now be used with an "accept" or a "discard" verb,
6389 but an error is generated for any others (because I can't see a useful way
6390 to define what should happen).
6392 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6393 are now processed in the same way as lines in the Exim configuration file.
6394 In particular, continuation lines are supported.
6396 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6398 15. Added -ti meaning -t -i.
6400 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6401 domains, and warn by logging if others are found.
6403 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6404 returned from GIFCONF when Exim is trying to find the list of interfaces on
6405 a host. The code in os.c has been modified to copy each ifreq to an aligned
6406 structure in all cases.
6408 Also, in some cases, the returned ifreq's were being copied to a 'struct
6409 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6410 means the last couple of bytes of an IPv6 address could be chopped if the
6411 ifreq contained only a normal sockaddr (14 bytes storage).
6413 18. Named domain lists were not supported in the hosts_treat_as_local option.
6414 An entry such as +xxxx was not recognized, and was treated as a literal
6417 19. Ensure that header lines added by a DATA ACL are included in the reject log
6418 if the ACL subsequently rejects the message.
6420 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6421 MD5 (which is deprecated).
6423 21. When testing a filter file using -bf, Exim was writing a message when it
6424 took the sender from a "From " line in the message, but it was not doing so
6425 when it took $return_path from a Return-Path: header line. It now does.
6427 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6428 with a valid header line field name (a series of printing characters
6429 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6431 23. Changed "disc" in the source to "disk" to conform to the documentation and
6432 the book and for uniformity.
6434 24. Ignore Sendmail's -Ooption=value command line item.
6436 25. When execve() failed while trying to run a command in a pipe transport,
6437 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6438 could be confused with a return value of 69 from the command itself. This
6439 has been changed to 127, the value the shell returns if it is asked to run
6440 a non-existent command. The wording for the related log line suggests a
6441 non-existent command as the problem.
6443 26. If received_header_text expands to an empty string, do not add a Received:
6444 header line to the message. (Well, it adds a token one on the spool, but
6445 marks it "old" so that it doesn't get used or transmitted.)
6447 27. Installed eximstats 1.28 (addition of -nt option).
6449 28. There was no check for failure on the call to getsockname() in the daemon
6450 code. This can fail if there is a shortage of resources on the system, with
6451 ENOMEM, for example. A temporary error is now given on failure.
6453 29. Contrary to the C standard, it seems that in some environments, the
6454 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6455 program. Exim now does this explicitly; it affects the formatting of
6456 timestamps using strftime().
6458 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6459 complaints. I've now initialized all the variables, to avoid this.
6461 32. Header lines added by a system filter were not being "seen" during
6462 transport-time rewrites.
6464 33. The info_callback() function passed to OpenSSL is set up with type void
6465 (*)(SSL *, int, int), as described somewhere. However, when calling the
6466 function (actually a macro) that sets it up, the type void(*)() is
6467 expected. I've put in a cast to prevent warnings from picky compilers.
6469 34. If a DNS black list lookup found a CNAME record, but there were no A
6470 records associated with the domain it pointed at, Exim crashed.
6472 35. If a DNS black list lookup returned more than one A record, Exim ignored
6473 all but the first. It now scans all returned addresses if a particular IP
6474 value is being sought. In this situation, the contents of the
6475 $dnslist_value variable are a list of all the addresses, separated by a
6478 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6479 to accept a host name and all its aliases if the forward lookup for any of
6480 them yielded the IP address of the incoming connection. Now it accepts only
6481 those names whose forward lookup yields the correct IP address. Any other
6482 names are discarded. This closes a loophole whereby a rogue DNS
6483 administrator could create reverse DNS records to break through a
6484 wildcarded host restriction in an ACL.
6486 37. If a user filter or a system filter that ran in a subprocess used any of
6487 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6488 the wrong values were passed to the pipe command ($thisaddress had the
6489 value of $0, $0 had the value of $1, etc). This bug was introduced by
6490 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6492 38. Improved the line breaking for long SMTP error messages from ACLs.
6493 Previously, if there was no break point between 40 and 75 characters, Exim
6494 left the rest of the message alone. Two changes have been made: (a) I've
6495 reduced the minimum length to 35 characters; (b) if it can't find a break
6496 point between 35 and 75 characters, it looks ahead and uses the first one
6497 that it finds. This may give the occasional overlong line, but at least the
6498 remaining text gets split now.
6500 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6501 file descriptors might be low, and that setting 1000 would always raise it.
6502 It turns out that in some environments, the limit is already over 1000 and
6503 that lowering it causes trouble. So now Exim takes care not to decrease it.
6505 40. When delivering a message, the value of $return_path is set to $sender_
6506 address at the start of routing (routers may change the value). By an
6507 oversight, this default was not being set up when an address was tested by
6508 -bt or -bv, which affected the outcome if any router or filter referred to
6511 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6512 log only when "message" or "log_message" are set. However, if one of the
6513 conditions was an address verification, or a call to a nested ACL, the
6514 messages generated by the underlying test were being passed through. This
6515 no longer happens. The underlying message is available in $acl_verify_
6516 message for both "message" and "log_message" expansions, so it can be
6517 passed through if needed.
6519 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6520 new expansion $bh_ to give the encoded byte string without charset
6521 translation. Translation happens only if iconv() is available; HAVE_ICONV
6522 indicates this at build time. HEADERS_CHARSET gives the charset to
6523 translate to; headers_charset can change it in the configuration, and
6524 "headers charset" can change it in an individual filter file.
6526 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6527 that creates RFC 2047 encoded "words" labels them as that charset instead
6528 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6529 expansion operator; (ii) when Exim creates a From: line for a local
6530 message; (iii) when a header line is rewritten to include a "phrase" part.
6532 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6533 buggy, causing it to skip the first lines of messages whose message ID
6534 ended in 'D'. This would not have bitten before Exim release 4.14, because
6535 message IDs were unlikely to end in 'D' before then. The effect was to have
6536 incorrect size information for certain domains.
6538 45. #include "config.h" was missing at the start of the crypt16.c module. This
6539 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6542 46. If there was a timeout during a "random" callout check, Exim treated it as
6543 a failure of the random address, and carried on sending RSET and the real
6544 address. If the delay was just some slowness somewhere, the response to the
6545 original RCPT would be taken as a response to RSET and so on, causing
6546 mayhem of various kinds.
6548 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6549 when I implemented it. It didn't allow for the fact that some option values
6550 may legitimately be negative (e.g. size_addition), and it didn't even do
6551 the right test for positive values.
6553 48. Domain names in DNS records are case-independent. Exim always looks them up
6554 in lower case. Some resolvers return domain names in exactly the case they
6555 appear in the zone file, that is, they may contain uppercase letters. Not
6556 all resolvers do this - some return always lower case. Exim was treating a
6557 change of case by a resolver as a change of domain, similar to a widening
6558 of a domain abbreviation. This triggered its re-routing code and so it was
6559 trying to route what was effectively the same domain again. This normally
6560 caused routing to fail (because the router wouldn't handle the domain
6561 twice). Now Exim checks for this case specially, and just changes the
6562 casing of the domain that it ultimately uses when it transmits the message
6565 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6568 50. If a filter generated a file delivery with a non-absolute name (possible if
6569 no home directory exists for the router), the forbid_file option was not
6572 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6573 the existing equality matching.
6575 52. Exim was using ints instead of ino_t variables in some places where it was
6576 dealing with inode numbers.
6578 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6579 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6580 called TMPDIR, and if it finds it is different, it changes its value.
6582 54. The smtp_printf() function is now made available to local_scan() so
6583 additional output lines can be written before returning. There is also an
6584 smtp_fflush() function to enable the detection of a dropped connection.
6585 The variables smtp_input and smtp_batched_input are exported to
6588 55. Changed the default runtime configuration: the message "Unknown user"
6589 has been removed from the ACL, and instead placed on the localuser router,
6590 using the cannot_route_message feature. This means that any verification
6591 failures that generate their own messages won't get overridden. Similarly,
6592 the "Unrouteable address" message that was in the ACL for unverifiable
6593 relay addresses has also been removed.
6595 56. Added hosts_avoid_esmtp to the smtp transport.
6597 57. The exicyclog script was not checking for the esoteric option
6598 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6599 will work only if exicyclog is run under the appropriate euid.
6601 58. Following a discussion on the list, the rules by which Exim recognises line
6602 endings on incoming messages have been changed. The -dropcr and drop_cr
6603 options are now no-ops, retained only for backwards compatibility. The
6604 following line terminators are recognized: LF CRLF CR. However, special
6605 processing applies to CR:
6607 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6608 nor a local message in the state where . is a terminator.
6610 (ii) If a bare CR is encountered in a header line, an extra space is added
6611 after the line terminator so as not to end the header. The reasoning
6612 behind this is that bare CRs in header lines are most likely either
6613 to be mistakes, or people trying to play silly games.
6615 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6616 was being incorrectly given as 18 bytes larger than it should have been.
6617 This is a VOB (very old bug).
6619 60. This may never have affected anything current, but just in case it has:
6620 When the local host is found other than at the start of a list of hosts,
6621 the local host, those with the same MX, and any that follow, are discarded.
6622 When the list in question was part of a longer list of hosts, the following
6623 hosts (not currently being processed) were also being discarded. This no
6624 longer happens. I'm not sure if this situation could ever has previously
6627 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6630 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6631 header. This is something that is recommended in a new Internet Draft, and
6632 is something that is documented as being done by Sendmail. There are two
6633 possible values. For messages generated by the autoreply transport, Exim
6636 Auto-Submitted: auto-replied
6638 whereas for all other generated messages (e.g. bounces) it adds
6640 Auto-Submitted: auto-generated
6642 63. The "personal" condition in filters now includes a test for the
6643 Auto-Submitted: header. If it contains the string "auto-" the message it
6644 not considered personal.
6646 64. Added rcpt_include_affixes as a generic transport option.
6648 65. Added queue_only_override (default true).
6650 66. Added the syslog_duplication option.
6652 67. If what should have been the first header line of a message consisted of
6653 a space followed by a colon, Exim was mis-interpreting it as a header line.
6654 It isn't of course - it is syntactically invalid and should therefore be
6655 treated as the start of the message body. The misbehaviour could have
6656 caused a number of strange effects, including loss of data in subsequent
6657 header lines, and spool format errors.
6659 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6660 client host had authenticated. This control can now be exercised by an ACL
6661 for more flexibility.
6663 69. By default, callouts do not happen when testing with -bh. There is now a
6664 variant, -bhc, which does actually run the callout code, including
6665 consulting and updating the callout cache.
6667 70. Added support for saslauthd authentication, courtesy of Alexander
6670 71. If statvfs() failed on the spool or log directories while checking their
6671 size for availability, Exim confusingly gave the error "space shortage".
6672 Furthermore, in debugging mode it crashed with a floating point exception.
6673 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6674 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6675 serious problem, Exim now writes to the main and panic logs when this
6676 happens, with details of the failure. It then refuses to accept the
6677 incoming message, giving the message "spool directory problem" or "log
6678 directory problem" with a 421 code for SMTP messages.
6680 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6681 0, 1, and 2 exist, because some OS complain for execs without them (see
6682 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6683 descriptors. However, the code omitted to check that the open succeeded,
6684 causing mysterious errors if for some reason the permissions on /dev/null
6685 got screwed. Now Exim writes a message to the main and panic logs, and
6686 bombs out if it can't open /dev/null.
6688 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6689 interact so that it is all more flexible. It is supposed to remain
6690 backwards compatible. Also added extra_local_interfaces.
6692 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6693 to bomb out with an assertion failure - to the client this appears as a
6694 connection drop. This problem occurs in the part of the code that was taken
6695 from the Samba project. Fortunately, the assertion is in a very simple
6696 function, so I have fixed this by reproducing the function inline in the
6697 one place where it is called, and arranging for authentication to fail
6698 instead of killing the process with assert().
6700 75. The SPA client code was not working when the server requested OEM rather
6701 than Unicode encoding.
6703 76. Added code to make require_files with a specific uid setting more usable in
6704 the case where statting the file as root fails - usually a non-root-mounted
6705 NFS file system. When this happens and the failure is EACCES, Exim now
6706 forks a subprocess and does the per-uid checking as the relevant uid.
6708 77. Added process_log_path.
6710 78. If log_file_path was not explicitly set, a setting of check_log_space or
6711 check_log_inodes was ignored.
6713 79. If a space check for the spool or log partitions fails, the incident is now
6714 logged. Of course, in the latter case the data may get lost...
6716 80. Added the %p formatting code to string_format() so that it can be used to
6717 print addresses in debug_print(). Adjusted all the address printing in the
6718 debugging in store.c to use %p rather than %d.
6720 81. There was a concern that a line of code in smtp_in.c could overflow a
6721 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6722 initially expressed, the concern was not well-founded, because trailing
6723 spaces are removed early. However, if the trailing spaces were followed by
6724 a NULL, they did not get removed, so the overflow was possible. Two fixes
6727 (a) I re-wrote the offending code in a cleaner fashion.
6728 (b) If an incoming SMTP command contains a NULL character, it is rejected
6731 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6732 runs initgroups(), so that if the Exim user is in any additional groups,
6733 they will be used during message reception.
6739 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob