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.
92 JH/17 Rework error string handling in TLS interface so that the caller in
93 more cases is responsible for logging. This permits library-sourced
94 string to be attached to addresses during delivery, and collapses
95 pairs of long lines into single ones.
101 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
102 supports it and a size is available (ie. the sending peer gave us one).
104 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
105 years of being deprecated). The replacements are the ACLs
106 acl_smtp_mime and acl_not_smtp_mime.
108 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
109 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
110 or even an in-clear connection were permitted. Now, if the host lookup
111 was dnssec and dane was requested then the host is only used if the
112 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
113 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
114 if one fails this test.
115 This means that a poorly-configured remote DNS will make it incommunicado;
116 but it protects against a DNS-interception attack on it.
118 JH/04 Bug 1810: make continued-use of an open smtp transport connection
119 non-noisy when a race steals the message being considered.
121 JH/05 If main configuration option tls_certificate is unset, generate a
122 self-signed certificate for inbound TLS connections.
124 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
125 in rewrites and routers.
127 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
128 and logged a warning sing 4.83; now they are a configuration file error.
130 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
131 (lacking @domain). Apply the same qualification processing as RCPT.
133 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
135 JH/10 Support ${sha256:} applied to a string (as well as the previous
138 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
139 a cutthrough deliver is pending, as we always want to make a connection.
140 This also avoids re-routing the message when later placing the cutthrough
141 connection after a verify cache hit.
142 Do not update it with the verify result either.
144 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
145 when routing results in more than one destination address.
147 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
148 signing (which inhibits the cutthrough capability). Previously only
149 the presence of an option was tested; now an expansion evaluating as
150 empty is permissible (obviously it should depend only on data available
151 when the cutthrough connection is made).
153 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
154 the relevant preceding SMTP command did not note the pipelining mode.
156 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
157 Previously they were not counted.
159 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
160 as one having no matching records. Previously we deferred the message
161 that needed the lookup.
163 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
164 distinguished as "(=".
166 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
167 for missing MX records. Previously it only worked for missing A records.
169 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
171 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
172 after the data-go-ahead and data-ack. Patch from Jason Betts.
174 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
175 even for a "none" policy. Patch from Tony Meyer.
177 JH/22 Fix continued use of a connection for further deliveries. If a port was
178 specified by a router, it must also match for the delivery to be
181 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
182 When one of the recipients of a message was unsuitable for the connection
183 (has no matching addresses), we lost track of needing to mark it
184 deferred. As a result mail would be lost.
186 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
188 JH/25 Decoding ACL controls is now done using a binary search; the source code
189 takes up less space and should be simpler to maintain. Merge the ACL
190 condition decode tables also, with similar effect.
192 JH/26 Fix problem with one_time used on a redirect router which returned the
193 parent address unchanged. A retry would see the parent address marked as
194 delivered, so not attempt the (identical) child. As a result mail would
197 JH/27 Fix a possible security hole, wherein a process operating with the Exim
198 UID can gain a root shell. Credit to http://www.halfdog.net/ for
199 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
202 JH/28 Enable {spool,log} filesystem space and inode checks as default.
203 Main config options check_{log,spool}_{inodes,space} are now
204 100 inodes, 10MB unless set otherwise in the configuration.
206 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
207 Previously it only applied to the main-section connection policy
210 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
212 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
213 by me. Added RFC7919 DH primes as an alternative.
215 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
216 are not in the system include path.
218 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
219 GnuTLS, when a session startup failed (eg because the client disconnected)
220 Exim did stdio operations after fclose. This was exposed by a recent
221 change which nulled out the file handle after the fclose.
223 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
224 signed directly by the cert-signing cert, rather than an intermediate
225 OCSP-signing cert. This is the model used by LetsEncrypt.
227 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
229 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
230 an incoming connection.
232 HS/02 Bug 1802: Do not half-close the connection after sending a request
235 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
236 fallback to "prime256v1".
238 JH/34 SECURITY: Use proper copy of DATA command in error message.
239 Could leak key material. Remotely exploitable. CVE-2016-9963.
245 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
246 and 3.4.4 - once the server is enabled to respond to an OCSP request
247 it does even when not requested, resulting in a stapling non-aware
248 client dropping the TLS connection.
250 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
251 support variable-length bit vectors. No functional change.
253 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
254 The I= interface field on outgoing lines is now after the H= remote
255 host field, same as incoming lines. There is a separate
256 outgoing_interface log selector which allows you to disable the
259 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
260 If not running log_selector +smtp_connection the mainlog would be held
261 open indefinitely after a "too many connections" event, including to a
262 deleted file after a log rotate. Leave the per net connection logging
263 leaving it open for efficiency as that will be quickly detected by the
264 check on the next write.
266 HS/01 Bug 1671: Fix post transport crash.
267 Processing the wait-<transport> messages could crash the delivery
268 process if the message IDs didn't exist for some reason. When
269 using 'split_spool_directory=yes' the construction of the spool
270 file name failed already, exposing the same netto behaviour.
272 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
273 mime_regex ACL conditions.
275 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
276 to DSN fail messages (bounces): remote IP, remote greeting, remote response
277 to HELO, local diagnostic string.
279 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
280 log line to debug. Even when configured with a tls authenticator many
281 client connections are expected to not authenticate in this way, so
282 an authenticate fail is not an error.
284 HS/02 Add the Exim version string to the process info. This way exiwhat
285 gives some more detail about the running daemon.
287 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
288 matter for fast-change records such as DNSBLs.
290 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
291 retry record, even if constant. There may be multiple transports with
292 different interface settings and the retry behaviour needs to be kept
295 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
297 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
299 JH/10 Bug 840: fix log_defer_output option of pipe transport
301 JH/11 Bug 830: use same host for all RCPTS of a message, even under
302 hosts_randomize. This matters a lot when combined with mua_wrapper.
304 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
305 ${quote_pgsql:<string>} operator.
307 JH/13 Bug 1708: avoid misaligned access in cached lookup.
309 JH/14 Change header file name for freeradius-client. Relevant if compiling
310 with Radius support; from the Gentoo tree and checked under Fedora.
312 JH/15 Bug 1712: Introduce $prdr_requested flag variable
314 JH/16 Bug 1714: Permit an empty string as expansion result for transport
315 option transport_filter, meaning no filtering.
317 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
319 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
320 defaults to "*" (all hosts). The variable is now available when not built
321 with TLS, default unset, mainly to enable keeping the testsuite sane.
322 If a server certificate is not supplied (via tls_certificate) an error is
323 logged, and clients will find TLS connections fail on startup. Presumably
324 they will retry in-clear.
325 Packagers of Exim are strongly encouraged to create a server certificate
326 at installation time.
328 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
329 with the $config_file variable.
331 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
332 in transport context, after the attempt, and per-recipient. The latter type
333 is per host attempted. The event data is the error message, and the errno
334 information encodes the lookup type (A vs. MX) used for the (first) host,
335 and the trailing two digits of the smtp 4xx response.
337 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
338 to write to mainlog (or rejectlog, paniclog) in the window between file
339 creation and permissions/ownership being changed. Particularly affects
340 installations where exicyclog is run as root, rather than exim user;
341 result is that the running daemon panics and dies.
343 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
345 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
346 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
347 "pri" and "weight". Note that the previous implicit priority given by the
348 list order is no longer honoured.
350 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
353 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
354 by defining SUPPORT_SOCKS.
356 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
357 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
358 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
359 variables are renamed to proxy_{local,external}_{address,port}.
361 JH/27 Move Internationalisation support from Experimental to mainline, enabled
362 for a build by defining SUPPORT_I18N
364 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
365 of the query string, and make ${quote_redis:} do that quoting.
367 JH/29 Move Events support from Experimental to mainline, enabled by default
368 and removable for a build by defining DISABLE_EVENT.
370 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
372 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
373 cached by the daemon.
375 JH/32 Move Redis support from Experimental to mainline, enabled for a build
376 by defining LOOKUP_REDIS. The libhiredis library is required.
378 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
379 keys are given for lookup.
381 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
382 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
383 only supported when built with TLS support. The PolarSSL SHA routines
384 are still used when the TLS library is too old for convenient support.
386 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
387 openssl_options), for security. OpenSSL forces this from version 1.1.0
388 server-side so match that on older versions.
390 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
391 allocation for $value could be released as the expansion processing
392 concluded, but leaving the global pointer active for it.
394 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
395 and to use the domains and local_parts ACL conditions.
397 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
398 incorrectly not doubled on cutthrough transmission, hence seen as a
399 body-termination at the receiving system - resulting in truncated mails.
400 Commonly the sender saw a TCP-level error, and retransmitted the message
401 via the normal store-and-forward channel. This could result in duplicates
402 received - but deduplicating mailstores were liable to retain only the
403 initial truncated version.
405 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
407 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
409 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
410 we're in there, support oversigning also; bug 1309.
412 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
414 HS/04 Add support for keep_environment and add_environment options.
416 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
417 either intentional arithmetic overflow during PRNG, or testing config-
420 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
421 delivery resulted in actual delivery. Cancel cutthrough before DATA
424 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
425 rejects a recipient: pass the reject to the originator.
427 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
428 Many were false-positives and ignorable, but it's worth fixing the
431 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
432 for the new environment-manipulation done at startup. Move the routines
433 from being local to tls.c to being global via the os.c file.
435 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
436 an extract embedded as result-arg for a map, the first arg for extract
437 is unavailable so we cannot tell if this is a numbered or keyed
438 extraction. Accept either.
444 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
447 JH/02 The smtp transport option "multi_domain" is now expanded.
449 JH/03 The smtp transport now requests PRDR by default, if the server offers
452 JH/04 Certificate name checking on server certificates, when exim is a client,
453 is now done by default. The transport option tls_verify_cert_hostnames
454 can be used to disable this per-host. The build option
455 EXPERIMENTAL_CERTNAMES is withdrawn.
457 JH/05 The value of the tls_verify_certificates smtp transport and main options
458 default to the word "system" to access the system default CA bundle.
459 For GnuTLS, only version 3.0.20 or later.
461 JH/06 Verification of the server certificate for a TLS connection is now tried
462 (but not required) by default. The verification status is now logged by
463 default, for both outbound TLS and client-certificate supplying inbound
466 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
469 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
470 Status Notification (bounce) messages are now MIME format per RFC 3464.
471 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
472 under the control of the dsn_advertise_hosts option, and routers may
473 have a dsn_lasthop option.
475 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
476 default, modifiable by a malware= option. The list separator for
477 the options can now be changed in the usual way. Bug 68.
479 JH/10 The smtp_receive_timeout main option is now expanded before use.
481 JH/11 The incoming_interface log option now also enables logging of the
482 local interface on delivery outgoing connections.
484 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
485 if the interface and destination host and port all match.
487 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
490 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
491 Patch from Andrew Lewis.
493 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
494 now supports optional time-restrictions, weighting, and priority
495 modifiers per server. Patch originally by <rommer@active.by>.
497 JH/16 The spamd_address main option now supports a mixed list of local
498 and remote servers. Remote servers can be IPv6 addresses, and
499 specify a port-range.
501 JH/17 Bug 68: The spamd_address main option now supports an optional
502 timeout value per server.
504 JH/18 Bug 1581: Router and transport options headers_add/remove can
505 now have the list separator specified.
507 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
510 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
513 JH/21 Support for the A6 type of dns record is withdrawn.
515 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
516 rather than the verbs used.
518 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
519 from 255 to 1024 chars.
521 JH/24 Verification callouts now attempt to use TLS by default.
523 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
524 are generic router options now. The defaults didn't change.
526 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
527 Original patch from Alexander Shikoff, worked over by JH.
529 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
530 files if ZCAT_COMMAND is not executable.
532 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
534 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
536 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
537 Normally benign, it bites when the pair was led to by a CNAME;
538 modern usage is to not canonicalize the domain to a CNAME target
539 (and we were inconsistent anyway for A-only vs AAAA+A).
541 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
543 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
544 when evaluating $sender_host_dnssec.
546 JH/31 Check the HELO verification lookup for DNSSEC, adding new
547 $sender_helo_dnssec variable.
549 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
551 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
553 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
555 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
556 documented as working, but never had. Support all but $spam_report.
558 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
559 added for tls authenticator.
561 HS/03 Add perl_taintmode main config option
567 TL/01 When running the test suite, the README says that variables such as
568 no_msglog_check are global and can be placed anywhere in a specific
569 test's script, however it was observed that placement needed to be near
570 the beginning for it to behave that way. Changed the runtest perl
571 script to read through the entire script once to detect and set these
572 variables, reset to the beginning of the script, and then run through
573 the script parsing/test process like normal.
575 TL/02 The BSD's have an arc4random API. One of the functions to induce
576 adding randomness was arc4random_stir(), but it has been removed in
577 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
578 function when detected.
580 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
581 cause callback expansion.
583 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
584 syntax errors in an expansion can be treated as a string instead of
585 logging or causing an error, due to the internal use of bool_lax
586 instead of bool when processing it.
588 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
589 server certificates when making smtp deliveries.
591 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
593 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
595 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
597 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
598 Merged patch from Sebastian Wiedenroth.
600 JH/05 Fix results-pipe from transport process. Several recipients, combined
601 with certificate use, exposed issues where response data items split
602 over buffer boundaries were not parsed properly. This eventually
603 resulted in duplicates being sent. This issue only became common enough
604 to notice due to the introduction of connection certificate information,
605 the item size being so much larger. Found and fixed by Wolfgang Breyha.
607 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
608 size buffer was used, resulting in syntax errors when an expansion
611 JH/07 Add support for directories of certificates when compiled with a GnuTLS
612 version 3.3.6 or later.
614 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
615 is EXPERIMENTAL_EVENT, the main-configuration and transport options
616 both become "event_action", the variables become $event_name, $event_data
617 and $event_defer_errno. There is a new variable $verify_mode, usable in
618 routers, transports and related events. The tls:cert event is now also
619 raised for inbound connections, if the main configuration event_action
622 TL/06 In test suite, disable OCSP for old versions of openssl which contained
623 early OCSP support, but no stapling (appears to be less than 1.0.0).
625 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
626 server certificate names available under the smtp transport option
627 "tls_verify_cert_hostname" now do not permit multi-component wildcard
630 JH/10 Time-related extraction expansions from certificates now use the main
631 option "timezone" setting for output formatting, and are consistent
632 between OpenSSL and GnuTLS compilations. Bug 1541.
634 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
635 encoded parameter in the incoming message. Bug 1558.
637 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
638 include certificate info, eximon was claiming there were spoolfile
641 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
643 JH/14 Log delivery-related information more consistently, using the sequence
644 "H=<name> [<ip>]" wherever possible.
646 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
647 are problematic for Debian distribution, omit them from the release
650 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
652 JH/16 Fix string representation of time values on 64bit time_t architectures.
655 JH/17 Fix a null-indirection in certextract expansions when a nondefault
656 output list separator was used.
661 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
662 checkers that were complaining about end of non-void function with no
665 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
666 This was a regression introduced in 4.83 by another bugfix.
668 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
670 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
671 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
677 TF/01 Correctly close the server side of TLS when forking for delivery.
679 When a message was received over SMTP with TLS, Exim failed to clear up
680 the incoming connection properly after forking off the child process to
681 deliver the message. In some situations the subsequent outgoing
682 delivery connection happened to have the same fd number as the incoming
683 connection previously had. Exim would try to use TLS and fail, logging
684 a "Bad file descriptor" error.
686 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
687 utilities have not been installed.
689 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
690 temporary space as the ACL may create new global variables.
692 TL/01 LDAP support uses per connection or global context settings, depending
693 upon the detected version of the libraries at build time.
695 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
696 to extract and use the src ip:port in logging and expansions as if it
697 were a direct connection from the outside internet. PPv2 support was
698 updated based on HAProxy spec change in May 2014.
700 JH/02 Add ${listextract {number}{list}{success}{fail}}.
702 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
703 Properly escape header and check for NULL return.
705 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
708 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
710 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
711 characters in header names, implemented as a verify condition.
712 Contributed by Michael Fischer v. Mollard.
714 TL/05 Rename SPF condition results err_perm and err_temp to standardized
715 results permerror and temperror. Previous values are deprecated but
716 still accepted. In a future release, err_perm and err_temp will be
717 completely removed, which will be a backward incompatibility if the
718 ACL tests for either of these two old results. Patch contributed by
719 user bes-internal on the mailing list.
721 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
723 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
724 selectors, in both main and reject logs.
726 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
729 JH/07 Add malware type "sock" for talking to simple daemon.
731 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
733 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
734 routers/transports under cutthrough routing.
736 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
737 numbers. Touch up "bool" conditional to keep the same definition.
739 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
741 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
743 JH/12 Expand items in router/transport headers_add or headers_remove lists
744 individually rather than the list as a whole. Bug 1452.
746 Required for reasonable handling of multiple headers_ options when
747 they may be empty; requires that headers_remove items with embedded
748 colons must have them doubled (or the list-separator changed).
750 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
751 view the policy declared in the DMARC record. Currently, $dmarc_status
752 is a combined value of both the record presence and the result of the
755 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
757 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
758 dnslookup router and the smtp transport (applying to the forward
761 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
762 of ldap servers used for a specific lookup. Patch provided by Heiko
765 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
766 New variable $lookup_dnssec_authenticated for observability.
768 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
769 Patch submitted by Lars Timman.
771 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
773 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
774 Requires trusted mode and valid format message id, aborts otherwise.
775 Patch contributed by Heiko Schlichting.
777 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
778 certextract with support for various fields. Bug 1358.
780 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
781 is requested by default, modifiable by smtp transport option
784 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
785 operate on certificate variables to give certificate fingerprints
786 Also new ${sha256:cert_variable}.
788 JH/23 The PRDR feature is moved from being Experimental into the mainline.
790 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
791 Christian Aistleitner.
793 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
795 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
796 file. Patch from Wolfgang Breyha.
798 JH/25 Expand the coverage of the delivery $host and $host_address to
799 client authenticators run in verify callout. Bug 1476.
801 JH/26 Port service names are now accepted for tls_on_connect_ports, to
802 align with daemon_smtp_ports. Bug 72.
804 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
805 support and error reporting did not work properly.
807 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
808 and is readable. Patch from Andrew Colin Kissa.
810 TL/14 Enhance documentation of ${run expansion and how it parses the
811 commandline after expansion, particularly in the case when an
812 unquoted variable expansion results in an empty value.
814 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
816 PP/02 Fix internal collision of T_APL on systems which support RFC3123
817 by renaming away from it. Addresses GH issue 15, reported by
820 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
822 TL/15 SECURITY: prevent double expansion in math comparison functions
823 (can expand unsanitized data). Not remotely exploitable.
830 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
832 PP/02 Make -n do something, by making it not do something.
833 When combined with -bP, the name of an option is not output.
835 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
838 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
839 $sender_host_name and config options to manage this, and basic check
842 PP/05 DSCP support for outbound connections and control modifier for inbound.
844 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
845 (Only plugin which currently uses this is kerberos4, which nobody should
846 be using, but we should make it available and other future plugins might
847 conceivably use it, even though it would break NAT; stuff *should* be
848 using channel bindings instead).
850 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
851 name; added for Sendmail compatibility; requires admin caller.
852 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
853 just ignore it); requires trusted caller.
854 Also parse but ignore: -Ac -Am -X<logfile>
857 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
859 TL/02 Add +smtp_confirmation as a default logging option.
861 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
862 Patch by Magnus Holmgren from 2007-02-20.
864 TL/04 Bugzilla 1281 - Spec typo.
865 Bugzilla 1283 - Spec typo.
866 Bugzilla 1290 - Spec grammar fixes.
868 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
870 TL/06 Add Experimental DMARC support using libopendmarc libraries.
872 TL/07 Fix an out of order global option causing a segfault. Reported to dev
873 mailing list by by Dmitry Isaikin.
875 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
877 JH/02 Support "G" suffix to numbers in ${if comparisons.
879 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
881 NM/01 Bugzilla 1197 - Spec typo
882 Bugzilla 1196 - Spec examples corrections
884 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
886 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
887 gnutls_enable_pkcs11, but renamed to more accurately indicate its
890 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
891 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
893 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
894 "acl {{name}{arg}...}", and optional args on acl condition
897 JH/05 Permit multiple router/transport headers_add/remove lines.
899 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
901 JH/07 Avoid using a waiting database for a single-message-only transport.
902 Performance patch from Paul Fisher. Bugzilla 1262.
904 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
907 JH/09 Add $headers_added variable, with content from use of ACL modifier
908 add_header (but not yet added to the message). Bugzilla 199.
910 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
911 Pulled from Bugzilla 817 by Wolfgang Breyha.
913 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
915 (nb: this is the same fix as in Exim 4.80.1)
917 JH/11 Add A= logging on delivery lines, and a client_set_id option on
920 JH/12 Add optional authenticated_sender logging to A= and a log_selector
923 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
925 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
926 advertise SMTP AUTH mechanism to us, instead of a generic
927 protocol violation error. Also, make Exim more robust to bad
928 data from the Dovecot auth socket.
930 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
932 When a queue runner is handling a message, Exim first routes the
933 recipient addresses, during which it prunes them based on the retry
934 hints database. After that it attempts to deliver the message to
935 any remaining recipients. It then updates the hints database using
938 So if a recipient address works intermittently, it can get repeatedly
939 deferred at routing time. The retry hints record remains fresh so the
940 address never reaches the final cutoff time.
942 This is a fairly common occurrence when a user is bumping up against
943 their storage quota. Exim had some logic in its local delivery code
944 to deal with this. However it did not apply to per-recipient defers
945 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
947 This change adds a proper retry rule check during routing so that the
948 final cutoff time is checked against the message's age. We only do
949 this check if there is an address retry record and there is not a
950 domain retry record; this implies that previous attempts to handle
951 the address had the retry_use_local_parts option turned on. We use
952 this as an approximation for the destination being like a local
953 delivery, as in LMTP.
955 I suspect this new check makes the old local delivery cutoff check
956 redundant, but I have not verified this so I left the code in place.
958 TF/02 Correct gecos expansion when From: is a prefix of the username.
960 Test 0254 submits a message to Exim with the header
964 When I ran the test suite under the user fanf2, Exim expanded
965 the header to contain my full name, whereas it should have added
966 a Resent-Sender: header. It erroneously treats any prefix of the
967 username as equal to the username.
969 This change corrects that bug.
971 GF/01 DCC debug and logging tidyup
972 Error conditions log to paniclog rather than rejectlog.
973 Debug lines prefixed by "DCC: " to remove any ambiguity.
975 TF/03 Avoid unnecessary rebuilds of lookup-related code.
977 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
978 Bug spotted by Jeremy Harris; was flawed since initial commit.
979 Would have resulted in OCSP responses post-SNI triggering an Exim
980 NULL dereference and crash.
982 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
984 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
985 Bug detection, analysis and fix by Samuel Thibault.
986 Bugzilla 1331, Debian bug #698092.
988 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
990 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
991 Server implementation by Todd Lyons, client by JH.
992 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
993 config variable "prdr_enable" controls whether the server
994 advertises the facility. If the client requests PRDR a new
995 acl_data_smtp_prdr ACL is called once for each recipient, after
996 the body content is received and before the acl_smtp_data ACL.
997 The client is controlled by both of: a hosts_try_prdr option
998 on the smtp transport, and the server advertisement.
999 Default client logging of deliveries and rejections involving
1000 PRDR are flagged with the string "PRDR".
1002 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1003 fclose(). Diagnosis by Todd Lyons.
1005 PP/17 Update configure.default to handle IPv6 localhost better.
1006 Patch by Alain Williams (plus minor tweaks).
1009 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1010 This is now consistent with GnuTLS, and is now documented: the
1011 previous undocumented portable approach to treating the option as
1012 unset was to force an expansion failure. That still works, and
1013 an empty string is now equivalent.
1015 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1016 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1017 not performing validation itself.
1019 PP/20 Added force_command boolean option to pipe transport.
1020 Patch from Nick Koston, of cPanel Inc.
1022 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1025 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1027 PP/21 Fix eximon continuous updating with timestamped log-files.
1028 Broken in a format-string cleanup in 4.80, missed when I repaired the
1029 other false fix of the same issue.
1030 Report and fix from Heiko Schlichting.
1033 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1034 Report from Prashanth Katuri.
1036 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1037 It's SecureTransport, so affects any MacOS clients which use the
1038 system-integrated TLS libraries, including email clients.
1040 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1041 using a MIME ACL for non-SMTP local injection.
1042 Report and assistance in diagnosis by Warren Baker.
1044 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1046 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1048 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1049 last id that failed so it may be referenced in subsequent ACL's.
1051 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1054 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1055 ldap library initialization, allowing self-signed CA's to be
1056 used. Also properly sets require_cert option later in code by
1057 using NULL (global ldap config) instead of ldap handle (per
1058 session). Bug diagnosis and testing by alxgomz.
1060 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1061 the src/util/ subdirectory.
1063 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1064 renamed to Transport Post Delivery Action by Jeremy Harris, as
1067 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1068 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1069 redis_servers = needs to be configured which will be used by the redis
1070 lookup. Patch from Warren Baker, of The Packet Hub.
1072 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1074 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1075 hostname or reverse DNS when processing a host list. Used suggestions
1076 from multiple comments on this bug.
1078 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1080 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1081 Missed a few lines, added it to make the runtest require no keyboard
1084 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1085 contains upper case chars. Make router use caseful_local_part.
1087 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1088 support when GnuTLS has been built with p11-kit.
1094 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1096 This, or similar/improved, will also be change PP/11 of 4.82.
1102 PP/01 Handle short writes when writing local log-files.
1103 In practice, only affects FreeBSD (8 onwards).
1104 Bugzilla 1053, with thanks to Dmitry Isaikin.
1106 NM/01 Bugzilla 949 - Documentation tweak
1108 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1111 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1113 PP/02 Implemented gsasl authenticator.
1115 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1117 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1118 `pkg-config foo` for cflags/libs.
1120 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1121 with rest of GSASL and with heimdal_gssapi.
1123 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1124 `pkg-config foo` for cflags/libs for the TLS implementation.
1126 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1127 properties get this fed in as external SSF. A number of robustness
1128 and debugging improvements to the cyrus_sasl authenticator.
1130 PP/08 cyrus_sasl server now expands the server_realm option.
1132 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1133 Patch by Jeremy Harris.
1135 PP/10 Added dbmjz lookup type.
1137 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1139 PP/12 MAIL args handles TAB as well as SP, for better interop with
1140 non-compliant senders.
1141 Analysis and variant patch by Todd Lyons.
1143 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1144 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1145 Patch from Dirk Mueller <dmueller@suse.com>
1147 PP/13 tls_peerdn now print-escaped for spool files.
1148 Observed some $tls_peerdn in wild which contained \n, which resulted
1149 in spool file corruption.
1151 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1152 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1153 or write after TLS renegotiation, which otherwise led to messages
1156 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1157 as a tracking header (ie: a signed header comes before the signature).
1158 Patch from Wolfgang Breyha.
1160 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1161 comma-sep list; embedded commas doubled.
1163 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1165 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1167 Report and patch from Dmitry Banschikov.
1169 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1170 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1171 failures. We appear to now support TLS1.1+ with Exim.
1173 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1174 lets Exim select keys and certificates based upon TLS SNI from client.
1175 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1176 before an outbound SMTP session. New log_selector, +tls_sni.
1178 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1179 NULL dereference. Report and patch from Alun Jones.
1181 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1182 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1183 Not seeing resolver debug output on NetBSD, but suspect this is a
1184 resolver implementation change.
1186 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1187 Left warnings. Added "eximon gdb" invocation mode.
1189 PP/21 Defaulting "accept_8bitmime" to true, not false.
1191 PP/22 Added -bw for inetd wait mode support.
1193 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1194 locate the relevant includes and libraries. Made this the default.
1196 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1197 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1199 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1200 This may cause build issues on older platforms.
1202 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1203 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1204 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1205 Added SNI support via GnuTLS too.
1206 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1208 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1210 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1211 Applied second patch from Janne, implementing suggestion to default
1212 multiple-strings-in-record handling to match SPF spec.
1214 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1216 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1217 read-only, out of scope).
1218 Patch from Wolfgang Breyha, report from Stuart Northfield.
1220 PP/29 Fix three issues highlighted by clang analyser static analysis.
1221 Only crash-plausible issue would require the Cambridge-specific
1222 iplookup router and a misconfiguration.
1223 Report from Marcin Mirosław.
1225 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1227 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1228 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1229 As part of this, removing so much warning spew let me fix some minor
1230 real issues in debug logging.
1232 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1233 assignment on my part. Fixed.
1235 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1236 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1237 Janne Snabb (who went above and beyond: thank you).
1239 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1240 string otherwise requires a connection and a bunch more work and it's
1241 relatively easy to get wrong. Should also expose TLS library linkage
1244 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1245 64-bit ${eval} (JH/03).
1247 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1248 GNU libc to support some of the 64-bit stuff, should not lead to
1249 conflicts. Defined before os.h is pulled in, so if a given platform
1250 needs to override this, it can.
1252 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1253 protection layer was required, which is not implemented.
1254 Bugzilla 1254, patch from Wolfgang Breyha.
1256 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1257 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1258 tls_dhparam take prime identifiers. Also unbreak combination of
1259 OpenSSL+DH_params+TLSSNI.
1261 PP/39 Disable SSLv2 by default in OpenSSL support.
1267 PP/01 Solaris build fix for Oracle's LDAP libraries.
1268 Bugzilla 1109, patch from Stephen Usher.
1270 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1272 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1275 TF/02 Fix a couple more cases where we did not log the error message
1276 when unlink() failed. See also change 4.74-TF/03.
1278 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1279 lock up or crash if it happened to be inside a call to libc when it
1280 got a SIGUSR1 from exiwhat.
1282 The SIGUSR1 handler appends the current process status to the process
1283 log which is later printed by exiwhat. It used to use the general
1284 purpose logging code to do this, but several functions it calls are
1285 not safe for signals.
1287 The new output code in the SIGUSR1 handler is specific to the process
1288 log, and simple enough that it's easy to inspect for signal safety.
1289 Removing some special cases also simplifies the general logging code.
1290 Removing the spurious timestamps from the process log simplifies
1293 TF/04 Improved ratelimit ACL condition.
1295 The /noupdate option has been deprecated in favour of /readonly which
1296 has clearer semantics. The /leaky, /strict, and /readonly update modes
1297 are mutually exclusive. The update mode is no longer included in the
1298 database key; it just determines when the database is updated. (This
1299 means that when you upgrade Exim will forget old rate measurements.)
1301 Exim now checks that the per_* options are used with an update mode that
1302 makes sense for the current ACL. For example, when Exim is processing a
1303 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1304 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1305 must specify per_mail/readonly. If you omit the update mode it defaults to
1306 /leaky where that makes sense (as before) or /readonly where required.
1308 The /noupdate option is now undocumented but still supported for
1309 backwards compatibility. It is equivalent to /readonly except that in
1310 ACLs where /readonly is required you may specify /leaky/noupdate or
1311 /strict/noupdate which are treated the same as /readonly.
1313 A useful new feature is the /count= option. This is a generalization
1314 of the per_byte option, so that you can measure the throughput of other
1315 aggregate values. For example, the per_byte option is now equivalent
1316 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1318 The per_rcpt option has been generalized using the /count= mechanism
1319 (though it's more complicated than the per_byte equivalence). When it is
1320 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1321 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1322 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1323 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1324 Note that using per_rcpt with a non-readonly update mode in more than
1325 one ACL will cause the recipients to be double-counted. (The per_mail
1326 and per_byte options don't have this problem.)
1328 The handling of very low rates has changed slightly. If the computed rate
1329 is less than the event's count (usually one) then this event is the first
1330 after a long gap. In this case the rate is set to the same as this event's
1331 count, so that the first message of a spam run is counted properly.
1333 The major new feature is a mechanism for counting the rate of unique
1334 events. The new per_addr option counts the number of different
1335 recipients that someone has sent messages to in the last time period. It
1336 behaves like per_rcpt if all the recipient addresses are different, but
1337 duplicate recipient addresses do not increase the measured rate. Like
1338 the /count= option this is a general mechanism, so the per_addr option
1339 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1340 example, measure the rate that a client uses different sender addresses
1341 with the options per_mail/unique=$sender_address. There are further
1342 details in the main documentation.
1344 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1346 TF/06 Removed a few PCRE remnants.
1348 TF/07 Automatically extract Exim's version number from tags in the git
1349 repository when doing development or release builds.
1351 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1352 Bugzilla 879. Patch from Paul Fisher.
1354 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1355 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1358 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1360 PP/05 Variable $av_failed, true if the AV scanner deferred.
1361 Bugzilla 1078. Patch from John Horne.
1363 PP/06 Stop make process more reliably on build failure.
1364 Bugzilla 1087. Patch from Heiko Schlittermann.
1366 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1367 Bugzilla 1089. Patch from Heiko Schlittermann.
1369 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1370 Bugzilla 1131. Patch from Holger Weiß.
1372 PP/09 Handle IPv6 addresses with SPF.
1373 Bugzilla 860. Patch from Wolfgang Breyha.
1375 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1377 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1380 PP/11 match_* no longer expand right-hand-side by default.
1381 New compile-time build option, EXPAND_LISTMATCH_RHS.
1382 New expansion conditions, "inlist", "inlisti".
1384 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1386 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1388 PP/14 fix log_write() format string regression from TF/03.
1389 Bugzilla 1152. Patch from Dmitry Isaikin.
1395 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1397 PP/02 Harmonised TLS library version reporting; only show if debugging.
1398 Layout now matches that introduced for other libraries in 4.74 PP/03.
1400 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1402 PP/04 New "dns_use_edns0" global option.
1404 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1407 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1408 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1410 TK/01 Updated PolarSSL code to 0.14.2.
1411 Bugzilla 1097. Patch from Andreas Metzler.
1413 PP/07 Catch divide-by-zero in ${eval:...}.
1414 Fixes bugzilla 1102.
1416 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1419 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1420 format-string attack -- SECURITY: remote arbitrary code execution.
1422 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1423 time unintentionally subject to list matching rules, letting the header
1424 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1425 arbitrary string expansion). This allowed for information disclosure.
1427 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1428 INT_MIN/-1 -- value coerced to INT_MAX.
1434 NM/01 Workaround for PCRE version dependency in version reporting
1437 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1438 This fixes portability to compilers other than gcc, notably
1439 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1441 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1442 makefiles for portability to HP-UX and POSIX correctness.
1444 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1445 Also via indented variable definition in the Makefile.
1446 (Debugging by Oliver Heesakkers).
1448 PP/02 Restore caching of spamd results with expanded spamd_address.
1449 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1451 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1452 Improves build reliability. Fix from: Frank Elsner
1454 NM/02 Fix wide character breakage in the rfc2047 coding
1455 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1457 NM/03 Allow underscore in dnslist lookups
1458 Fixes bug 1026. Patch from Graeme Fowler
1460 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1461 Code patches from Adam Ciarcinski of NetBSD.
1463 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1466 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1467 is logged, to avoid truncation. Patch from John Horne.
1469 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1470 Patch from Jakob Hirsch.
1472 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1473 SQL string expansion failure details.
1474 Patch from Andrey Oktyabrski.
1476 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1477 Patch from Simon Arlott.
1479 PP/09 New lookups functionality failed to compile on old gcc which rejects
1480 extern declarations in function scope.
1481 Patch from Oliver Fleischmann
1483 PP/10 Use sig_atomic_t for flags set from signal handlers.
1484 Check getgroups() return and improve debugging.
1485 Fixed developed for diagnosis in bug 927 (which turned out to be
1488 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1489 Patch from Mark Zealey.
1491 PP/12 Bugzilla 1056: Improved spamd server selection.
1492 Patch from Mark Zealey.
1494 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1495 Based on patch from Heiko Schlittermann.
1497 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1498 Patch from Uwe Doering, sign-off by Michael Haardt.
1500 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1501 variable declaration deep within a block. Bug and patch from
1504 PP/15 lookups-Makefile IRIX compatibility coercion.
1506 PP/16 Make DISABLE_DKIM build knob functional.
1508 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1509 Patch by Simon Arlott
1511 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1512 variable argument macros. Our copy now differs from upstream.
1518 TF/01 Failure to get a lock on a hints database can have serious
1519 consequences so log it to the panic log.
1521 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1522 controlled using the smtp_confirmation log selector.
1524 TF/03 Include the error message when we fail to unlink a spool file.
1526 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1527 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1528 for maintaining out-of-tree patches for some time.
1530 PP/01 Bugzilla 139: Documentation and portability issues.
1531 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1532 Handle per-OS dynamic-module compilation flags.
1534 PP/02 Let /dev/null have normal permissions.
1535 The 4.73 fixes were a little too stringent and complained about the
1536 permissions on /dev/null. Exempt it from some checks.
1537 Reported by Andreas M. Kirchwitz.
1539 PP/03 Report version information for many libraries, including
1540 Exim version information for dynamically loaded libraries. Created
1541 version.h, now support a version extension string for distributors
1542 who patch heavily. Dynamic module ABI change.
1544 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1545 privilege escalation vulnerability whereby the Exim run-time user
1546 can cause root to append content of the attacker's choosing to
1549 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1552 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1553 If dropping privileges for untrusted macros, we disabled normal logging
1554 on the basis that it would fail; for the Exim run-time user, this is not
1555 the case, and it resulted in successful deliveries going unlogged.
1556 Fixed. Reported by Andreas Metzler.
1562 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1563 only prepend for the Resent-* case. Fixes regression introduced in
1564 Exim 4.70 by NM/22 for Bugzilla 607.
1566 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1567 increasing numbers of administrators be bitten by this.
1569 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1571 PP/03 Bugzilla 994: added openssl_options main configuration option.
1573 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1575 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1577 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1579 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1580 without a peer certificate, leading to a segfault because of an
1581 assumption that peers always have certificates. Be a little more
1582 paranoid. Problem reported by Martin Tscholak.
1584 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1585 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1586 NB: ClamAV planning to remove STREAM in "middle of 2010".
1587 CL also introduces -bmalware, various -d+acl logging additions and
1588 more caution in buffer sizes.
1590 PP/09 Implemented reverse_ip expansion operator.
1592 PP/10 Bugzilla 937: provide a "debug" ACL control.
1594 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1596 PP/12 Bugzilla 973: Implement --version.
1598 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1600 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1602 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1604 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1605 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1606 ignore trailing whitespace.
1608 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1610 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1613 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1614 Notification from Dr Andrew Aitchison.
1616 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1617 ExtendedDetectionInfo response format.
1618 Notification from John Horne.
1620 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1623 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1624 XSL and documented dependency on system catalogs, with examples of how
1627 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1630 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1631 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1632 configuration file which is writeable by the Exim user or group.
1634 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1635 of configuration files to cover files specified with the -C option if
1636 they are going to be used with root privileges, not just the default
1639 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1640 option (effectively making it always true).
1642 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1643 files to be used while preserving root privileges.
1645 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1646 that rogue child processes cannot use them.
1648 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1649 run-time user, instead of root.
1651 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1652 Exim run-time user without dropping privileges.
1654 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1655 result string, instead of calling string_vformat() twice with the same
1658 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1659 for other users. Others should always drop root privileges if they use
1660 -C on the command line, even for a whitelisted configure file.
1662 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1664 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1670 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1671 $data_path, and $header_path variables; fixed documentation bugs and
1674 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1675 exipick to access non-standard spools, including the "frozen" queue
1678 NM/01 Bugzilla 965: Support mysql stored procedures.
1679 Patch from Alain Williams
1681 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1683 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1684 Patch from Andreas Metzler
1686 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1687 Patch from Kirill Miazine
1689 NM/05 Bugzilla 671: Added umask to procmail example.
1691 JJ/03 installed exipick 20100323.0, fixing doc bug
1693 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1694 directory. Notification and patch from Dan Rosenberg.
1696 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1698 TK/02 Improve log output when DKIM signing operation fails.
1700 MH/01 Treat the transport option dkim_domain as a colon separated
1701 list, not as a single string, and sign the message with each element,
1702 omitting multiple occurences of the same signer.
1704 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1705 Bugzilla 985, 986. Patch by Simon Arlott
1707 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1708 Patch by Simon Arlott
1710 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1711 MBX locking. Notification from Dan Rosenberg.
1717 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1719 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1721 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1723 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1725 NM/04 Bugzilla 915: Flag broken perl installation during build.
1731 TK/01 Added patch by Johannes Berg that expands the main option
1732 "spamd_address" if it starts with a dollar sign.
1734 TK/02 Write list of recipients to X-Envelope-Sender header when building
1735 the mbox-format spool file for content scanning (suggested by Jakob
1738 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1739 (http://www.dcc-servers.net/) support via dccifd. Activated by
1740 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1742 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1743 by Mark Daniel Reidel <mr@df.eu>.
1745 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1746 When building exim an external PCRE library is now needed -
1747 PCRE is a system library on the majority of modern systems.
1748 See entry on PCRE_LIBS in EDITME file.
1750 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1751 conversation. Added nologin parameter to request.
1752 Patch contributed by Kirill Miazine.
1754 TF/01 Do not log submission mode rewrites if they do not change the address.
1756 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1758 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1759 log files in place. Contributed by Roberto Lima.
1761 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1763 TF/03 Bugzilla 615: When checking the local_parts router precondition
1764 after a local_part_suffix or local_part_prefix option, Exim now
1765 does not use the address's named list lookup cache, since this
1766 contains cached lookups for the whole local part.
1768 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1769 Robert Millan. Documentation is in experimental-spec.txt.
1771 TF/04 Bugzilla 668: Fix parallel build (make -j).
1773 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1775 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1776 Patch provided by Jan Srzednicki.
1778 TF/05 Leading white space used to be stripped from $spam_report which
1779 wrecked the formatting. Now it is preserved.
1781 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1782 that they are available at delivery time.
1784 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1786 TF/08 TLS error reporting now respects the incoming_interface and
1787 incoming_port log selectors.
1789 TF/09 Produce a more useful error message if an SMTP transport's hosts
1790 setting expands to an empty string.
1792 NM/06 Bugzilla 744: EXPN did not work under TLS.
1793 Patch provided by Phil Pennock.
1795 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1796 Patch provided by Richard Godbee.
1798 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1799 acl_smtp_notquit, added index entry.
1801 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1802 Patch provided by Eugene Bujak.
1804 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1805 accept(). Patch provided by Maxim Dounin.
1807 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1808 Patch provided by Phil Pennock.
1810 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1812 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1813 Patch provided by Brad "anomie" Jorsch.
1815 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1816 Patch provided by Dean Brooks.
1818 TK/05 Add native DKIM support (does not depend on external libraries).
1820 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1821 Patch provided by Graeme Fowler.
1823 NM/16 Bugzilla 851: Documentation example syntax fix.
1825 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1827 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1830 NM/19 Bugzilla 745: TLS version reporting.
1831 Patch provided by Phil Pennock.
1833 NM/20 Bugzilla 167: bool: condition support.
1834 Patch provided by Phil Pennock.
1836 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1837 clients. Patch provided by Phil Pennock.
1839 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1840 Patch provided by Brad "anomie" Jorsch.
1842 NM/23 Bugzilla 687: Fix misparses in eximstats.
1843 Patch provided by Heiko Schlittermann.
1845 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1846 Patch provided by Heiko Schlittermann.
1848 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1849 plus update to original patch.
1851 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1853 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1854 Patch provided by David Brownlee.
1856 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1858 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1860 NM/30 Bugzilla 888: TLS documentation bugfixes.
1862 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1864 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1865 Unlike the original bugzilla I have changed all shell scripts in src tree.
1867 NM/33 Bugzilla 898: Transport filter timeout fix.
1868 Patch by Todd Rinaldo.
1870 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1871 Patch by Serge Demonchaux.
1873 NM/35 Bugzilla 39: Base64 decode bug fixes.
1874 Patch by Jakob Hirsch.
1876 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1878 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1880 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1882 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1888 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1889 ALT-N's libdkim that I have put here:
1890 http://duncanthrax.net/exim-experimental/
1892 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1893 were called 'true' and it seems that C99 defines that as a reserved
1894 keyword to be used with 'bool' variable types. That means you could
1895 not include C99-style headers which use bools without triggering
1896 build errors in sieve.c.
1898 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1899 as mailq or other aliases. Changed the --help handling significantly
1900 to do whats expected. exim_usage() emits usage/help information.
1902 SC/01 Added the -bylocaldomain option to eximstats.
1904 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1906 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1908 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1914 PH/01 Another patch from the Sieve maintainer.
1916 PH/02 When an IPv6 address is converted to a string for single-key lookup
1917 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1918 dots are used instead of colons so that keys in lsearch files need not
1919 contain colons. This was done some time before quoting was made available
1920 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1921 (notated using the quote facility) so as to distinguish them from IPv4
1922 keys. This meant that lookups for IP addresses in host lists did not work
1923 for iplsearch lookups.
1925 This has been fixed by arranging for IPv6 addresses to be expressed with
1926 colons if the lookup type is iplsearch. This is not incompatible, because
1927 previously such lookups could never work.
1929 The situation is now rather anomalous, since one *can* have colons in
1930 ordinary lsearch keys. However, making the change in all cases is
1931 incompatible and would probably break a number of configurations.
1933 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1936 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1937 conversion specification without a maximum field width, thereby enabling
1938 a rogue spamd server to cause a buffer overflow. While nobody in their
1939 right mind would setup Exim to query an untrusted spamd server, an
1940 attacker that gains access to a server running spamd could potentially
1941 exploit this vulnerability to run arbitrary code as the Exim user.
1943 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1944 $primary_hostname instead of what libspf2 thinks the hosts name is.
1946 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1947 a directory entry by the name of the lookup key. Previously, if a
1948 symlink pointed to a non-existing file or a file in a directory that
1949 Exim lacked permissions to read, a lookup for a key matching that
1950 symlink would fail. Now it is enough that a matching directory entry
1951 exists, symlink or not. (Bugzilla 503.)
1953 PH/03 The body_linecount and body_zerocount variables are now exported in the
1956 PH/04 Added the $dnslist_matched variable.
1958 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1959 This means they are set thereafter only if the connection becomes
1962 PH/06 Added the client_condition to authenticators so that some can be skipped
1963 by clients under certain conditions.
1965 PH/07 The error message for a badly-placed control=no_multiline_responses left
1966 "_responses" off the end of the name.
1968 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1970 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1971 (without spaces) instead of just copying the configuration text.
1973 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1975 PH/11 Added $max_received_linelength.
1977 PH/12 Added +ignore_defer and +include_defer to host lists.
1979 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1980 way in which PCRE > 7.0 is built.
1982 PH/14 Implemented queue_only_load_latch.
1984 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1985 MAIL command. The effect was to mangle the value on 64-bit systems.
1987 PH/16 Another patch from the Sieve maintainer.
1989 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1991 PH/18 If a system quota error occurred while trying to create the file for
1992 a maildir delivery, the message "Mailbox is full" was not appended to the
1993 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1994 only to a quota excession during the actual writing of the file.
1996 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1997 characters?) which causes problems in log lines. The DN values are now
1998 passed through string_printing() before being added to log lines.
2000 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2001 and InterBase are left for another time.)
2003 PH/21 Added message_body_newlines option.
2005 PH/22 Guard against possible overflow in moan_check_errorcopy().
2007 PH/23 POSIX allows open() to be a macro; guard against that.
2009 PH/24 If the recipient of an error message contained an @ in the local part
2010 (suitably quoted, of course), incorrect values were put in $domain and
2011 $local_part during the evaluation of errors_copy.
2017 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2018 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2021 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2022 issue a MAIL command.
2024 PH/02 In an ACL statement such as
2026 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2028 if a client was not listed at all, or was listed with a value other than
2029 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2030 the condition was not true (as it should be), so access was not denied.
2031 The bug was that the ! inversion was incorrectly passed on to the second
2032 item. This has been fixed.
2034 PH/03 Added additional dnslists conditions == and =& which are different from
2035 = and & when the dns lookup returns more than one IP address.
2037 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2038 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2040 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2041 FSYNC, which compiles an option called disable_fsync that allows for
2042 bypassing fsync(). The documentation is heavily laced with warnings.
2044 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2046 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2047 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2048 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2049 including adding "make clean"; (3) Added -fPIC when compiling the test
2050 dynamically loaded module, to get rid of a warning.
2052 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2053 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2054 = 0s. The bug is otherwise harmless.
2056 PH/07 There was a bug in the dovecot authenticator such that the value of
2057 $auth1 could be overwritten, and so not correctly preserved, after a
2058 successful authentication. This usually meant that the value preserved by
2059 the server_setid option was incorrect.
2061 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2063 PH/09 Installed PCRE release 7.0.
2065 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2066 run for batched SMTP input. It is now run at the start of every message
2067 in the batch. While fixing this I discovered that the process information
2068 (output by running exiwhat) was not always getting set for -bs and -bS
2069 input. This is fixed, and it now also says "batched" for BSMTP.
2071 PH/11 Added control=no_pipelining.
2073 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2074 patch, slightly modified), and move the expansion of helo_data till after
2075 the connection is made in the smtp transport (so it can use these
2078 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2080 PH/14 Added log_selector = +pid.
2082 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2084 PH/16 Add ${if forany and ${if forall.
2086 PH/17 Added dsn_from option to vary the From: line in DSNs.
2088 PH/18 Flush SMTP output before performing a callout, unless control =
2089 no_callout_flush is set.
2091 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2092 was true (the default) a successful delivery failed to delete the retry
2093 item, thus causing premature timeout of the address. The bug is now
2096 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2098 PH/21 Long custom messages for fakedefer and fakereject are now split up
2099 into multiline responses in the same way that messages for "deny" and
2100 other ACL rejections are.
2102 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2103 with slight modification.
2105 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2106 draft, changing the syntax and factoring some duplicate code".
2108 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2109 for deliveries of the second and subsequent messages over the same SMTP
2112 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2113 ${reduce, with only minor "tidies".
2115 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2117 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2118 expansion side effects.
2120 PH/27 When a message times out after an over-quota error from an Exim-imposed
2121 quota, the bounce message says "mailbox is full". This message was not
2122 being given when it was a system quota that was exceeded. It now should
2125 MH/03 Made $recipients available in local_scan(). local_scan() already has
2126 better access to the recipient list through recipients_list[], but
2127 $recipients can be useful in postmaster-provided expansion strings.
2129 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2130 in the case of a MAIL command with additional options following the
2131 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2132 were accidentally chopped off.
2134 PH/29 SMTP synchronization checks are implemented when a command is read -
2135 there is a check that no more input is waiting when there shouldn't be
2136 any. However, for some commands, a delay in an ACL can mean that it is
2137 some time before the response is written. In this time, more input might
2138 arrive, invalidly. So now there are extra checks after an ACL has run for
2139 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2140 pipelining has not been advertised.
2142 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2144 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2145 returned in the SMTP response when smtp_return_error_details was false.
2146 This has been fixed.
2148 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2149 instead of the C I/O that was originally supplied, because problems were
2150 reported on Solaris.
2152 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2153 Exim which did not show up earlier: it was assuming that a call to
2154 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2155 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2156 was a macro that became an assignment, so it seemed to work. This has
2157 changed to a proper function call with a void return, hence the compile
2158 error. Exim's code has been fixed.
2160 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2163 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2165 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2167 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2168 because it turns out to be too big or there is a timeout) there is no
2169 "Completed" line in the log. When some messages of this type were
2170 selected by exigrep, they were listed as "not completed". Others were
2171 picked up by some special patterns. I have improved the selection
2172 criteria to be more general.
2174 PH/38 The host_find_failed option in the manualroute router can now be set
2175 to "ignore", to completely ignore a host whose IP address cannot be
2176 found. If all hosts are ignored, the behaviour is controlled by the new
2177 host_all_ignored option.
2179 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2180 homing or because of multiple MX records with /mx) generated more than
2181 one IP address, and the following item turned out to be the local host,
2182 all the secondary addresses of the first item were incorrectly removed
2183 from the list, along with the local host and any following hosts (which
2184 is what is supposed to happen).
2186 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2187 whoever called Exim into the -H file. In the case of the daemon it was
2188 behaving confusingly. When first started, it used values for whoever
2189 started the daemon, but after a SIGHUP it used the Exim user (because it
2190 calls itself on a restart). I have changed the code so that it now always
2193 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2194 message are rejected with the same error (e.g. no authentication or bad
2195 sender address), and a DATA command is nevertheless sent (as can happen
2196 with PIPELINING or a stupid MUA), the error message that was given to the
2197 RCPT commands is included in the rejection of the DATA command. This is
2198 intended to be helpful for MUAs that show only the final error to their
2201 PH/42 Another patch from the Sieve maintainer.
2203 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2204 Eximstats - Fixed some broken HTML links and added missing column headers
2206 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2207 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2209 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2211 PH/43 Yet another patch from the Sieve maintainer.
2213 PH/44 I found a way to check for a TCP/IP connection going away before sending
2214 the response to the final '.' that terminates a message, but only in the
2215 case where the client has not sent further data following the '.'
2216 (unfortunately, this is allowed). However, in many cases there won't be
2217 any further data because there won't be any more messages to send. A call
2218 to select() can be used: if it shows that the input is "ready", there is
2219 either input waiting, or the socket has been closed. An attempt to read
2220 the next input character can distinguish the two cases. Previously, Exim
2221 would have sent an OK response which the client would never have see.
2222 This could lead to message repetition. This fix should cure that, at
2223 least in a lot of common cases.
2225 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2226 advertised in response to EHLO.
2232 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2233 fixed by 4.65/MH/01 (is this a record?) are fixed:
2235 (i) An empty string was always treated as zero by the numeric comparison
2236 operators. This behaviour has been restored.
2238 (ii) It is documented that the numeric comparison operators always treat
2239 their arguments as decimal numbers. This was broken in that numbers
2240 starting with 0 were being interpreted as octal.
2242 While fixing these problems I realized that there was another issue that
2243 hadn't been noticed. Values of message_size_limit (both the global option
2244 and the transport option) were treated as octal if they started with 0.
2245 The documentation was vague. These values are now always treated as
2246 decimal, and I will make that clear in the documentation.
2252 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2253 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2256 MH/01 Don't check that the operands of numeric comparison operators are
2257 integers when their expansion is in "skipping" mode (fixes bug
2258 introduced by 4.64-PH/07).
2260 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2261 child addresses, Exim now panics and dies. Previously, because the count
2262 is held in a short int, deliveries were likely to be lost. As such a
2263 large number of recipients for a single message is ridiculous
2264 (performance will be very, very poor), I have chosen to impose a limit
2265 rather than extend the field.
2271 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2272 leftover -K file (the existence of which was triggered by #402).
2273 While we were at it, introduced process PID as part of the -K
2274 filename. This should rule out race conditions when creating
2277 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2278 processing considerably. Previous code took too long for large mails,
2279 triggering a timeout which in turn triggers #401.
2281 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2282 in the DK code in transports.c. sendfile() is not really portable,
2283 hence the _LINUX specificness.
2285 TF/01 In the add_headers option to the mail command in an Exim filter,
2286 there was a bug that Exim would claim a syntax error in any
2287 header after the first one which had an odd number of characters
2290 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2291 callout verification, Exim cached a "reject" for the entire domain. This
2292 is correct for most verifications, but it is not correct for a recipient
2293 verification with use_sender or use_postmaster set, because in that case
2294 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2295 case of MAIL FROM:<> rejection from other early rejections (e.g.
2296 rejection of HELO). When verifying a recipient using a non-null MAIL
2297 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2298 Whatever the result of the callout, the value of the domain cache is
2299 left unchanged (for any other kind of callout, getting as far as trying
2300 RCPT means that the domain itself is ok).
2302 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2305 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2306 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2307 session, the daemon ignores ECONNECT errors and logs others; it now
2308 ignores EPIPE as well.
2310 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2311 (quoted-printable decoding).
2313 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2314 later the small subsequent patch to fix an introduced bug.
2316 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2318 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2320 PH/08 An error is now given if message_size_limit is specified negative.
2322 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2323 to be given (somewhat) arbitrary names.
2325 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2328 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2329 miscellaneous code fixes
2331 PH/10 Added the log_reject_target ACL modifier to specify where to log
2334 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2335 hostname. This is wrong, because it relates to the incoming message (and
2336 probably the interface on which it is arriving) and not to the outgoing
2337 callout (which could be using a different interface). This has been
2338 changed to use the value of the helo_data option from the smtp transport
2339 instead - this is what is used when a message is actually being sent. If
2340 there is no remote transport (possible with a router that sets up host
2341 addresses), $smtp_active_hostname is used.
2343 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2344 tweaks were necessary in order to get it to work (see also 21 below):
2345 (a) The code assumed that strncpy() returns a negative number on buffer
2346 overflow, which isn't the case. Replaced with Exim's string_format()
2348 (b) There were several signed/unsigned issues. I just did the minimum
2349 hacking in of casts. There is scope for a larger refactoring.
2350 (c) The code used strcasecmp() which is not a standard C function.
2351 Replaced with Exim's strcmpic() function.
2352 (d) The code set only $1; it now sets $auth1 as well.
2353 (e) A simple test gave the error "authentication client didn't specify
2354 service in request". It would seem that Dovecot has changed its
2355 interface. Fortunately there's a specification; I followed it and
2356 changed what the client sends and it appears to be working now.
2358 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2361 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2362 address A is aliased to B and C, where B exists and C does not. Without
2363 -v the output is "A verified" because verification stops after a
2364 successful redirection if more than one address is generated. However,
2365 with -v the child addresses are also verified. Exim was outputting "A
2366 failed to verify" and then showing the successful verification for C,
2367 with its parentage. It now outputs "B failed to verify", showing B's
2368 parentage before showing the successful verification of C.
2370 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2371 look up a TXT record in a specific list after matching in a combined
2374 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2375 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2376 they consult the DNS. I had assumed they would set it the way they
2377 wanted; and indeed my experiments on Linux seem to show that in some
2378 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2379 To be on the safe side, however, I have now made the interface to
2380 host_find_byname() similar to host_find_bydns(), with an argument
2381 containing the DNS resolver options. The host_find_byname() function now
2382 sets these options at its start, just as host_find_bydns() does. The smtp
2383 transport options dns_qualify_single and dns_search_parents are passed to
2384 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2385 of host_find_byname() use the default settings of RES_DEFNAMES
2386 (qualify_single) but not RES_DNSRCH (search_parents).
2388 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2389 spool_read_header() do less string testing, by means of a preliminary
2390 switch on the second character of optional "-foo" lines. (This is
2391 overdue, caused by the large number of possibilities that now exist.
2392 Originally there were few.) While I was there, I also converted the
2393 str(n)cmp tests so they don't re-test the leading "-" and the first
2394 character, in the hope this might squeeze out yet more improvement.
2396 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2397 flag allowing group syntax was set by the header_syntax check but not
2398 turned off, possible causing trouble later; (2) The flag was not being
2399 set at all for the header_verify test, causing "group"-style headers to
2400 be rejected. I have now set it in this case, and also caused header_
2401 verify to ignore an empty address taken from a group. While doing this, I
2402 came across some other cases where the code for allowing group syntax
2403 while scanning a header line wasn't quite right (mostly, not resetting
2404 the flag correctly in the right place). These bugs could have caused
2405 trouble for malformed header lines. I hope it is now all correct.
2407 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2408 with the "reply" argument non-NULL. The code, however (which originally
2409 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2410 but it didn't always do it. This confused somebody who was copying the
2411 code for some other use. I have removed all the tests.
2413 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2414 feature that was used to support insecure browsers during the U.S. crypto
2415 embargo. It requires special client support, and Exim is probably the
2416 only MTA that supported it -- and would never use it because real RSA is
2417 always available. This code has been removed, because it had the bad
2418 effect of slowing Exim down by computing (never used) parameters for the
2419 RSA_EXPORT functionality.
2421 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2422 authenticator to fail if there's a tab character in the incoming data
2423 (there should never be unless someone is messing about, as it's supposed
2424 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2425 if the connection is using TLS or if the remote IP is the same as the
2426 local IP, and the "valid-client-cert option" if a client certificate has
2429 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2430 authenticators. This can be used for authorization after authentication
2431 succeeds. (In the case of plaintext, it servers for both authentication
2434 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2435 if any retry times were supplied.
2437 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2438 connection, where there is no client IP address to check. In this
2439 situation, the verify now always succeeds.
2441 PH/25 Applied John Jetmore's -Mset patch.
2443 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2445 PH/27 In a string expansion for a processed (not raw) header when multiple
2446 headers of the same name were present, leading whitespace was being
2447 removed from all of them, but trailing whitespace was being removed only
2448 from the last one. Now trailing whitespace is removed from each header
2449 before concatenation. Completely empty headers in a concatenation (as
2450 before) are ignored.
2452 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2453 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2455 PH/29 [Removed. This was a change that I later backed out, and forgot to
2456 correct the ChangeLog entry (that I had efficiently created) before
2457 committing the later change.]
2459 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2460 address errors (4xx response to RCPT) over the same connection as other
2461 messages routed to the same hosts. Such deliveries are always "forced",
2462 so retry times are not inspected. This resulted in far too many retries
2463 for the affected addresses. The effect occurred only when there were more
2464 hosts than the hosts_max_try setting in the smtp transport when it had
2465 the 4xx errors. Those hosts that it had tried were not added to the list
2466 of hosts for which the message was waiting, so if all were tried, there
2467 was no problem. Two fixes have been applied:
2469 (i) If there are any address or message errors in an SMTP delivery, none
2470 of the hosts (tried or untried) are now added to the list of hosts
2471 for which the message is waiting, so the message should not be a
2472 candidate for sending over the same connection that was used for a
2473 successful delivery of some other message. This seems entirely
2474 reasonable: after all the message is NOT "waiting for some host".
2475 This is so "obvious" that I'm not sure why it wasn't done
2476 previously. Hope I haven't missed anything, but it can't do any
2477 harm, as the worst effect is to miss an optimization.
2479 (ii) If, despite (i), such a delivery is accidentally attempted, the
2480 routing retry time is respected, so at least it doesn't keep
2481 hammering the server.
2483 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2484 in ${readsocket because some servers need this prod.
2486 PH/32 Added some extra debug output when updating a wait-xxx database.
2488 PH/33 The hint "could be header name not terminated by colon", which has been
2489 given for certain expansion errors for a long time, was not being given
2490 for the ${if def:h_colon_omitted{... case.
2492 PH/34 The spec says: "With one important exception, whenever a domain list is
2493 being scanned, $domain contains the subject domain." There was at least
2494 one case where this was not true.
2496 PH/35 The error "getsockname() failed: connection reset by peer" was being
2497 written to the panic log as well as the main log, but it isn't really
2498 panic-worthy as it just means the connection died rather early on. I have
2499 removed the panic log writing for the ECONNRESET error when getsockname()
2502 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2503 runs only) independently of the message's sender address. This meant
2504 that, if the 4xx error was in fact related to the sender, a different
2505 message to the same recipient with a different sender could confuse
2506 things. In particular, this can happen when sending to a greylisting
2507 server, but other circumstances could also provoke similar problems.
2508 I have changed the default so that the retry time for these errors is now
2509 based a combination of the sender and recipient addresses. This change
2510 can be overridden by setting address_retry_include_sender=false in the
2513 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2514 remote server are returned as part of bounce messages. This was not
2515 happening for LMTP over a pipe (the lmtp transport), but now it is the
2516 same for both kinds of LMTP.
2518 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2519 in header lines that were in fact CNAMEs. This is no longer the case.
2521 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2522 and queue runs started by the daemon processed all messages. This has
2523 been fixed so that -R and -S can now usefully be given with -q<time>.
2525 PH/40 Import PCRE release 6.7 (fixes some bugs).
2527 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2529 PH/42 Give an error if -q is specified more than once.
2531 PH/43 Renamed the variables $interface_address and $interface_port as
2532 $received_ip_address and $received_port, to make it clear that these
2533 values apply to message reception, and not to the outgoing interface when
2534 a message is delivered. (The old names remain recognized, of course.)
2536 PH/44 There was no timeout on the connect() call when using a Unix domain
2537 socket in the ${readsocket expansion. There now is.
2539 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2540 be meaningful with "accept".
2542 SC/01 Eximstats V1.43
2543 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2545 SC/02 Eximstats V1.44
2546 Use a glob alias rather than an array ref in the generated
2547 parser. This improves both readability and performance.
2549 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2550 Collect SpamAssassin and rejection statistics.
2551 Don't display local sender or destination tables unless
2552 there is data to show.
2553 Added average volumes into the top table text output.
2555 SC/04 Eximstats V1.46
2556 Collect data on the number of addresses (recipients)
2557 as well as the number of messages.
2559 SC/05 Eximstats V1.47
2560 Added 'Message too big' to the list of mail rejection
2561 reasons (thanks to Marco Gaiarin).
2563 SC/06 Eximstats V1.48
2564 Mainlog lines which have GMT offsets and are too short to
2565 have a flag are now skipped.
2567 SC/07 Eximstats V1.49 (Alain Williams)
2568 Added the -emptyok flag.
2570 SC/08 Eximstats V1.50
2571 Fixes for obtaining the IP address from reject messages.
2573 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2574 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2575 whitespace changes from 4.64-PH/27
2577 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2580 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2581 are found, allow negative numbers in numeric criteria)
2583 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2585 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2586 to match changes made in 4.64-PH/43
2588 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2590 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2591 "domain" is actually an IP address, and force a failure. This locks out
2592 those revolvers/nameservers that support "A-for-A" lookups, in
2593 contravention of the specifications.
2595 PH/48 When a host name was looked up from an IP address, and the subsequent
2596 forward lookup of the name timed out, the host name was left in
2597 $sender_host_name, contrary to the specification.
2599 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2600 restricted to single-key lookups, Exim was not diagnosing an error if
2601 * or *@ was used with a query-style lookup.
2603 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2605 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2606 long ago, but noone interested enough thought of it. Let's just say that
2607 the "1.1" means that there are some new functions that weren't there at
2608 some point in the past.
2610 PH/51 Error processing for expansion failure of helo_data from an smtp
2611 transport during callout processing was broken.
2613 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2614 tested/used via the -bh/-bhc/-bs options.
2616 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2617 bug, fixed in subsequent PCRE releases).
2619 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2620 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2622 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2628 SC/01 Use a glob alias rather than an array ref in eximstats generated
2629 parser. This improves both readability and performance.
2631 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2632 Don't display local sender or destination tables in eximstats unless
2633 there is data to show.
2634 Added average volumes into the eximstats top table text output.
2636 SC/03 Collect data on the number of addresses (recipients) as well
2637 as the number of messages in eximstats.
2639 TF/01 Correct an error in the documentation for the redirect router. Exim
2640 does (usually) call initgroups() when daemonizing.
2642 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2643 with consistent privilege compared to when running as a daemon.
2645 TF/03 Note in the spec that $authenticated_id is not set for local
2646 submissions from trusted users.
2648 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2649 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2651 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2652 by adding some example configuration directives to the default
2653 configuration file. A little bit of work is required to uncomment the
2654 directives and define how usernames and passwords are checked, but
2655 there is now a framework to start from.
2657 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2658 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2659 without this. I don't know how relevant this is to other LDAP libraries.
2661 PH/02 Add the verb name to the "unknown ACL verb" error.
2663 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2665 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2667 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2668 directory not expanded when it should be if an expanded home directory
2669 was set for the address (which is overridden by the transport).
2671 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2674 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2675 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2676 because it is too late at that time, and has no effect.
2678 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2679 security issue with \' (bugzilla #107). I could not use the
2680 PQescapeStringConn() function, because it needs a PGconn value as one of
2683 PH/08 When testing addresses using -bt, indicate those final addresses that
2684 are duplicates that would not cause an additional delivery. At least one
2685 person was confused, thinking that -bt output corresponded to deliveries.
2686 (Suppressing duplicates isn't a good idea as you lose the information
2687 about possibly different redirections that led to the duplicates.)
2689 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2690 systems where poll() doesn't work, in particular OS X.
2692 PH/10 Added more information to debugging output for retry time not reached.
2694 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2695 operations in malware.c.
2697 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2700 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2701 syslog_duplication set false, log lines that would normally be written
2702 both the the main log and to the reject log were not written to syslog at
2705 PH/14 In the default configuration, change the use of "message" in ACL warn
2706 statements to "add_header".
2708 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2709 not followed by a command (e.g. "seen endif").
2711 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2712 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2715 PH/17 Added extra conditions to the default value of delay_warning_condition
2719 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2720 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2721 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2724 The Auto-Submitted: and various List- headers are standardised, whereas I
2725 don't think Precedence: ever was.
2727 PH/18 Refactored debugging code in route_finduser() to show more information,
2728 in particular, the error code if getpwnam() issues one.
2730 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2731 This is apparently needed in addition to the PH/07 change above to avoid
2732 any possible encoding problems.
2734 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2735 but not after initializing Perl.
2737 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2738 output them only if debugging. By default they are written stderr,
2739 apparently, which is not desirable.
2741 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2744 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2747 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2749 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2750 authenticated or an ident call has been made. Suppress the default
2751 values for $authenticated_id and $authenticated_sender (but permit -oMai
2752 and -oMas) when testing with -bh.
2754 PH/24 Re-jigged the order of the tests in the default configuration so that the
2755 tests for valid domains and recipients precede the DNS black list and CSA
2756 tests, on the grounds that those ones are more expensive.
2758 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2759 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2760 This bug exists in every version of Exim that I still have, right back to
2763 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2764 However, an attempt to turn on case-sensitivity in a regex key by
2765 including (?-i) didn't work because the subject string was already
2766 lowercased, and the effects were non-intuitive. It turns out that a
2767 one-line patch can be used to allow (?-i) to work as expected.
2773 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2774 other effects) broke the use of negated acl sub-conditions.
2776 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2779 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2780 "Deny" causes Exim to reject the incoming connection with a 554 error.
2781 Unfortunately, if there is a major crisis, such as a disk failure,
2782 tcp-wrappers gives "deny", whereas what one would like would be some
2783 kind of temporary error. A kludge has been added to help with this.
2784 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2785 554 error is used if errno is still zero or contains ENOENT (which occurs
2786 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2789 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2791 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2792 errors. Otherwise a message that provokes a temporary error (when other
2793 messages do not) can cause a whole host to time out.
2795 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2796 addresses were routed directly to files or pipes from a redirect router.
2797 File deliveries just didn't batch; pipe deliveries might have suffered
2800 PH/06 A failure to get a lock for a hints database would erroneously always say
2801 "Failed to get write lock", even when it was really a read lock.
2803 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2804 of 0600. This has been changed to use the value of the lockfile_mode
2805 option (which defaults to 0600).
2807 PH/08 Applied small patch from the Sieve maintainer.
2809 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2810 folder from quota calculations, a direct delivery into this folder messed
2811 up the contents of the maildirsize file. This was because the regex was
2812 used only to exclude .Trash (or whatever) when the size of the mailbox
2813 was calculated. There was no check that a delivery was happening into an
2814 excluded directory. This bug has been fixed by ignoring all quota
2815 processing for deliveries into excluded directories.
2817 PH/10 Added the maildirfolder_create_regex option to appendfile.
2823 PH/01 The code for finding all the local interface addresses on a FreeBSD
2824 system running IPv6 was broken. This may well have applied to all BSD
2825 systems, as well as to others that have similar system calls. The broken
2826 code found IPv4 interfaces correctly, but gave incorrect values for the
2827 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2828 that it would not match correctly against @[] and not recognize the IPv6
2831 PH/02 The ipliteral router was not recognizing addresses of the form user@
2832 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2834 PH/03 Added disable_ipv6.
2836 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2837 decoded form, because it is most often used to construct To: headers
2838 lines in autoreplies, and the decoded form may well be syntactically
2839 invalid. However, $reply_address has leading white space removed, and all
2840 newlines turned into spaces so that the autoreply transport does not
2843 PH/05 If group was specified without a user on a router, and no group or user
2844 was specified on a transport, the group from the router was ignored.
2846 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2847 for visible compile-time settings that can be used to change these
2848 numbers, for those that want even more. Backwards compatibility with old
2849 spool files has been maintained. However, going back to a previous Exim
2850 release will lost any variables that are in spool files.
2852 PH/07 Two small changes when running in the test harness: increase delay when
2853 passing a TCP/IP connection to a new process, in case the original
2854 process has to generate a bounce, and remove special handling of
2855 127.0.0.2 (sic), which is no longer necessary.
2857 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2858 be the same on different OS.
2860 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2863 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2864 whether --show-vars was specified or not
2866 JJ/02 exipick: Added support for new ACL variable spool format introduced
2869 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2870 syntactically invalid From: or Reply-to: line, and a filter used this to
2871 generate an autoreply, and therefore failed to obtain an address for the
2872 autoreply, Exim could try to deliver to a non-existent relative file
2873 name, causing unrelated and misleading errors. What now happens is that
2874 it logs this as a hard delivery error, but does not attempt to create a
2877 PH/11 The exinext utility has a -C option for testing purposes, but although
2878 the given file was scanned by exinext itself; it wasn't being passed on
2879 when Exim was called.
2881 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2882 an end-of-file indication when reading a command response.
2884 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2885 compiled. In many other places in Exim, IPv6 addresses are always
2886 recognized, so I have changed this. It also means that IPv4 domain
2887 literals of the form [IPV4:n.n.n.n] are now always recognized.
2889 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2890 used if the router is not running as root, for example, when verifying at
2891 ACL time, or when using -bh. The debugging output from this situation was
2892 non-existent - all you got was a failure to exec. I have made two
2895 (a) Failures to set uid/gid, the current directory, or a process leader
2896 in a subprocess such as that created by queryprogram now generate
2897 suitable debugging output when -d is set.
2899 (b) The queryprogram router detects when it is not running as root,
2900 outputs suitable debugging information if -d is set, and then runs
2901 the subprocess without attempting to change uid/gid.
2903 PH/15 Minor change to Makefile for building test_host (undocumented testing
2906 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2907 additional section of a DNS packet that returns MX or SRV records.
2908 Instead, it always explicitly searches for A/AAAA records. This avoids
2909 major problems that occur when a DNS server includes only records of one
2910 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2911 fixed another bug: if SRV records were looked up and the corresponding
2912 address records were *not* found in the additional section, the port
2913 values from the SRV records were lost.
2915 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2916 using the correct key (the original address) when searching the retry
2917 rules in order to find which one to use for generating the retry hint.
2919 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2920 adding the default one. Similarly, if it contains a Reply-To: header, the
2921 errors_reply_to option, if set, is not used.
2923 PH/19 When calculating a retry time, Exim used to measure the "time since
2924 failure" by looking at the "first failed" field in the retry record. Now
2925 it does not use this if it is later than than the arrival time of the
2926 message. Instead it uses the arrival time. This makes for better
2927 behaviour in cases where some deliveries succeed, thus re-setting the
2928 "first failed" field. An example is a quota failure for a huge message
2929 when small messages continue to be delivered. Without this change, the
2930 "time since failure" will always be short, possible causing more frequent
2931 delivery attempts for the huge message than are intended.
2932 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2934 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2935 $1, $2, $3) because the numerical variables can be reset during some
2936 expansion items (e.g. "match"), thereby losing the authentication data.
2938 PH/21 Make -bV show the size of off_t variables so that the test suite can
2939 decide whether to run tests for quotas > 2G.
2941 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2942 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2943 filecount value is greater than 2G or if a quota value is greater than 2G
2944 on a system where the size of off_t is not greater than 4, a panic error
2947 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2948 never match. The debug and -bh output now contains an explicit error
2949 message indicating a malformed IPv4 address or mask.
2951 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2952 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2953 PH/23 above applies.
2955 PH/25 Do not write to syslog when running in the test harness. The only
2956 occasion when this arises is a failure to open the main or panic logs
2957 (for which there is an explicit test).
2959 PH/26 Added the /no_tell option to "control=freeze".
2961 PH/27 If a host name lookup failed very early in a connection, for example, if
2962 the IP address matched host_lookup and the reverse lookup yielded a name
2963 that did not have a forward lookup, an error message of the form "no IP
2964 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2965 could be logged. Now it outputs the IP address instead of "NULL".
2967 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2968 allows the sender and the authenticated sender to be set when
2969 submitting a message from within Exim. Since child_open_exim() is
2970 documented for local_scan(), the new function should be too.
2972 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2973 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2974 results in an empty string is now treated as unset.
2976 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2978 PH/31 Added sender_verify_fail logging option.
2980 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2981 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2982 tidied the source and removed it altogether.
2984 PH/33 When a queue run was abandoned because the load average was too high, a
2985 log line was always written; now it is written only if the queue_run log
2986 selector is set. In addition, the log line for abandonment now contains
2987 information about the queue run such as the pid. This is always present
2988 in "start" and "stop" lines but was omitted from the "abandon" line.
2990 PH/34 Omit spaces between a header name and the colon in the error message that
2991 is given when verify = headers_syntax fails (if there are lots of them,
2992 the message gets confusing).
2994 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2995 names, as there are now some PTR records that contain slashes. This check
2996 is only to protect against broken name servers that fall over on strange
2997 characters, so the fact that it applies to all lookups doesn't matter.
2999 PH/36 Now that the new test suite is complete, we can remove some of the
3000 special code in Exim that was needed for the old test suite. For example,
3001 sorting DNS records because real resolvers return them in an arbitrary
3002 order. The new test suite's fake resolver always returns records in the
3005 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3006 bounces) except when queue_only is set, to avoid logging races between
3007 the different processes.
3009 PH/38 Panic-die if .include specifies a non-absolute path.
3011 PH/39 A tweak to the "H" retry rule from its user.
3013 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3014 a label. They prevented compilation on older perls.
3016 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3017 a warning to be raised on newish perls.
3019 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3020 on queue. Changes to match documented behaviour of showing count of
3021 messages matching specified criteria.
3023 PH/40 Changed the default ident timeout from 30s to 5s.
3025 PH/41 Added support for the use of login_cap features, on those BSD systems
3026 that have them, for controlling the resources used by pipe deliveries.
3028 PH/42 The content-scanning code uses fopen() to create files in which to put
3029 message data. Previously it was not paying any attention to the mode of
3030 the files. Exim runs with umask(0) because the rest of the code creates
3031 files with open(), and sets the required mode explicitly. Thus, these
3032 files were ending up world-writeable. This was not a big issue, because,
3033 being within the spool directory, they were not world-accessible. I have
3034 created a function called modefopen, which takes an additional mode
3035 argument. It sets umask(777), creates the file, chmods it to the required
3036 mode, then resets the umask. All the relevant calls to fopen() in the
3037 content scanning code have been changed to use this function.
3039 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3040 to 24 hours. This avoids potential overflow problems when processing G
3041 and H retry rules. I suspect nobody ever tinkers with this value.
3043 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3045 PH/45 When the plaintext authenticator is running as a client, the server's
3046 challenges are checked to ensure they are valid base64 strings. By
3047 default, the authentication attempt is cancelled if an invalid string is
3048 received. Setting client_ignore_invalid_base64 true ignores these errors.
3049 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3050 they are received. Thus, the responses can be made to depend on the
3051 challenges. If an invalid string is ignored, an empty string is placed in
3054 PH/46 Messages that are created by the autoreply transport now contains a
3055 References: header, in accordance with RFCs 2822 and 3834.
3057 PH/47 Added authenticated_sender_force to the smtp transport.
3059 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3061 PH/49 Installed latest patch from the Sieve maintainer.
3063 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3064 was also set, the appendfile transport was unnecessarily scanning a
3065 directory of message files (e.g. for maildir delivery) to find the count
3066 of files (along with the size), even though it did not need this
3067 information. It now does the scan only if it needs to find either the
3068 size of the count of files.
3070 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3072 PH/52 Two bugs concerned with error handling when the smtp transport is
3075 (i) Exim was not creating retry information for temporary errors given
3076 for individual recipients after the DATA command when the smtp transport
3077 was used in LMTP mode. This meant that they could be retried too
3078 frequently, and not timed out correctly.
3080 (ii) Exim was setting the flag that allows error details to be returned
3081 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3082 recipients that were returned after the DATA command.
3084 PH/53 This is related to PH/52, but is more general: for any failing address,
3085 when detailed error information was permitted to be returned to the
3086 sender, but the error was temporary, then after the final timeout, only
3087 "retry timeout exceeded" was returned. Now it returns the full error as
3088 well as "retry timeout exceeded".
3090 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3091 do this, and (what is worse) MTAs that accept it.
3093 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3094 will now be deprecated.
3096 PH/56 New os.c-cygwin from the Cygwin maintainer.
3098 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3099 formats (previously only available in exim formats via -bpr, -bpru,
3100 and -bpra. Now also available in native and exiqgrep formats)
3102 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3103 with very large, slow to parse queues
3105 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3107 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3109 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3110 responses to authentication challenges, though it was showing the
3111 challenges; (ii) I've removed the CR characters from the debug output for
3114 PH/58 Allow for the insertion of a newline as well as a space when a string
3115 is turned into more than one encoded-word during RFC 2047 encoding. The
3116 Sieve code now uses this.
3118 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3119 data_4xx, lost_connection, tls_required.
3121 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3122 message was being sent as an SMTP response.
3124 PH/61 Add -l and -k options to exicyclog.
3126 PH/62 When verifying, if an address was redirected to one new address, so that
3127 verification continued, and the new address failed or deferred after
3128 having set something in $address_data, the value of $address_data was not
3129 passed back to the ACL. This was different to the case when no
3130 redirection occurred. The value is now passed back in both cases.
3132 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3133 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3134 use login_cap.h, so on its own it isn't the distinguishing feature. The
3135 new name refers directly to the setclassresources() function.
3137 PH/65 Added configuration files for NetBSD3.
3139 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3141 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3142 is preferred over IPv4.
3144 PH/68 The bounce_return_message and bounce_return_body options were not being
3145 honoured for bounces generated during the reception of non-SMTP messages.
3146 In particular, this applied to messages rejected by the ACL. This bug has
3147 been fixed. However, if bounce_return_message is true and bounce_return_
3148 body is false, the headers that are returned for a non-SMTP message
3149 include only those that have been read before the error was detected.
3150 (In the case of an ACL rejection, they have all been read.)
3152 PH/69 The HTML version of the specification is now built in a directory called
3153 spec_html instead of spec.html, because the latter looks like a path with
3154 a MIME-type, and this confuses some software.
3156 PH/70 Catch two compiler warnings in sieve.c.
3158 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3159 function verify_get_ident() calls ip_connect() to connect a socket, but
3160 if the "connect()" function timed out, ip_connect() used to close the
3161 socket. However, verify_get_ident() also closes the socket later, and in
3162 between Exim writes to the log, which may get opened at this point. When
3163 the socket was closed in ip_connect(), the log could get the same file
3164 descriptor number as the socket. This naturally causes chaos. The fix is
3165 not to close the socket in ip_connect(); the socket should be closed by
3166 the function that creates it. There was only one place in the code where
3167 this was missing, in the iplookup router, which I don't think anybody now
3168 uses, but I've fixed it anyway.
3170 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3171 well as to direct DNS lookups. Otherwise the handling of names in host
3172 lists is inconsistent and therefore confusing.
3178 PH/01 Two changes to the default runtime configuration:
3180 (1) Move the checks for relay_from_hosts and authenticated clients from
3181 after to before the (commented out) DNS black list checks.
3183 (2) Add control=submission to the relay_from_hosts and authenticated
3184 clients checks, on the grounds that messages accepted by these
3185 statements are most likely to be submissions.
3187 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3189 (1) Generate an error if the third argument for the ${prvs expansion is
3192 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3195 (3) Reset the variables that are obtained from the first argument of
3196 ${prvscheck and used in the second argument before leaving the code,
3197 because their memory is reclaimed, so using them afterwards may do
3200 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3201 one (it's much easier than Tom thought :-).
3203 (5) Because of (4), we can now allow for the use of $prvscheck_result
3204 inside the third argument.
3206 PH/03 For some reason, the default setting of PATH when running a command from
3207 a pipe transport was just "/usr/bin". I have changed it to
3210 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3211 anything to be listed in the output from -bV.
3213 PH/05 When a filter generated an autoreply, the entire To: header line was
3214 quoted in the delivery log line, like this:
3216 => >A.N.Other <ano@some.domain> <original@ddress> ...
3218 This has been changed so that it extracts the operative address. There
3219 may be more than one such address. If so, they are comma-separated, like
3222 => >ano@some.domain,ona@other.domain <original@ddress> ...
3224 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3225 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3226 not being looked up in the rDNS to get a host name, Exim was showing the
3227 IP address twice in Received: lines, even though the IP addresses were
3228 identical. For example:
3230 Received: from [1.2.3.4] (helo=[1.2.3.4])
3232 However, if the real host name was known, it was omitting the HELO data
3233 if it matched the actual IP address. This has been tidied up so that it
3234 doesn't show the same IP address twice.
3236 PH/07 When both +timestamp and +memory debugging was on, the value given by
3237 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3238 called by the debug printing, thereby overwriting the timestamp buffer.
3239 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3241 PH/08 When the original message was included in an autoreply transport, it
3242 always said "this is a copy of the message, including all the headers",
3243 even if body_only or headers_only was set. It now gives an appropriate
3246 PH/09 Applied a patch from the Sieve maintainer which:
3248 o fixes some comments
3249 o adds the (disabled) notify extension core
3250 o adds some debug output for the result of if/elsif tests
3251 o points to the current vacation draft in the documentation
3252 and documents the missing references header update
3256 o fixes a bug in processing the envelope test (when testing
3257 multiple envelope elements, the last element determined the
3260 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3261 Electronic Mail") by including:
3263 Auto-submitted: auto-generated
3265 in the messages that it generates (bounce messages and others, such as
3266 warnings). In the case of bounce messages for non-SMTP messages, there was
3267 also a typo: it was using "Auto_submitted" (underscore instead of
3268 hyphen). Since every message generated by Exim is necessarily in response
3269 to another message, thes have all been changed to:
3271 Auto-Submitted: auto-replied
3273 in accordance with these statements in the RFC:
3275 The auto-replied keyword:
3277 - SHOULD be used on messages sent in direct response to another
3278 message by an automatic process,
3280 - MUST NOT be used on manually-generated messages,
3282 - MAY be used on Delivery Status Notifications (DSNs) and Message
3283 Disposition Notifications (MDNs),
3285 - MUST NOT be used on messages generated by automatic or periodic
3286 processes, except for messages which are automatic responses to
3289 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3290 to the default Received: header definition.
3292 PH/12 Added log selector acl_warn_skipped (default on).
3294 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3295 variables because (a) they are in the wrong storage pool and (b) even if
3296 they were copied, it wouldn't work properly because of the caching.
3298 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3299 checking when decoding. Apparently there are clients that generate
3300 overlong encoded strings. Why am I not surprised?
3302 PH/15 If the first argument of "${if match_address" was not empty, but did not
3303 contain an "@" character, Exim crashed. Now it writes a panic log message
3304 and treats the condition as false.
3306 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3308 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3309 extension "envelope-auth". The code is finished and in agreement with
3310 other implementations, but there is no documentation so far and in fact,
3311 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3312 not changing the active code.
3314 Print executed "if" and "elsif" statements when debugging is used. This
3315 helps a great deal to understand what a filter does.
3317 Document more things not specified clearly in RFC3028. I had all this
3318 sorted out, when out of a sudden new issues came to my mind. Oops."
3320 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3323 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3324 canonical form (as documented). However, after a host name lookup from
3325 the IP address, check_host() was doing a simple string comparison with
3326 addresses acquired from the DNS when checking that the found name did
3327 have the original IP as one of its addresses. Since any found IPv6
3328 addresses are likely to be in abbreviated form, the comparison could
3329 fail. Luckily, there already exists a function for doing the comparison
3330 by converting both addresses to binary, so now that is used instead of
3331 the text comparison.
3333 PH/20 There was another similar case to PH/19, when a complete host name was
3334 given in a host list; looking up its IP address could give an abbreviated
3335 form, whereas the current host's name might or might not be abbreviated.
3336 The same fix has been applied.
3342 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3343 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3346 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3347 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3349 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3351 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3352 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3353 Also, for "Auto-Submitted", treat anything other than "no" as
3354 non-personal, in accordance with RFC 3834. (Previously it treated
3355 anything starting "auto-" as non-personal.)
3357 TF/01 The control=submission/name=... option had a problem with syntax
3358 errors if the name included a slash character. The /name= option
3359 now slurps the rest of the string, so it can include any characters
3360 but it must come last in the list of options (after /sender_retain
3363 PH/05 Some modifications to the interface to the fake nameserver for the new
3371 TK/01 Added the "success_on_redirect" address verification option. See
3372 NewStuff for rationale and an example.
3374 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3376 PH/02 Patch to exigrep to allow it to work on syslog lines.
3378 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3379 fread() to skip over the body file's header line, because in Cygwin the
3380 header line is locked and is inaccessible.
3382 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3383 co-exist for some time) to make it clear that it is the Exim ID that is
3384 referenced, not the Message-ID: header line.
3386 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3387 string_format() function, because snprintf() does not exist on all
3390 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3391 ${stat: expansion item.
3393 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3394 protocol synchronization error", to keep the pedants happy.
3396 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3397 well as for IRIX systems, when gcc is being used. See the host.c source
3400 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3402 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3405 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3406 transport and to the smtp transport in LMTP mode.
3408 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3410 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3411 run a filter in a subprocess. This could lead to confusion in subsequent
3412 lookups in the parent process. There should also be a search_tidyup() at
3413 the end of the subprocess.
3415 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3416 only if the host matched helo_try_verify_hosts, which caused the
3417 verification to occur when the EHLO/HELO command was issued. The ACL just
3418 tested the remembered result. Now, if a previous verification attempt has
3419 not happened, "verify = helo" does it there and then.
3421 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3423 TK/03 Fix log output including CR from clamd.
3425 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3426 exist provoked a memory error which could cause a segfault.
3428 PH/15 Installed PCRE 6.2
3430 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3432 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3433 of the problem. Specifically, suggested +O2 rather than +O1 for the
3436 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3438 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3439 record, the port was not correctly logged when the outgoing_port log
3440 selector was set (it logged the transort's default port).
3442 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3443 fallback_hosts, and "hosts" in the smtp transport.
3445 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3446 host errors such as "Connection refused".
3448 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3449 authentication with radiusclient 0.4.9:
3451 - Error returned from rc_read_config was caught wrongly
3452 - Username/password not passed on to radius server due to wrong length.
3454 The presumption is that some radiusclient API changes for 4.51/PH/17
3455 were not taken care of correctly. The code is still untested by me (my
3456 Linux distribution still has 0.3.2 of radiusclient), but it was
3457 contributed by a Radius user.
3459 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3460 expanding the "port" option of the smtp transport.
3462 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3463 while reading a MIME header. Thanks to Tom Hughes for a patch.
3465 PH/24 Include config.h inside local_scan.h so that configuration settings are
3468 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3469 that in an ACL for RCPT (for example), you can examine exactly what was
3472 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3473 commands, but it was not correctly comparing the address with the actual
3474 client host address. Thus, it would show the EHLO address in Received:
3475 header lines when this was not necessary.
3477 PH/27 Added the % operator to ${eval:}.
3479 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3480 it should be ignoring failures (because with -C, for example, it has lost
3481 privilege). It wasn't ignoring creation failures other than "already
3484 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3487 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3488 because an input line was too long, either on its own, or by virtue of
3489 too many continuations, the temporary file was not being removed, and the
3490 return code was incorrect.
3492 PH/31 Missing "BOOL" in function definition in filtertest.c.
3494 PH/32 Applied Sieve patches from the maintainer.
3496 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3498 PH/33 Added "verify = not_blind".
3500 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3501 Local/Makefile (with some defaults set). These are used in built scripts
3502 such as exicyclog, but they have never been used in the exim_install
3503 script (though there are many overriding facilities there). I have
3504 arranged that the exim_install script now takes note of these two
3507 PH/35 Installed configuration files for Dragonfly.
3509 PH/36 When a locally submitted message by a trusted user did not contain a
3510 From: header, and the sender address was obtained from -f or from an SMTP
3511 MAIL command, and the trusted user did not use -F to supply a sender
3512 name, $originator_name was incorrectly used when constructing a From:
3513 header. Furthermore, $originator_name was used for submission mode
3514 messages from external hosts without From: headers in a similar way,
3515 which is clearly wrong.
3517 PH/37 Added control=suppress_local_fixups.
3519 PH/38 When log_selector = +received_sender was set, and the addition of the
3520 sender made the log line's construction buffer exactly full, or one byte
3521 less than full, an overflow happened when the terminating "\n" was
3524 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3525 when the result of a list match is failure because a DNS lookup failed.
3527 PH/40 RM_COMMAND is now used in the building process.
3529 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3530 the "build-* directories that it finds.
3532 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3533 address in a domain literal was a prefix of an interface address.
3535 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3536 when verifying a sender address, unless rewrite_headers is false.
3538 PH/44 Wrote a long comment about why errors_to addresses are verified as
3539 recipients, not senders.
3541 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3542 the ratelimit ACL was added.
3544 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3546 PH/46 Added extra information about PostgreSQL errors to the error string.
3548 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3549 avoiding the need to install special zones in a real server. This is
3550 backwards compatible; if it can't find the fake resolver, it drops back.
3551 Thus, both old and new test suites can be run.
3553 TF/02 Added util/ratelimit.pl
3555 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3556 clock is set back in time.
3558 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3559 Candler <B.Candler@pobox.com>.
3561 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3562 OK for addresses that are the result of redirections.
3564 PH/48 A number of further additions for the benefit of the new test suite,
3565 including a fake gethostbyname() that interfaces to the fake DNS resolver
3568 TF/06 The fix for widen_domains has also been applied to qualify_single and
3569 search_parents which are the other dnslookup options that can cause
3572 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3575 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3577 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3578 service" if it encountered an MX record with an empty target hostname.
3579 The message is now "an MX or SRV record indicated no SMTP service".
3581 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3582 if the DNS of the sending site is misconfigured. This is quite a
3583 common situation. This change restores the behaviour of treating a
3584 helo verification defer as a failure.
3586 PH/51 If self=fail was set on a router, the bounce message did not include the
3587 actual error message.
3593 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3595 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3596 command itself ended in error, the underlying message about the transport
3597 filter timeout was being overwritten with the pipe command error. Now the
3598 underlying error message should be appended to the second error message.
3600 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3602 PH/02 Reduce the amount of output that "make" produces by default. Full output
3603 can still be requested.
3605 PH/03 The warning log line about a condition test deferring for a "warn" verb
3606 was being output only once per connection, rather than after each
3607 occurrence (because it was using the same function as for successful
3608 "warn" verbs). This seems wrong, so I have changed it.
3610 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3611 it should not have, which might have caused a crash in the right
3612 circumstances, but probably never did.
3614 PH/04 Installed a modified version of Tony Finch's patch to make submission
3615 mode fix the return path as well as the Sender: header line, and to
3616 add a /name= option so that you can make the user's friendly name appear
3619 TF/03 Added the control = fakedefer ACL modifier.
3621 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3622 Mark Lowes for thorough testing.
3624 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3626 TK/03 Merged latest SRS patch from Miles Wilton.
3628 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3629 with the definition in sysexits.h (which is #included earlier).
3630 Fortunately, Exim does not actually use EX_OK. The code used to try to
3631 preserve the sysexits.h value, by assuming that macro definitions were
3632 scanned for macro replacements. I have been disabused of this notion,
3633 so now the code just undefines EX_OK before #including unistd.h.
3635 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3636 in the smtp transport. When a block could not be written in a single
3637 write() function, the timeout was being re-applied to each part-write.
3638 This seems wrong - if the receiver was accepting one byte at a time it
3639 would take for ever. The timeout is now adjusted when this happens. It
3640 doesn't have to be particularly precise.
3642 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3643 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3645 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3646 without a domain if the check to <postmaster@domain> fails.
3648 SC/01 Eximstats: added -xls and the ability to specify output files
3649 (patch written by Frank Heydlauf).
3651 SC/02 Eximstats: use FileHandles for outputting results.
3653 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3655 SC/04 Eximstats: fixed display of large numbers with -nvr option
3657 SC/05 Eximstats: fixed merging of reports with empty tables.
3659 SC/06 Eximstats: added the -include_original_destination flag
3661 SC/07 Eximstats: removed tabs and trailing whitespace.
3663 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3665 TK/06 MBOX spool code: Add real "From " MBOX separator line
3666 so the .eml file is really in mbox format (even though
3667 most programs do not really care). Patch from Alex Miller.
3669 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3670 The latter is generated from $received_to and is only set if the
3671 message has one envelope recipient. SA can use these headers,
3672 obviously out-of-the-box. Patch from Alex Miller.
3674 PH/08 The ${def test on a variable was returning false if the variable's
3675 value was "0", contrary to what the specification has always said!
3676 The result should be true unless the variable is empty.
3678 PH/09 The syntax error of a character other than { following "${if
3679 def:variable_name" (after optional whitespace) was not being diagnosed.
3680 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3681 accidental colon was present, for example, could give incorrect results.
3683 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3684 result is used (not including appendfile, where other changes are about
3687 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3688 This involved changing a lot of size variables from int to off_t. It
3689 should work with maildirs and everything.
3691 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3692 spamd dying while we are connected to it.
3694 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3697 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3698 function for BDB 4.3.
3700 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3702 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3703 the sources for PCRE have been re-arranged and more files are now
3706 PH/15 The code I had for printing potentially long long variables in PH/11
3707 above was not the best (it lost precision). The length of off_t variables
3708 is now inspected at build time, and an appropriate printing format (%ld
3709 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3710 to be "long long int" or "long int". This is needed for the internal
3711 formatting function string_vformat().
3713 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3714 the configuration file to be ":syslog", then the script "guesses" where
3715 the logs files are, rather than using the compiled in default. In our
3716 case the guess is not the same as the compiled default, so the script
3717 suddenly stopped working when I started to use syslog. The patch checks
3718 to see if log_file_path is "". If so, it attempts to read it from exim
3719 with no configuration file to get the compiled in version, before it
3720 falls back to the previous guessing code."
3722 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3723 implementing BATV in an Exim configuration. See NewStuff for the gory
3726 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3727 Makefile that are specific to HP-UX.
3729 PH/18 If the "use_postmaster" option was set for a recipient callout together
3730 with the "random" option, the postmaster address was used as the MAIL
3731 FROM address for the random test, but not for the subsequent recipient
3732 test. It is now used for both.
3734 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3735 patch removes a few documentation additions to RFC 3028, because the
3736 latest draft now contains them. It adds the new en;ascii-case comparator
3737 and a new error check for 8bit text in MIME parts. Comparator and
3738 require names are now matched exactly. I enabled the subaddress
3739 extension, but it is not well tested yet (read: it works for me)."
3741 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3742 rework some of the code of TK/09 above to avoid the hardwired use of
3743 "%lld" and "long long". Replaced the call to snprintf() with a call to
3746 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3747 records point to non-existent hosts", "retry timeout exceeded", and
3748 "retry time not reached for any host after a long failure period".
3750 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3751 experimental DomainKeys support:
3753 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3754 (2) On an error such as an illegally used "control", the wrong name for
3755 the control was given.
3757 These problems did NOT occur unless DomainKeys support was compiled.
3759 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3761 PH/24 Added ${if match_ip condition.
3763 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3764 databases so that it will be absolutely obvious if a crash occurs in the
3765 DB library. This is a regular occurrence (often caused by mis-matched
3768 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3769 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3770 picked up on a user's system that detects such things. There doesn't seem
3771 to be a gcc warning option for this - only an attribute that has to be
3772 put on the function's prototype. It seems that in Fedora Core 4 they have
3773 set this on a number of new functions. No doubt there will be more in due
3776 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3777 specify a transport. However, if an address that was verified by such a
3778 router was the subject of a callout, Exim crashed because it tried to
3779 read the rcpt_include_affixes from the non-existent transport. Now it
3780 just assumes that the setting of that option is false. This bug was
3781 introduced by 4.51/PH/31.
3783 PH/28 Changed -d+all to exclude +memory, because that information is very
3784 rarely of interest, but it makes the output a lot bigger. People tend to
3785 do -d+all out of habit.
3787 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3788 code in os-type was giving problems when libc.so lives in lib64, like on
3791 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3792 aren't the modern standard, and it seems that some systems' include files
3793 don't always have them. Exim was already checking for some of the newer
3794 ones like T_AAAA, and defining it itself. I've added checks for all the
3795 record types that Exim uses.
3797 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3798 not automatically generating a new one, as it is supposed to. This
3799 prevented TLS from working. If the file did exist, but contained invalid
3800 data, a new version was generated, as expected. It was only the case of a
3801 non-existent file that was broken.
3803 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3804 with a change in libdomainkeys > 0.64.
3806 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3807 from DNS. If the selector record carries the flag, it now has
3808 precedence over the domain-wide flag.
3810 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3812 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3813 the use of a port name that isn't defined in /etc/services) occurred, the
3814 message was deferred as in a normal delivery, and thus remained on the
3815 spool, instead of being failed because of the mua_wrapper setting. This
3816 is now fixed, and I tidied up some of the mua_wrapper messages at the
3819 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3820 the messages in a hash of arrays rather than using individual hashes.
3821 This is a bit cleaner and results in dramatic memory savings, albeit
3822 at a slight CPU cost.
3824 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3825 as requested by Marc Sherman.
3827 SC/10 Eximstats: added histograms for user specified patterns as requested
3830 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3832 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3833 fopen() in the content-scanning modules that did not already have it.
3839 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3840 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3842 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3844 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3846 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3847 to test Sieve filters that use "vacation".
3849 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3850 that changes the way the GnuTLS parameters are stored in the cache file.
3851 The new format can be generated externally. For backward compatibility,
3852 if the data in the cache doesn't make sense, Exim assumes it has read an
3853 old-format file, and it generates new data and writes a new file. This
3854 means that you can't go back to an older release without removing the
3857 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3858 work if there are any delivery delays because "one_time" forces the
3859 parent to be marked "delivered", so its unseen clone is never tried
3860 again. For this reason, Exim now forbids the simultaneous setting of
3863 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3864 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3865 entry if you want to know the details. The fix, however, neglected to
3866 consider the case when local delivery batching is involved. The test for
3867 "previously delivered" was not happening when checking to see if an
3868 address could be batched with a previous (undelivered) one; under
3869 certain circumstances this could lead to multiple deliveries to the same
3872 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3873 in its include files, and this causes problems building Exim.
3875 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3876 header_syntax) but Exim was just ignoring anything given after a slash.
3877 In particular, this caused confusion with an attempt to use "verify =
3878 reverse_host_lookup/defer_ok". An error is now given when options are
3879 supplied for verify items that do not have them. (Maybe reverse_host_
3880 lookup should have a defer_ok option, but that's a different point.)
3882 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3883 defined by RFC 821) to 2048, because there were problems with some AUTH
3884 commands, and RFC 1869 says the size should be increased for extended
3885 SMTP commands that take arguments.
3887 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3890 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3891 "unknown" error; now it says that the functionality isn't in the binary.
3893 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3894 an address' error message when a string expansion fails (syntax or
3895 whatever). Otherwise the password may appear in the log. Following change
3896 PH/42 below, there is no longer a chance of it appearing in a bounce
3899 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3901 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3902 of its addresses was ever tried. (Bugzilla bug #2.)
3904 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3905 the result incorrectly in the debug output. (It correctly added a newline
3906 to what was transported.)
3908 TF/01 Added $received_time.
3910 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3911 commented out examples of how to interface to a virus scanner and to
3912 SpamAssassin. Also added commented examples of av_scanner and
3913 spamd_address settings.
3915 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3916 and controls are allowed in which ACLs. There were a couple of minor
3917 errors. Some of the entries in the conditions table (which is a table of
3918 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3919 negation of where the condition IS allowed.
3921 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3923 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3924 header file does not have a version number, so I've had to invent a new
3925 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3926 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3927 radiusclient), but it was contributed by a Radius user.
3929 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3930 files or directories, for OpenSSL.
3932 PH/19 When an Exim process that is running as root has to create an Exim log
3933 file, it does so in a subprocess that runs as exim:exim so as to get the
3934 ownership right at creation (otherwise, other Exim processes might see
3935 the file with the wrong ownership). There was no test for failure of this
3936 fork() call, which would lead to the process getting stuck as it waited
3937 for a non-existent subprocess. Forks do occasionally fail when resources
3938 run out. I reviewed all the other calls to fork(); they all seem to check
3941 PH/20 When checking for unexpected SMTP input at connect time (before writing
3942 the banner), Exim was not dealing correctly with a non-positive return
3943 from the read() function. If the client had disconnected by this time,
3944 the result was a log entry for a synchronization error with an empty
3945 string after "input=" when read() returned zero. If read() returned -1
3946 (an event I could not check), uninitialized data bytes were printed.
3947 There were reports of junk text (parts of files, etc) appearing after
3950 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3952 PH/22 Added support for macro redefinition, and (re)definition in between
3953 driver and ACL definitions.
3955 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3956 forgetting to use the resulting value; it was using the unexpanded value.
3958 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3959 hadn't been configured. The fix is from Juergen Kreileder, who
3960 understands it better than I do:
3962 "Here's what I see happening with three configured cyrus_sasl
3963 authenticators configured (plain, login, cram-md5):
3965 On startup auth_cyrus_sasl_init() gets called for each of these.
3966 This means three calls to sasl_listmech() without a specified mech_list.
3967 => SASL tests which mechs of all available mechs actually work
3968 => three warnings about OTP not working
3969 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3971 With the patch, sasl_listmech() also gets called three times. But now
3972 SASL's mech_list option is set to the server_mech specified in the the
3973 authenticator. Or in other words, the answer from sasl_listmech()
3974 gets limited to just the mech you're testing for (which is different
3976 => the return list contains just 'plain' or 'login', 'cram-md5' or
3977 nothing depending on the value of ob->server_mech.
3979 I've just tested the patch: Authentication still works fine,
3980 unavailable mechs specified in the exim configuration are still
3981 caught, and the auth.log warnings about OTP are gone."
3983 PH/25 When debugging is enabled, the contents of the command line are added
3984 to the debugging output, even when log_selector=+arguments is not
3987 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3988 answer is "GNU", and only if the return is "GNU/something" is the answer
3991 PH/27 $acl_verify_message is now set immediately after the failure of a
3992 verification in an ACL, and so is available in subsequent modifiers. In
3993 particular, the message can be preserved by coding like this:
3995 warn !verify = sender
3996 set acl_m0 = $acl_verify_message
3998 Previously, $acl_verify_message was set only while expanding "message"
3999 and "log_message" when a very denied access.
4001 PH/28 Modified OS/os.c-Linux with
4003 -#ifndef OS_LOAD_AVERAGE
4004 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4006 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4007 nomenclature these days.)
4009 PH/29 Installed patch from the Sieve maintainer that adds the options
4010 sieve_useraddress and sieve_subaddress to the redirect router.
4012 PH/30 In these circumstances:
4013 . Two addresses routed to the same list of hosts;
4014 . First host does not offer TLS;
4015 . First host accepts first address;
4016 . First host gives temporary error to second address;
4017 . Second host offers TLS and a TLS session is established;
4018 . Second host accepts second address.
4019 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4020 and peerdn, if requested) that were in fact used only for the second
4023 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4024 attention to any local part prefix or suffix that was matched by the
4025 router that accepted the address. It now behaves in the same way as it
4026 does for delivery: the affixes are removed from the local part unless
4027 rcpt_include_affixes is set on the transport.
4029 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4030 timeout during the DATA phase of an incoming message.
4032 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4033 applied a patch sanctioned by the Sieve maintainer.
4035 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4036 the uid or gid is negative. A case of a negative gid caused this to be
4037 noticed. The fix allows for either to be negative.
4039 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4040 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4043 PH/36 The change PH/12 above was broken. Fixed it.
4045 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4046 the grounds that routing the same address twice would always produce the
4047 same answer. This might have been true once, but it is certainly no
4048 longer true now. Routing a child address may depend on the previous
4049 routing that produced that child. Some complicated redirection strategies
4050 went wrong when messages had multiple recipients, and made Exim's
4051 behaviour dependent on the order in which the addresses were given.
4053 I have moved the duplicate checking until after the routing is complete.
4054 Exim scans the addresses that are assigned to local and remote
4055 transports, and removes any duplicates. This means that more work will be
4056 done, as duplicates will always all be routed, but duplicates are
4057 presumably rare, so I don't expect this is of any significance.
4059 For deliveries to pipes, files, and autoreplies, the duplicate checking
4060 still happens during the routing process, since they are not going to be
4063 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4064 It corrects a timeout issue with spamd. This is Ian's comment: "The
4065 background is that sometimes spamd either never reads data from a
4066 connection it has accepted, or it never writes response data. The exiscan
4067 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4068 blindly assumes that writes won't block so it may never time out."
4070 PH/39 Allow G after quota size as well as K and M.
4072 PH/40 The value set for $authenticated_id in an authenticator may not contain
4073 binary zeroes or newlines because the value is written to log lines and
4074 to spool files. There was no check on this. Now the value is run through
4075 the string_printing() function so that such characters are converted to
4076 printable escape sequences.
4078 PH/41 $message_linecount is a new variable that contains the total number of
4079 lines in the message. Compare $body_linecount, which is the count for the
4082 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4083 bounce and delay warning messages, except in certain special cases, which
4086 (a) An SMTP error message from a remote host;
4087 (b) A message specified in a :fail: redirection;
4088 (c) A message specified in a "fail" command in a system filter;
4089 (d) A message specified in a FAIL return from the queryprogram router;
4090 (e) A message specified by the cannot_route_message router option.
4092 In these cases only, Exim does include the error details in bounce and
4093 warning messages. There are also a few cases where bland messages such
4094 as "unrouteable address" or "local delivery error" are given.
4096 PH/43 $value is now also set for the "else" part of a ${run expansion.
4098 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4099 being worked on, but at least Exim now implements the latest version to
4102 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4103 process to complete was treated as a delivery failure, a timeout while
4104 writing the message to the pipe was logged, but erroneously treated as a
4105 successful delivery. Such timeouts include transport filter timeouts. For
4106 consistency with the overall process timeout, these timeouts are now
4107 treated as errors, giving rise to delivery failures by default. However,
4108 there is now a new Boolean option for the pipe transport called
4109 timeout_defer, which, if set TRUE, converts the failures into defers for
4110 both kinds of timeout. A transport filter timeout is now identified in
4113 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4114 systems where "make" and "gmake" are different, calling "gmake" at top
4115 level broke things. I've arranged for the value of $(MAKE) to be passed
4116 from the Makefile to this script so that it can call the same version of
4120 A note about Exim versions 4.44 and 4.50
4121 ----------------------------------------
4123 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4124 changes of various kinds. As a consequence, a big documentation update was
4125 needed. This delayed the release for rather longer than seemed good, especially
4126 in the light of a couple of (minor) security issues. Therefore, the changes
4127 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4128 release. So 4.44 and 4.50 are in effect two different branches that both start
4131 I have left the 4.50 change log unchanged; it contains all the changes since
4132 4.43. The change log for 4.44 is below; many of its items are identical to
4133 those for 4.50. This seems to be the most sensible way to preserve the
4134 historical information.
4140 1. Minor wording change to the doc/README.SIEVE file.
4142 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4143 computation of the current number of files was incorrect.
4145 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4146 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4149 4. Give more explanation in the error message when the command for a transport
4150 filter fails to execute.
4152 5. There are several places where Exim runs a non-Exim command in a
4153 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4154 was being done only for the command run by the queryprogram router. It is
4155 now done for all such subprocesses. The other cases are: ${run, transport
4156 filters, and the commands run by the lmtp and pipe transports.
4158 6. Added CONFIGURE_GROUP build-time option.
4160 7. Some older OS have a limit of 256 on the maximum number of file
4161 descriptors. Exim was using setrlimit() to set 1000 as a large value
4162 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4163 systems. I've change it so that if it can't get 1000, it tries for 256.
4165 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4166 was an oversight, and furthermore, ever since the addition of extra
4167 controls (e.g. 4.43/32), the checks on when to allow different forms of
4168 "control" were broken. There should now be diagnostics for all cases when a
4169 control that does not make sense is encountered.
4171 9. Added the /retain_sender option to "control=submission".
4173 10. $recipients is now available in the predata ACL (oversight).
4175 11. Tidy the search cache before the fork to do a delivery from a message
4176 received from the command line. Otherwise the child will trigger a lookup
4177 failure and thereby defer the delivery if it tries to use (for example) a
4178 cached ldap connection that the parent has called unbind on.
4180 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4181 of $address_data from the recipient verification was clobbered by the
4182 sender verification.
4184 13. The value of address_data from a sender verification is now available in
4185 $sender_address_data in subsequent conditions in the ACL statement.
4187 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4189 15. Added a new option "connect=<time>" to callout options, to set a different
4192 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4193 was its contents. (It was OK if the option was not defined at all.)
4195 17. A "Completed" log line is now written for messages that are removed from
4196 the spool by the -Mrm option.
4198 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4199 information about exactly what failed.
4201 19. Added -dd to debug only the daemon process.
4203 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4204 handles timeouts, both on the server side and network timeouts. Renamed the
4205 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4207 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4208 It is now set to "smtps".
4210 22. $host_address is now set to the target address during the checking of
4211 ignore_target_hosts.
4213 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4214 being passed; this would have caused $sender_host_name to have been used if
4215 matching the list had actually called for a host name (not very likely,
4216 since this list is usually IP addresses). A host name is now passed as
4219 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4220 code that specifies a non-restarting handler (typically sigaction() in
4221 modern systems) in an attempt to fix a rare and obscure crash bug.
4223 25. Narrowed the window for a race in the daemon that could cause it to ignore
4224 SIGCHLD signals. This is not a major problem, because they are used only to
4225 wake it up if nothing else does.
4227 26. A malformed maildirsize file could cause Exim to calculate negative values
4228 for the mailbox size or file count. Odd effects could occur as a result.
4229 The maildirsize information is now recalculated if the size or filecount
4232 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4233 support for a long time. Removed HAVE_SYS_VFS_H.
4235 28. Installed the latest version of exipick from John Jetmore.
4237 29. In an address list, if the pattern was not a regular expression, an empty
4238 subject address (from a bounce message) matched only if the pattern was an
4239 empty string. Non-empty patterns were not even tested. This was the wrong
4240 because it is perfectly reasonable to use an empty address as part of a
4241 database query. An empty address is now tested by patterns that are
4242 lookups. However, all the other forms of pattern expect the subject to
4243 contain a local part and a domain, and therefore, for them, an empty
4244 address still always fails if the pattern is not itself empty.
4246 30. Exim went into a mad DNS loop when attempting to do a callout where the
4247 host was specified on an smtp transport, and looking it up yielded more
4248 than one IP address.
4250 31. Re-factored the code for checking spool and log partition space into a
4251 function that finds that data and another that does the check. The former
4252 is then used to implement four new variables: $spool_space, $log_space,
4253 $spool_inodes, and $log_inodes.
4255 32. The RFC2047 encoding function was originally intended for short strings
4256 such as real names; it was not keeping to the 75-character limit for
4257 encoded words that the RFC imposes. It now respects the limit, and
4258 generates multiple encoded words if necessary. To be on the safe side, I
4259 have increased the buffer size for the ${rfc2047: expansion operator from
4262 33. It is now permitted to omit both strings after an "if" condition; if the
4263 condition is true, the result is "true". As before, when the second string
4264 is omitted, a false condition yields an empty string. This makes it less
4265 cumbersome to write custom ACL and router conditions.
4267 34. Failure to deliver a bounce message always caused it to be frozen, even if
4268 there was an errors_to setting on the router. The errors_to setting is now
4271 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4272 canonical form (fully expanded) before being placed in
4273 $sender_host_address.
4275 36. The table in the code that translates DNS record types into text (T_A to
4276 "A" for instance) was missing entries for NS and CNAME. It is just possible
4277 that this could have caused confusion if both these types were looked up
4278 for the same domain, because the text type is used as part of Exim's
4279 per-process caching. But the chance of anyone hitting this buglet seems
4282 37. The dnsdb lookup has been extended in a number of ways.
4284 (1) There is a new type, "zns", which walks up the domain tree until it
4285 finds some nameserver records. It should be used with care.
4287 (2) There is a new type, "mxh", which is like "mx" except that it returns
4288 just the host names, not the priorities.
4290 (3) It is now possible to give a list of domains (or IP addresses) to be
4291 looked up. The behaviour when one of the lookups defers can be
4292 controlled by a keyword.
4294 (4) It is now possible to specify the separator character for use when
4295 multiple records are returned.
4297 38. The dnslists ACL condition has been extended: it is now possible to supply
4298 a list of IP addresses and/or domains to be looked up in a particular DNS
4301 39. Added log_selector=+queue_time_overall.
4303 40. When running the queue in the test harness, wait just a tad after forking a
4304 delivery process, to get repeatability of debugging output.
4306 41. Include certificate and key file names in error message when GnuTLS fails
4307 to set them up, because the GnuTLS error message doesn't include the name
4308 of the failing file when there is a problem reading it.
4310 42. Allow both -bf and -bF in the same test run.
4312 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4314 44. The "Exiscan patch" is now merged into the mainline Exim source.
4316 45. Sometimes the final signoff response after QUIT could fail to get
4317 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4318 < 0 before doing a fflush(). This bug looks as though it goes back to the
4319 introduction of TLS in release 3.20, but "sometimes" must have been rare
4320 because the tests only now provoked it.
4322 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4323 (this can affect the format of dates).
4325 47. exim_tidydb, when checking for the continued existence of a message for
4326 which it has found a message-specific retry record, was not finding
4327 messages that were in split spool directories. Consequently, it was
4328 deleting retry records that should have stayed in existence.
4330 48. Steve fixed some bugs in eximstats.
4332 49. The SPA authentication driver was not abandoning authentication and moving
4333 on to the next authenticator when an expansion was forced to fail,
4334 contradicting the general specification for all authenticators. Instead it
4335 was generating a temporary error. It now behaves as specified.
4337 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4338 (the order specifies the preference for clients). The order is now AES256,
4339 AES128, 3DES, ARCFOUR128.
4341 51. Small patch to Sieve code - explicitly set From: when generating an
4344 52. Exim crashed if a remote delivery caused a very long error message to be
4345 recorded - for instance if somebody sent an entire SpamAssassin report back
4346 as a large number of 550 error lines. This bug was coincidentally fixed by
4347 increasing the size of one of Exim's internal buffers (big_buffer) that
4348 happened as part of the Exiscan merge. However, to be on the safe side, I
4349 have made the code more robust (and fixed the comments that describe what
4352 53. Now that there can be additional text after "Completed" in log lines (if
4353 the queue_time_overall log selector is set), a one-byte patch to exigrep
4354 was needed to allow it to recognize "Completed" as not the last thing in
4357 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4358 patch that reportedly fixes this has been added. I am not expert enough to
4359 create a test for it. This is what the patch creator wrote:
4361 "I found a little strange behaviour of ldap code when working with
4362 Windows 2003 AD Domain, where users was placed in more than one
4363 Organization Units. When I tried to give exim partial DN, the exit code
4364 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4365 But simultaneously result of request was absolutely normal ldap result,
4366 so I produce this patch..."
4368 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4369 REFERENCE, so I have modified the code to exclude the patch when that macro
4372 55. Some experimental protocols are using DNS PTR records for new purposes. The
4373 keys for these records are domain names, not reversed IP addresses. The
4374 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4375 leaves it alone. Component reversal etc. now happens only for IP addresses.
4378 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4380 57. Double the size of the debug message buffer (to 2048) so that more of very
4381 long debug lines gets shown.
4383 58. The exicyclog utility now does better if the number of log files to keep
4384 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4386 59. Two changes related to the smtp_active_hostname option:
4388 (1) $smtp_active_hostname is now available as a variable.
4389 (2) The default for smtp_banner uses $smtp_active_hostname instead
4390 of $primary_hostname.
4392 60. The host_aton() function is supposed to be passed a string that is known
4393 to be a valid IP address. However, in the case of IPv6 addresses, it was
4394 not checking this. This is a hostage to fortune. Exim now panics and dies
4395 if the condition is not met. A case was found where this could be provoked
4396 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4397 components; fortuitously, this particular loophole had already been fixed
4398 by change 4.50/55 above.
4400 If there are any other similar loopholes, the new check in host_aton()
4401 itself should stop them being exploited. The report I received stated that
4402 data on the command line could provoke the exploit when Exim was running as
4403 exim, but did not say which command line option was involved. All I could
4404 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4405 running as the user.
4408 61. There was a buffer overflow vulnerability in the SPA authentication code
4409 (which came originally from the Samba project). I have added a test to the
4410 spa_base64_to_bits() function which I hope fixes it.
4413 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4414 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4416 63. The daemon start-up calls getloadavg() while still root for those OS that
4417 need the first call to be done as root, but it missed one case: when
4418 deliver_queue_load_max is set with deliver_drop_privilege. This is
4419 necessary for the benefit of the queue runner, because there is no re-exec
4420 when deliver_drop_privilege is set.
4422 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4423 This has been fixed.
4425 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4426 was in use, was not putting the data itself into the right store pool;
4427 consequently, it could be overwritten for a subsequent message in the same
4428 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4431 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4433 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4434 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4435 the function were treating the return as a boolean value, which happened to
4436 work because 0=false and not-0=true, but is not correct code.
4438 68. The host_aton() function was not handling scoped IPv6 addresses (those
4439 with, for example, "%eth0" on the end) correctly.
4441 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4442 negated items (that is, ~something) in unsigned ints. Some compilers
4443 apparently mutter when there is no cast.
4445 70. If an address verification called from an ACL failed, and did not produce a
4446 user-specific message (i.e. there was only a "system" message), nothing was
4447 put in $acl_verify_message. In this situation, it now puts the system
4450 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4451 session; change 4.31/43 added the unwanted input to the log line - except
4452 that it did not do this in the start of session case. It now does.
4454 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4455 This is wrong and can cause the other end to generate a synchronization
4456 error if it is another Exim or anything else that does the synchronization
4457 check. A QUIT command is no longer sent after a timeout.
4459 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4460 during host lookups.
4462 74. The defer_ok option of callout verification was not working if it was used
4463 when verifying addresses in header lines, that is, for this case:
4465 verify = header_sender/callout=defer_ok
4467 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4468 those file descriptors could be used for SMTP connections. If anything
4469 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4470 could be sent to the SMTP client, causing chaos. The daemon now opens
4471 stdin, stdout, and stderr to /dev/null when it puts itself into the
4474 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4475 log by default. The user can override this with suitable Perl magic.
4477 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4478 the log message when discard triggers, was not working for the DATA ACL or
4479 for the non-SMTP ACL.
4481 78. Error message wording change in sieve.c.
4483 79. If smtp_accept_max_per_host was set, the number of connections could be
4484 restricted to fewer than expected, because the daemon was trying to set up
4485 a new connection before checking whether the processes handling previous
4486 connections had finished. The check for completed processes is now done
4487 earlier. On busy systems, this bug wouldn't be noticed because something
4488 else would have woken the daemon, and it would have reaped the completed
4491 80. If a message was submitted locally by a user whose login name contained one
4492 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4493 It caused a spool format error. I have fixed the spool reading code. A
4494 related problem was that the "from" clause in the Received: line became
4495 illegal because of the space(s). It is now covered by ${quote_local_part.
4497 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4500 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4502 83. Patch from Sieve maintainer to fix unterminated string problem in
4503 "vacation" handling.
4505 84. Some minor changes to the Linux configuration files to help with other
4506 OS variants using glibc.
4508 85. One more patch for Sieve to update vacation handling to latest spec.
4511 ----------------------------------------------------
4512 See the note above about the 4.44 and 4.50 releases.
4513 ----------------------------------------------------
4519 1. Change 4.43/35 introduced a bug that caused file counts to be
4520 incorrectly computed when quota_filecount was set in an appendfile
4523 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4524 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4527 3. Give more explanation in the error message when the command for a transport
4528 filter fails to execute.
4530 4. There are several places where Exim runs a non-Exim command in a
4531 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4532 was being done only for the command run by the queryprogram router. It is
4533 now done for all such subprocesses. The other cases are: ${run, transport
4534 filters, and the commands run by the lmtp and pipe transports.
4536 5. Some older OS have a limit of 256 on the maximum number of file
4537 descriptors. Exim was using setrlimit() to set 1000 as a large value
4538 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4539 systems. I've change it so that if it can't get 1000, it tries for 256.
4541 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4542 was an oversight, and furthermore, ever since the addition of extra
4543 controls (e.g. 4.43/32), the checks on when to allow different forms of
4544 "control" were broken. There should now be diagnostics for all cases when a
4545 control that does not make sense is encountered.
4547 7. $recipients is now available in the predata ACL (oversight).
4549 8. Tidy the search cache before the fork to do a delivery from a message
4550 received from the command line. Otherwise the child will trigger a lookup
4551 failure and thereby defer the delivery if it tries to use (for example) a
4552 cached ldap connection that the parent has called unbind on.
4554 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4555 of $address_data from the recipient verification was clobbered by the
4556 sender verification.
4558 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4559 was its contents. (It was OK if the option was not defined at all.)
4561 11. A "Completed" log line is now written for messages that are removed from
4562 the spool by the -Mrm option.
4564 12. $host_address is now set to the target address during the checking of
4565 ignore_target_hosts.
4567 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4568 being passed; this would have caused $sender_host_name to have been used if
4569 matching the list had actually called for a host name (not very likely,
4570 since this list is usually IP addresses). A host name is now passed as
4573 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4574 code that specifies a non-restarting handler (typically sigaction() in
4575 modern systems) in an attempt to fix a rare and obscure crash bug.
4577 15. Narrowed the window for a race in the daemon that could cause it to ignore
4578 SIGCHLD signals. This is not a major problem, because they are used only to
4579 wake it up if nothing else does.
4581 16. A malformed maildirsize file could cause Exim to calculate negative values
4582 for the mailbox size or file count. Odd effects could occur as a result.
4583 The maildirsize information is now recalculated if the size or filecount
4586 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4587 support for a long time. Removed HAVE_SYS_VFS_H.
4589 18. Updated exipick to current release from John Jetmore.
4591 19. Allow an empty sender to be matched against a lookup in an address list.
4592 Previously the only cases considered were a regular expression, or an
4595 20. Exim went into a mad DNS lookup loop when doing a callout where the
4596 host was specified on the transport, if the DNS lookup yielded more than
4599 21. The RFC2047 encoding function was originally intended for short strings
4600 such as real names; it was not keeping to the 75-character limit for
4601 encoded words that the RFC imposes. It now respects the limit, and
4602 generates multiple encoded words if necessary. To be on the safe side, I
4603 have increased the buffer size for the ${rfc2047: expansion operator from
4606 22. Failure to deliver a bounce message always caused it to be frozen, even if
4607 there was an errors_to setting on the router. The errors_to setting is now
4610 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4611 canonical form (fully expanded) before being placed in
4612 $sender_host_address.
4614 24. Updated eximstats to version 1.33
4616 25. Include certificate and key file names in error message when GnuTLS fails
4617 to set them up, because the GnuTLS error message doesn't include the name
4618 of the failing file when there is a problem reading it.
4620 26. Expand error message when OpenSSL has problems setting up cert/key files.
4623 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4624 (this can affect the format of dates).
4626 28. exim_tidydb, when checking for the continued existence of a message for
4627 which it has found a message-specific retry record, was not finding
4628 messages that were in split spool directories. Consequently, it was
4629 deleting retry records that should have stayed in existence.
4631 29. eximstats updated to version 1.35
4632 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4633 1.35 - bugfix such that pie charts by volume are generated correctly
4635 30. The SPA authentication driver was not abandoning authentication and moving
4636 on to the next authenticator when an expansion was forced to fail,
4637 contradicting the general specification for all authenticators. Instead it
4638 was generating a temporary error. It now behaves as specified.
4640 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4641 (the order specifies the preference for clients). The order is now AES256,
4642 AES128, 3DES, ARCFOUR128.
4644 31. Small patch to Sieve code - explicitly set From: when generating an
4647 32. Exim crashed if a remote delivery caused a very long error message to be
4648 recorded - for instance if somebody sent an entire SpamAssassin report back
4649 as a large number of 550 error lines. This bug was coincidentally fixed by
4650 increasing the size of one of Exim's internal buffers (big_buffer) that
4651 happened as part of the Exiscan merge. However, to be on the safe side, I
4652 have made the code more robust (and fixed the comments that describe what
4655 33. Some experimental protocols are using DNS PTR records for new purposes. The
4656 keys for these records are domain names, not reversed IP addresses. The
4657 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4658 leaves it alone. Component reversal etc. now happens only for IP addresses.
4661 34. The host_aton() function is supposed to be passed a string that is known
4662 to be a valid IP address. However, in the case of IPv6 addresses, it was
4663 not checking this. This is a hostage to fortune. Exim now panics and dies
4664 if the condition is not met. A case was found where this could be provoked
4665 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4666 components; fortuitously, this particular loophole had already been fixed
4667 by change 4.50/55 or 4.44/33 above.
4669 If there are any other similar loopholes, the new check in host_aton()
4670 itself should stop them being exploited. The report I received stated that
4671 data on the command line could provoke the exploit when Exim was running as
4672 exim, but did not say which command line option was involved. All I could
4673 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4674 running as the user.
4677 35. There was a buffer overflow vulnerability in the SPA authentication code
4678 (which came originally from the Samba project). I have added a test to the
4679 spa_base64_to_bits() function which I hope fixes it.
4682 36. The daemon start-up calls getloadavg() while still root for those OS that
4683 need the first call to be done as root, but it missed one case: when
4684 deliver_queue_load_max is set with deliver_drop_privilege. This is
4685 necessary for the benefit of the queue runner, because there is no re-exec
4686 when deliver_drop_privilege is set.
4688 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4689 was in use, was not putting the data itself into the right store pool;
4690 consequently, it could be overwritten for a subsequent message in the same
4691 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4694 38. Sometimes the final signoff response after QUIT could fail to get
4695 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4696 < 0 before doing a fflush(). This bug looks as though it goes back to the
4697 introduction of TLS in release 3.20, but "sometimes" must have been rare
4698 because the tests only now provoked it.
4704 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4705 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4706 Now it returns an integer. A number of calls were still expecting a T/F
4707 return. Fortuitously, in all cases, the tests worked in OK situations,
4708 which is the norm. However, things would have gone wrong on any write
4709 failures on the smtp file descriptor. This function is used when sending
4710 messages over SMTP and also when doing verify callouts.
4712 2. When Exim is called to do synchronous delivery of a locally submitted
4713 message (the -odf or -odi options), it no longer closes stderr before doing
4716 3. Implemented the mua_wrapper option.
4718 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4720 5. Implemented the functions header_remove(), header_testname(),
4721 header_add_at_position(), and receive_remove_recipient(), and exported them
4724 6. If an ACL "warn" statement specified the addition of headers, Exim already
4725 inserted X-ACL-Warn: at the start if there was no header name. However, it
4726 was not making this test for the second and subsequent header lines if
4727 there were newlines in the string. This meant that an invalid header could
4728 be inserted if Exim was badly configured.
4730 7. Allow an ACL "warn" statement to add header lines at the start or after all
4731 the Received: headers, as well as at the end.
4733 8. Added the rcpt_4xx retry error code.
4735 9. Added postmaster_mailfrom=xxx to callout verification option.
4737 10. Added mailfrom=xxxx to the callout verification option, for verify=
4740 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4741 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4743 12. Inserted some casts to stop certain compilers warning when using pointer
4744 differences as field lengths or precisions in printf-type calls (mostly
4745 affecting debugging statements).
4747 13. Added optional readline() support for -be (dynamically loaded).
4749 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4750 same clock tick as a message's arrival, so that its received time was the
4751 same as the "first fail" time on the retry record, and that message
4752 remained on the queue past the ultimate address timeout, every queue runner
4753 would try a delivery (because it was past the ultimate address timeout) but
4754 after another failure, the ultimate address timeout, which should have then
4755 bounced the address, did not kick in. This was a "< instead of <=" error;
4756 in most cases the first failure would have been in the next clock tick
4757 after the received time, and all would be well.
4759 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4760 being recognized when the domain list was tested by the match_domain
4761 condition in an expansion string.
4763 16. Added the ${str2b64: operator.
4765 17. Exim was always calling setrlimit() to set a large limit for the number of
4766 processes, without checking whether the existing limit was already
4767 adequate. (It did check for the limit on file descriptors.) Furthermore,
4768 errors from getrlimit() and setrlimit() were being ignored. Now they are
4769 logged to the main and panic logs, but Exim does carry on, to try to do its
4770 job under whatever limits there are.
4772 18. Imported PCRE 5.0.
4774 19. Trivial typo in log message " temporarily refused connection" (the leading
4777 20. If the log selector return_path_on_delivery was set and an address was
4778 redirected to /dev/null, the delivery process crashed because it assumed
4779 that a return path would always be set for a "successful" delivery. In this
4780 case, the whole delivery is bypassed as an optimization, and therefore no
4783 21. Internal re-arrangement: the function for sending a challenge and reading
4784 a response while authentication was assuming a zero-terminated challenge
4785 string. It's now changed to take a pointer and a length, to allow for
4786 binary data in such strings.
4788 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4790 23. Exim was not respecting finduser_retries when seeking the login of the
4791 uid under which it was called; it was always trying 10 times. (The default
4792 setting of finduser_retries is zero.) Also, it was sleeping after the final
4793 failure, which is pointless.
4795 24. Implemented tls_on_connect_ports.
4797 25. Implemented acl_smtp_predata.
4799 26. If the domain in control=submission is set empty, Exim assumes that the
4800 authenticated id is a complete email address when it generates From: or
4801 Sender: header lines.
4803 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4804 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4805 chown and chgrp in /bin and hostname in /usr/bin.
4807 28. Exim was keeping the "process log" file open after each use, just as it
4808 does for the main log. This opens the possibility of it remaining open for
4809 long periods when the USR1 signal hits a daemon. Occasional processlog
4810 errors were reported, that could have been caused by this. Anyway, it seems
4811 much more sensible not to leave this file open at all, so that is what now
4814 29. The long-running daemon process does not normally write to the log once it
4815 has entered its main loop, and it closes the log before doing so. This is
4816 so that log files can straightforwardly be renamed and moved. However,
4817 there are a couple of unusual error situations where the daemon does write
4818 log entries, and I had neglected to close the log afterwards.
4820 30. The text of an SMTP error response that was received during a remote
4821 delivery was being truncated at 512 bytes. This is too short for some of
4822 the long messages that one sometimes sees. I've increased the limit to
4825 31. It is now possible to make retry rules that apply only when a message has a
4826 specific sender, in particular, an empty sender.
4828 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4829 it possible to be selective about when SMTP synchronization is enforced.
4831 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4833 32. Implemented hosts_connection_nolog.
4835 33. Added an ACL for QUIT.
4837 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4840 35. Added mailbox_size and mailbox_filecount to appendfile.
4842 36. Added control = no_multiline_responses to ACLs.
4844 37. There was a bug in the logic of the code that waits for the clock to tick
4845 in the case where the clock went backwards by a substantial amount such
4846 that the microsecond fraction of "now" was more than the microsecond
4847 fraction of "then" (but the whole seconds number was less).
4849 38. Added support for the libradius Radius client library this is found on
4850 FreeBSD (previously only the radiusclient library was supported).
4856 1. When certain lookups returned multiple values in the form name=value, the
4857 quoting of the values was not always being done properly. Specifically:
4858 (a) If the value started with a double quote, but contained no whitespace,
4860 (b) If the value contained whitespace other than a space character (i.e.
4861 tabs or newlines or carriage returns) it was not quoted.
4862 This fix has been applied to the mysql and pgsql lookups by writing a
4863 separate quoting function and calling it from the lookup code. The fix
4864 should probably also be applied to nisplus, ibase and oracle lookups, but
4865 since I cannot test any of those, I have not disturbed their existing code.
4867 2. A hit in the callout cache for a specific address caused a log line with no
4868 reason for rejecting RCPT. Now it says "Previous (cached) callout
4869 verification failure".
4871 3. There was an off-by-one bug in the queryprogram router. An over-long
4872 return line was truncated at 256 instead of 255 characters, thereby
4873 overflowing its buffer with the terminating zero. As well as fixing this, I
4874 have increased the buffer size to 1024 (and made a note to document this).
4876 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4877 when Exim is waiting for an SMTP response from a remote server, Exim
4878 restarts its select() call on the socket, thereby resetting its timeout.
4879 This is not a problem when such interrupts are rare. Somebody set up a cron
4880 job to run exiwhat every 2 minutes, which is less than the normal select()
4881 timeout (5 or 10 minutes). This meant that the select() timeout never
4882 kicked in because it was always reset. I have fixed this by comparing the
4883 time when an interrupt arrives with the time at the start of the first call
4884 to select(). If more time than the timeout has elapsed, the interrupt is
4885 treated as a timeout.
4887 5. Some internal re-factoring in preparation for the addition of Sieve
4888 extensions (by MH). In particular, the "personal" test is moved to a
4889 separate function, and given an option for scanning Cc: and Bcc: (which is
4890 not set for Exim filters).
4892 6. When Exim created an email address using the login of the caller as the
4893 local part (e.g. when creating a From: or Sender: header line), it was not
4894 quoting the local part when it contained special characters such as @.
4896 7. Installed new OpenBSD configuration files.
4898 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4899 try to make them clearer.
4901 9. Callout options, other than the timeout value, were being ignored when
4902 verifying sender addresses in header lines. For example, when using
4904 verify = header_sender/callout=no_cache
4906 the cache was (incorrectly) being used.
4908 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4909 only the Cygwin environment.
4911 11. When return_path_on_delivery was set as a log selector, if different remote
4912 addresses in the same message used different return paths and parallel
4913 remote delivery occurred, the wrong values would sometimes be logged.
4914 (Whenever a remote delivery process finished, the return path value from
4915 the most recently started remote delivery process was logged.)
4917 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4918 lines when AUTH and/or TLS are in use. This is the "received protocol"
4919 field. Exim used to use "asmtp" for authenticated SMTP, without any
4920 indication (in the protocol name) for TLS use. Now it follows the RFC and
4921 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4922 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4923 names appear in log lines as well as in Received: header lines.
4925 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4926 extensions, and comparison tests, and to fix some bugs.
4928 14. Changes to the "personal" filter test:
4930 (1) The test was buggy in that it was just doing the equivalent of
4931 "contains" tests on header lines. For example, if a user's address was
4932 anne@some.where, the "personal" test would incorrectly be true for
4934 To: susanne@some.where
4936 This test is now done by extracting each address from the header in turn,
4937 and checking the entire address. Other tests that are part of "personal"
4938 are now done using regular expressions (for example, to check local parts
4939 of addresses in From: header lines).
4941 (2) The list of non-personal local parts in From: addresses has been
4942 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4943 taken from the Sieve specification recommendations.
4945 (3) If the message contains any header line starting with "List-" it is
4946 treated as non-personal.
4948 (4) The test for "circular" in the Subject: header line has been removed
4949 because it now seems ill-conceived.
4951 15. Minor typos in src/EDITME comments corrected.
4953 16. Installed latest exipick from John Jetmore.
4955 17. If headers_add on a router specified a text string that was too long for
4956 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4957 of string_sprintf() is now avoided.
4959 18. $message_body_size was not set (it was always zero) when running the DATA
4960 ACL and the local_scan() function.
4962 19. For the "mail" command in an Exim filter, no default was being set for
4963 the once_repeat time, causing a random time value to be used if "once" was
4964 specified. (If the value happened to be <= 0, no repeat happened.) The
4965 default is now 0s, meaning "never repeat". The "vacation" command was OK
4966 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4967 (I found it when inspecting the code).
4969 20. There is now an overall timeout for performing a callout verification. It
4970 defaults to 4 times the callout timeout, which applies to individual SMTP
4971 commands during the callout. The overall timeout applies when there is more
4972 than one host that can be tried. The timeout is checked before trying the
4973 next host. This prevents very long delays if there are a large number of
4974 hosts and all are timing out (e.g. when the network connections are timing
4975 out). The value of the overall timeout can be changed by specifying an
4976 additional sub-option for "callout", called "maxwait". For example:
4978 verify = sender/callout=5s,maxwait=20s
4980 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4981 to the end before writing, but this should make it even safer).
4983 22. Exim was forgetting that it had advertised PIPELINING for the second and
4984 subsequent messages on an SMTP connection. It was also not resetting its
4985 memory on STARTTLS and an internal HELO.
4987 23. When Exim logs an SMTP synchronization error within a session, it now
4988 records whether PIPELINING has been advertised or not.
4990 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4991 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4992 rather than long int.
4994 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4996 26. Added the never_mail option to autoreply.
5002 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5003 crash if the getsockname() call failed; this can happen if a connection is
5004 closed very soon after it is established. The problem was simply in the
5005 order in which certain operations were done, causing Exim to try to write
5006 to the SMTP stream before it had set up the file descriptor. The bug has
5007 been fixed by making things happen in the correct order.
5013 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5014 before the connection was closed, thus losing the rejection response.
5016 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5017 some early Solaris releases, but causes trouble in current releases where
5018 socklen_t is defined.
5020 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5023 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5026 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5027 that releasing the top of it at the end releases what was used for sub-
5028 expansions (unless the block got too big). However, discard this block if
5029 the first thing is a variable or header, so that we can use its block when
5030 it is dynamic (useful for very large $message_headers, for example).
5032 6. Lookups now cache *every* query, not just the most recent. A new, separate
5033 store pool is used for this. It can be recovered when all lookup caches are
5034 flushed. Lookups now release memory at the end of their result strings.
5035 This has involved some general refactoring of the lookup sources.
5037 7. Some code has been added to the store_xxx() functions to reduce the amount
5038 of flapping under certain conditions.
5040 8. log_incoming_interface used to affect only the <= reception log lines. Now
5041 it causes the local interface and port to be added to several more SMTP log
5042 lines, for example "SMTP connection from", and rejection lines.
5044 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5046 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5048 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5049 could be overwritten at the end of the current message (or the start of a
5050 new message if it was set in a HELO ACL). The value is now preserved for
5051 the duration of the SMTP connection.
5053 12. If a transport had a headers_rewrite setting, and a matching header line
5054 contained an unqualified address, that address was qualified, even if it
5055 did not match any rewriting rules. The underlying bug was that the values
5056 of the flags that permit the existence of unqualified sender and recipient
5057 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5058 non-local messages, and by -bnq for local messages) were not being
5059 preserved with the message after it was received.
5061 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5062 "next input=" as part of the text comprising the host identity instead of
5063 the correct text. The code was using the same buffer for two different
5064 strings. However, depending on which order the printing function evaluated
5065 its arguments, the bug did not always show up. Under Linux, for example, my
5066 test suite worked just fine.
5068 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5069 doesn't work with some older versions of Perl. It has been changed to "my",
5070 which in any case is probably the better facility to use.
5072 15. A really picky compiler found some instances of statements for creating
5073 error messages that either had too many or two few arguments for the format
5076 16. The size of the buffer for calls to the DNS resolver has been increased
5077 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5078 for addresses that have a lot of PTR records. This alleviates a problem; it
5079 does not fully solve it.
5081 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5082 buffer now truncates the list and logs the incident, which is the same
5083 action as happens when Exim is looking up a host name and its aliases.
5084 Previously in this situation something unpredictable would happen;
5085 sometimes it was "internal error: store_reset failed".
5087 18. If a server dropped the connection unexpectedly when an Exim client was
5088 using GnuTLS and trying to read a response, the client delivery process
5089 crashed while trying to generate an error log message.
5091 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5092 string, for example:
5094 warn message = H1: something\nH2: something
5096 the text was added as a single header line from Exim's point of view
5097 though it ended up OK in the delivered message. However, searching for the
5098 second and subsequent header lines using $h_h2: did not work. This has been
5099 fixed. Similarly, if a system filter added multiple headers in this way,
5100 the routers could not see them.
5102 20. Expanded the error message when iplsearch is called with an invalid key to
5103 suggest using net-iplsearch in a host list.
5105 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5106 ACLs are no longer actually imposed (and a message to that effect is
5109 22. If a "gecos" field in a passwd entry contained escaped characters, in
5110 particular, if it contained a \" sequence, Exim got it wrong when building
5111 a From: or a Sender: header from that name. A second bug also caused
5112 incorrect handling when an unquoted " was present following a character
5113 that needed quoting.
5115 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5116 was not being matched caselessly.
5118 24. Arranged for all hyphens in the exim.8 source to be escaped with
5121 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5122 properly. Recipient callout cache records were still being keyed to include
5123 the sender, even when use_sender was set false. This led to far more
5124 callouts that were necessary. The sender is no longer included in the key
5125 when use_sender is false.
5127 26. Added "control = submission" modifier to ACLs.
5129 27. Added the ${base62d: operator to decode base 62 numbers.
5131 28. dnsdb lookups can now access SRV records.
5133 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5134 the configuration file.
5136 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5137 (-v) mode. This makes the output for a verbose queue run more intelligible.
5139 31. Added a use_postmaster feature to recipient callouts.
5141 32. Added the $body_zerocount variable, containing the number of binary zero
5142 bytes in the message body.
5144 33. The time of last modification of the "new" subdirectory is now used as the
5145 "mailbox time last read" when there is a quota error for a maildir
5148 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5150 35. Added +ignore_unknown as a special item in host lists.
5152 36. Code for decoding IPv6 addresses in host lists is now included, even if
5153 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5154 address was recognized as an IP address, but was then not correctly decoded
5155 into binary, causing unexpected and incorrect effects when compared with
5162 1. Very minor rewording of debugging text in manualroute to say "list of
5163 hosts" instead of "hostlist".
5165 2. If verify=header_syntax was set, and a header line with an unqualified
5166 address (no domain) and a large number of spaces between the end of the
5167 name and the colon was received, the reception process suffered a buffer
5168 overflow, and (when I tested it) crashed. This was caused by some obsolete
5169 code that should have been removed. The fix is to remove it!
5171 3. When running in the test harness, delay a bit after writing a bounce
5172 message to get a bit more predictability in the log output.
5174 4. Added a call to search_tidyup() just before forking a reception process. In
5175 theory, someone could use a lookup in the expansion of smtp_accept_max_
5176 per_host which, without the tidyup, could leave open a database connection.
5178 5. Added the variables $recipient_data and $sender_data which get set from a
5179 lookup success in an ACL "recipients" or "senders" condition, or a router
5180 "senders" option, similar to $domain_data and $local_part_data.
5182 6. Moved the writing of debug_print from before to after the "senders" test
5185 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5186 problems for message scanning, either using a data ACL, or using
5187 local_scan() because the Received: header was not generated till after they
5188 were called (in order to set the time as the time of reception completion).
5189 I have revised the way this works. The header is now generated after the
5190 body is received, but before the ACL or local_scan() are called. After they
5191 are run, the timestamp in the header is updated.
5197 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5198 before starting a queue runner without re-exec. This happened only when
5199 deliver_drop_privilege was set or when the Exim user was set to root. The
5200 effect of the bug was that timeouts during subsequent deliveries caused
5201 crashes instead of being properly handled. The handler is now left at its
5202 default (and expected) setting.
5204 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5205 message, again when deliver_drop_privilege is set or Exim is run as root.
5206 The bug described in (1) was not present in this case, but the tidying up
5207 of the other signals was missing. I have made the two cases consistent.
5209 3. The ignore_target_hosts setting on a manualroute router was being ignored
5210 for hosts that were looked up using the /MX notation.
5212 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5215 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5216 operated on the sender address. After changing the $sender_address to <>
5217 for the sender address verify, Exim was re-instated it as the original
5218 (before rewriting) address, but remembering that it had rewritten it, so it
5219 wasn't rewriting it again. This bug also had the effect of breaking the
5220 sender address verification caching when the sender address was rewritten.
5222 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5223 This has been changed so that if the ip literal address matches
5224 ignore_target_hosts, the router declines.
5226 7. Added expansion conditions match_domain, match_address, and match_local_
5227 part (NOT match_host).
5229 8. The placeholder for the Received: header didn't have a length field set.
5231 9. Added code to Exim itself and to exim_lock to test for a specific race
5232 condition that could lead to file corruption when using MBX delivery. The
5233 issue is with the lockfile that is created in /tmp. If this file is removed
5234 after a process has opened it but before that process has acquired a lock,
5235 there is the potential for a second process to recreate the file and also
5236 acquire a lock. This could lead to two Exim processes writing to the file
5237 at the same time. The added code performs the same test as UW imapd; it
5238 checks after acquiring the lock that its file descriptor still refers to
5239 the same named file.
5241 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5242 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5245 11. Added the smtp_active_hostname option. If used, this will typically be made
5246 to depend on the incoming interface address. Because $interface_address is
5247 not set up until the daemon has forked a reception process, error responses
5248 that can happen earlier (such as "too many connections") no longer contain
5251 12. If an expansion in a condition on a "warn" statement fails because a lookup
5252 defers, the "warn" statement is abandoned, and the next ACL statement is
5253 processed. Previously this caused the whole ACL to be aborted.
5255 13. Added the iplsearch lookup type.
5257 14. Added ident_timeout as a log selector.
5259 15. Added tls_certificate_verified as a log selector.
5261 16. Added a global option tls_require_ciphers (compare the smtp transport
5262 option of the same name). This controls incoming TLS connections.
5264 17. I finally figured out how to make tls_require_ciphers do a similar thing
5265 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5266 before starting the TLS session.
5268 18. Tabs are now shown as \t in -bP output.
5270 19. If the log selector return_path_on_delivery was set, Exim crashed when
5271 bouncing a message because it had too many Received: header lines.
5273 20. If two routers both had headers_remove settings, and the first one included
5274 a superfluous trailing colon, the final name in the first list and the
5275 first name in the second list were incorrectly joined into one item (with a
5276 colon in the middle).
5282 1. Added -C and -D options to the exinext utility, mainly to make it easier
5283 to include in the automated testing, but these could be helpful when
5284 multiple configurations are in use.
5286 2. The exinext utility was not formatting the output nicely when there was
5287 an alternate port involved in the retry record key, nor when there was a
5288 message id as well (for retries that were specific to a specific message
5289 and a specific host). It was also confused by IPv6 addresses, because of
5290 the additional colons they contain. I have fixed the IPv4 problem, and
5291 patched it up to do a reasonable job for IPv6.
5293 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5294 delivery, the log line now contains "pipelined" if PIPELINING was used.
5296 4. An SMTP transport process used to panic and die if the bind() call to set
5297 an explicit outgoing interface failed. This has been changed; it is now
5298 treated in the same way as a connect() failure.
5300 5. A reference to $sender_host_name in the part of a conditional expansion
5301 that was being skipped was still causing a DNS lookup. This no longer
5304 6. The def: expansion condition was not recognizing references to header lines
5305 that used bh_ and bheader_.
5307 7. Added the _cache feature to named lists.
5309 8. The code for checking quota_filecount in the appendfile transport was
5310 allowing one more file than it should have been.
5312 9. For compatibility with Sendmail, the command line option
5320 and sets the incoming protocol and host name (for trusted callers). The
5321 host name and its colon can be omitted when only the protocol is to be set.
5322 Note the Exim already has two private options, -pd and -ps, that refer to
5323 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5324 "s", but I don't think that's a major issue.
5326 10. A number of refactoring changes to the code, none of which should affect
5329 (a) The number of logging options was getting close to filling up the
5330 32-bit word that was used as a bit map. I have split them into two classes:
5331 those that are passed in the argument to log_write(), and those that are
5332 only ever tested independently outside of that function. These are now in
5333 separate 32-bit words, so there is plenty of room for expansion again.
5334 There is no change in the user interface or the logging behaviour.
5336 (b) When building, for example, log lines, the code previously used a
5337 macro that called string_cat() twice, in order to add two strings. This is
5338 not really sufficiently general. Furthermore, there was one instance where
5339 it was actually wrong because one of the argument was used twice, and in
5340 one call a function was used. (As it happened, calling the function twice
5341 did not affect the overall behaviour.) The macro has been replaced by a
5342 function that can join an arbitrary number of extra strings onto a growing
5345 (c) The code for expansion conditions now uses a table and a binary chop
5346 instead of a serial search (which was left over from when there were very
5347 few conditions). Also, it now recognizes conditions like "pam" even when
5348 the relevant support is not compiled in: a suitably worded error message is
5349 given if an attempt is made to use such a condition.
5351 11. Added ${time_interval:xxxxx}.
5353 12. A bug was causing one of the ddress fields not to be passed back correctly
5354 from remote delivery subprocesses. The field in question was not being
5355 subsequently used, so this caused to problems in practice.
5357 13. Added new log selectors queue_time and deliver_time.
5359 14. Might have fixed a bug in maildirsizefile handling that threw up
5360 "unexpected character" debug warnings, and recalculated the data
5361 unnecessarily. In any case, I expanded the warning message to give more
5364 15. Added the message "Restricted characters in address" to the statements in
5365 the default ACL that block characters like @ and % in local parts.
5367 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5368 Three changes have been made:
5370 (a) There was a serious bug; a negative response to MAIL caused the whole
5371 recipient domain to be cached as invalid, thereby blocking all messages
5372 to all local parts at the same domain, from all senders. This bug has
5373 been fixed. The domain is no longer cached after a negative response to
5374 MAIL if the sender used is not empty.
5376 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5379 (c) A new callout option, "use_sender" has been added for people who want
5380 the modified behaviour.
5386 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5389 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5390 indeed breaks things for older releases.
5392 3. Added additional logging to the case where there is a problem reading data
5393 from a filter that is running in a subprocess using a pipe, in order to
5394 try to track down a specific problem.
5396 4. Testing facility fudge: when running in the test harness and attempting
5397 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5398 getting "No route to host". Convert this to a timeout.
5400 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5403 6. Some OS don't have socklen_t but use size_t instead. This affects the
5404 fifth argument of getsockopt() amongst other things. This is now
5405 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5406 can be set for individual OS. I have set it for SunOS5, OSF1, and
5407 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5408 some earlier ones do not.
5410 7. Change 4.30/15 was not doing the test caselessly.
5412 8. The standard form for an IPv6 address literal was being rejected by address
5413 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5414 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5415 this, as well as the form without the "IPv6" on the front (but only when
5416 address literals are enabled, of course).
5418 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5420 10. Exim crashed if a message with an empty sender address specified by -f
5421 encountered a router with an errors_to setting. This could be provoked only
5422 by a command such as
5426 where an empty string was supplied; "<>" did not hit this bug.
5428 11. Installed PCRE release 4.5.
5430 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5431 remained set. It is now erased.
5433 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5434 times from message ids (which are base 36 rather than the normal 62).
5436 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5437 were being counted as actual protocol errors, and logged if the log
5438 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5439 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5440 and DATA following a set of rejected RCPTs do not count as protocol errors.
5441 In other words, Exim assumes they were pipelined, though this may not
5442 actually be the case. Of course, in all cases the client gets an
5443 appropriate error code.
5445 15. If a lookup fails in an ACL condition, a message about the failure may
5446 be available; it is used if testing the ACL cannot continue, because most
5447 such messages specify what the cause of the deferral is. However, some
5448 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5449 that caused an old message to be retained and used if a later statement
5450 caused a defer, replacing the real cause of the deferral.
5452 16. If an IP address had so many PTR records that the DNS lookup buffer
5453 was not large enough to hold them, Exim could crash while trying to process
5454 the truncated data. It now detects and logs this case.
5456 17. Further to 4.21/58, another change has been made: if (and only if) the
5457 first line of a message (the first header line) ends with CRLF, a bare LF
5458 in a subsequent header line has a space inserted after it, so as not to
5459 terminate the header.
5461 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5462 unnecessarily, used atoi() instead of strtol(), and didn't check the
5463 termination when getting file sizes from file names by regex.
5465 19. Completely re-implemented the support for maildirsize files, in the light
5466 of a number of problems with the previous contributed implementation
5467 (4.30/29). In particular:
5469 . If the quota is zero, the maildirsize file is maintained, but no quota is
5472 . If the maildir directory does not exist, it is created before any attempt
5473 to write a maildirsize file.
5475 . The quota value in the file is just a cache; if the quota is changed in
5476 the transport, the new value overrides.
5478 . A regular expression is available for excluding directories from the
5481 20. The autoreply transport checks the characters in options that define the
5482 message's headers; it allows continued headers, but it was checking with
5483 isspace() after an embedded newline instead of explicitly looking for a
5486 21. If all the "regular" hosts to which an address was routed had passed their
5487 expiry times, and had not reached their retry times, the address was
5488 bounced, even if fallback hosts were defined. Now Exim should go on to try
5491 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5492 equivalent code in the SMTP transport. Some hosts send humungous responses
5493 to HELO/EHLO, more than 1024 it seems.
5495 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5496 gives compiler warnings in some environments. I've now done it "properly",
5499 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5500 (because of problems with the built-in one) was declared to return uschar *
5501 instead of char *, causing compiler failure.
5503 25. Fixed a file descriptor leak when processing alias/forward files.
5505 26. Fixed a minor format string issue in dbfn.c.
5507 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5509 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5510 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5513 29. When a sender address is verified, it is cached, to save repeating the test
5514 when there is more than one recipient in a message. However, when the
5515 verification involves a callout, it is possible for different callout
5516 options to be set for different recipients. It is too complicated to keep
5517 track of this in the cache, so now Exim always runs a verification when a
5518 callout is required, relying on the callout cache for the optimization.
5519 The overhead is duplication of the address routing, but this should not be
5522 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5523 to be verified with callout=postmaster, and the main callout worked but the
5524 postmaster check failed, the verification correctly failed. However, if a
5525 subsequent RCPT command asked for sender verification *without* the
5526 postmaster check, incorrect caching caused this verification also to fail,
5529 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5530 it was not caching the DNS options (qualify_single, search_parents) that
5531 were used when the lookup failed. A subsequent lookup with different
5532 options therefore always gave the same answer, though there were cases
5533 where it should not have. (Example: a "domains = !$mx_any" option on a
5534 dnslookup router: the "domains" option is always processed without any
5535 widening, but the router might have qualify_single set.) Now Exim uses the
5536 cached value only when the same options are set.
5538 32. Added John Jetmore's "exipick" utility to the distribution.
5540 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5541 than a timeout (e.g. a certificate is required, and is not provided), an
5542 Exim server now closes the connection immediately. Previously it waited for
5543 the client to close - but if the client is SSL, it seems that they each
5544 wait for each other, leading to a delay before one of them times out.
5546 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5547 maintained 0.8.x compatibility because I don't think many are using it, and
5548 it is clearly obsolete.
5550 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5553 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5554 client certificate was expired. A simple patch fixes this, though I don't
5555 understand the full logic of why the verify callback is called multiple
5558 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5559 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5560 which causes problems with some clients (such as the Certicom SSL Plus
5561 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5562 disables the coutermeasure allowing Eudora to connect."
5564 38. Exim was not checking that a write() to a log file succeeded. This could
5565 lead to Bad Things if a log got too big, in particular if it hit a file
5566 size limit. Exim now panics and dies if it cannot write to a log file, just
5567 as it does if it cannot open a log file.
5569 39. Modified OS/Makefile-Linux so that it now contains
5571 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5573 The two -D definitions ensure that Exim is compiled with large file
5574 support, which makes it possible to handle log files that are bigger than
5577 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5578 instance) a domain was checked against a named list that involved a lookup,
5579 causing $domain_data to be set, then another domain was checked against the
5580 same list, then the first domain was re-checked, the value of $domain_data
5581 after the final check could be wrong. In particular, if the second check
5582 failed, it could be set empty. This bug probably also applied to
5585 41. The strip_trailing_dot option was not being applied to the address given
5586 with the -f command-line option.
5588 42. The code for reading a message's header from the spool was incrementing
5589 $received_count, but never initializing it. This meant that the value was
5590 incorrect (doubled) while delivering a message in the same process in which
5591 it was received. In the most common configuration of Exim, this never
5592 happens - a fresh exec is done - but it can happen when
5593 deliver_drop_privilege is set.
5595 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5596 it now includes up to 150 characters of the unexpected data in the log
5599 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5600 and building data strings. The size of both of these buffers was 10 000
5601 bytes - far larger than anybody would *ever* want, thought I. Needless to
5602 say, somebody hit the limit. I have increased the maximum line length to
5603 20 000 and the maximum data length of concatenated lines to 100 000. I have
5604 also fixed two bugs, because there was no checking on these buffers. Tsk,
5605 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5606 buffer is too small.
5608 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5609 lsearch lookups. Now it does.
5611 46. When parsing a route_list item in a manualroute router, a fixed-length
5612 buffer was used for the list of hosts. I made this 1024 bytes long,
5613 thinking that nobody would ever have a list of hosts that long. Wrong.
5614 Somebody had a whole pile of complicated expansion conditions, and the
5615 string was silently truncated, leading to an expansion error. It turns out
5616 that it is easier to change to an unlimited length (owing to other changes
5617 that have happened since this code was originally written) than to build
5618 structure for giving a limitation error. The length of the item that
5619 expands into the list of hosts is now unlimited.
5621 47. The lsearch lookup could not handle data where the length of text line was
5622 more than 4095 characters. Such lines were truncated, leading to shortened
5623 data being returned. It should now handle lines of any length.
5625 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5626 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5629 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5630 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5631 no longer happens in comment lines. A list of replacements is now placed
5632 at the head of all of the source files, except those whose only change is
5633 to replace PERL_COMMAND in the very first #! line.
5635 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5636 messages on the queue, with a bottom-up merge sort, using code contributed
5637 by Michael Haardt. This should make operations like -bp somewhat faster on
5638 large queues. It won't affect queue runners, except when queue_run_in_order
5641 51. Installed eximstats 1.31 in the distribution.
5643 52. Added support for SRV lookups to the dnslookup router.
5645 53. If an ACL referred to $message_body or $message_body_end, the value was not
5646 reset for any messages that followed in the same SMTP session.
5648 54. The store-handling optimization for building very long strings was not
5649 differentiating between the different store pools. I don't think this
5650 actually made any difference in practice, but I've tidied it.
5652 55. While running the routers to verify a sender address, $sender_address
5653 was still set to the sender address. This is wrong, because when routing to
5654 send a bounce to the sender, it would be empty. Therefore, I have changed
5655 it so that, while verifying a sender address, $sender_address is set to <>.
5656 (There is no change to what happens when verifying a recipient address.)
5658 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5659 target A or AAAA records (if not already returned) without resetting the
5660 qualify_single or search_parents options of the DNS resolver. These are
5661 inappropriate in this case because the targets of MX and SRV records must
5662 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5663 target that, when qualified, matched something in the local domain. These
5664 two options are now turned off when doing these lookups.
5666 57. It seems that at least some releases of Reiserfs (which does not have the
5667 concept of a fixed number of inodes) returns zero and not -1 for the
5668 number of available inodes. This interacted badly with check_spool_inodes,
5669 which assumed that -1 was the "no such thing" setting. What I have done is
5670 to check that the total number of inodes is greater than zero before doing
5671 the test of how many are available.
5673 58. When a "warn" ACL statement has a log_message modifier, the message is
5674 remembered, and not repeated. This is to avoid a lot of repetition when a
5675 message has many recipients that cause the same warning to be written.
5676 However, Exim was preserving the list of already written lines for an
5677 entire SMTP session, which doesn't seem right. The memory is now reset if a
5678 new message is started.
5680 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5681 debugging output unless log_rewrite was also set.
5683 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5684 of (int)(handle) when we know that handle contains (void *)(-1).
5686 61. The Exim daemon panic-logs an error return when it closes the incoming
5687 connection. However "connection reset by peer" seems to be common, and
5688 isn't really an error worthy of noting specially, so that particular error
5691 62. When Exim is trying to find all the local interfaces, it used to panic and
5692 die if the ioctl to get the interface flags failed. However, it seems that
5693 on at least one OS (Solaris 9) it is possible to have an interface that is
5694 included in the list of interfaces, but for which you get a failure error
5695 for this call. This happens when the interface is not "plumbed" into a
5696 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5697 failure of the "get flags" call assumes that the interface is down.
5699 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5700 makes life easier for people who are doing arithmetic on fields extracted
5701 from dates, where you often get leading zeros that should not be
5702 interpreted as octal.
5704 64. Added qualify_domain to the redirect router, to override the global
5707 65. If a pathologically long header line contained very many addresses (the
5708 report of this problem mentioned 10 000) and each of them was rewritten,
5709 Exim could use up a very large amount of memory. (It kept on making new
5710 copies of the header line as it rewrote, and never released the old ones.)
5711 At the expense of a bit more processing, the header rewriting function has
5712 been changed so that it no longer eats memory in this way.
5714 66. The generation of the Received: header has been moved from the time that a
5715 message starts to be received, to the time that it finishes. The timestamp
5716 in the Received: header should now be very close to that of the <= log
5717 line. There are two side-effects of this change:
5719 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5720 logged header lines no longer include the local Received: line, because
5721 it has not yet been created. The same applies to a copy of the message
5722 that is returned to a non-SMTP sender when a message is rejected.
5724 (b) When a filter file is tested using -bf, no additional Received: header
5725 is added to the test message. After some thought, I decided that this
5728 This change does not affect the value of $received_for. It is still set
5729 after address rewriting, but before local_scan() is called.
5731 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5733 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5734 gave an unhelpful panic error message, and a defer error. I have managed to
5735 change this behaviour so that it now rejects any supplied certificate,
5736 which seems right, as the list of acceptable certificates is empty.
5738 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5739 gave an unhelpful defer error. I have not managed to make this reject any
5740 supplied certificates, but the error message it gives is "no certificate
5741 supplied", which is not helpful.
5743 70. exigrep's output now also includes lines that are not associated with any
5744 message, but which match the given pattern. Implemented by a patch from
5745 Martin Sluka, which also tidied up the Perl a bit.
5747 71. Recipient callout verification, like sender verification, was using <> in
5748 the MAIL FROM command. This isn't really the right thing, since the actual
5749 sender may affect whether the remote host accepts the recipient or not. I
5750 have changed it to use the actual sender in the callout; this means that
5751 the cache record is now keyed on a recipient/sender pair, not just the
5752 recipient address. There doesn't seem to be a real danger of callout loops,
5753 since a callout by the remote host to check the sender would use <>.
5754 [SEE ABOVE: changed after hitting problems.]
5756 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5757 temporary errors. However, in the case of such a code being given after
5758 the end of a data transmission (i.e. after ".") Exim was failing to write
5759 a retry record for the message. (Yes, there was some broken host that was
5760 actually sending 8xx at this point.)
5762 73. An unknown lookup type in a host list could cause Exim to panic-die when
5763 the list was checked. (An example that provoked this was putting <; in the
5764 middle of a list instead of at the start.) If this happened during a DATA
5765 ACL check, a -D file could be left lying around. This kind of configuration
5766 error no longer causes Exim to die; instead it causes a defer error. The
5767 incident is still logged to the main and panic logs.
5769 74. Buglet left over from Exim 3 conversion. The message "too many messages
5770 in one connection" was written to the rejectlog but not the mainlog, except
5771 when address rewriting (yes!) was being logged.
5773 75. Added write_rejectlog option.
5775 76. When a system filter was run not as root (that is, when system_filter_user
5776 was set), the values of the $n variables were not being returned to the
5777 main process; thus, they were not subsequently available in the $sn
5780 77. Added +return_path_on_delivery log selector.
5782 78. A connection timeout was being treated differently from recipients deferred
5783 when testing hosts_max_try with a message that was older than the host's
5784 retry timeout. (The host should not be counted, thus allowing all hosts to
5785 be tried at least once before bouncing.) This may have been the cause of an
5786 occasionally reported bug whereby a message would remain on the queue
5787 longer than the retry timeout, but would be bounced if a delivery was
5788 forced. I say "may" because I never totally pinned down the problem;
5789 setting up timeout/retry tests is difficult. See also the next item.
5791 79. The ultimate address timeout was not being applied to errors that involved
5792 a combination of host plus message (for example, a timeout on a MAIL
5793 command). When an address resolved to a number of possible hosts, and they
5794 were not all tried for each delivery (e.g. because of hosts_max_try), a
5795 message could remain on the queue longer than the retry timeout.
5797 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5800 81. Fixed an obscure SMTP outgoing bug which required at least the following
5801 conditions: (a) there was another message waiting for the same server;
5802 (b) the server returned 5xx to all RCPT commands in the first message so
5803 that the message was not completed; (c) the server dropped the connection
5804 or gave a negative response to the RSET that Exim sends to abort the
5805 transaction. The observed case was a dropped connection after DATA that had
5806 been sent in pipelining mode. That is, the server had advertised PIPELINING
5807 but was not implementing it correctly. The effect of the bug was incorrect
5808 behaviour, such as trying another host, and this could lead to a crash.
5814 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5815 and daemon.c were passed as pointers to ints; they should have been
5816 pointers to socklen_t variables (which are typically unsigned ints).
5818 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5821 3. Fixed a really odd bug that affected only the testing scheme; patching a
5822 certain fixed string in the binary changed the value of another string that
5823 happened to be identical to the end of the original first string.
5825 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5826 name", it returns that address as the IP address. On some operating
5827 systems (e.g. Solaris), it also passes back the IP address string as the
5828 "host name". However, on others (e.g. Linux), it passes back an empty
5829 string. Exim wasn't checking for this, and was changing the host name to an
5830 empty string, assuming it had been canonicalized.
5832 5. Although rare, it is permitted to have more than one PTR record for a given
5833 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5834 all the names associated with an address, because they do in Solaris.
5835 However, it seems that they do not in Linux for data that comes from the
5836 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5837 I found this out when I moved to a new Linux workstation and tried to run
5838 the Exim test suite.
5840 To get round this problem I have changed the code so that it now does its
5841 own call to the DNS to look up PTR records when searching for a host name.
5842 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5843 addresses that are only in /etc/hosts are still found.
5845 This behaviour is, however, controlled by an option called host_lookup_
5846 order, which defaults to "bydns:byaddr". If people want to use the other
5847 order, or indeed, just use one or the other means of lookup, they can
5848 specify it in this variable.
5850 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5851 some operating systems, this comes back from gethostbyaddr() as an empty
5852 string, and this is what Exim used to test for. However, it seems that in
5853 other systems, "." is yielded. Exim now tests for this case too.
5855 7. The values of check_spool_space and check_log_space are now held internally
5856 as a number of kilobytes instead of an absolute number of bytes. If a
5857 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5858 kilobyte. This means that much larger values can be stored.
5860 8. Exim monitor: an attempt to get the action menu when not actually pointing
5861 at a message produces an empty menu entitled "No message selected". This
5862 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5863 no entries in it ("Shell widget menu has zero width and/or height"). So I
5864 have added a single, blank menu entry in this case.
5866 9. Added ${quote_local_part.
5868 10. MIME decoding is now applied to the contents of Subject: header lines when
5871 11. Now that a reference to $sender_host_address automatically causes a reverse
5872 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5873 host lookup before query-style lookups in lists that might use this
5874 variable. This has therefore been abolished, and the "net-" prefix is no
5875 longer necessary for query-style lookups.
5877 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5878 have been a typo for LFLAGS, so it has been changed.
5880 13. The install script calls Exim with "-C /dev/null" in order to find the
5881 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5882 to be output. However, since Exim outputs its version number before the
5883 error, it didn't break the script. It just looked ugly. I fixed this by
5884 always allowing "-C /dev/null" if the caller is root.
5886 14. Ignore overlarge ACL variable number when reading spool file - insurance
5887 against a later release with more variables having written the file.
5889 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5890 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5891 this, as well as the form without the "IPv6" on the front.
5893 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5894 OS/Makefile-Darwin file.
5896 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5897 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5899 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5900 variable. However, some OS use ints for inodes. I've added cast to long int
5901 to get rid of the compiler warning.
5903 19. I had forgotten to lock out "/../" in configuration file names when
5904 ALT_CONFIG_PREFIX was set.
5906 20. Routers used for verification do not need to specify transports. However,
5907 if such a router generated a host list, and callout was configured, Exim
5908 crashed, because it could not find a port number from the (non-existent)
5909 transport. It now assumes port 25 in this circumstance.
5911 21. Added the -t option to exigrep.
5913 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5914 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5915 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5916 them is set, LOOKUP_LSEARCH is forced.
5918 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5920 24. Added sender and host information to the "rejected by local_scan()" log
5921 line; previously there was no indication of these.
5923 25. Added .include_if_exists.
5925 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5926 Linux. It turns out that not all file systems support this. Apparently some
5927 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5928 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5929 not supported on the file descriptor, is now ignored when Exim is trying to
5930 sync a directory. This applies only to Linux.
5932 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5934 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5935 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5938 29. Added support for maildirsize files from supplied patch (modified a bit).
5940 30. The use of :fail: followed by an empty string could lead Exim to respond to
5941 sender verification failures with (e.g.):
5943 550 Verification failed for <xxx>
5944 550 Sender verify failed
5946 where the first response line was missing the '-' that indicates it is not
5947 the final line of the response.
5949 31. The loop for finding the name of the user that called Exim had a hardwired
5950 limit of 10; it now uses the value of finduser_retries, which is used for
5951 all other user lookups.
5953 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5956 33. Exim was neglecting to zero errno before one call of strtol() when
5957 expanding a string and expecting an integer value. On some systems this
5958 resulted in spurious "integer overflow" errors. Also, it was casting the
5959 result into an int without checking.
5961 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5962 did not work. The code looks as if it has *never* worked, though it appears
5963 to have been documented since at least release 1.62. I have made it work.
5965 35. The "timeout_DNS" error in retry rules, also documented since at least
5966 1.62, also never worked. As it isn't clear exactly what this means, and
5967 clearly it isn't a major issue, I have abolished the feature by treating it
5968 as "timeout", and writing a warning to the main and panic logs.
5970 36. The display of retry rules for -brt wasn't always showing the error code
5973 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5974 timeout_connect_A, timeout_connect_MX.
5976 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5977 to the empty sender.
5979 39. The daemon was not analysing the content of -oX till after it had closed
5980 stderr and disconnected from the controlling terminal. This meant that any
5981 syntax errors were only noted on the panic log, and the return code from
5982 the command was 0. By re-arranging the code a little, I've made the
5983 decoding happen first, so such errors now appear on stderr, and the return
5984 code is 1. However, the actual setting up of the sockets still happens in
5985 the disconnected process, so errors there are still only recorded on the
5988 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5989 connections (as happens on some IP stacks) was logged at start up time as
5990 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5991 it from "IPv6 and IPv4", which means that two separate sockets are being
5994 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5995 says whether AF_INET or AF_INET6 was passed as an argument.
5997 42. Exiwhat output was messed up when time zones were included in log
6000 43. Exiwhat now gives more information about the daemon's listening ports,
6001 and whether -tls-on-connect was used.
6003 44. The "port" option of the smtp transport is now expanded.
6005 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6006 silently ignored. Now an error message is written to the main and panic
6009 46. There's a new ACL modifier called "logwrite" which writes to a log file
6010 as soon as it is encountered.
6012 47. Added $local_user_uid and $local_user_gid at routing time.
6014 48. Exim crashed when trying to verify a sender address that was being
6017 49. Exim was recognizing only a space character after ".include". It now also
6018 recognizes a tab character.
6020 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6021 extracting the relevant information from the specification. The man page no
6022 longer contains scrambled data for the -d option, and I've added a section
6023 at the front about calling Exim under different names.
6025 51. Added "extra_headers" argument to the "mail" command in filter files.
6027 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6030 53. Installed eximstats 1.29.
6032 54. Added transport_filter_timeout as a generic transport option.
6034 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6035 Cc: header lines. This was required by RFC 822, but it not required by RFC
6038 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6039 incoming messages that did not have them. Now it does so only if the
6040 message originates locally, that is, if there is no associated remote host
6041 address. When Resent- header lines are present, this applies to the Resent-
6042 lines rather than the non-Resent- lines.
6044 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6045 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6047 58. Messages for configuration errors now include the name of the main
6048 configuration file - useful now that there may be more than one file in a
6049 list (.included file names were always shown).
6051 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6052 for those rare installations that do not start the daemon as root or run it
6053 setuid root. I've cut out the call to initgroups() if the daemon is not
6056 60. The Exim user and group can now be bound into the binary as text strings
6057 that are looked up at the start of Exim's processing.
6059 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6061 62. Added $mailstore_basename variable.
6063 63. Installed patch to sieve.c from Michael Haardt.
6065 64. When Exim failed to open the panic log after failing to open the main log,
6066 the original message it was trying to log was written to stderr and debug
6067 output, but if they were not available (the usual case in production), it
6068 was lost. Now it is written to syslog before the two lines that record the
6069 failures to open the logs.
6071 65. Users' Exim filters run in subprocesses under the user's uid. It is
6072 possible for a "deliver" command or an alias in a "personal" command to
6073 provoke an address rewrite. If logging of address rewriting is configured,
6074 this fails because the process is not running as root or exim. There may be
6075 a better way of dealing with this, but for the moment (because 4.30 needs
6076 to be released), I have disabled address rewrite logging when running a
6077 filter in a non-root, non-exim process.
6083 1. The buildconfig auxiliary program wasn't quoting the value set for
6084 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6085 not defined. This bug was masked in 4.22 by the effect that was fixed in
6088 2. Some messages that were rejected after a message id was allocated were
6089 shown as "incomplete" by exigrep. It no longer does this for messages that
6090 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6092 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6093 have allow_domain_literals set, the ID did not get logged in the <= line.
6094 Domain literals are now always recognized in Message-ID: header lines.
6096 4. The first argument for a ${extract expansion item is the key name or field
6097 number. Leading and trailing spaces in this item were not being ignored,
6098 causing some misleading effects.
6100 5. When deliver_drop_privilege was set, single queue runner processes started
6101 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6102 same command in the process it spins off) were not dropping privilege.
6104 6. When the daemon running as "exim" started a queue runner, it always
6105 re-executed Exim in the spun-off process. This is a waste of effort when
6106 deliver_drop_privilege is set. The new process now just calls the
6107 queue-runner function directly.
6113 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6116 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6117 ignored. Though the use of -oP was forcing the writing of a pid file, it
6118 was always written to the default place.
6120 3. If the message "no IP address found for host xxxx" is generated during
6121 incoming verification, it is now followed by identification of the incoming
6122 connection (so you can more easily find what provoked it).
6124 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6126 5. Added some features to "harden" Exim a bit more against certain attacks:
6128 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6129 be put in Local/Makefile. This is like the never_users runtime option,
6130 but it cannot be overridden. The default setting is "root".
6132 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6133 prefix string with which any file named in a -C command line option
6136 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6137 is retained for -C and -D only if the caller of Exim is root. Without
6138 it, the exim user may also use -C and -D and retain privilege.
6140 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6141 command line option is disabled.
6143 6. Macro names set by the -D option must start with an upper case letter, just
6144 like macro names defined in the configuration file.
6146 7. Added "dereference=" facility to LDAP.
6148 8. Two instances of the typo "uknown" in the source files are fixed.
6150 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6151 the Configure-Makefile script screwed up while processing it.
6153 10. Incorporated PCRE 4.4.
6155 11. The SMTP synchronization check was not operating right at the start of an
6156 SMTP session. For example, it could not catch a HELO sent before the client
6157 waited for the greeting. There is now a check for outstanding input at the
6158 point when the greeting is written. Because of the duplex, asynchronous
6159 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6160 way, but not yet received, when the check is performed.
6162 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6163 on TCP/IP sockets, because this apparently causes some broken clients to
6166 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6167 unchanged) from the Cygwin maintainer.
6169 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6170 was supported instead of testing for mbx. Effectively a typo.
6172 15. The spa authenticator server code was not checking that the input it
6173 received was valid base64.
6175 16. The debug output line for the "set" modifier in ACLs was not showing the
6176 name of the variable that was being set.
6178 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6180 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6181 lookup on its own. Something else was needed to trigger the lookup. For
6182 example, a match in host_lookup or the need for a host name in a host list.
6183 Now, if $sender_host_name is referenced and the host name has not yet been
6184 looked up, a lookup is performed. If the lookup fails, the variable remains
6185 empty, and $host_lookup_failed is set to "1".
6187 19. Added "eqi" as a case-independent comparison operator.
6189 20. The saslauthd authentication condition could segfault if neither service
6190 nor realm was specified.
6192 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6193 error message that was logged was misleading, and incoming SMTP
6194 connections were dropped. The message is now more accurate, and temporary
6195 errors are given to SMTP connections.
6197 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6198 (except RSET) with a 421 error, until QUIT is received. However, it was
6199 failing to send a response to QUIT.
6201 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6202 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6203 ACL is now run after the helo_try_verify_hosts code.
6205 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6206 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6207 case-independent, so other case variants are also recognized). Apparently
6208 some systems use these upper case variants.
6210 25. If more than two messages were waiting for the same host, and a transport
6211 filter was specified for the transport, Exim sent two messages over the
6212 same TCP/IP connection, and then failed with "socket operation on non-
6213 socket" when it tried to send the third.
6215 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6217 27. The extern definition of crypt16() in expand.c was not being excluded when
6218 the OS had its own crypt16() function.
6220 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6221 as a preferred synonym for return_size_limit, both as an option and as an
6224 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6226 30. Changed the default configuration ACL to relax the local part checking rule
6227 for addresses that are not in any local domains. For these addresses,
6228 slashes and pipe symbols are allowed within local parts, but the sequence
6229 /../ is explicitly forbidden.
6231 31. SPA server authentication was not clearing the challenge buffer before
6234 32. log_message in a "warn" ACL statement was writing to the reject log as
6235 well as to the main log, which contradicts the documentation and doesn't
6236 seem right (because no rejection is happening). So I have stopped it.
6238 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6239 However, I am unable to do any testing of this.
6241 34. Fixed an infelicity in the appendfile transport. When checking directories
6242 for a mailbox, to see if any needed to be created, it was accidentally
6243 using path names with one or more superfluous leading slashes; tracing
6244 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6246 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6247 contents are added to the log line that is written for every discarded
6248 recipient. (Previously a log_message setting was ignored.)
6250 36. The ${quote: operator now quotes the string if it is empty.
6252 37. The install script runs exim in order to find its version number. If for
6253 some reason other than non-existence or emptiness, which it checks, it
6254 could not run './exim', it was installing it with an empty version number,
6255 i.e. as "exim-". This error state is now caught, and the installation is
6258 38. An argument was missing from the function that creates an error message
6259 when Exim fails to connect to the socket for saslauthd authentication.
6260 This could cause Exim to crash, or give a corrupted message.
6262 39. Added isip, isip4, and isip6 to ${if conditions.
6264 40. The ACL variables $acl_xx are now saved with the message, and can be
6265 accessed later in routers, transports, and filters.
6267 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6268 strings in the file are not string-expanded.
6270 42. If a MAIL command specified a SIZE value that was too large to fit into an
6271 int variable, the check against message_size_limit failed. Such values are
6272 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6273 day this will have to be increased, but I don't think I want to be around
6274 when emails are that large.
6281 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6282 iconv() is not standard in FreeBSD.
6284 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6285 IPv6 enabled. The observed symptom was a segmentation fault on return from
6286 the function os_common_find_running_interfaces() in src/os.c.
6288 3. In the check_special_case() function in daemon.c I had used "errno" as an
6289 argument name, which causes warnings on some systems. This was basically a
6290 typo, since it was named "eno" in the comments!
6292 4. The code that waits for the clock to tick (at a resolution of some fraction
6293 of a second) so as to ensure message-id uniqueness was always waiting for
6294 at least one whole tick, when it could have waited for less. [This is
6295 almost certainly not relevant at current processor speeds, where it is
6296 unlikely to ever wait at all. But we try to future-proof.]
6298 5. The function that sleeps for a time interval that includes fractions of a
6299 second contained a race. It did not block SIGALRM between setting the
6300 timer, and suspending (a couple of lines later). If the interval was short
6301 and the sigsuspend() was delayed until after it had expired, the suspension
6302 never ended. On busy systems this could lead to processes getting stuck for
6305 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6306 process, before it forks any delivery processes. The open lookup caching
6307 mechanism meant that the open file or database connection was passed into
6308 the delivery process. The problem was that delivery processes always tidy
6309 up cached lookup data. This could cause a problem for the next delivery
6310 process started by the queue runner, because the external queue runner
6311 process does not know about the closure. So the next delivery process
6312 still has data in the lookup cache. In the case of a file lookup, there was
6313 no problem because closing a file descriptor in a subprocess doesn't affect
6314 the parent. However, if the lookup was caching a connection to a database,
6315 the connection was closed, and the second delivery process was likely to
6316 see errors such as "PGSQL: query failed: server closed the connection
6317 unexpectedly". The problem has been fixed by closing all cached lookups
6318 in a queue runner before running a delivery process.
6320 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6321 seem to have the "const" qualifier which it has on other OS. I've
6324 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6325 *of the same type* (client or server) with the same public name that an
6326 error should be diagnosed.
6328 9. When Exim looked up a host name for an IP address, but failed to find the
6329 original IP address when looking up the host name (a safety check), it
6330 output the message "<ip address> does not match any IP for NULL", which was
6331 confusing, to say the least. The bug was that the host name should have
6332 appeared instead of "NULL".
6334 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6335 user that is built into the binary, and the -C or -D options is used, root
6336 privilege is dropped before the configuration file is read. In addition,
6337 logging is switched to stderr instead of the normal log files. If the
6338 configuration then re-defines the Exim user, the unprivileged environment
6339 is probably not what is expected, so Exim logs a panic warning message (but
6342 However, if deliver_drop_privilege is set, the unprivileged state may well
6343 be exactly what is intended, so the warning has been cut out in that case,
6344 and Exim is allowed to try to write to its normal log files.
6350 1. smtp_return_error_details was not giving details for temporary sender
6351 or receiver verification errors.
6353 2. Diagnose a configuration error if two authenticators have the same public
6356 3. Exim used not to create the message log file for a message until the first
6357 delivery attempt. This could be confusing when incoming messages were held
6358 for policy or load reasons. The message log file is now created at the time
6359 the message is received, and an initial "Received" line is written to it.
6361 4. The automatically generated man page for command line options had a minor
6362 bug that caused no ill effects; however, a more serious problem was that
6363 the procedure for building the man page automatically didn't always
6364 operate. Consequently, release 4.20 contains an out-of-date version. This
6365 shouldn't happen again.
6367 5. When building Exim with embedded Perl support, the script that builds the
6368 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6369 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6371 6. The freeze_tell option was not being used for messages that were frozen on
6372 arrival, either by an ACL or by local_scan().
6374 7. Added the smtp_incomplete_transaction log selector.
6376 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6377 was accepting AUTH without a new EHLO.
6379 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6380 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6383 10. Logging of TCP/IP connections (when configured) now happens in the main
6384 daemon process instead of the child process, so that the TCP/IP connection
6385 count is more accurate (but it can never be perfect).
6387 11. The use of "drop" in a nested ACL was not being handled correctly in the
6388 outer ACL. Now, if condition failure induced by the nested "drop" causes
6389 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6390 or "require"), the connection is dropped.
6392 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6393 that yield "discard" can now be used with an "accept" or a "discard" verb,
6394 but an error is generated for any others (because I can't see a useful way
6395 to define what should happen).
6397 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6398 are now processed in the same way as lines in the Exim configuration file.
6399 In particular, continuation lines are supported.
6401 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6403 15. Added -ti meaning -t -i.
6405 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6406 domains, and warn by logging if others are found.
6408 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6409 returned from GIFCONF when Exim is trying to find the list of interfaces on
6410 a host. The code in os.c has been modified to copy each ifreq to an aligned
6411 structure in all cases.
6413 Also, in some cases, the returned ifreq's were being copied to a 'struct
6414 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6415 means the last couple of bytes of an IPv6 address could be chopped if the
6416 ifreq contained only a normal sockaddr (14 bytes storage).
6418 18. Named domain lists were not supported in the hosts_treat_as_local option.
6419 An entry such as +xxxx was not recognized, and was treated as a literal
6422 19. Ensure that header lines added by a DATA ACL are included in the reject log
6423 if the ACL subsequently rejects the message.
6425 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6426 MD5 (which is deprecated).
6428 21. When testing a filter file using -bf, Exim was writing a message when it
6429 took the sender from a "From " line in the message, but it was not doing so
6430 when it took $return_path from a Return-Path: header line. It now does.
6432 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6433 with a valid header line field name (a series of printing characters
6434 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6436 23. Changed "disc" in the source to "disk" to conform to the documentation and
6437 the book and for uniformity.
6439 24. Ignore Sendmail's -Ooption=value command line item.
6441 25. When execve() failed while trying to run a command in a pipe transport,
6442 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6443 could be confused with a return value of 69 from the command itself. This
6444 has been changed to 127, the value the shell returns if it is asked to run
6445 a non-existent command. The wording for the related log line suggests a
6446 non-existent command as the problem.
6448 26. If received_header_text expands to an empty string, do not add a Received:
6449 header line to the message. (Well, it adds a token one on the spool, but
6450 marks it "old" so that it doesn't get used or transmitted.)
6452 27. Installed eximstats 1.28 (addition of -nt option).
6454 28. There was no check for failure on the call to getsockname() in the daemon
6455 code. This can fail if there is a shortage of resources on the system, with
6456 ENOMEM, for example. A temporary error is now given on failure.
6458 29. Contrary to the C standard, it seems that in some environments, the
6459 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6460 program. Exim now does this explicitly; it affects the formatting of
6461 timestamps using strftime().
6463 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6464 complaints. I've now initialized all the variables, to avoid this.
6466 32. Header lines added by a system filter were not being "seen" during
6467 transport-time rewrites.
6469 33. The info_callback() function passed to OpenSSL is set up with type void
6470 (*)(SSL *, int, int), as described somewhere. However, when calling the
6471 function (actually a macro) that sets it up, the type void(*)() is
6472 expected. I've put in a cast to prevent warnings from picky compilers.
6474 34. If a DNS black list lookup found a CNAME record, but there were no A
6475 records associated with the domain it pointed at, Exim crashed.
6477 35. If a DNS black list lookup returned more than one A record, Exim ignored
6478 all but the first. It now scans all returned addresses if a particular IP
6479 value is being sought. In this situation, the contents of the
6480 $dnslist_value variable are a list of all the addresses, separated by a
6483 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6484 to accept a host name and all its aliases if the forward lookup for any of
6485 them yielded the IP address of the incoming connection. Now it accepts only
6486 those names whose forward lookup yields the correct IP address. Any other
6487 names are discarded. This closes a loophole whereby a rogue DNS
6488 administrator could create reverse DNS records to break through a
6489 wildcarded host restriction in an ACL.
6491 37. If a user filter or a system filter that ran in a subprocess used any of
6492 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6493 the wrong values were passed to the pipe command ($thisaddress had the
6494 value of $0, $0 had the value of $1, etc). This bug was introduced by
6495 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6497 38. Improved the line breaking for long SMTP error messages from ACLs.
6498 Previously, if there was no break point between 40 and 75 characters, Exim
6499 left the rest of the message alone. Two changes have been made: (a) I've
6500 reduced the minimum length to 35 characters; (b) if it can't find a break
6501 point between 35 and 75 characters, it looks ahead and uses the first one
6502 that it finds. This may give the occasional overlong line, but at least the
6503 remaining text gets split now.
6505 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6506 file descriptors might be low, and that setting 1000 would always raise it.
6507 It turns out that in some environments, the limit is already over 1000 and
6508 that lowering it causes trouble. So now Exim takes care not to decrease it.
6510 40. When delivering a message, the value of $return_path is set to $sender_
6511 address at the start of routing (routers may change the value). By an
6512 oversight, this default was not being set up when an address was tested by
6513 -bt or -bv, which affected the outcome if any router or filter referred to
6516 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6517 log only when "message" or "log_message" are set. However, if one of the
6518 conditions was an address verification, or a call to a nested ACL, the
6519 messages generated by the underlying test were being passed through. This
6520 no longer happens. The underlying message is available in $acl_verify_
6521 message for both "message" and "log_message" expansions, so it can be
6522 passed through if needed.
6524 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6525 new expansion $bh_ to give the encoded byte string without charset
6526 translation. Translation happens only if iconv() is available; HAVE_ICONV
6527 indicates this at build time. HEADERS_CHARSET gives the charset to
6528 translate to; headers_charset can change it in the configuration, and
6529 "headers charset" can change it in an individual filter file.
6531 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6532 that creates RFC 2047 encoded "words" labels them as that charset instead
6533 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6534 expansion operator; (ii) when Exim creates a From: line for a local
6535 message; (iii) when a header line is rewritten to include a "phrase" part.
6537 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6538 buggy, causing it to skip the first lines of messages whose message ID
6539 ended in 'D'. This would not have bitten before Exim release 4.14, because
6540 message IDs were unlikely to end in 'D' before then. The effect was to have
6541 incorrect size information for certain domains.
6543 45. #include "config.h" was missing at the start of the crypt16.c module. This
6544 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6547 46. If there was a timeout during a "random" callout check, Exim treated it as
6548 a failure of the random address, and carried on sending RSET and the real
6549 address. If the delay was just some slowness somewhere, the response to the
6550 original RCPT would be taken as a response to RSET and so on, causing
6551 mayhem of various kinds.
6553 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6554 when I implemented it. It didn't allow for the fact that some option values
6555 may legitimately be negative (e.g. size_addition), and it didn't even do
6556 the right test for positive values.
6558 48. Domain names in DNS records are case-independent. Exim always looks them up
6559 in lower case. Some resolvers return domain names in exactly the case they
6560 appear in the zone file, that is, they may contain uppercase letters. Not
6561 all resolvers do this - some return always lower case. Exim was treating a
6562 change of case by a resolver as a change of domain, similar to a widening
6563 of a domain abbreviation. This triggered its re-routing code and so it was
6564 trying to route what was effectively the same domain again. This normally
6565 caused routing to fail (because the router wouldn't handle the domain
6566 twice). Now Exim checks for this case specially, and just changes the
6567 casing of the domain that it ultimately uses when it transmits the message
6570 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6573 50. If a filter generated a file delivery with a non-absolute name (possible if
6574 no home directory exists for the router), the forbid_file option was not
6577 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6578 the existing equality matching.
6580 52. Exim was using ints instead of ino_t variables in some places where it was
6581 dealing with inode numbers.
6583 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6584 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6585 called TMPDIR, and if it finds it is different, it changes its value.
6587 54. The smtp_printf() function is now made available to local_scan() so
6588 additional output lines can be written before returning. There is also an
6589 smtp_fflush() function to enable the detection of a dropped connection.
6590 The variables smtp_input and smtp_batched_input are exported to
6593 55. Changed the default runtime configuration: the message "Unknown user"
6594 has been removed from the ACL, and instead placed on the localuser router,
6595 using the cannot_route_message feature. This means that any verification
6596 failures that generate their own messages won't get overridden. Similarly,
6597 the "Unrouteable address" message that was in the ACL for unverifiable
6598 relay addresses has also been removed.
6600 56. Added hosts_avoid_esmtp to the smtp transport.
6602 57. The exicyclog script was not checking for the esoteric option
6603 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6604 will work only if exicyclog is run under the appropriate euid.
6606 58. Following a discussion on the list, the rules by which Exim recognises line
6607 endings on incoming messages have been changed. The -dropcr and drop_cr
6608 options are now no-ops, retained only for backwards compatibility. The
6609 following line terminators are recognized: LF CRLF CR. However, special
6610 processing applies to CR:
6612 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6613 nor a local message in the state where . is a terminator.
6615 (ii) If a bare CR is encountered in a header line, an extra space is added
6616 after the line terminator so as not to end the header. The reasoning
6617 behind this is that bare CRs in header lines are most likely either
6618 to be mistakes, or people trying to play silly games.
6620 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6621 was being incorrectly given as 18 bytes larger than it should have been.
6622 This is a VOB (very old bug).
6624 60. This may never have affected anything current, but just in case it has:
6625 When the local host is found other than at the start of a list of hosts,
6626 the local host, those with the same MX, and any that follow, are discarded.
6627 When the list in question was part of a longer list of hosts, the following
6628 hosts (not currently being processed) were also being discarded. This no
6629 longer happens. I'm not sure if this situation could ever has previously
6632 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6635 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6636 header. This is something that is recommended in a new Internet Draft, and
6637 is something that is documented as being done by Sendmail. There are two
6638 possible values. For messages generated by the autoreply transport, Exim
6641 Auto-Submitted: auto-replied
6643 whereas for all other generated messages (e.g. bounces) it adds
6645 Auto-Submitted: auto-generated
6647 63. The "personal" condition in filters now includes a test for the
6648 Auto-Submitted: header. If it contains the string "auto-" the message it
6649 not considered personal.
6651 64. Added rcpt_include_affixes as a generic transport option.
6653 65. Added queue_only_override (default true).
6655 66. Added the syslog_duplication option.
6657 67. If what should have been the first header line of a message consisted of
6658 a space followed by a colon, Exim was mis-interpreting it as a header line.
6659 It isn't of course - it is syntactically invalid and should therefore be
6660 treated as the start of the message body. The misbehaviour could have
6661 caused a number of strange effects, including loss of data in subsequent
6662 header lines, and spool format errors.
6664 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6665 client host had authenticated. This control can now be exercised by an ACL
6666 for more flexibility.
6668 69. By default, callouts do not happen when testing with -bh. There is now a
6669 variant, -bhc, which does actually run the callout code, including
6670 consulting and updating the callout cache.
6672 70. Added support for saslauthd authentication, courtesy of Alexander
6675 71. If statvfs() failed on the spool or log directories while checking their
6676 size for availability, Exim confusingly gave the error "space shortage".
6677 Furthermore, in debugging mode it crashed with a floating point exception.
6678 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6679 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6680 serious problem, Exim now writes to the main and panic logs when this
6681 happens, with details of the failure. It then refuses to accept the
6682 incoming message, giving the message "spool directory problem" or "log
6683 directory problem" with a 421 code for SMTP messages.
6685 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6686 0, 1, and 2 exist, because some OS complain for execs without them (see
6687 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6688 descriptors. However, the code omitted to check that the open succeeded,
6689 causing mysterious errors if for some reason the permissions on /dev/null
6690 got screwed. Now Exim writes a message to the main and panic logs, and
6691 bombs out if it can't open /dev/null.
6693 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6694 interact so that it is all more flexible. It is supposed to remain
6695 backwards compatible. Also added extra_local_interfaces.
6697 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6698 to bomb out with an assertion failure - to the client this appears as a
6699 connection drop. This problem occurs in the part of the code that was taken
6700 from the Samba project. Fortunately, the assertion is in a very simple
6701 function, so I have fixed this by reproducing the function inline in the
6702 one place where it is called, and arranging for authentication to fail
6703 instead of killing the process with assert().
6705 75. The SPA client code was not working when the server requested OEM rather
6706 than Unicode encoding.
6708 76. Added code to make require_files with a specific uid setting more usable in
6709 the case where statting the file as root fails - usually a non-root-mounted
6710 NFS file system. When this happens and the failure is EACCES, Exim now
6711 forks a subprocess and does the per-uid checking as the relevant uid.
6713 77. Added process_log_path.
6715 78. If log_file_path was not explicitly set, a setting of check_log_space or
6716 check_log_inodes was ignored.
6718 79. If a space check for the spool or log partitions fails, the incident is now
6719 logged. Of course, in the latter case the data may get lost...
6721 80. Added the %p formatting code to string_format() so that it can be used to
6722 print addresses in debug_print(). Adjusted all the address printing in the
6723 debugging in store.c to use %p rather than %d.
6725 81. There was a concern that a line of code in smtp_in.c could overflow a
6726 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6727 initially expressed, the concern was not well-founded, because trailing
6728 spaces are removed early. However, if the trailing spaces were followed by
6729 a NULL, they did not get removed, so the overflow was possible. Two fixes
6732 (a) I re-wrote the offending code in a cleaner fashion.
6733 (b) If an incoming SMTP command contains a NULL character, it is rejected
6736 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6737 runs initgroups(), so that if the Exim user is in any additional groups,
6738 they will be used during message reception.
6744 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob