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 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
53 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
54 than -2003 did; needs libidn2 in addition to libidn.
56 JH/02 The path option on a pipe transport is now expanded before use.
58 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
59 Patch provided by "Björn", documentation fix added too.
61 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
62 missing a wire-to-host endian conversion.
64 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
65 close after a BDAT command line could be taken as a following command,
66 giving a synch failure. Fix by only checking for synch immediately
67 before acknowledging the chunk.
69 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
70 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
71 macro. Patches provided by Josh Soref.
73 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
74 Previously we did not; the RFC seems ambiguous and VRFY is not listed
75 by IANA as a service extension. However, John Klensin suggests that we
78 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
79 the dkim code may be unix-mode line endings rather than smtp wire-format
80 CRLF, so prepend a CR to any bare LF.
82 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
83 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
85 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
86 the first were themselves being wrongly included in the feed into dkim
87 processing; with most chunk sizes in use this resulted in an incorrect
88 body hash calculated value.
90 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
91 DKIM signature block, for verification. Although advised against by
92 standards it is specifically not ruled illegal.
94 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
96 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
97 missing a body hash (the bh= tag).
99 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
100 It seems that HAProxy sends the Proxy Protocol information in clear and
101 only then does a TLS startup, so do the same.
103 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
104 TCP connections (such as for Spamd) unless the daemon successfully set
105 Fast Open mode on its listening sockets. This fixes breakage seen on
106 too-old kernels or those not configured for Fast Open, at the cost of
107 requiring both directions being enabled for TFO, and TFO never being used
108 by non-daemon-related Exim processes.
110 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
111 endings, at least on the first header line. Try to canonify any that get
112 past that check, despite the cost.
114 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
115 now limited to an arbitrary five deep, while parsing addresses with the
116 strip_excess_angle_brackets option enabled.
118 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
119 instead leave the unprompted TLS handshake in socket buffer for the
120 TLS library to consume.
122 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
124 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
126 JH/16 Drop variables when they go out of scope. Memory management drops a whole
127 region in one operation, for speed, and this leaves assigned pointers
128 dangling. Add checks run only under the testsuite which checks all
129 variables at a store-reset and panics on a dangling pointer; add code
130 explicitly nulling out all the variables discovered. Fixes one known
131 bug: a transport crash, where a dangling pointer for $sending_ip_address
132 originally assigned in a verify callout, is re-used.
134 PP/06 Drop '.' from @INC in various Perl scripts.
136 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
138 PP/08 Reduce a number of compilation warnings under clang; building with
139 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
140 should be warning-free.
142 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
144 HS/01 Fix portability problems introduced by PP/08 for platforms where
145 realloc(NULL) is not equivalent to malloc() [SunOS et al].
147 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
148 chunk. This allows us to accept broken chunked messages. We need a more
149 general solution here.
151 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
152 already-broken messages in the queue.
154 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
156 JH/19 Fix reference counting bug in routing-generated-address tracking.
162 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
163 supports it and a size is available (ie. the sending peer gave us one).
165 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
166 years of being deprecated). The replacements are the ACLs
167 acl_smtp_mime and acl_not_smtp_mime.
169 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
170 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
171 or even an in-clear connection were permitted. Now, if the host lookup
172 was dnssec and dane was requested then the host is only used if the
173 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
174 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
175 if one fails this test.
176 This means that a poorly-configured remote DNS will make it incommunicado;
177 but it protects against a DNS-interception attack on it.
179 JH/04 Bug 1810: make continued-use of an open smtp transport connection
180 non-noisy when a race steals the message being considered.
182 JH/05 If main configuration option tls_certificate is unset, generate a
183 self-signed certificate for inbound TLS connections.
185 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
186 in rewrites and routers.
188 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
189 and logged a warning sing 4.83; now they are a configuration file error.
191 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
192 (lacking @domain). Apply the same qualification processing as RCPT.
194 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
196 JH/10 Support ${sha256:} applied to a string (as well as the previous
199 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
200 a cutthrough deliver is pending, as we always want to make a connection.
201 This also avoids re-routing the message when later placing the cutthrough
202 connection after a verify cache hit.
203 Do not update it with the verify result either.
205 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
206 when routing results in more than one destination address.
208 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
209 signing (which inhibits the cutthrough capability). Previously only
210 the presence of an option was tested; now an expansion evaluating as
211 empty is permissible (obviously it should depend only on data available
212 when the cutthrough connection is made).
214 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
215 the relevant preceding SMTP command did not note the pipelining mode.
217 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
218 Previously they were not counted.
220 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
221 as one having no matching records. Previously we deferred the message
222 that needed the lookup.
224 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
225 distinguished as "(=".
227 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
228 for missing MX records. Previously it only worked for missing A records.
230 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
232 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
233 after the data-go-ahead and data-ack. Patch from Jason Betts.
235 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
236 even for a "none" policy. Patch from Tony Meyer.
238 JH/22 Fix continued use of a connection for further deliveries. If a port was
239 specified by a router, it must also match for the delivery to be
242 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
243 When one of the recipients of a message was unsuitable for the connection
244 (has no matching addresses), we lost track of needing to mark it
245 deferred. As a result mail would be lost.
247 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
249 JH/25 Decoding ACL controls is now done using a binary search; the source code
250 takes up less space and should be simpler to maintain. Merge the ACL
251 condition decode tables also, with similar effect.
253 JH/26 Fix problem with one_time used on a redirect router which returned the
254 parent address unchanged. A retry would see the parent address marked as
255 delivered, so not attempt the (identical) child. As a result mail would
258 JH/27 Fix a possible security hole, wherein a process operating with the Exim
259 UID can gain a root shell. Credit to http://www.halfdog.net/ for
260 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
263 JH/28 Enable {spool,log} filesystem space and inode checks as default.
264 Main config options check_{log,spool}_{inodes,space} are now
265 100 inodes, 10MB unless set otherwise in the configuration.
267 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
268 Previously it only applied to the main-section connection policy
271 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
273 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
274 by me. Added RFC7919 DH primes as an alternative.
276 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
277 are not in the system include path.
279 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
280 GnuTLS, when a session startup failed (eg because the client disconnected)
281 Exim did stdio operations after fclose. This was exposed by a recent
282 change which nulled out the file handle after the fclose.
284 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
285 signed directly by the cert-signing cert, rather than an intermediate
286 OCSP-signing cert. This is the model used by LetsEncrypt.
288 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
290 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
291 an incoming connection.
293 HS/02 Bug 1802: Do not half-close the connection after sending a request
296 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
297 fallback to "prime256v1".
299 JH/34 SECURITY: Use proper copy of DATA command in error message.
300 Could leak key material. Remotely exploitable. CVE-2016-9963.
306 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
307 and 3.4.4 - once the server is enabled to respond to an OCSP request
308 it does even when not requested, resulting in a stapling non-aware
309 client dropping the TLS connection.
311 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
312 support variable-length bit vectors. No functional change.
314 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
315 The I= interface field on outgoing lines is now after the H= remote
316 host field, same as incoming lines. There is a separate
317 outgoing_interface log selector which allows you to disable the
320 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
321 If not running log_selector +smtp_connection the mainlog would be held
322 open indefinitely after a "too many connections" event, including to a
323 deleted file after a log rotate. Leave the per net connection logging
324 leaving it open for efficiency as that will be quickly detected by the
325 check on the next write.
327 HS/01 Bug 1671: Fix post transport crash.
328 Processing the wait-<transport> messages could crash the delivery
329 process if the message IDs didn't exist for some reason. When
330 using 'split_spool_directory=yes' the construction of the spool
331 file name failed already, exposing the same netto behaviour.
333 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
334 mime_regex ACL conditions.
336 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
337 to DSN fail messages (bounces): remote IP, remote greeting, remote response
338 to HELO, local diagnostic string.
340 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
341 log line to debug. Even when configured with a tls authenticator many
342 client connections are expected to not authenticate in this way, so
343 an authenticate fail is not an error.
345 HS/02 Add the Exim version string to the process info. This way exiwhat
346 gives some more detail about the running daemon.
348 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
349 matter for fast-change records such as DNSBLs.
351 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
352 retry record, even if constant. There may be multiple transports with
353 different interface settings and the retry behaviour needs to be kept
356 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
358 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
360 JH/10 Bug 840: fix log_defer_output option of pipe transport
362 JH/11 Bug 830: use same host for all RCPTS of a message, even under
363 hosts_randomize. This matters a lot when combined with mua_wrapper.
365 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
366 ${quote_pgsql:<string>} operator.
368 JH/13 Bug 1708: avoid misaligned access in cached lookup.
370 JH/14 Change header file name for freeradius-client. Relevant if compiling
371 with Radius support; from the Gentoo tree and checked under Fedora.
373 JH/15 Bug 1712: Introduce $prdr_requested flag variable
375 JH/16 Bug 1714: Permit an empty string as expansion result for transport
376 option transport_filter, meaning no filtering.
378 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
380 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
381 defaults to "*" (all hosts). The variable is now available when not built
382 with TLS, default unset, mainly to enable keeping the testsuite sane.
383 If a server certificate is not supplied (via tls_certificate) an error is
384 logged, and clients will find TLS connections fail on startup. Presumably
385 they will retry in-clear.
386 Packagers of Exim are strongly encouraged to create a server certificate
387 at installation time.
389 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
390 with the $config_file variable.
392 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
393 in transport context, after the attempt, and per-recipient. The latter type
394 is per host attempted. The event data is the error message, and the errno
395 information encodes the lookup type (A vs. MX) used for the (first) host,
396 and the trailing two digits of the smtp 4xx response.
398 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
399 to write to mainlog (or rejectlog, paniclog) in the window between file
400 creation and permissions/ownership being changed. Particularly affects
401 installations where exicyclog is run as root, rather than exim user;
402 result is that the running daemon panics and dies.
404 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
406 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
407 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
408 "pri" and "weight". Note that the previous implicit priority given by the
409 list order is no longer honoured.
411 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
414 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
415 by defining SUPPORT_SOCKS.
417 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
418 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
419 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
420 variables are renamed to proxy_{local,external}_{address,port}.
422 JH/27 Move Internationalisation support from Experimental to mainline, enabled
423 for a build by defining SUPPORT_I18N
425 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
426 of the query string, and make ${quote_redis:} do that quoting.
428 JH/29 Move Events support from Experimental to mainline, enabled by default
429 and removable for a build by defining DISABLE_EVENT.
431 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
433 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
434 cached by the daemon.
436 JH/32 Move Redis support from Experimental to mainline, enabled for a build
437 by defining LOOKUP_REDIS. The libhiredis library is required.
439 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
440 keys are given for lookup.
442 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
443 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
444 only supported when built with TLS support. The PolarSSL SHA routines
445 are still used when the TLS library is too old for convenient support.
447 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
448 openssl_options), for security. OpenSSL forces this from version 1.1.0
449 server-side so match that on older versions.
451 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
452 allocation for $value could be released as the expansion processing
453 concluded, but leaving the global pointer active for it.
455 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
456 and to use the domains and local_parts ACL conditions.
458 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
459 incorrectly not doubled on cutthrough transmission, hence seen as a
460 body-termination at the receiving system - resulting in truncated mails.
461 Commonly the sender saw a TCP-level error, and retransmitted the message
462 via the normal store-and-forward channel. This could result in duplicates
463 received - but deduplicating mailstores were liable to retain only the
464 initial truncated version.
466 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
468 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
470 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
471 we're in there, support oversigning also; bug 1309.
473 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
475 HS/04 Add support for keep_environment and add_environment options.
477 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
478 either intentional arithmetic overflow during PRNG, or testing config-
481 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
482 delivery resulted in actual delivery. Cancel cutthrough before DATA
485 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
486 rejects a recipient: pass the reject to the originator.
488 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
489 Many were false-positives and ignorable, but it's worth fixing the
492 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
493 for the new environment-manipulation done at startup. Move the routines
494 from being local to tls.c to being global via the os.c file.
496 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
497 an extract embedded as result-arg for a map, the first arg for extract
498 is unavailable so we cannot tell if this is a numbered or keyed
499 extraction. Accept either.
505 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
508 JH/02 The smtp transport option "multi_domain" is now expanded.
510 JH/03 The smtp transport now requests PRDR by default, if the server offers
513 JH/04 Certificate name checking on server certificates, when exim is a client,
514 is now done by default. The transport option tls_verify_cert_hostnames
515 can be used to disable this per-host. The build option
516 EXPERIMENTAL_CERTNAMES is withdrawn.
518 JH/05 The value of the tls_verify_certificates smtp transport and main options
519 default to the word "system" to access the system default CA bundle.
520 For GnuTLS, only version 3.0.20 or later.
522 JH/06 Verification of the server certificate for a TLS connection is now tried
523 (but not required) by default. The verification status is now logged by
524 default, for both outbound TLS and client-certificate supplying inbound
527 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
530 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
531 Status Notification (bounce) messages are now MIME format per RFC 3464.
532 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
533 under the control of the dsn_advertise_hosts option, and routers may
534 have a dsn_lasthop option.
536 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
537 default, modifiable by a malware= option. The list separator for
538 the options can now be changed in the usual way. Bug 68.
540 JH/10 The smtp_receive_timeout main option is now expanded before use.
542 JH/11 The incoming_interface log option now also enables logging of the
543 local interface on delivery outgoing connections.
545 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
546 if the interface and destination host and port all match.
548 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
551 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
552 Patch from Andrew Lewis.
554 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
555 now supports optional time-restrictions, weighting, and priority
556 modifiers per server. Patch originally by <rommer@active.by>.
558 JH/16 The spamd_address main option now supports a mixed list of local
559 and remote servers. Remote servers can be IPv6 addresses, and
560 specify a port-range.
562 JH/17 Bug 68: The spamd_address main option now supports an optional
563 timeout value per server.
565 JH/18 Bug 1581: Router and transport options headers_add/remove can
566 now have the list separator specified.
568 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
571 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
574 JH/21 Support for the A6 type of dns record is withdrawn.
576 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
577 rather than the verbs used.
579 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
580 from 255 to 1024 chars.
582 JH/24 Verification callouts now attempt to use TLS by default.
584 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
585 are generic router options now. The defaults didn't change.
587 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
588 Original patch from Alexander Shikoff, worked over by JH.
590 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
591 files if ZCAT_COMMAND is not executable.
593 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
595 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
597 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
598 Normally benign, it bites when the pair was led to by a CNAME;
599 modern usage is to not canonicalize the domain to a CNAME target
600 (and we were inconsistent anyway for A-only vs AAAA+A).
602 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
604 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
605 when evaluating $sender_host_dnssec.
607 JH/31 Check the HELO verification lookup for DNSSEC, adding new
608 $sender_helo_dnssec variable.
610 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
612 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
614 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
616 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
617 documented as working, but never had. Support all but $spam_report.
619 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
620 added for tls authenticator.
622 HS/03 Add perl_taintmode main config option
628 TL/01 When running the test suite, the README says that variables such as
629 no_msglog_check are global and can be placed anywhere in a specific
630 test's script, however it was observed that placement needed to be near
631 the beginning for it to behave that way. Changed the runtest perl
632 script to read through the entire script once to detect and set these
633 variables, reset to the beginning of the script, and then run through
634 the script parsing/test process like normal.
636 TL/02 The BSD's have an arc4random API. One of the functions to induce
637 adding randomness was arc4random_stir(), but it has been removed in
638 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
639 function when detected.
641 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
642 cause callback expansion.
644 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
645 syntax errors in an expansion can be treated as a string instead of
646 logging or causing an error, due to the internal use of bool_lax
647 instead of bool when processing it.
649 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
650 server certificates when making smtp deliveries.
652 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
654 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
656 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
658 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
659 Merged patch from Sebastian Wiedenroth.
661 JH/05 Fix results-pipe from transport process. Several recipients, combined
662 with certificate use, exposed issues where response data items split
663 over buffer boundaries were not parsed properly. This eventually
664 resulted in duplicates being sent. This issue only became common enough
665 to notice due to the introduction of connection certificate information,
666 the item size being so much larger. Found and fixed by Wolfgang Breyha.
668 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
669 size buffer was used, resulting in syntax errors when an expansion
672 JH/07 Add support for directories of certificates when compiled with a GnuTLS
673 version 3.3.6 or later.
675 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
676 is EXPERIMENTAL_EVENT, the main-configuration and transport options
677 both become "event_action", the variables become $event_name, $event_data
678 and $event_defer_errno. There is a new variable $verify_mode, usable in
679 routers, transports and related events. The tls:cert event is now also
680 raised for inbound connections, if the main configuration event_action
683 TL/06 In test suite, disable OCSP for old versions of openssl which contained
684 early OCSP support, but no stapling (appears to be less than 1.0.0).
686 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
687 server certificate names available under the smtp transport option
688 "tls_verify_cert_hostname" now do not permit multi-component wildcard
691 JH/10 Time-related extraction expansions from certificates now use the main
692 option "timezone" setting for output formatting, and are consistent
693 between OpenSSL and GnuTLS compilations. Bug 1541.
695 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
696 encoded parameter in the incoming message. Bug 1558.
698 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
699 include certificate info, eximon was claiming there were spoolfile
702 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
704 JH/14 Log delivery-related information more consistently, using the sequence
705 "H=<name> [<ip>]" wherever possible.
707 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
708 are problematic for Debian distribution, omit them from the release
711 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
713 JH/16 Fix string representation of time values on 64bit time_t architectures.
716 JH/17 Fix a null-indirection in certextract expansions when a nondefault
717 output list separator was used.
722 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
723 checkers that were complaining about end of non-void function with no
726 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
727 This was a regression introduced in 4.83 by another bugfix.
729 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
731 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
732 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
738 TF/01 Correctly close the server side of TLS when forking for delivery.
740 When a message was received over SMTP with TLS, Exim failed to clear up
741 the incoming connection properly after forking off the child process to
742 deliver the message. In some situations the subsequent outgoing
743 delivery connection happened to have the same fd number as the incoming
744 connection previously had. Exim would try to use TLS and fail, logging
745 a "Bad file descriptor" error.
747 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
748 utilities have not been installed.
750 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
751 temporary space as the ACL may create new global variables.
753 TL/01 LDAP support uses per connection or global context settings, depending
754 upon the detected version of the libraries at build time.
756 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
757 to extract and use the src ip:port in logging and expansions as if it
758 were a direct connection from the outside internet. PPv2 support was
759 updated based on HAProxy spec change in May 2014.
761 JH/02 Add ${listextract {number}{list}{success}{fail}}.
763 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
764 Properly escape header and check for NULL return.
766 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
769 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
771 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
772 characters in header names, implemented as a verify condition.
773 Contributed by Michael Fischer v. Mollard.
775 TL/05 Rename SPF condition results err_perm and err_temp to standardized
776 results permerror and temperror. Previous values are deprecated but
777 still accepted. In a future release, err_perm and err_temp will be
778 completely removed, which will be a backward incompatibility if the
779 ACL tests for either of these two old results. Patch contributed by
780 user bes-internal on the mailing list.
782 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
784 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
785 selectors, in both main and reject logs.
787 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
790 JH/07 Add malware type "sock" for talking to simple daemon.
792 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
794 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
795 routers/transports under cutthrough routing.
797 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
798 numbers. Touch up "bool" conditional to keep the same definition.
800 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
802 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
804 JH/12 Expand items in router/transport headers_add or headers_remove lists
805 individually rather than the list as a whole. Bug 1452.
807 Required for reasonable handling of multiple headers_ options when
808 they may be empty; requires that headers_remove items with embedded
809 colons must have them doubled (or the list-separator changed).
811 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
812 view the policy declared in the DMARC record. Currently, $dmarc_status
813 is a combined value of both the record presence and the result of the
816 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
818 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
819 dnslookup router and the smtp transport (applying to the forward
822 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
823 of ldap servers used for a specific lookup. Patch provided by Heiko
826 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
827 New variable $lookup_dnssec_authenticated for observability.
829 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
830 Patch submitted by Lars Timman.
832 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
834 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
835 Requires trusted mode and valid format message id, aborts otherwise.
836 Patch contributed by Heiko Schlichting.
838 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
839 certextract with support for various fields. Bug 1358.
841 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
842 is requested by default, modifiable by smtp transport option
845 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
846 operate on certificate variables to give certificate fingerprints
847 Also new ${sha256:cert_variable}.
849 JH/23 The PRDR feature is moved from being Experimental into the mainline.
851 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
852 Christian Aistleitner.
854 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
856 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
857 file. Patch from Wolfgang Breyha.
859 JH/25 Expand the coverage of the delivery $host and $host_address to
860 client authenticators run in verify callout. Bug 1476.
862 JH/26 Port service names are now accepted for tls_on_connect_ports, to
863 align with daemon_smtp_ports. Bug 72.
865 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
866 support and error reporting did not work properly.
868 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
869 and is readable. Patch from Andrew Colin Kissa.
871 TL/14 Enhance documentation of ${run expansion and how it parses the
872 commandline after expansion, particularly in the case when an
873 unquoted variable expansion results in an empty value.
875 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
877 PP/02 Fix internal collision of T_APL on systems which support RFC3123
878 by renaming away from it. Addresses GH issue 15, reported by
881 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
883 TL/15 SECURITY: prevent double expansion in math comparison functions
884 (can expand unsanitized data). Not remotely exploitable.
891 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
893 PP/02 Make -n do something, by making it not do something.
894 When combined with -bP, the name of an option is not output.
896 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
899 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
900 $sender_host_name and config options to manage this, and basic check
903 PP/05 DSCP support for outbound connections and control modifier for inbound.
905 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
906 (Only plugin which currently uses this is kerberos4, which nobody should
907 be using, but we should make it available and other future plugins might
908 conceivably use it, even though it would break NAT; stuff *should* be
909 using channel bindings instead).
911 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
912 name; added for Sendmail compatibility; requires admin caller.
913 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
914 just ignore it); requires trusted caller.
915 Also parse but ignore: -Ac -Am -X<logfile>
918 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
920 TL/02 Add +smtp_confirmation as a default logging option.
922 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
923 Patch by Magnus Holmgren from 2007-02-20.
925 TL/04 Bugzilla 1281 - Spec typo.
926 Bugzilla 1283 - Spec typo.
927 Bugzilla 1290 - Spec grammar fixes.
929 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
931 TL/06 Add Experimental DMARC support using libopendmarc libraries.
933 TL/07 Fix an out of order global option causing a segfault. Reported to dev
934 mailing list by by Dmitry Isaikin.
936 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
938 JH/02 Support "G" suffix to numbers in ${if comparisons.
940 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
942 NM/01 Bugzilla 1197 - Spec typo
943 Bugzilla 1196 - Spec examples corrections
945 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
947 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
948 gnutls_enable_pkcs11, but renamed to more accurately indicate its
951 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
952 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
954 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
955 "acl {{name}{arg}...}", and optional args on acl condition
958 JH/05 Permit multiple router/transport headers_add/remove lines.
960 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
962 JH/07 Avoid using a waiting database for a single-message-only transport.
963 Performance patch from Paul Fisher. Bugzilla 1262.
965 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
968 JH/09 Add $headers_added variable, with content from use of ACL modifier
969 add_header (but not yet added to the message). Bugzilla 199.
971 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
972 Pulled from Bugzilla 817 by Wolfgang Breyha.
974 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
976 (nb: this is the same fix as in Exim 4.80.1)
978 JH/11 Add A= logging on delivery lines, and a client_set_id option on
981 JH/12 Add optional authenticated_sender logging to A= and a log_selector
984 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
986 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
987 advertise SMTP AUTH mechanism to us, instead of a generic
988 protocol violation error. Also, make Exim more robust to bad
989 data from the Dovecot auth socket.
991 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
993 When a queue runner is handling a message, Exim first routes the
994 recipient addresses, during which it prunes them based on the retry
995 hints database. After that it attempts to deliver the message to
996 any remaining recipients. It then updates the hints database using
999 So if a recipient address works intermittently, it can get repeatedly
1000 deferred at routing time. The retry hints record remains fresh so the
1001 address never reaches the final cutoff time.
1003 This is a fairly common occurrence when a user is bumping up against
1004 their storage quota. Exim had some logic in its local delivery code
1005 to deal with this. However it did not apply to per-recipient defers
1006 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1008 This change adds a proper retry rule check during routing so that the
1009 final cutoff time is checked against the message's age. We only do
1010 this check if there is an address retry record and there is not a
1011 domain retry record; this implies that previous attempts to handle
1012 the address had the retry_use_local_parts option turned on. We use
1013 this as an approximation for the destination being like a local
1014 delivery, as in LMTP.
1016 I suspect this new check makes the old local delivery cutoff check
1017 redundant, but I have not verified this so I left the code in place.
1019 TF/02 Correct gecos expansion when From: is a prefix of the username.
1021 Test 0254 submits a message to Exim with the header
1025 When I ran the test suite under the user fanf2, Exim expanded
1026 the header to contain my full name, whereas it should have added
1027 a Resent-Sender: header. It erroneously treats any prefix of the
1028 username as equal to the username.
1030 This change corrects that bug.
1032 GF/01 DCC debug and logging tidyup
1033 Error conditions log to paniclog rather than rejectlog.
1034 Debug lines prefixed by "DCC: " to remove any ambiguity.
1036 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1038 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1039 Bug spotted by Jeremy Harris; was flawed since initial commit.
1040 Would have resulted in OCSP responses post-SNI triggering an Exim
1041 NULL dereference and crash.
1043 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1045 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1046 Bug detection, analysis and fix by Samuel Thibault.
1047 Bugzilla 1331, Debian bug #698092.
1049 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1051 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1052 Server implementation by Todd Lyons, client by JH.
1053 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1054 config variable "prdr_enable" controls whether the server
1055 advertises the facility. If the client requests PRDR a new
1056 acl_data_smtp_prdr ACL is called once for each recipient, after
1057 the body content is received and before the acl_smtp_data ACL.
1058 The client is controlled by both of: a hosts_try_prdr option
1059 on the smtp transport, and the server advertisement.
1060 Default client logging of deliveries and rejections involving
1061 PRDR are flagged with the string "PRDR".
1063 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1064 fclose(). Diagnosis by Todd Lyons.
1066 PP/17 Update configure.default to handle IPv6 localhost better.
1067 Patch by Alain Williams (plus minor tweaks).
1070 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1071 This is now consistent with GnuTLS, and is now documented: the
1072 previous undocumented portable approach to treating the option as
1073 unset was to force an expansion failure. That still works, and
1074 an empty string is now equivalent.
1076 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1077 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1078 not performing validation itself.
1080 PP/20 Added force_command boolean option to pipe transport.
1081 Patch from Nick Koston, of cPanel Inc.
1083 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1086 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1088 PP/21 Fix eximon continuous updating with timestamped log-files.
1089 Broken in a format-string cleanup in 4.80, missed when I repaired the
1090 other false fix of the same issue.
1091 Report and fix from Heiko Schlichting.
1094 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1095 Report from Prashanth Katuri.
1097 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1098 It's SecureTransport, so affects any MacOS clients which use the
1099 system-integrated TLS libraries, including email clients.
1101 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1102 using a MIME ACL for non-SMTP local injection.
1103 Report and assistance in diagnosis by Warren Baker.
1105 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1107 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1109 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1110 last id that failed so it may be referenced in subsequent ACL's.
1112 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1115 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1116 ldap library initialization, allowing self-signed CA's to be
1117 used. Also properly sets require_cert option later in code by
1118 using NULL (global ldap config) instead of ldap handle (per
1119 session). Bug diagnosis and testing by alxgomz.
1121 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1122 the src/util/ subdirectory.
1124 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1125 renamed to Transport Post Delivery Action by Jeremy Harris, as
1128 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1129 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1130 redis_servers = needs to be configured which will be used by the redis
1131 lookup. Patch from Warren Baker, of The Packet Hub.
1133 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1135 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1136 hostname or reverse DNS when processing a host list. Used suggestions
1137 from multiple comments on this bug.
1139 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1141 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1142 Missed a few lines, added it to make the runtest require no keyboard
1145 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1146 contains upper case chars. Make router use caseful_local_part.
1148 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1149 support when GnuTLS has been built with p11-kit.
1155 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1157 This, or similar/improved, will also be change PP/11 of 4.82.
1163 PP/01 Handle short writes when writing local log-files.
1164 In practice, only affects FreeBSD (8 onwards).
1165 Bugzilla 1053, with thanks to Dmitry Isaikin.
1167 NM/01 Bugzilla 949 - Documentation tweak
1169 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1172 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1174 PP/02 Implemented gsasl authenticator.
1176 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1178 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1179 `pkg-config foo` for cflags/libs.
1181 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1182 with rest of GSASL and with heimdal_gssapi.
1184 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1185 `pkg-config foo` for cflags/libs for the TLS implementation.
1187 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1188 properties get this fed in as external SSF. A number of robustness
1189 and debugging improvements to the cyrus_sasl authenticator.
1191 PP/08 cyrus_sasl server now expands the server_realm option.
1193 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1194 Patch by Jeremy Harris.
1196 PP/10 Added dbmjz lookup type.
1198 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1200 PP/12 MAIL args handles TAB as well as SP, for better interop with
1201 non-compliant senders.
1202 Analysis and variant patch by Todd Lyons.
1204 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1205 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1206 Patch from Dirk Mueller <dmueller@suse.com>
1208 PP/13 tls_peerdn now print-escaped for spool files.
1209 Observed some $tls_peerdn in wild which contained \n, which resulted
1210 in spool file corruption.
1212 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1213 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1214 or write after TLS renegotiation, which otherwise led to messages
1217 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1218 as a tracking header (ie: a signed header comes before the signature).
1219 Patch from Wolfgang Breyha.
1221 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1222 comma-sep list; embedded commas doubled.
1224 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1226 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1228 Report and patch from Dmitry Banschikov.
1230 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1231 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1232 failures. We appear to now support TLS1.1+ with Exim.
1234 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1235 lets Exim select keys and certificates based upon TLS SNI from client.
1236 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1237 before an outbound SMTP session. New log_selector, +tls_sni.
1239 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1240 NULL dereference. Report and patch from Alun Jones.
1242 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1243 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1244 Not seeing resolver debug output on NetBSD, but suspect this is a
1245 resolver implementation change.
1247 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1248 Left warnings. Added "eximon gdb" invocation mode.
1250 PP/21 Defaulting "accept_8bitmime" to true, not false.
1252 PP/22 Added -bw for inetd wait mode support.
1254 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1255 locate the relevant includes and libraries. Made this the default.
1257 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1258 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1260 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1261 This may cause build issues on older platforms.
1263 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1264 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1265 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1266 Added SNI support via GnuTLS too.
1267 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1269 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1271 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1272 Applied second patch from Janne, implementing suggestion to default
1273 multiple-strings-in-record handling to match SPF spec.
1275 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1277 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1278 read-only, out of scope).
1279 Patch from Wolfgang Breyha, report from Stuart Northfield.
1281 PP/29 Fix three issues highlighted by clang analyser static analysis.
1282 Only crash-plausible issue would require the Cambridge-specific
1283 iplookup router and a misconfiguration.
1284 Report from Marcin Mirosław.
1286 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1288 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1289 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1290 As part of this, removing so much warning spew let me fix some minor
1291 real issues in debug logging.
1293 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1294 assignment on my part. Fixed.
1296 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1297 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1298 Janne Snabb (who went above and beyond: thank you).
1300 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1301 string otherwise requires a connection and a bunch more work and it's
1302 relatively easy to get wrong. Should also expose TLS library linkage
1305 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1306 64-bit ${eval} (JH/03).
1308 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1309 GNU libc to support some of the 64-bit stuff, should not lead to
1310 conflicts. Defined before os.h is pulled in, so if a given platform
1311 needs to override this, it can.
1313 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1314 protection layer was required, which is not implemented.
1315 Bugzilla 1254, patch from Wolfgang Breyha.
1317 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1318 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1319 tls_dhparam take prime identifiers. Also unbreak combination of
1320 OpenSSL+DH_params+TLSSNI.
1322 PP/39 Disable SSLv2 by default in OpenSSL support.
1328 PP/01 Solaris build fix for Oracle's LDAP libraries.
1329 Bugzilla 1109, patch from Stephen Usher.
1331 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1333 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1336 TF/02 Fix a couple more cases where we did not log the error message
1337 when unlink() failed. See also change 4.74-TF/03.
1339 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1340 lock up or crash if it happened to be inside a call to libc when it
1341 got a SIGUSR1 from exiwhat.
1343 The SIGUSR1 handler appends the current process status to the process
1344 log which is later printed by exiwhat. It used to use the general
1345 purpose logging code to do this, but several functions it calls are
1346 not safe for signals.
1348 The new output code in the SIGUSR1 handler is specific to the process
1349 log, and simple enough that it's easy to inspect for signal safety.
1350 Removing some special cases also simplifies the general logging code.
1351 Removing the spurious timestamps from the process log simplifies
1354 TF/04 Improved ratelimit ACL condition.
1356 The /noupdate option has been deprecated in favour of /readonly which
1357 has clearer semantics. The /leaky, /strict, and /readonly update modes
1358 are mutually exclusive. The update mode is no longer included in the
1359 database key; it just determines when the database is updated. (This
1360 means that when you upgrade Exim will forget old rate measurements.)
1362 Exim now checks that the per_* options are used with an update mode that
1363 makes sense for the current ACL. For example, when Exim is processing a
1364 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1365 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1366 must specify per_mail/readonly. If you omit the update mode it defaults to
1367 /leaky where that makes sense (as before) or /readonly where required.
1369 The /noupdate option is now undocumented but still supported for
1370 backwards compatibility. It is equivalent to /readonly except that in
1371 ACLs where /readonly is required you may specify /leaky/noupdate or
1372 /strict/noupdate which are treated the same as /readonly.
1374 A useful new feature is the /count= option. This is a generalization
1375 of the per_byte option, so that you can measure the throughput of other
1376 aggregate values. For example, the per_byte option is now equivalent
1377 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1379 The per_rcpt option has been generalized using the /count= mechanism
1380 (though it's more complicated than the per_byte equivalence). When it is
1381 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1382 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1383 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1384 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1385 Note that using per_rcpt with a non-readonly update mode in more than
1386 one ACL will cause the recipients to be double-counted. (The per_mail
1387 and per_byte options don't have this problem.)
1389 The handling of very low rates has changed slightly. If the computed rate
1390 is less than the event's count (usually one) then this event is the first
1391 after a long gap. In this case the rate is set to the same as this event's
1392 count, so that the first message of a spam run is counted properly.
1394 The major new feature is a mechanism for counting the rate of unique
1395 events. The new per_addr option counts the number of different
1396 recipients that someone has sent messages to in the last time period. It
1397 behaves like per_rcpt if all the recipient addresses are different, but
1398 duplicate recipient addresses do not increase the measured rate. Like
1399 the /count= option this is a general mechanism, so the per_addr option
1400 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1401 example, measure the rate that a client uses different sender addresses
1402 with the options per_mail/unique=$sender_address. There are further
1403 details in the main documentation.
1405 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1407 TF/06 Removed a few PCRE remnants.
1409 TF/07 Automatically extract Exim's version number from tags in the git
1410 repository when doing development or release builds.
1412 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1413 Bugzilla 879. Patch from Paul Fisher.
1415 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1416 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1419 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1421 PP/05 Variable $av_failed, true if the AV scanner deferred.
1422 Bugzilla 1078. Patch from John Horne.
1424 PP/06 Stop make process more reliably on build failure.
1425 Bugzilla 1087. Patch from Heiko Schlittermann.
1427 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1428 Bugzilla 1089. Patch from Heiko Schlittermann.
1430 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1431 Bugzilla 1131. Patch from Holger Weiß.
1433 PP/09 Handle IPv6 addresses with SPF.
1434 Bugzilla 860. Patch from Wolfgang Breyha.
1436 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1438 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1441 PP/11 match_* no longer expand right-hand-side by default.
1442 New compile-time build option, EXPAND_LISTMATCH_RHS.
1443 New expansion conditions, "inlist", "inlisti".
1445 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1447 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1449 PP/14 fix log_write() format string regression from TF/03.
1450 Bugzilla 1152. Patch from Dmitry Isaikin.
1456 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1458 PP/02 Harmonised TLS library version reporting; only show if debugging.
1459 Layout now matches that introduced for other libraries in 4.74 PP/03.
1461 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1463 PP/04 New "dns_use_edns0" global option.
1465 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1468 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1469 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1471 TK/01 Updated PolarSSL code to 0.14.2.
1472 Bugzilla 1097. Patch from Andreas Metzler.
1474 PP/07 Catch divide-by-zero in ${eval:...}.
1475 Fixes bugzilla 1102.
1477 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1480 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1481 format-string attack -- SECURITY: remote arbitrary code execution.
1483 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1484 time unintentionally subject to list matching rules, letting the header
1485 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1486 arbitrary string expansion). This allowed for information disclosure.
1488 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1489 INT_MIN/-1 -- value coerced to INT_MAX.
1495 NM/01 Workaround for PCRE version dependency in version reporting
1498 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1499 This fixes portability to compilers other than gcc, notably
1500 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1502 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1503 makefiles for portability to HP-UX and POSIX correctness.
1505 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1506 Also via indented variable definition in the Makefile.
1507 (Debugging by Oliver Heesakkers).
1509 PP/02 Restore caching of spamd results with expanded spamd_address.
1510 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1512 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1513 Improves build reliability. Fix from: Frank Elsner
1515 NM/02 Fix wide character breakage in the rfc2047 coding
1516 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1518 NM/03 Allow underscore in dnslist lookups
1519 Fixes bug 1026. Patch from Graeme Fowler
1521 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1522 Code patches from Adam Ciarcinski of NetBSD.
1524 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1527 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1528 is logged, to avoid truncation. Patch from John Horne.
1530 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1531 Patch from Jakob Hirsch.
1533 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1534 SQL string expansion failure details.
1535 Patch from Andrey Oktyabrski.
1537 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1538 Patch from Simon Arlott.
1540 PP/09 New lookups functionality failed to compile on old gcc which rejects
1541 extern declarations in function scope.
1542 Patch from Oliver Fleischmann
1544 PP/10 Use sig_atomic_t for flags set from signal handlers.
1545 Check getgroups() return and improve debugging.
1546 Fixed developed for diagnosis in bug 927 (which turned out to be
1549 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1550 Patch from Mark Zealey.
1552 PP/12 Bugzilla 1056: Improved spamd server selection.
1553 Patch from Mark Zealey.
1555 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1556 Based on patch from Heiko Schlittermann.
1558 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1559 Patch from Uwe Doering, sign-off by Michael Haardt.
1561 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1562 variable declaration deep within a block. Bug and patch from
1565 PP/15 lookups-Makefile IRIX compatibility coercion.
1567 PP/16 Make DISABLE_DKIM build knob functional.
1569 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1570 Patch by Simon Arlott
1572 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1573 variable argument macros. Our copy now differs from upstream.
1579 TF/01 Failure to get a lock on a hints database can have serious
1580 consequences so log it to the panic log.
1582 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1583 controlled using the smtp_confirmation log selector.
1585 TF/03 Include the error message when we fail to unlink a spool file.
1587 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1588 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1589 for maintaining out-of-tree patches for some time.
1591 PP/01 Bugzilla 139: Documentation and portability issues.
1592 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1593 Handle per-OS dynamic-module compilation flags.
1595 PP/02 Let /dev/null have normal permissions.
1596 The 4.73 fixes were a little too stringent and complained about the
1597 permissions on /dev/null. Exempt it from some checks.
1598 Reported by Andreas M. Kirchwitz.
1600 PP/03 Report version information for many libraries, including
1601 Exim version information for dynamically loaded libraries. Created
1602 version.h, now support a version extension string for distributors
1603 who patch heavily. Dynamic module ABI change.
1605 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1606 privilege escalation vulnerability whereby the Exim run-time user
1607 can cause root to append content of the attacker's choosing to
1610 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1613 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1614 If dropping privileges for untrusted macros, we disabled normal logging
1615 on the basis that it would fail; for the Exim run-time user, this is not
1616 the case, and it resulted in successful deliveries going unlogged.
1617 Fixed. Reported by Andreas Metzler.
1623 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1624 only prepend for the Resent-* case. Fixes regression introduced in
1625 Exim 4.70 by NM/22 for Bugzilla 607.
1627 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1628 increasing numbers of administrators be bitten by this.
1630 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1632 PP/03 Bugzilla 994: added openssl_options main configuration option.
1634 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1636 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1638 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1640 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1641 without a peer certificate, leading to a segfault because of an
1642 assumption that peers always have certificates. Be a little more
1643 paranoid. Problem reported by Martin Tscholak.
1645 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1646 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1647 NB: ClamAV planning to remove STREAM in "middle of 2010".
1648 CL also introduces -bmalware, various -d+acl logging additions and
1649 more caution in buffer sizes.
1651 PP/09 Implemented reverse_ip expansion operator.
1653 PP/10 Bugzilla 937: provide a "debug" ACL control.
1655 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1657 PP/12 Bugzilla 973: Implement --version.
1659 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1661 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1663 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1665 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1666 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1667 ignore trailing whitespace.
1669 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1671 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1674 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1675 Notification from Dr Andrew Aitchison.
1677 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1678 ExtendedDetectionInfo response format.
1679 Notification from John Horne.
1681 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1684 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1685 XSL and documented dependency on system catalogs, with examples of how
1688 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1691 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1692 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1693 configuration file which is writeable by the Exim user or group.
1695 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1696 of configuration files to cover files specified with the -C option if
1697 they are going to be used with root privileges, not just the default
1700 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1701 option (effectively making it always true).
1703 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1704 files to be used while preserving root privileges.
1706 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1707 that rogue child processes cannot use them.
1709 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1710 run-time user, instead of root.
1712 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1713 Exim run-time user without dropping privileges.
1715 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1716 result string, instead of calling string_vformat() twice with the same
1719 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1720 for other users. Others should always drop root privileges if they use
1721 -C on the command line, even for a whitelisted configure file.
1723 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1725 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1731 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1732 $data_path, and $header_path variables; fixed documentation bugs and
1735 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1736 exipick to access non-standard spools, including the "frozen" queue
1739 NM/01 Bugzilla 965: Support mysql stored procedures.
1740 Patch from Alain Williams
1742 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1744 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1745 Patch from Andreas Metzler
1747 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1748 Patch from Kirill Miazine
1750 NM/05 Bugzilla 671: Added umask to procmail example.
1752 JJ/03 installed exipick 20100323.0, fixing doc bug
1754 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1755 directory. Notification and patch from Dan Rosenberg.
1757 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1759 TK/02 Improve log output when DKIM signing operation fails.
1761 MH/01 Treat the transport option dkim_domain as a colon separated
1762 list, not as a single string, and sign the message with each element,
1763 omitting multiple occurences of the same signer.
1765 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1766 Bugzilla 985, 986. Patch by Simon Arlott
1768 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1769 Patch by Simon Arlott
1771 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1772 MBX locking. Notification from Dan Rosenberg.
1778 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1780 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1782 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1784 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1786 NM/04 Bugzilla 915: Flag broken perl installation during build.
1792 TK/01 Added patch by Johannes Berg that expands the main option
1793 "spamd_address" if it starts with a dollar sign.
1795 TK/02 Write list of recipients to X-Envelope-Sender header when building
1796 the mbox-format spool file for content scanning (suggested by Jakob
1799 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1800 (http://www.dcc-servers.net/) support via dccifd. Activated by
1801 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1803 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1804 by Mark Daniel Reidel <mr@df.eu>.
1806 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1807 When building exim an external PCRE library is now needed -
1808 PCRE is a system library on the majority of modern systems.
1809 See entry on PCRE_LIBS in EDITME file.
1811 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1812 conversation. Added nologin parameter to request.
1813 Patch contributed by Kirill Miazine.
1815 TF/01 Do not log submission mode rewrites if they do not change the address.
1817 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1819 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1820 log files in place. Contributed by Roberto Lima.
1822 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1824 TF/03 Bugzilla 615: When checking the local_parts router precondition
1825 after a local_part_suffix or local_part_prefix option, Exim now
1826 does not use the address's named list lookup cache, since this
1827 contains cached lookups for the whole local part.
1829 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1830 Robert Millan. Documentation is in experimental-spec.txt.
1832 TF/04 Bugzilla 668: Fix parallel build (make -j).
1834 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1836 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1837 Patch provided by Jan Srzednicki.
1839 TF/05 Leading white space used to be stripped from $spam_report which
1840 wrecked the formatting. Now it is preserved.
1842 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1843 that they are available at delivery time.
1845 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1847 TF/08 TLS error reporting now respects the incoming_interface and
1848 incoming_port log selectors.
1850 TF/09 Produce a more useful error message if an SMTP transport's hosts
1851 setting expands to an empty string.
1853 NM/06 Bugzilla 744: EXPN did not work under TLS.
1854 Patch provided by Phil Pennock.
1856 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1857 Patch provided by Richard Godbee.
1859 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1860 acl_smtp_notquit, added index entry.
1862 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1863 Patch provided by Eugene Bujak.
1865 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1866 accept(). Patch provided by Maxim Dounin.
1868 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1869 Patch provided by Phil Pennock.
1871 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1873 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1874 Patch provided by Brad "anomie" Jorsch.
1876 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1877 Patch provided by Dean Brooks.
1879 TK/05 Add native DKIM support (does not depend on external libraries).
1881 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1882 Patch provided by Graeme Fowler.
1884 NM/16 Bugzilla 851: Documentation example syntax fix.
1886 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1888 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1891 NM/19 Bugzilla 745: TLS version reporting.
1892 Patch provided by Phil Pennock.
1894 NM/20 Bugzilla 167: bool: condition support.
1895 Patch provided by Phil Pennock.
1897 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1898 clients. Patch provided by Phil Pennock.
1900 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1901 Patch provided by Brad "anomie" Jorsch.
1903 NM/23 Bugzilla 687: Fix misparses in eximstats.
1904 Patch provided by Heiko Schlittermann.
1906 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1907 Patch provided by Heiko Schlittermann.
1909 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1910 plus update to original patch.
1912 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1914 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1915 Patch provided by David Brownlee.
1917 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1919 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1921 NM/30 Bugzilla 888: TLS documentation bugfixes.
1923 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1925 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1926 Unlike the original bugzilla I have changed all shell scripts in src tree.
1928 NM/33 Bugzilla 898: Transport filter timeout fix.
1929 Patch by Todd Rinaldo.
1931 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1932 Patch by Serge Demonchaux.
1934 NM/35 Bugzilla 39: Base64 decode bug fixes.
1935 Patch by Jakob Hirsch.
1937 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1939 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1941 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1943 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1949 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1950 ALT-N's libdkim that I have put here:
1951 http://duncanthrax.net/exim-experimental/
1953 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1954 were called 'true' and it seems that C99 defines that as a reserved
1955 keyword to be used with 'bool' variable types. That means you could
1956 not include C99-style headers which use bools without triggering
1957 build errors in sieve.c.
1959 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1960 as mailq or other aliases. Changed the --help handling significantly
1961 to do whats expected. exim_usage() emits usage/help information.
1963 SC/01 Added the -bylocaldomain option to eximstats.
1965 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1967 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1969 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1975 PH/01 Another patch from the Sieve maintainer.
1977 PH/02 When an IPv6 address is converted to a string for single-key lookup
1978 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1979 dots are used instead of colons so that keys in lsearch files need not
1980 contain colons. This was done some time before quoting was made available
1981 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1982 (notated using the quote facility) so as to distinguish them from IPv4
1983 keys. This meant that lookups for IP addresses in host lists did not work
1984 for iplsearch lookups.
1986 This has been fixed by arranging for IPv6 addresses to be expressed with
1987 colons if the lookup type is iplsearch. This is not incompatible, because
1988 previously such lookups could never work.
1990 The situation is now rather anomalous, since one *can* have colons in
1991 ordinary lsearch keys. However, making the change in all cases is
1992 incompatible and would probably break a number of configurations.
1994 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1997 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1998 conversion specification without a maximum field width, thereby enabling
1999 a rogue spamd server to cause a buffer overflow. While nobody in their
2000 right mind would setup Exim to query an untrusted spamd server, an
2001 attacker that gains access to a server running spamd could potentially
2002 exploit this vulnerability to run arbitrary code as the Exim user.
2004 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2005 $primary_hostname instead of what libspf2 thinks the hosts name is.
2007 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2008 a directory entry by the name of the lookup key. Previously, if a
2009 symlink pointed to a non-existing file or a file in a directory that
2010 Exim lacked permissions to read, a lookup for a key matching that
2011 symlink would fail. Now it is enough that a matching directory entry
2012 exists, symlink or not. (Bugzilla 503.)
2014 PH/03 The body_linecount and body_zerocount variables are now exported in the
2017 PH/04 Added the $dnslist_matched variable.
2019 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2020 This means they are set thereafter only if the connection becomes
2023 PH/06 Added the client_condition to authenticators so that some can be skipped
2024 by clients under certain conditions.
2026 PH/07 The error message for a badly-placed control=no_multiline_responses left
2027 "_responses" off the end of the name.
2029 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2031 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2032 (without spaces) instead of just copying the configuration text.
2034 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2036 PH/11 Added $max_received_linelength.
2038 PH/12 Added +ignore_defer and +include_defer to host lists.
2040 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2041 way in which PCRE > 7.0 is built.
2043 PH/14 Implemented queue_only_load_latch.
2045 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2046 MAIL command. The effect was to mangle the value on 64-bit systems.
2048 PH/16 Another patch from the Sieve maintainer.
2050 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2052 PH/18 If a system quota error occurred while trying to create the file for
2053 a maildir delivery, the message "Mailbox is full" was not appended to the
2054 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2055 only to a quota excession during the actual writing of the file.
2057 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2058 characters?) which causes problems in log lines. The DN values are now
2059 passed through string_printing() before being added to log lines.
2061 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2062 and InterBase are left for another time.)
2064 PH/21 Added message_body_newlines option.
2066 PH/22 Guard against possible overflow in moan_check_errorcopy().
2068 PH/23 POSIX allows open() to be a macro; guard against that.
2070 PH/24 If the recipient of an error message contained an @ in the local part
2071 (suitably quoted, of course), incorrect values were put in $domain and
2072 $local_part during the evaluation of errors_copy.
2078 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2079 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2082 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2083 issue a MAIL command.
2085 PH/02 In an ACL statement such as
2087 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2089 if a client was not listed at all, or was listed with a value other than
2090 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2091 the condition was not true (as it should be), so access was not denied.
2092 The bug was that the ! inversion was incorrectly passed on to the second
2093 item. This has been fixed.
2095 PH/03 Added additional dnslists conditions == and =& which are different from
2096 = and & when the dns lookup returns more than one IP address.
2098 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2099 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2101 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2102 FSYNC, which compiles an option called disable_fsync that allows for
2103 bypassing fsync(). The documentation is heavily laced with warnings.
2105 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2107 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2108 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2109 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2110 including adding "make clean"; (3) Added -fPIC when compiling the test
2111 dynamically loaded module, to get rid of a warning.
2113 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2114 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2115 = 0s. The bug is otherwise harmless.
2117 PH/07 There was a bug in the dovecot authenticator such that the value of
2118 $auth1 could be overwritten, and so not correctly preserved, after a
2119 successful authentication. This usually meant that the value preserved by
2120 the server_setid option was incorrect.
2122 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2124 PH/09 Installed PCRE release 7.0.
2126 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2127 run for batched SMTP input. It is now run at the start of every message
2128 in the batch. While fixing this I discovered that the process information
2129 (output by running exiwhat) was not always getting set for -bs and -bS
2130 input. This is fixed, and it now also says "batched" for BSMTP.
2132 PH/11 Added control=no_pipelining.
2134 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2135 patch, slightly modified), and move the expansion of helo_data till after
2136 the connection is made in the smtp transport (so it can use these
2139 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2141 PH/14 Added log_selector = +pid.
2143 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2145 PH/16 Add ${if forany and ${if forall.
2147 PH/17 Added dsn_from option to vary the From: line in DSNs.
2149 PH/18 Flush SMTP output before performing a callout, unless control =
2150 no_callout_flush is set.
2152 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2153 was true (the default) a successful delivery failed to delete the retry
2154 item, thus causing premature timeout of the address. The bug is now
2157 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2159 PH/21 Long custom messages for fakedefer and fakereject are now split up
2160 into multiline responses in the same way that messages for "deny" and
2161 other ACL rejections are.
2163 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2164 with slight modification.
2166 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2167 draft, changing the syntax and factoring some duplicate code".
2169 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2170 for deliveries of the second and subsequent messages over the same SMTP
2173 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2174 ${reduce, with only minor "tidies".
2176 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2178 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2179 expansion side effects.
2181 PH/27 When a message times out after an over-quota error from an Exim-imposed
2182 quota, the bounce message says "mailbox is full". This message was not
2183 being given when it was a system quota that was exceeded. It now should
2186 MH/03 Made $recipients available in local_scan(). local_scan() already has
2187 better access to the recipient list through recipients_list[], but
2188 $recipients can be useful in postmaster-provided expansion strings.
2190 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2191 in the case of a MAIL command with additional options following the
2192 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2193 were accidentally chopped off.
2195 PH/29 SMTP synchronization checks are implemented when a command is read -
2196 there is a check that no more input is waiting when there shouldn't be
2197 any. However, for some commands, a delay in an ACL can mean that it is
2198 some time before the response is written. In this time, more input might
2199 arrive, invalidly. So now there are extra checks after an ACL has run for
2200 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2201 pipelining has not been advertised.
2203 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2205 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2206 returned in the SMTP response when smtp_return_error_details was false.
2207 This has been fixed.
2209 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2210 instead of the C I/O that was originally supplied, because problems were
2211 reported on Solaris.
2213 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2214 Exim which did not show up earlier: it was assuming that a call to
2215 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2216 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2217 was a macro that became an assignment, so it seemed to work. This has
2218 changed to a proper function call with a void return, hence the compile
2219 error. Exim's code has been fixed.
2221 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2224 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2226 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2228 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2229 because it turns out to be too big or there is a timeout) there is no
2230 "Completed" line in the log. When some messages of this type were
2231 selected by exigrep, they were listed as "not completed". Others were
2232 picked up by some special patterns. I have improved the selection
2233 criteria to be more general.
2235 PH/38 The host_find_failed option in the manualroute router can now be set
2236 to "ignore", to completely ignore a host whose IP address cannot be
2237 found. If all hosts are ignored, the behaviour is controlled by the new
2238 host_all_ignored option.
2240 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2241 homing or because of multiple MX records with /mx) generated more than
2242 one IP address, and the following item turned out to be the local host,
2243 all the secondary addresses of the first item were incorrectly removed
2244 from the list, along with the local host and any following hosts (which
2245 is what is supposed to happen).
2247 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2248 whoever called Exim into the -H file. In the case of the daemon it was
2249 behaving confusingly. When first started, it used values for whoever
2250 started the daemon, but after a SIGHUP it used the Exim user (because it
2251 calls itself on a restart). I have changed the code so that it now always
2254 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2255 message are rejected with the same error (e.g. no authentication or bad
2256 sender address), and a DATA command is nevertheless sent (as can happen
2257 with PIPELINING or a stupid MUA), the error message that was given to the
2258 RCPT commands is included in the rejection of the DATA command. This is
2259 intended to be helpful for MUAs that show only the final error to their
2262 PH/42 Another patch from the Sieve maintainer.
2264 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2265 Eximstats - Fixed some broken HTML links and added missing column headers
2267 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2268 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2270 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2272 PH/43 Yet another patch from the Sieve maintainer.
2274 PH/44 I found a way to check for a TCP/IP connection going away before sending
2275 the response to the final '.' that terminates a message, but only in the
2276 case where the client has not sent further data following the '.'
2277 (unfortunately, this is allowed). However, in many cases there won't be
2278 any further data because there won't be any more messages to send. A call
2279 to select() can be used: if it shows that the input is "ready", there is
2280 either input waiting, or the socket has been closed. An attempt to read
2281 the next input character can distinguish the two cases. Previously, Exim
2282 would have sent an OK response which the client would never have see.
2283 This could lead to message repetition. This fix should cure that, at
2284 least in a lot of common cases.
2286 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2287 advertised in response to EHLO.
2293 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2294 fixed by 4.65/MH/01 (is this a record?) are fixed:
2296 (i) An empty string was always treated as zero by the numeric comparison
2297 operators. This behaviour has been restored.
2299 (ii) It is documented that the numeric comparison operators always treat
2300 their arguments as decimal numbers. This was broken in that numbers
2301 starting with 0 were being interpreted as octal.
2303 While fixing these problems I realized that there was another issue that
2304 hadn't been noticed. Values of message_size_limit (both the global option
2305 and the transport option) were treated as octal if they started with 0.
2306 The documentation was vague. These values are now always treated as
2307 decimal, and I will make that clear in the documentation.
2313 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2314 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2317 MH/01 Don't check that the operands of numeric comparison operators are
2318 integers when their expansion is in "skipping" mode (fixes bug
2319 introduced by 4.64-PH/07).
2321 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2322 child addresses, Exim now panics and dies. Previously, because the count
2323 is held in a short int, deliveries were likely to be lost. As such a
2324 large number of recipients for a single message is ridiculous
2325 (performance will be very, very poor), I have chosen to impose a limit
2326 rather than extend the field.
2332 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2333 leftover -K file (the existence of which was triggered by #402).
2334 While we were at it, introduced process PID as part of the -K
2335 filename. This should rule out race conditions when creating
2338 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2339 processing considerably. Previous code took too long for large mails,
2340 triggering a timeout which in turn triggers #401.
2342 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2343 in the DK code in transports.c. sendfile() is not really portable,
2344 hence the _LINUX specificness.
2346 TF/01 In the add_headers option to the mail command in an Exim filter,
2347 there was a bug that Exim would claim a syntax error in any
2348 header after the first one which had an odd number of characters
2351 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2352 callout verification, Exim cached a "reject" for the entire domain. This
2353 is correct for most verifications, but it is not correct for a recipient
2354 verification with use_sender or use_postmaster set, because in that case
2355 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2356 case of MAIL FROM:<> rejection from other early rejections (e.g.
2357 rejection of HELO). When verifying a recipient using a non-null MAIL
2358 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2359 Whatever the result of the callout, the value of the domain cache is
2360 left unchanged (for any other kind of callout, getting as far as trying
2361 RCPT means that the domain itself is ok).
2363 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2366 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2367 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2368 session, the daemon ignores ECONNECT errors and logs others; it now
2369 ignores EPIPE as well.
2371 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2372 (quoted-printable decoding).
2374 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2375 later the small subsequent patch to fix an introduced bug.
2377 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2379 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2381 PH/08 An error is now given if message_size_limit is specified negative.
2383 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2384 to be given (somewhat) arbitrary names.
2386 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2389 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2390 miscellaneous code fixes
2392 PH/10 Added the log_reject_target ACL modifier to specify where to log
2395 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2396 hostname. This is wrong, because it relates to the incoming message (and
2397 probably the interface on which it is arriving) and not to the outgoing
2398 callout (which could be using a different interface). This has been
2399 changed to use the value of the helo_data option from the smtp transport
2400 instead - this is what is used when a message is actually being sent. If
2401 there is no remote transport (possible with a router that sets up host
2402 addresses), $smtp_active_hostname is used.
2404 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2405 tweaks were necessary in order to get it to work (see also 21 below):
2406 (a) The code assumed that strncpy() returns a negative number on buffer
2407 overflow, which isn't the case. Replaced with Exim's string_format()
2409 (b) There were several signed/unsigned issues. I just did the minimum
2410 hacking in of casts. There is scope for a larger refactoring.
2411 (c) The code used strcasecmp() which is not a standard C function.
2412 Replaced with Exim's strcmpic() function.
2413 (d) The code set only $1; it now sets $auth1 as well.
2414 (e) A simple test gave the error "authentication client didn't specify
2415 service in request". It would seem that Dovecot has changed its
2416 interface. Fortunately there's a specification; I followed it and
2417 changed what the client sends and it appears to be working now.
2419 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2422 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2423 address A is aliased to B and C, where B exists and C does not. Without
2424 -v the output is "A verified" because verification stops after a
2425 successful redirection if more than one address is generated. However,
2426 with -v the child addresses are also verified. Exim was outputting "A
2427 failed to verify" and then showing the successful verification for C,
2428 with its parentage. It now outputs "B failed to verify", showing B's
2429 parentage before showing the successful verification of C.
2431 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2432 look up a TXT record in a specific list after matching in a combined
2435 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2436 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2437 they consult the DNS. I had assumed they would set it the way they
2438 wanted; and indeed my experiments on Linux seem to show that in some
2439 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2440 To be on the safe side, however, I have now made the interface to
2441 host_find_byname() similar to host_find_bydns(), with an argument
2442 containing the DNS resolver options. The host_find_byname() function now
2443 sets these options at its start, just as host_find_bydns() does. The smtp
2444 transport options dns_qualify_single and dns_search_parents are passed to
2445 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2446 of host_find_byname() use the default settings of RES_DEFNAMES
2447 (qualify_single) but not RES_DNSRCH (search_parents).
2449 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2450 spool_read_header() do less string testing, by means of a preliminary
2451 switch on the second character of optional "-foo" lines. (This is
2452 overdue, caused by the large number of possibilities that now exist.
2453 Originally there were few.) While I was there, I also converted the
2454 str(n)cmp tests so they don't re-test the leading "-" and the first
2455 character, in the hope this might squeeze out yet more improvement.
2457 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2458 flag allowing group syntax was set by the header_syntax check but not
2459 turned off, possible causing trouble later; (2) The flag was not being
2460 set at all for the header_verify test, causing "group"-style headers to
2461 be rejected. I have now set it in this case, and also caused header_
2462 verify to ignore an empty address taken from a group. While doing this, I
2463 came across some other cases where the code for allowing group syntax
2464 while scanning a header line wasn't quite right (mostly, not resetting
2465 the flag correctly in the right place). These bugs could have caused
2466 trouble for malformed header lines. I hope it is now all correct.
2468 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2469 with the "reply" argument non-NULL. The code, however (which originally
2470 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2471 but it didn't always do it. This confused somebody who was copying the
2472 code for some other use. I have removed all the tests.
2474 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2475 feature that was used to support insecure browsers during the U.S. crypto
2476 embargo. It requires special client support, and Exim is probably the
2477 only MTA that supported it -- and would never use it because real RSA is
2478 always available. This code has been removed, because it had the bad
2479 effect of slowing Exim down by computing (never used) parameters for the
2480 RSA_EXPORT functionality.
2482 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2483 authenticator to fail if there's a tab character in the incoming data
2484 (there should never be unless someone is messing about, as it's supposed
2485 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2486 if the connection is using TLS or if the remote IP is the same as the
2487 local IP, and the "valid-client-cert option" if a client certificate has
2490 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2491 authenticators. This can be used for authorization after authentication
2492 succeeds. (In the case of plaintext, it servers for both authentication
2495 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2496 if any retry times were supplied.
2498 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2499 connection, where there is no client IP address to check. In this
2500 situation, the verify now always succeeds.
2502 PH/25 Applied John Jetmore's -Mset patch.
2504 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2506 PH/27 In a string expansion for a processed (not raw) header when multiple
2507 headers of the same name were present, leading whitespace was being
2508 removed from all of them, but trailing whitespace was being removed only
2509 from the last one. Now trailing whitespace is removed from each header
2510 before concatenation. Completely empty headers in a concatenation (as
2511 before) are ignored.
2513 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2514 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2516 PH/29 [Removed. This was a change that I later backed out, and forgot to
2517 correct the ChangeLog entry (that I had efficiently created) before
2518 committing the later change.]
2520 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2521 address errors (4xx response to RCPT) over the same connection as other
2522 messages routed to the same hosts. Such deliveries are always "forced",
2523 so retry times are not inspected. This resulted in far too many retries
2524 for the affected addresses. The effect occurred only when there were more
2525 hosts than the hosts_max_try setting in the smtp transport when it had
2526 the 4xx errors. Those hosts that it had tried were not added to the list
2527 of hosts for which the message was waiting, so if all were tried, there
2528 was no problem. Two fixes have been applied:
2530 (i) If there are any address or message errors in an SMTP delivery, none
2531 of the hosts (tried or untried) are now added to the list of hosts
2532 for which the message is waiting, so the message should not be a
2533 candidate for sending over the same connection that was used for a
2534 successful delivery of some other message. This seems entirely
2535 reasonable: after all the message is NOT "waiting for some host".
2536 This is so "obvious" that I'm not sure why it wasn't done
2537 previously. Hope I haven't missed anything, but it can't do any
2538 harm, as the worst effect is to miss an optimization.
2540 (ii) If, despite (i), such a delivery is accidentally attempted, the
2541 routing retry time is respected, so at least it doesn't keep
2542 hammering the server.
2544 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2545 in ${readsocket because some servers need this prod.
2547 PH/32 Added some extra debug output when updating a wait-xxx database.
2549 PH/33 The hint "could be header name not terminated by colon", which has been
2550 given for certain expansion errors for a long time, was not being given
2551 for the ${if def:h_colon_omitted{... case.
2553 PH/34 The spec says: "With one important exception, whenever a domain list is
2554 being scanned, $domain contains the subject domain." There was at least
2555 one case where this was not true.
2557 PH/35 The error "getsockname() failed: connection reset by peer" was being
2558 written to the panic log as well as the main log, but it isn't really
2559 panic-worthy as it just means the connection died rather early on. I have
2560 removed the panic log writing for the ECONNRESET error when getsockname()
2563 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2564 runs only) independently of the message's sender address. This meant
2565 that, if the 4xx error was in fact related to the sender, a different
2566 message to the same recipient with a different sender could confuse
2567 things. In particular, this can happen when sending to a greylisting
2568 server, but other circumstances could also provoke similar problems.
2569 I have changed the default so that the retry time for these errors is now
2570 based a combination of the sender and recipient addresses. This change
2571 can be overridden by setting address_retry_include_sender=false in the
2574 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2575 remote server are returned as part of bounce messages. This was not
2576 happening for LMTP over a pipe (the lmtp transport), but now it is the
2577 same for both kinds of LMTP.
2579 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2580 in header lines that were in fact CNAMEs. This is no longer the case.
2582 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2583 and queue runs started by the daemon processed all messages. This has
2584 been fixed so that -R and -S can now usefully be given with -q<time>.
2586 PH/40 Import PCRE release 6.7 (fixes some bugs).
2588 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2590 PH/42 Give an error if -q is specified more than once.
2592 PH/43 Renamed the variables $interface_address and $interface_port as
2593 $received_ip_address and $received_port, to make it clear that these
2594 values apply to message reception, and not to the outgoing interface when
2595 a message is delivered. (The old names remain recognized, of course.)
2597 PH/44 There was no timeout on the connect() call when using a Unix domain
2598 socket in the ${readsocket expansion. There now is.
2600 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2601 be meaningful with "accept".
2603 SC/01 Eximstats V1.43
2604 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2606 SC/02 Eximstats V1.44
2607 Use a glob alias rather than an array ref in the generated
2608 parser. This improves both readability and performance.
2610 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2611 Collect SpamAssassin and rejection statistics.
2612 Don't display local sender or destination tables unless
2613 there is data to show.
2614 Added average volumes into the top table text output.
2616 SC/04 Eximstats V1.46
2617 Collect data on the number of addresses (recipients)
2618 as well as the number of messages.
2620 SC/05 Eximstats V1.47
2621 Added 'Message too big' to the list of mail rejection
2622 reasons (thanks to Marco Gaiarin).
2624 SC/06 Eximstats V1.48
2625 Mainlog lines which have GMT offsets and are too short to
2626 have a flag are now skipped.
2628 SC/07 Eximstats V1.49 (Alain Williams)
2629 Added the -emptyok flag.
2631 SC/08 Eximstats V1.50
2632 Fixes for obtaining the IP address from reject messages.
2634 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2635 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2636 whitespace changes from 4.64-PH/27
2638 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2641 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2642 are found, allow negative numbers in numeric criteria)
2644 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2646 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2647 to match changes made in 4.64-PH/43
2649 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2651 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2652 "domain" is actually an IP address, and force a failure. This locks out
2653 those revolvers/nameservers that support "A-for-A" lookups, in
2654 contravention of the specifications.
2656 PH/48 When a host name was looked up from an IP address, and the subsequent
2657 forward lookup of the name timed out, the host name was left in
2658 $sender_host_name, contrary to the specification.
2660 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2661 restricted to single-key lookups, Exim was not diagnosing an error if
2662 * or *@ was used with a query-style lookup.
2664 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2666 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2667 long ago, but noone interested enough thought of it. Let's just say that
2668 the "1.1" means that there are some new functions that weren't there at
2669 some point in the past.
2671 PH/51 Error processing for expansion failure of helo_data from an smtp
2672 transport during callout processing was broken.
2674 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2675 tested/used via the -bh/-bhc/-bs options.
2677 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2678 bug, fixed in subsequent PCRE releases).
2680 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2681 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2683 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2689 SC/01 Use a glob alias rather than an array ref in eximstats generated
2690 parser. This improves both readability and performance.
2692 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2693 Don't display local sender or destination tables in eximstats unless
2694 there is data to show.
2695 Added average volumes into the eximstats top table text output.
2697 SC/03 Collect data on the number of addresses (recipients) as well
2698 as the number of messages in eximstats.
2700 TF/01 Correct an error in the documentation for the redirect router. Exim
2701 does (usually) call initgroups() when daemonizing.
2703 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2704 with consistent privilege compared to when running as a daemon.
2706 TF/03 Note in the spec that $authenticated_id is not set for local
2707 submissions from trusted users.
2709 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2710 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2712 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2713 by adding some example configuration directives to the default
2714 configuration file. A little bit of work is required to uncomment the
2715 directives and define how usernames and passwords are checked, but
2716 there is now a framework to start from.
2718 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2719 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2720 without this. I don't know how relevant this is to other LDAP libraries.
2722 PH/02 Add the verb name to the "unknown ACL verb" error.
2724 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2726 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2728 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2729 directory not expanded when it should be if an expanded home directory
2730 was set for the address (which is overridden by the transport).
2732 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2735 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2736 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2737 because it is too late at that time, and has no effect.
2739 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2740 security issue with \' (bugzilla #107). I could not use the
2741 PQescapeStringConn() function, because it needs a PGconn value as one of
2744 PH/08 When testing addresses using -bt, indicate those final addresses that
2745 are duplicates that would not cause an additional delivery. At least one
2746 person was confused, thinking that -bt output corresponded to deliveries.
2747 (Suppressing duplicates isn't a good idea as you lose the information
2748 about possibly different redirections that led to the duplicates.)
2750 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2751 systems where poll() doesn't work, in particular OS X.
2753 PH/10 Added more information to debugging output for retry time not reached.
2755 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2756 operations in malware.c.
2758 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2761 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2762 syslog_duplication set false, log lines that would normally be written
2763 both the the main log and to the reject log were not written to syslog at
2766 PH/14 In the default configuration, change the use of "message" in ACL warn
2767 statements to "add_header".
2769 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2770 not followed by a command (e.g. "seen endif").
2772 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2773 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2776 PH/17 Added extra conditions to the default value of delay_warning_condition
2780 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2781 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2782 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2785 The Auto-Submitted: and various List- headers are standardised, whereas I
2786 don't think Precedence: ever was.
2788 PH/18 Refactored debugging code in route_finduser() to show more information,
2789 in particular, the error code if getpwnam() issues one.
2791 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2792 This is apparently needed in addition to the PH/07 change above to avoid
2793 any possible encoding problems.
2795 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2796 but not after initializing Perl.
2798 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2799 output them only if debugging. By default they are written stderr,
2800 apparently, which is not desirable.
2802 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2805 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2808 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2810 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2811 authenticated or an ident call has been made. Suppress the default
2812 values for $authenticated_id and $authenticated_sender (but permit -oMai
2813 and -oMas) when testing with -bh.
2815 PH/24 Re-jigged the order of the tests in the default configuration so that the
2816 tests for valid domains and recipients precede the DNS black list and CSA
2817 tests, on the grounds that those ones are more expensive.
2819 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2820 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2821 This bug exists in every version of Exim that I still have, right back to
2824 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2825 However, an attempt to turn on case-sensitivity in a regex key by
2826 including (?-i) didn't work because the subject string was already
2827 lowercased, and the effects were non-intuitive. It turns out that a
2828 one-line patch can be used to allow (?-i) to work as expected.
2834 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2835 other effects) broke the use of negated acl sub-conditions.
2837 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2840 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2841 "Deny" causes Exim to reject the incoming connection with a 554 error.
2842 Unfortunately, if there is a major crisis, such as a disk failure,
2843 tcp-wrappers gives "deny", whereas what one would like would be some
2844 kind of temporary error. A kludge has been added to help with this.
2845 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2846 554 error is used if errno is still zero or contains ENOENT (which occurs
2847 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2850 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2852 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2853 errors. Otherwise a message that provokes a temporary error (when other
2854 messages do not) can cause a whole host to time out.
2856 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2857 addresses were routed directly to files or pipes from a redirect router.
2858 File deliveries just didn't batch; pipe deliveries might have suffered
2861 PH/06 A failure to get a lock for a hints database would erroneously always say
2862 "Failed to get write lock", even when it was really a read lock.
2864 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2865 of 0600. This has been changed to use the value of the lockfile_mode
2866 option (which defaults to 0600).
2868 PH/08 Applied small patch from the Sieve maintainer.
2870 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2871 folder from quota calculations, a direct delivery into this folder messed
2872 up the contents of the maildirsize file. This was because the regex was
2873 used only to exclude .Trash (or whatever) when the size of the mailbox
2874 was calculated. There was no check that a delivery was happening into an
2875 excluded directory. This bug has been fixed by ignoring all quota
2876 processing for deliveries into excluded directories.
2878 PH/10 Added the maildirfolder_create_regex option to appendfile.
2884 PH/01 The code for finding all the local interface addresses on a FreeBSD
2885 system running IPv6 was broken. This may well have applied to all BSD
2886 systems, as well as to others that have similar system calls. The broken
2887 code found IPv4 interfaces correctly, but gave incorrect values for the
2888 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2889 that it would not match correctly against @[] and not recognize the IPv6
2892 PH/02 The ipliteral router was not recognizing addresses of the form user@
2893 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2895 PH/03 Added disable_ipv6.
2897 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2898 decoded form, because it is most often used to construct To: headers
2899 lines in autoreplies, and the decoded form may well be syntactically
2900 invalid. However, $reply_address has leading white space removed, and all
2901 newlines turned into spaces so that the autoreply transport does not
2904 PH/05 If group was specified without a user on a router, and no group or user
2905 was specified on a transport, the group from the router was ignored.
2907 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2908 for visible compile-time settings that can be used to change these
2909 numbers, for those that want even more. Backwards compatibility with old
2910 spool files has been maintained. However, going back to a previous Exim
2911 release will lost any variables that are in spool files.
2913 PH/07 Two small changes when running in the test harness: increase delay when
2914 passing a TCP/IP connection to a new process, in case the original
2915 process has to generate a bounce, and remove special handling of
2916 127.0.0.2 (sic), which is no longer necessary.
2918 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2919 be the same on different OS.
2921 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2924 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2925 whether --show-vars was specified or not
2927 JJ/02 exipick: Added support for new ACL variable spool format introduced
2930 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2931 syntactically invalid From: or Reply-to: line, and a filter used this to
2932 generate an autoreply, and therefore failed to obtain an address for the
2933 autoreply, Exim could try to deliver to a non-existent relative file
2934 name, causing unrelated and misleading errors. What now happens is that
2935 it logs this as a hard delivery error, but does not attempt to create a
2938 PH/11 The exinext utility has a -C option for testing purposes, but although
2939 the given file was scanned by exinext itself; it wasn't being passed on
2940 when Exim was called.
2942 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2943 an end-of-file indication when reading a command response.
2945 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2946 compiled. In many other places in Exim, IPv6 addresses are always
2947 recognized, so I have changed this. It also means that IPv4 domain
2948 literals of the form [IPV4:n.n.n.n] are now always recognized.
2950 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2951 used if the router is not running as root, for example, when verifying at
2952 ACL time, or when using -bh. The debugging output from this situation was
2953 non-existent - all you got was a failure to exec. I have made two
2956 (a) Failures to set uid/gid, the current directory, or a process leader
2957 in a subprocess such as that created by queryprogram now generate
2958 suitable debugging output when -d is set.
2960 (b) The queryprogram router detects when it is not running as root,
2961 outputs suitable debugging information if -d is set, and then runs
2962 the subprocess without attempting to change uid/gid.
2964 PH/15 Minor change to Makefile for building test_host (undocumented testing
2967 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2968 additional section of a DNS packet that returns MX or SRV records.
2969 Instead, it always explicitly searches for A/AAAA records. This avoids
2970 major problems that occur when a DNS server includes only records of one
2971 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2972 fixed another bug: if SRV records were looked up and the corresponding
2973 address records were *not* found in the additional section, the port
2974 values from the SRV records were lost.
2976 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2977 using the correct key (the original address) when searching the retry
2978 rules in order to find which one to use for generating the retry hint.
2980 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2981 adding the default one. Similarly, if it contains a Reply-To: header, the
2982 errors_reply_to option, if set, is not used.
2984 PH/19 When calculating a retry time, Exim used to measure the "time since
2985 failure" by looking at the "first failed" field in the retry record. Now
2986 it does not use this if it is later than than the arrival time of the
2987 message. Instead it uses the arrival time. This makes for better
2988 behaviour in cases where some deliveries succeed, thus re-setting the
2989 "first failed" field. An example is a quota failure for a huge message
2990 when small messages continue to be delivered. Without this change, the
2991 "time since failure" will always be short, possible causing more frequent
2992 delivery attempts for the huge message than are intended.
2993 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2995 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2996 $1, $2, $3) because the numerical variables can be reset during some
2997 expansion items (e.g. "match"), thereby losing the authentication data.
2999 PH/21 Make -bV show the size of off_t variables so that the test suite can
3000 decide whether to run tests for quotas > 2G.
3002 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3003 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3004 filecount value is greater than 2G or if a quota value is greater than 2G
3005 on a system where the size of off_t is not greater than 4, a panic error
3008 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3009 never match. The debug and -bh output now contains an explicit error
3010 message indicating a malformed IPv4 address or mask.
3012 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3013 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3014 PH/23 above applies.
3016 PH/25 Do not write to syslog when running in the test harness. The only
3017 occasion when this arises is a failure to open the main or panic logs
3018 (for which there is an explicit test).
3020 PH/26 Added the /no_tell option to "control=freeze".
3022 PH/27 If a host name lookup failed very early in a connection, for example, if
3023 the IP address matched host_lookup and the reverse lookup yielded a name
3024 that did not have a forward lookup, an error message of the form "no IP
3025 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3026 could be logged. Now it outputs the IP address instead of "NULL".
3028 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3029 allows the sender and the authenticated sender to be set when
3030 submitting a message from within Exim. Since child_open_exim() is
3031 documented for local_scan(), the new function should be too.
3033 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3034 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3035 results in an empty string is now treated as unset.
3037 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3039 PH/31 Added sender_verify_fail logging option.
3041 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3042 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3043 tidied the source and removed it altogether.
3045 PH/33 When a queue run was abandoned because the load average was too high, a
3046 log line was always written; now it is written only if the queue_run log
3047 selector is set. In addition, the log line for abandonment now contains
3048 information about the queue run such as the pid. This is always present
3049 in "start" and "stop" lines but was omitted from the "abandon" line.
3051 PH/34 Omit spaces between a header name and the colon in the error message that
3052 is given when verify = headers_syntax fails (if there are lots of them,
3053 the message gets confusing).
3055 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3056 names, as there are now some PTR records that contain slashes. This check
3057 is only to protect against broken name servers that fall over on strange
3058 characters, so the fact that it applies to all lookups doesn't matter.
3060 PH/36 Now that the new test suite is complete, we can remove some of the
3061 special code in Exim that was needed for the old test suite. For example,
3062 sorting DNS records because real resolvers return them in an arbitrary
3063 order. The new test suite's fake resolver always returns records in the
3066 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3067 bounces) except when queue_only is set, to avoid logging races between
3068 the different processes.
3070 PH/38 Panic-die if .include specifies a non-absolute path.
3072 PH/39 A tweak to the "H" retry rule from its user.
3074 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3075 a label. They prevented compilation on older perls.
3077 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3078 a warning to be raised on newish perls.
3080 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3081 on queue. Changes to match documented behaviour of showing count of
3082 messages matching specified criteria.
3084 PH/40 Changed the default ident timeout from 30s to 5s.
3086 PH/41 Added support for the use of login_cap features, on those BSD systems
3087 that have them, for controlling the resources used by pipe deliveries.
3089 PH/42 The content-scanning code uses fopen() to create files in which to put
3090 message data. Previously it was not paying any attention to the mode of
3091 the files. Exim runs with umask(0) because the rest of the code creates
3092 files with open(), and sets the required mode explicitly. Thus, these
3093 files were ending up world-writeable. This was not a big issue, because,
3094 being within the spool directory, they were not world-accessible. I have
3095 created a function called modefopen, which takes an additional mode
3096 argument. It sets umask(777), creates the file, chmods it to the required
3097 mode, then resets the umask. All the relevant calls to fopen() in the
3098 content scanning code have been changed to use this function.
3100 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3101 to 24 hours. This avoids potential overflow problems when processing G
3102 and H retry rules. I suspect nobody ever tinkers with this value.
3104 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3106 PH/45 When the plaintext authenticator is running as a client, the server's
3107 challenges are checked to ensure they are valid base64 strings. By
3108 default, the authentication attempt is cancelled if an invalid string is
3109 received. Setting client_ignore_invalid_base64 true ignores these errors.
3110 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3111 they are received. Thus, the responses can be made to depend on the
3112 challenges. If an invalid string is ignored, an empty string is placed in
3115 PH/46 Messages that are created by the autoreply transport now contains a
3116 References: header, in accordance with RFCs 2822 and 3834.
3118 PH/47 Added authenticated_sender_force to the smtp transport.
3120 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3122 PH/49 Installed latest patch from the Sieve maintainer.
3124 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3125 was also set, the appendfile transport was unnecessarily scanning a
3126 directory of message files (e.g. for maildir delivery) to find the count
3127 of files (along with the size), even though it did not need this
3128 information. It now does the scan only if it needs to find either the
3129 size of the count of files.
3131 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3133 PH/52 Two bugs concerned with error handling when the smtp transport is
3136 (i) Exim was not creating retry information for temporary errors given
3137 for individual recipients after the DATA command when the smtp transport
3138 was used in LMTP mode. This meant that they could be retried too
3139 frequently, and not timed out correctly.
3141 (ii) Exim was setting the flag that allows error details to be returned
3142 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3143 recipients that were returned after the DATA command.
3145 PH/53 This is related to PH/52, but is more general: for any failing address,
3146 when detailed error information was permitted to be returned to the
3147 sender, but the error was temporary, then after the final timeout, only
3148 "retry timeout exceeded" was returned. Now it returns the full error as
3149 well as "retry timeout exceeded".
3151 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3152 do this, and (what is worse) MTAs that accept it.
3154 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3155 will now be deprecated.
3157 PH/56 New os.c-cygwin from the Cygwin maintainer.
3159 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3160 formats (previously only available in exim formats via -bpr, -bpru,
3161 and -bpra. Now also available in native and exiqgrep formats)
3163 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3164 with very large, slow to parse queues
3166 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3168 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3170 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3171 responses to authentication challenges, though it was showing the
3172 challenges; (ii) I've removed the CR characters from the debug output for
3175 PH/58 Allow for the insertion of a newline as well as a space when a string
3176 is turned into more than one encoded-word during RFC 2047 encoding. The
3177 Sieve code now uses this.
3179 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3180 data_4xx, lost_connection, tls_required.
3182 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3183 message was being sent as an SMTP response.
3185 PH/61 Add -l and -k options to exicyclog.
3187 PH/62 When verifying, if an address was redirected to one new address, so that
3188 verification continued, and the new address failed or deferred after
3189 having set something in $address_data, the value of $address_data was not
3190 passed back to the ACL. This was different to the case when no
3191 redirection occurred. The value is now passed back in both cases.
3193 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3194 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3195 use login_cap.h, so on its own it isn't the distinguishing feature. The
3196 new name refers directly to the setclassresources() function.
3198 PH/65 Added configuration files for NetBSD3.
3200 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3202 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3203 is preferred over IPv4.
3205 PH/68 The bounce_return_message and bounce_return_body options were not being
3206 honoured for bounces generated during the reception of non-SMTP messages.
3207 In particular, this applied to messages rejected by the ACL. This bug has
3208 been fixed. However, if bounce_return_message is true and bounce_return_
3209 body is false, the headers that are returned for a non-SMTP message
3210 include only those that have been read before the error was detected.
3211 (In the case of an ACL rejection, they have all been read.)
3213 PH/69 The HTML version of the specification is now built in a directory called
3214 spec_html instead of spec.html, because the latter looks like a path with
3215 a MIME-type, and this confuses some software.
3217 PH/70 Catch two compiler warnings in sieve.c.
3219 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3220 function verify_get_ident() calls ip_connect() to connect a socket, but
3221 if the "connect()" function timed out, ip_connect() used to close the
3222 socket. However, verify_get_ident() also closes the socket later, and in
3223 between Exim writes to the log, which may get opened at this point. When
3224 the socket was closed in ip_connect(), the log could get the same file
3225 descriptor number as the socket. This naturally causes chaos. The fix is
3226 not to close the socket in ip_connect(); the socket should be closed by
3227 the function that creates it. There was only one place in the code where
3228 this was missing, in the iplookup router, which I don't think anybody now
3229 uses, but I've fixed it anyway.
3231 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3232 well as to direct DNS lookups. Otherwise the handling of names in host
3233 lists is inconsistent and therefore confusing.
3239 PH/01 Two changes to the default runtime configuration:
3241 (1) Move the checks for relay_from_hosts and authenticated clients from
3242 after to before the (commented out) DNS black list checks.
3244 (2) Add control=submission to the relay_from_hosts and authenticated
3245 clients checks, on the grounds that messages accepted by these
3246 statements are most likely to be submissions.
3248 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3250 (1) Generate an error if the third argument for the ${prvs expansion is
3253 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3256 (3) Reset the variables that are obtained from the first argument of
3257 ${prvscheck and used in the second argument before leaving the code,
3258 because their memory is reclaimed, so using them afterwards may do
3261 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3262 one (it's much easier than Tom thought :-).
3264 (5) Because of (4), we can now allow for the use of $prvscheck_result
3265 inside the third argument.
3267 PH/03 For some reason, the default setting of PATH when running a command from
3268 a pipe transport was just "/usr/bin". I have changed it to
3271 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3272 anything to be listed in the output from -bV.
3274 PH/05 When a filter generated an autoreply, the entire To: header line was
3275 quoted in the delivery log line, like this:
3277 => >A.N.Other <ano@some.domain> <original@ddress> ...
3279 This has been changed so that it extracts the operative address. There
3280 may be more than one such address. If so, they are comma-separated, like
3283 => >ano@some.domain,ona@other.domain <original@ddress> ...
3285 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3286 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3287 not being looked up in the rDNS to get a host name, Exim was showing the
3288 IP address twice in Received: lines, even though the IP addresses were
3289 identical. For example:
3291 Received: from [1.2.3.4] (helo=[1.2.3.4])
3293 However, if the real host name was known, it was omitting the HELO data
3294 if it matched the actual IP address. This has been tidied up so that it
3295 doesn't show the same IP address twice.
3297 PH/07 When both +timestamp and +memory debugging was on, the value given by
3298 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3299 called by the debug printing, thereby overwriting the timestamp buffer.
3300 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3302 PH/08 When the original message was included in an autoreply transport, it
3303 always said "this is a copy of the message, including all the headers",
3304 even if body_only or headers_only was set. It now gives an appropriate
3307 PH/09 Applied a patch from the Sieve maintainer which:
3309 o fixes some comments
3310 o adds the (disabled) notify extension core
3311 o adds some debug output for the result of if/elsif tests
3312 o points to the current vacation draft in the documentation
3313 and documents the missing references header update
3317 o fixes a bug in processing the envelope test (when testing
3318 multiple envelope elements, the last element determined the
3321 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3322 Electronic Mail") by including:
3324 Auto-submitted: auto-generated
3326 in the messages that it generates (bounce messages and others, such as
3327 warnings). In the case of bounce messages for non-SMTP messages, there was
3328 also a typo: it was using "Auto_submitted" (underscore instead of
3329 hyphen). Since every message generated by Exim is necessarily in response
3330 to another message, thes have all been changed to:
3332 Auto-Submitted: auto-replied
3334 in accordance with these statements in the RFC:
3336 The auto-replied keyword:
3338 - SHOULD be used on messages sent in direct response to another
3339 message by an automatic process,
3341 - MUST NOT be used on manually-generated messages,
3343 - MAY be used on Delivery Status Notifications (DSNs) and Message
3344 Disposition Notifications (MDNs),
3346 - MUST NOT be used on messages generated by automatic or periodic
3347 processes, except for messages which are automatic responses to
3350 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3351 to the default Received: header definition.
3353 PH/12 Added log selector acl_warn_skipped (default on).
3355 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3356 variables because (a) they are in the wrong storage pool and (b) even if
3357 they were copied, it wouldn't work properly because of the caching.
3359 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3360 checking when decoding. Apparently there are clients that generate
3361 overlong encoded strings. Why am I not surprised?
3363 PH/15 If the first argument of "${if match_address" was not empty, but did not
3364 contain an "@" character, Exim crashed. Now it writes a panic log message
3365 and treats the condition as false.
3367 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3369 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3370 extension "envelope-auth". The code is finished and in agreement with
3371 other implementations, but there is no documentation so far and in fact,
3372 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3373 not changing the active code.
3375 Print executed "if" and "elsif" statements when debugging is used. This
3376 helps a great deal to understand what a filter does.
3378 Document more things not specified clearly in RFC3028. I had all this
3379 sorted out, when out of a sudden new issues came to my mind. Oops."
3381 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3384 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3385 canonical form (as documented). However, after a host name lookup from
3386 the IP address, check_host() was doing a simple string comparison with
3387 addresses acquired from the DNS when checking that the found name did
3388 have the original IP as one of its addresses. Since any found IPv6
3389 addresses are likely to be in abbreviated form, the comparison could
3390 fail. Luckily, there already exists a function for doing the comparison
3391 by converting both addresses to binary, so now that is used instead of
3392 the text comparison.
3394 PH/20 There was another similar case to PH/19, when a complete host name was
3395 given in a host list; looking up its IP address could give an abbreviated
3396 form, whereas the current host's name might or might not be abbreviated.
3397 The same fix has been applied.
3403 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3404 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3407 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3408 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3410 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3412 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3413 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3414 Also, for "Auto-Submitted", treat anything other than "no" as
3415 non-personal, in accordance with RFC 3834. (Previously it treated
3416 anything starting "auto-" as non-personal.)
3418 TF/01 The control=submission/name=... option had a problem with syntax
3419 errors if the name included a slash character. The /name= option
3420 now slurps the rest of the string, so it can include any characters
3421 but it must come last in the list of options (after /sender_retain
3424 PH/05 Some modifications to the interface to the fake nameserver for the new
3432 TK/01 Added the "success_on_redirect" address verification option. See
3433 NewStuff for rationale and an example.
3435 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3437 PH/02 Patch to exigrep to allow it to work on syslog lines.
3439 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3440 fread() to skip over the body file's header line, because in Cygwin the
3441 header line is locked and is inaccessible.
3443 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3444 co-exist for some time) to make it clear that it is the Exim ID that is
3445 referenced, not the Message-ID: header line.
3447 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3448 string_format() function, because snprintf() does not exist on all
3451 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3452 ${stat: expansion item.
3454 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3455 protocol synchronization error", to keep the pedants happy.
3457 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3458 well as for IRIX systems, when gcc is being used. See the host.c source
3461 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3463 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3466 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3467 transport and to the smtp transport in LMTP mode.
3469 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3471 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3472 run a filter in a subprocess. This could lead to confusion in subsequent
3473 lookups in the parent process. There should also be a search_tidyup() at
3474 the end of the subprocess.
3476 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3477 only if the host matched helo_try_verify_hosts, which caused the
3478 verification to occur when the EHLO/HELO command was issued. The ACL just
3479 tested the remembered result. Now, if a previous verification attempt has
3480 not happened, "verify = helo" does it there and then.
3482 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3484 TK/03 Fix log output including CR from clamd.
3486 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3487 exist provoked a memory error which could cause a segfault.
3489 PH/15 Installed PCRE 6.2
3491 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3493 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3494 of the problem. Specifically, suggested +O2 rather than +O1 for the
3497 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3499 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3500 record, the port was not correctly logged when the outgoing_port log
3501 selector was set (it logged the transort's default port).
3503 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3504 fallback_hosts, and "hosts" in the smtp transport.
3506 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3507 host errors such as "Connection refused".
3509 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3510 authentication with radiusclient 0.4.9:
3512 - Error returned from rc_read_config was caught wrongly
3513 - Username/password not passed on to radius server due to wrong length.
3515 The presumption is that some radiusclient API changes for 4.51/PH/17
3516 were not taken care of correctly. The code is still untested by me (my
3517 Linux distribution still has 0.3.2 of radiusclient), but it was
3518 contributed by a Radius user.
3520 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3521 expanding the "port" option of the smtp transport.
3523 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3524 while reading a MIME header. Thanks to Tom Hughes for a patch.
3526 PH/24 Include config.h inside local_scan.h so that configuration settings are
3529 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3530 that in an ACL for RCPT (for example), you can examine exactly what was
3533 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3534 commands, but it was not correctly comparing the address with the actual
3535 client host address. Thus, it would show the EHLO address in Received:
3536 header lines when this was not necessary.
3538 PH/27 Added the % operator to ${eval:}.
3540 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3541 it should be ignoring failures (because with -C, for example, it has lost
3542 privilege). It wasn't ignoring creation failures other than "already
3545 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3548 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3549 because an input line was too long, either on its own, or by virtue of
3550 too many continuations, the temporary file was not being removed, and the
3551 return code was incorrect.
3553 PH/31 Missing "BOOL" in function definition in filtertest.c.
3555 PH/32 Applied Sieve patches from the maintainer.
3557 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3559 PH/33 Added "verify = not_blind".
3561 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3562 Local/Makefile (with some defaults set). These are used in built scripts
3563 such as exicyclog, but they have never been used in the exim_install
3564 script (though there are many overriding facilities there). I have
3565 arranged that the exim_install script now takes note of these two
3568 PH/35 Installed configuration files for Dragonfly.
3570 PH/36 When a locally submitted message by a trusted user did not contain a
3571 From: header, and the sender address was obtained from -f or from an SMTP
3572 MAIL command, and the trusted user did not use -F to supply a sender
3573 name, $originator_name was incorrectly used when constructing a From:
3574 header. Furthermore, $originator_name was used for submission mode
3575 messages from external hosts without From: headers in a similar way,
3576 which is clearly wrong.
3578 PH/37 Added control=suppress_local_fixups.
3580 PH/38 When log_selector = +received_sender was set, and the addition of the
3581 sender made the log line's construction buffer exactly full, or one byte
3582 less than full, an overflow happened when the terminating "\n" was
3585 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3586 when the result of a list match is failure because a DNS lookup failed.
3588 PH/40 RM_COMMAND is now used in the building process.
3590 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3591 the "build-* directories that it finds.
3593 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3594 address in a domain literal was a prefix of an interface address.
3596 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3597 when verifying a sender address, unless rewrite_headers is false.
3599 PH/44 Wrote a long comment about why errors_to addresses are verified as
3600 recipients, not senders.
3602 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3603 the ratelimit ACL was added.
3605 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3607 PH/46 Added extra information about PostgreSQL errors to the error string.
3609 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3610 avoiding the need to install special zones in a real server. This is
3611 backwards compatible; if it can't find the fake resolver, it drops back.
3612 Thus, both old and new test suites can be run.
3614 TF/02 Added util/ratelimit.pl
3616 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3617 clock is set back in time.
3619 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3620 Candler <B.Candler@pobox.com>.
3622 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3623 OK for addresses that are the result of redirections.
3625 PH/48 A number of further additions for the benefit of the new test suite,
3626 including a fake gethostbyname() that interfaces to the fake DNS resolver
3629 TF/06 The fix for widen_domains has also been applied to qualify_single and
3630 search_parents which are the other dnslookup options that can cause
3633 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3636 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3638 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3639 service" if it encountered an MX record with an empty target hostname.
3640 The message is now "an MX or SRV record indicated no SMTP service".
3642 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3643 if the DNS of the sending site is misconfigured. This is quite a
3644 common situation. This change restores the behaviour of treating a
3645 helo verification defer as a failure.
3647 PH/51 If self=fail was set on a router, the bounce message did not include the
3648 actual error message.
3654 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3656 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3657 command itself ended in error, the underlying message about the transport
3658 filter timeout was being overwritten with the pipe command error. Now the
3659 underlying error message should be appended to the second error message.
3661 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3663 PH/02 Reduce the amount of output that "make" produces by default. Full output
3664 can still be requested.
3666 PH/03 The warning log line about a condition test deferring for a "warn" verb
3667 was being output only once per connection, rather than after each
3668 occurrence (because it was using the same function as for successful
3669 "warn" verbs). This seems wrong, so I have changed it.
3671 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3672 it should not have, which might have caused a crash in the right
3673 circumstances, but probably never did.
3675 PH/04 Installed a modified version of Tony Finch's patch to make submission
3676 mode fix the return path as well as the Sender: header line, and to
3677 add a /name= option so that you can make the user's friendly name appear
3680 TF/03 Added the control = fakedefer ACL modifier.
3682 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3683 Mark Lowes for thorough testing.
3685 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3687 TK/03 Merged latest SRS patch from Miles Wilton.
3689 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3690 with the definition in sysexits.h (which is #included earlier).
3691 Fortunately, Exim does not actually use EX_OK. The code used to try to
3692 preserve the sysexits.h value, by assuming that macro definitions were
3693 scanned for macro replacements. I have been disabused of this notion,
3694 so now the code just undefines EX_OK before #including unistd.h.
3696 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3697 in the smtp transport. When a block could not be written in a single
3698 write() function, the timeout was being re-applied to each part-write.
3699 This seems wrong - if the receiver was accepting one byte at a time it
3700 would take for ever. The timeout is now adjusted when this happens. It
3701 doesn't have to be particularly precise.
3703 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3704 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3706 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3707 without a domain if the check to <postmaster@domain> fails.
3709 SC/01 Eximstats: added -xls and the ability to specify output files
3710 (patch written by Frank Heydlauf).
3712 SC/02 Eximstats: use FileHandles for outputting results.
3714 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3716 SC/04 Eximstats: fixed display of large numbers with -nvr option
3718 SC/05 Eximstats: fixed merging of reports with empty tables.
3720 SC/06 Eximstats: added the -include_original_destination flag
3722 SC/07 Eximstats: removed tabs and trailing whitespace.
3724 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3726 TK/06 MBOX spool code: Add real "From " MBOX separator line
3727 so the .eml file is really in mbox format (even though
3728 most programs do not really care). Patch from Alex Miller.
3730 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3731 The latter is generated from $received_to and is only set if the
3732 message has one envelope recipient. SA can use these headers,
3733 obviously out-of-the-box. Patch from Alex Miller.
3735 PH/08 The ${def test on a variable was returning false if the variable's
3736 value was "0", contrary to what the specification has always said!
3737 The result should be true unless the variable is empty.
3739 PH/09 The syntax error of a character other than { following "${if
3740 def:variable_name" (after optional whitespace) was not being diagnosed.
3741 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3742 accidental colon was present, for example, could give incorrect results.
3744 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3745 result is used (not including appendfile, where other changes are about
3748 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3749 This involved changing a lot of size variables from int to off_t. It
3750 should work with maildirs and everything.
3752 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3753 spamd dying while we are connected to it.
3755 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3758 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3759 function for BDB 4.3.
3761 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3763 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3764 the sources for PCRE have been re-arranged and more files are now
3767 PH/15 The code I had for printing potentially long long variables in PH/11
3768 above was not the best (it lost precision). The length of off_t variables
3769 is now inspected at build time, and an appropriate printing format (%ld
3770 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3771 to be "long long int" or "long int". This is needed for the internal
3772 formatting function string_vformat().
3774 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3775 the configuration file to be ":syslog", then the script "guesses" where
3776 the logs files are, rather than using the compiled in default. In our
3777 case the guess is not the same as the compiled default, so the script
3778 suddenly stopped working when I started to use syslog. The patch checks
3779 to see if log_file_path is "". If so, it attempts to read it from exim
3780 with no configuration file to get the compiled in version, before it
3781 falls back to the previous guessing code."
3783 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3784 implementing BATV in an Exim configuration. See NewStuff for the gory
3787 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3788 Makefile that are specific to HP-UX.
3790 PH/18 If the "use_postmaster" option was set for a recipient callout together
3791 with the "random" option, the postmaster address was used as the MAIL
3792 FROM address for the random test, but not for the subsequent recipient
3793 test. It is now used for both.
3795 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3796 patch removes a few documentation additions to RFC 3028, because the
3797 latest draft now contains them. It adds the new en;ascii-case comparator
3798 and a new error check for 8bit text in MIME parts. Comparator and
3799 require names are now matched exactly. I enabled the subaddress
3800 extension, but it is not well tested yet (read: it works for me)."
3802 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3803 rework some of the code of TK/09 above to avoid the hardwired use of
3804 "%lld" and "long long". Replaced the call to snprintf() with a call to
3807 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3808 records point to non-existent hosts", "retry timeout exceeded", and
3809 "retry time not reached for any host after a long failure period".
3811 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3812 experimental DomainKeys support:
3814 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3815 (2) On an error such as an illegally used "control", the wrong name for
3816 the control was given.
3818 These problems did NOT occur unless DomainKeys support was compiled.
3820 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3822 PH/24 Added ${if match_ip condition.
3824 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3825 databases so that it will be absolutely obvious if a crash occurs in the
3826 DB library. This is a regular occurrence (often caused by mis-matched
3829 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3830 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3831 picked up on a user's system that detects such things. There doesn't seem
3832 to be a gcc warning option for this - only an attribute that has to be
3833 put on the function's prototype. It seems that in Fedora Core 4 they have
3834 set this on a number of new functions. No doubt there will be more in due
3837 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3838 specify a transport. However, if an address that was verified by such a
3839 router was the subject of a callout, Exim crashed because it tried to
3840 read the rcpt_include_affixes from the non-existent transport. Now it
3841 just assumes that the setting of that option is false. This bug was
3842 introduced by 4.51/PH/31.
3844 PH/28 Changed -d+all to exclude +memory, because that information is very
3845 rarely of interest, but it makes the output a lot bigger. People tend to
3846 do -d+all out of habit.
3848 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3849 code in os-type was giving problems when libc.so lives in lib64, like on
3852 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3853 aren't the modern standard, and it seems that some systems' include files
3854 don't always have them. Exim was already checking for some of the newer
3855 ones like T_AAAA, and defining it itself. I've added checks for all the
3856 record types that Exim uses.
3858 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3859 not automatically generating a new one, as it is supposed to. This
3860 prevented TLS from working. If the file did exist, but contained invalid
3861 data, a new version was generated, as expected. It was only the case of a
3862 non-existent file that was broken.
3864 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3865 with a change in libdomainkeys > 0.64.
3867 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3868 from DNS. If the selector record carries the flag, it now has
3869 precedence over the domain-wide flag.
3871 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3873 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3874 the use of a port name that isn't defined in /etc/services) occurred, the
3875 message was deferred as in a normal delivery, and thus remained on the
3876 spool, instead of being failed because of the mua_wrapper setting. This
3877 is now fixed, and I tidied up some of the mua_wrapper messages at the
3880 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3881 the messages in a hash of arrays rather than using individual hashes.
3882 This is a bit cleaner and results in dramatic memory savings, albeit
3883 at a slight CPU cost.
3885 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3886 as requested by Marc Sherman.
3888 SC/10 Eximstats: added histograms for user specified patterns as requested
3891 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3893 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3894 fopen() in the content-scanning modules that did not already have it.
3900 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3901 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3903 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3905 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3907 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3908 to test Sieve filters that use "vacation".
3910 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3911 that changes the way the GnuTLS parameters are stored in the cache file.
3912 The new format can be generated externally. For backward compatibility,
3913 if the data in the cache doesn't make sense, Exim assumes it has read an
3914 old-format file, and it generates new data and writes a new file. This
3915 means that you can't go back to an older release without removing the
3918 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3919 work if there are any delivery delays because "one_time" forces the
3920 parent to be marked "delivered", so its unseen clone is never tried
3921 again. For this reason, Exim now forbids the simultaneous setting of
3924 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3925 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3926 entry if you want to know the details. The fix, however, neglected to
3927 consider the case when local delivery batching is involved. The test for
3928 "previously delivered" was not happening when checking to see if an
3929 address could be batched with a previous (undelivered) one; under
3930 certain circumstances this could lead to multiple deliveries to the same
3933 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3934 in its include files, and this causes problems building Exim.
3936 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3937 header_syntax) but Exim was just ignoring anything given after a slash.
3938 In particular, this caused confusion with an attempt to use "verify =
3939 reverse_host_lookup/defer_ok". An error is now given when options are
3940 supplied for verify items that do not have them. (Maybe reverse_host_
3941 lookup should have a defer_ok option, but that's a different point.)
3943 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3944 defined by RFC 821) to 2048, because there were problems with some AUTH
3945 commands, and RFC 1869 says the size should be increased for extended
3946 SMTP commands that take arguments.
3948 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3951 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3952 "unknown" error; now it says that the functionality isn't in the binary.
3954 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3955 an address' error message when a string expansion fails (syntax or
3956 whatever). Otherwise the password may appear in the log. Following change
3957 PH/42 below, there is no longer a chance of it appearing in a bounce
3960 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3962 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3963 of its addresses was ever tried. (Bugzilla bug #2.)
3965 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3966 the result incorrectly in the debug output. (It correctly added a newline
3967 to what was transported.)
3969 TF/01 Added $received_time.
3971 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3972 commented out examples of how to interface to a virus scanner and to
3973 SpamAssassin. Also added commented examples of av_scanner and
3974 spamd_address settings.
3976 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3977 and controls are allowed in which ACLs. There were a couple of minor
3978 errors. Some of the entries in the conditions table (which is a table of
3979 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3980 negation of where the condition IS allowed.
3982 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3984 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3985 header file does not have a version number, so I've had to invent a new
3986 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3987 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3988 radiusclient), but it was contributed by a Radius user.
3990 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3991 files or directories, for OpenSSL.
3993 PH/19 When an Exim process that is running as root has to create an Exim log
3994 file, it does so in a subprocess that runs as exim:exim so as to get the
3995 ownership right at creation (otherwise, other Exim processes might see
3996 the file with the wrong ownership). There was no test for failure of this
3997 fork() call, which would lead to the process getting stuck as it waited
3998 for a non-existent subprocess. Forks do occasionally fail when resources
3999 run out. I reviewed all the other calls to fork(); they all seem to check
4002 PH/20 When checking for unexpected SMTP input at connect time (before writing
4003 the banner), Exim was not dealing correctly with a non-positive return
4004 from the read() function. If the client had disconnected by this time,
4005 the result was a log entry for a synchronization error with an empty
4006 string after "input=" when read() returned zero. If read() returned -1
4007 (an event I could not check), uninitialized data bytes were printed.
4008 There were reports of junk text (parts of files, etc) appearing after
4011 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4013 PH/22 Added support for macro redefinition, and (re)definition in between
4014 driver and ACL definitions.
4016 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4017 forgetting to use the resulting value; it was using the unexpanded value.
4019 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4020 hadn't been configured. The fix is from Juergen Kreileder, who
4021 understands it better than I do:
4023 "Here's what I see happening with three configured cyrus_sasl
4024 authenticators configured (plain, login, cram-md5):
4026 On startup auth_cyrus_sasl_init() gets called for each of these.
4027 This means three calls to sasl_listmech() without a specified mech_list.
4028 => SASL tests which mechs of all available mechs actually work
4029 => three warnings about OTP not working
4030 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4032 With the patch, sasl_listmech() also gets called three times. But now
4033 SASL's mech_list option is set to the server_mech specified in the the
4034 authenticator. Or in other words, the answer from sasl_listmech()
4035 gets limited to just the mech you're testing for (which is different
4037 => the return list contains just 'plain' or 'login', 'cram-md5' or
4038 nothing depending on the value of ob->server_mech.
4040 I've just tested the patch: Authentication still works fine,
4041 unavailable mechs specified in the exim configuration are still
4042 caught, and the auth.log warnings about OTP are gone."
4044 PH/25 When debugging is enabled, the contents of the command line are added
4045 to the debugging output, even when log_selector=+arguments is not
4048 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4049 answer is "GNU", and only if the return is "GNU/something" is the answer
4052 PH/27 $acl_verify_message is now set immediately after the failure of a
4053 verification in an ACL, and so is available in subsequent modifiers. In
4054 particular, the message can be preserved by coding like this:
4056 warn !verify = sender
4057 set acl_m0 = $acl_verify_message
4059 Previously, $acl_verify_message was set only while expanding "message"
4060 and "log_message" when a very denied access.
4062 PH/28 Modified OS/os.c-Linux with
4064 -#ifndef OS_LOAD_AVERAGE
4065 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4067 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4068 nomenclature these days.)
4070 PH/29 Installed patch from the Sieve maintainer that adds the options
4071 sieve_useraddress and sieve_subaddress to the redirect router.
4073 PH/30 In these circumstances:
4074 . Two addresses routed to the same list of hosts;
4075 . First host does not offer TLS;
4076 . First host accepts first address;
4077 . First host gives temporary error to second address;
4078 . Second host offers TLS and a TLS session is established;
4079 . Second host accepts second address.
4080 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4081 and peerdn, if requested) that were in fact used only for the second
4084 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4085 attention to any local part prefix or suffix that was matched by the
4086 router that accepted the address. It now behaves in the same way as it
4087 does for delivery: the affixes are removed from the local part unless
4088 rcpt_include_affixes is set on the transport.
4090 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4091 timeout during the DATA phase of an incoming message.
4093 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4094 applied a patch sanctioned by the Sieve maintainer.
4096 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4097 the uid or gid is negative. A case of a negative gid caused this to be
4098 noticed. The fix allows for either to be negative.
4100 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4101 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4104 PH/36 The change PH/12 above was broken. Fixed it.
4106 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4107 the grounds that routing the same address twice would always produce the
4108 same answer. This might have been true once, but it is certainly no
4109 longer true now. Routing a child address may depend on the previous
4110 routing that produced that child. Some complicated redirection strategies
4111 went wrong when messages had multiple recipients, and made Exim's
4112 behaviour dependent on the order in which the addresses were given.
4114 I have moved the duplicate checking until after the routing is complete.
4115 Exim scans the addresses that are assigned to local and remote
4116 transports, and removes any duplicates. This means that more work will be
4117 done, as duplicates will always all be routed, but duplicates are
4118 presumably rare, so I don't expect this is of any significance.
4120 For deliveries to pipes, files, and autoreplies, the duplicate checking
4121 still happens during the routing process, since they are not going to be
4124 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4125 It corrects a timeout issue with spamd. This is Ian's comment: "The
4126 background is that sometimes spamd either never reads data from a
4127 connection it has accepted, or it never writes response data. The exiscan
4128 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4129 blindly assumes that writes won't block so it may never time out."
4131 PH/39 Allow G after quota size as well as K and M.
4133 PH/40 The value set for $authenticated_id in an authenticator may not contain
4134 binary zeroes or newlines because the value is written to log lines and
4135 to spool files. There was no check on this. Now the value is run through
4136 the string_printing() function so that such characters are converted to
4137 printable escape sequences.
4139 PH/41 $message_linecount is a new variable that contains the total number of
4140 lines in the message. Compare $body_linecount, which is the count for the
4143 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4144 bounce and delay warning messages, except in certain special cases, which
4147 (a) An SMTP error message from a remote host;
4148 (b) A message specified in a :fail: redirection;
4149 (c) A message specified in a "fail" command in a system filter;
4150 (d) A message specified in a FAIL return from the queryprogram router;
4151 (e) A message specified by the cannot_route_message router option.
4153 In these cases only, Exim does include the error details in bounce and
4154 warning messages. There are also a few cases where bland messages such
4155 as "unrouteable address" or "local delivery error" are given.
4157 PH/43 $value is now also set for the "else" part of a ${run expansion.
4159 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4160 being worked on, but at least Exim now implements the latest version to
4163 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4164 process to complete was treated as a delivery failure, a timeout while
4165 writing the message to the pipe was logged, but erroneously treated as a
4166 successful delivery. Such timeouts include transport filter timeouts. For
4167 consistency with the overall process timeout, these timeouts are now
4168 treated as errors, giving rise to delivery failures by default. However,
4169 there is now a new Boolean option for the pipe transport called
4170 timeout_defer, which, if set TRUE, converts the failures into defers for
4171 both kinds of timeout. A transport filter timeout is now identified in
4174 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4175 systems where "make" and "gmake" are different, calling "gmake" at top
4176 level broke things. I've arranged for the value of $(MAKE) to be passed
4177 from the Makefile to this script so that it can call the same version of
4181 A note about Exim versions 4.44 and 4.50
4182 ----------------------------------------
4184 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4185 changes of various kinds. As a consequence, a big documentation update was
4186 needed. This delayed the release for rather longer than seemed good, especially
4187 in the light of a couple of (minor) security issues. Therefore, the changes
4188 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4189 release. So 4.44 and 4.50 are in effect two different branches that both start
4192 I have left the 4.50 change log unchanged; it contains all the changes since
4193 4.43. The change log for 4.44 is below; many of its items are identical to
4194 those for 4.50. This seems to be the most sensible way to preserve the
4195 historical information.
4201 1. Minor wording change to the doc/README.SIEVE file.
4203 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4204 computation of the current number of files was incorrect.
4206 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4207 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4210 4. Give more explanation in the error message when the command for a transport
4211 filter fails to execute.
4213 5. There are several places where Exim runs a non-Exim command in a
4214 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4215 was being done only for the command run by the queryprogram router. It is
4216 now done for all such subprocesses. The other cases are: ${run, transport
4217 filters, and the commands run by the lmtp and pipe transports.
4219 6. Added CONFIGURE_GROUP build-time option.
4221 7. Some older OS have a limit of 256 on the maximum number of file
4222 descriptors. Exim was using setrlimit() to set 1000 as a large value
4223 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4224 systems. I've change it so that if it can't get 1000, it tries for 256.
4226 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4227 was an oversight, and furthermore, ever since the addition of extra
4228 controls (e.g. 4.43/32), the checks on when to allow different forms of
4229 "control" were broken. There should now be diagnostics for all cases when a
4230 control that does not make sense is encountered.
4232 9. Added the /retain_sender option to "control=submission".
4234 10. $recipients is now available in the predata ACL (oversight).
4236 11. Tidy the search cache before the fork to do a delivery from a message
4237 received from the command line. Otherwise the child will trigger a lookup
4238 failure and thereby defer the delivery if it tries to use (for example) a
4239 cached ldap connection that the parent has called unbind on.
4241 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4242 of $address_data from the recipient verification was clobbered by the
4243 sender verification.
4245 13. The value of address_data from a sender verification is now available in
4246 $sender_address_data in subsequent conditions in the ACL statement.
4248 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4250 15. Added a new option "connect=<time>" to callout options, to set a different
4253 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4254 was its contents. (It was OK if the option was not defined at all.)
4256 17. A "Completed" log line is now written for messages that are removed from
4257 the spool by the -Mrm option.
4259 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4260 information about exactly what failed.
4262 19. Added -dd to debug only the daemon process.
4264 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4265 handles timeouts, both on the server side and network timeouts. Renamed the
4266 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4268 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4269 It is now set to "smtps".
4271 22. $host_address is now set to the target address during the checking of
4272 ignore_target_hosts.
4274 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4275 being passed; this would have caused $sender_host_name to have been used if
4276 matching the list had actually called for a host name (not very likely,
4277 since this list is usually IP addresses). A host name is now passed as
4280 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4281 code that specifies a non-restarting handler (typically sigaction() in
4282 modern systems) in an attempt to fix a rare and obscure crash bug.
4284 25. Narrowed the window for a race in the daemon that could cause it to ignore
4285 SIGCHLD signals. This is not a major problem, because they are used only to
4286 wake it up if nothing else does.
4288 26. A malformed maildirsize file could cause Exim to calculate negative values
4289 for the mailbox size or file count. Odd effects could occur as a result.
4290 The maildirsize information is now recalculated if the size or filecount
4293 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4294 support for a long time. Removed HAVE_SYS_VFS_H.
4296 28. Installed the latest version of exipick from John Jetmore.
4298 29. In an address list, if the pattern was not a regular expression, an empty
4299 subject address (from a bounce message) matched only if the pattern was an
4300 empty string. Non-empty patterns were not even tested. This was the wrong
4301 because it is perfectly reasonable to use an empty address as part of a
4302 database query. An empty address is now tested by patterns that are
4303 lookups. However, all the other forms of pattern expect the subject to
4304 contain a local part and a domain, and therefore, for them, an empty
4305 address still always fails if the pattern is not itself empty.
4307 30. Exim went into a mad DNS loop when attempting to do a callout where the
4308 host was specified on an smtp transport, and looking it up yielded more
4309 than one IP address.
4311 31. Re-factored the code for checking spool and log partition space into a
4312 function that finds that data and another that does the check. The former
4313 is then used to implement four new variables: $spool_space, $log_space,
4314 $spool_inodes, and $log_inodes.
4316 32. The RFC2047 encoding function was originally intended for short strings
4317 such as real names; it was not keeping to the 75-character limit for
4318 encoded words that the RFC imposes. It now respects the limit, and
4319 generates multiple encoded words if necessary. To be on the safe side, I
4320 have increased the buffer size for the ${rfc2047: expansion operator from
4323 33. It is now permitted to omit both strings after an "if" condition; if the
4324 condition is true, the result is "true". As before, when the second string
4325 is omitted, a false condition yields an empty string. This makes it less
4326 cumbersome to write custom ACL and router conditions.
4328 34. Failure to deliver a bounce message always caused it to be frozen, even if
4329 there was an errors_to setting on the router. The errors_to setting is now
4332 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4333 canonical form (fully expanded) before being placed in
4334 $sender_host_address.
4336 36. The table in the code that translates DNS record types into text (T_A to
4337 "A" for instance) was missing entries for NS and CNAME. It is just possible
4338 that this could have caused confusion if both these types were looked up
4339 for the same domain, because the text type is used as part of Exim's
4340 per-process caching. But the chance of anyone hitting this buglet seems
4343 37. The dnsdb lookup has been extended in a number of ways.
4345 (1) There is a new type, "zns", which walks up the domain tree until it
4346 finds some nameserver records. It should be used with care.
4348 (2) There is a new type, "mxh", which is like "mx" except that it returns
4349 just the host names, not the priorities.
4351 (3) It is now possible to give a list of domains (or IP addresses) to be
4352 looked up. The behaviour when one of the lookups defers can be
4353 controlled by a keyword.
4355 (4) It is now possible to specify the separator character for use when
4356 multiple records are returned.
4358 38. The dnslists ACL condition has been extended: it is now possible to supply
4359 a list of IP addresses and/or domains to be looked up in a particular DNS
4362 39. Added log_selector=+queue_time_overall.
4364 40. When running the queue in the test harness, wait just a tad after forking a
4365 delivery process, to get repeatability of debugging output.
4367 41. Include certificate and key file names in error message when GnuTLS fails
4368 to set them up, because the GnuTLS error message doesn't include the name
4369 of the failing file when there is a problem reading it.
4371 42. Allow both -bf and -bF in the same test run.
4373 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4375 44. The "Exiscan patch" is now merged into the mainline Exim source.
4377 45. Sometimes the final signoff response after QUIT could fail to get
4378 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4379 < 0 before doing a fflush(). This bug looks as though it goes back to the
4380 introduction of TLS in release 3.20, but "sometimes" must have been rare
4381 because the tests only now provoked it.
4383 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4384 (this can affect the format of dates).
4386 47. exim_tidydb, when checking for the continued existence of a message for
4387 which it has found a message-specific retry record, was not finding
4388 messages that were in split spool directories. Consequently, it was
4389 deleting retry records that should have stayed in existence.
4391 48. Steve fixed some bugs in eximstats.
4393 49. The SPA authentication driver was not abandoning authentication and moving
4394 on to the next authenticator when an expansion was forced to fail,
4395 contradicting the general specification for all authenticators. Instead it
4396 was generating a temporary error. It now behaves as specified.
4398 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4399 (the order specifies the preference for clients). The order is now AES256,
4400 AES128, 3DES, ARCFOUR128.
4402 51. Small patch to Sieve code - explicitly set From: when generating an
4405 52. Exim crashed if a remote delivery caused a very long error message to be
4406 recorded - for instance if somebody sent an entire SpamAssassin report back
4407 as a large number of 550 error lines. This bug was coincidentally fixed by
4408 increasing the size of one of Exim's internal buffers (big_buffer) that
4409 happened as part of the Exiscan merge. However, to be on the safe side, I
4410 have made the code more robust (and fixed the comments that describe what
4413 53. Now that there can be additional text after "Completed" in log lines (if
4414 the queue_time_overall log selector is set), a one-byte patch to exigrep
4415 was needed to allow it to recognize "Completed" as not the last thing in
4418 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4419 patch that reportedly fixes this has been added. I am not expert enough to
4420 create a test for it. This is what the patch creator wrote:
4422 "I found a little strange behaviour of ldap code when working with
4423 Windows 2003 AD Domain, where users was placed in more than one
4424 Organization Units. When I tried to give exim partial DN, the exit code
4425 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4426 But simultaneously result of request was absolutely normal ldap result,
4427 so I produce this patch..."
4429 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4430 REFERENCE, so I have modified the code to exclude the patch when that macro
4433 55. Some experimental protocols are using DNS PTR records for new purposes. The
4434 keys for these records are domain names, not reversed IP addresses. The
4435 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4436 leaves it alone. Component reversal etc. now happens only for IP addresses.
4439 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4441 57. Double the size of the debug message buffer (to 2048) so that more of very
4442 long debug lines gets shown.
4444 58. The exicyclog utility now does better if the number of log files to keep
4445 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4447 59. Two changes related to the smtp_active_hostname option:
4449 (1) $smtp_active_hostname is now available as a variable.
4450 (2) The default for smtp_banner uses $smtp_active_hostname instead
4451 of $primary_hostname.
4453 60. The host_aton() function is supposed to be passed a string that is known
4454 to be a valid IP address. However, in the case of IPv6 addresses, it was
4455 not checking this. This is a hostage to fortune. Exim now panics and dies
4456 if the condition is not met. A case was found where this could be provoked
4457 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4458 components; fortuitously, this particular loophole had already been fixed
4459 by change 4.50/55 above.
4461 If there are any other similar loopholes, the new check in host_aton()
4462 itself should stop them being exploited. The report I received stated that
4463 data on the command line could provoke the exploit when Exim was running as
4464 exim, but did not say which command line option was involved. All I could
4465 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4466 running as the user.
4469 61. There was a buffer overflow vulnerability in the SPA authentication code
4470 (which came originally from the Samba project). I have added a test to the
4471 spa_base64_to_bits() function which I hope fixes it.
4474 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4475 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4477 63. The daemon start-up calls getloadavg() while still root for those OS that
4478 need the first call to be done as root, but it missed one case: when
4479 deliver_queue_load_max is set with deliver_drop_privilege. This is
4480 necessary for the benefit of the queue runner, because there is no re-exec
4481 when deliver_drop_privilege is set.
4483 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4484 This has been fixed.
4486 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4487 was in use, was not putting the data itself into the right store pool;
4488 consequently, it could be overwritten for a subsequent message in the same
4489 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4492 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4494 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4495 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4496 the function were treating the return as a boolean value, which happened to
4497 work because 0=false and not-0=true, but is not correct code.
4499 68. The host_aton() function was not handling scoped IPv6 addresses (those
4500 with, for example, "%eth0" on the end) correctly.
4502 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4503 negated items (that is, ~something) in unsigned ints. Some compilers
4504 apparently mutter when there is no cast.
4506 70. If an address verification called from an ACL failed, and did not produce a
4507 user-specific message (i.e. there was only a "system" message), nothing was
4508 put in $acl_verify_message. In this situation, it now puts the system
4511 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4512 session; change 4.31/43 added the unwanted input to the log line - except
4513 that it did not do this in the start of session case. It now does.
4515 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4516 This is wrong and can cause the other end to generate a synchronization
4517 error if it is another Exim or anything else that does the synchronization
4518 check. A QUIT command is no longer sent after a timeout.
4520 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4521 during host lookups.
4523 74. The defer_ok option of callout verification was not working if it was used
4524 when verifying addresses in header lines, that is, for this case:
4526 verify = header_sender/callout=defer_ok
4528 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4529 those file descriptors could be used for SMTP connections. If anything
4530 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4531 could be sent to the SMTP client, causing chaos. The daemon now opens
4532 stdin, stdout, and stderr to /dev/null when it puts itself into the
4535 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4536 log by default. The user can override this with suitable Perl magic.
4538 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4539 the log message when discard triggers, was not working for the DATA ACL or
4540 for the non-SMTP ACL.
4542 78. Error message wording change in sieve.c.
4544 79. If smtp_accept_max_per_host was set, the number of connections could be
4545 restricted to fewer than expected, because the daemon was trying to set up
4546 a new connection before checking whether the processes handling previous
4547 connections had finished. The check for completed processes is now done
4548 earlier. On busy systems, this bug wouldn't be noticed because something
4549 else would have woken the daemon, and it would have reaped the completed
4552 80. If a message was submitted locally by a user whose login name contained one
4553 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4554 It caused a spool format error. I have fixed the spool reading code. A
4555 related problem was that the "from" clause in the Received: line became
4556 illegal because of the space(s). It is now covered by ${quote_local_part.
4558 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4561 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4563 83. Patch from Sieve maintainer to fix unterminated string problem in
4564 "vacation" handling.
4566 84. Some minor changes to the Linux configuration files to help with other
4567 OS variants using glibc.
4569 85. One more patch for Sieve to update vacation handling to latest spec.
4572 ----------------------------------------------------
4573 See the note above about the 4.44 and 4.50 releases.
4574 ----------------------------------------------------
4580 1. Change 4.43/35 introduced a bug that caused file counts to be
4581 incorrectly computed when quota_filecount was set in an appendfile
4584 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4585 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4588 3. Give more explanation in the error message when the command for a transport
4589 filter fails to execute.
4591 4. There are several places where Exim runs a non-Exim command in a
4592 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4593 was being done only for the command run by the queryprogram router. It is
4594 now done for all such subprocesses. The other cases are: ${run, transport
4595 filters, and the commands run by the lmtp and pipe transports.
4597 5. Some older OS have a limit of 256 on the maximum number of file
4598 descriptors. Exim was using setrlimit() to set 1000 as a large value
4599 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4600 systems. I've change it so that if it can't get 1000, it tries for 256.
4602 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4603 was an oversight, and furthermore, ever since the addition of extra
4604 controls (e.g. 4.43/32), the checks on when to allow different forms of
4605 "control" were broken. There should now be diagnostics for all cases when a
4606 control that does not make sense is encountered.
4608 7. $recipients is now available in the predata ACL (oversight).
4610 8. Tidy the search cache before the fork to do a delivery from a message
4611 received from the command line. Otherwise the child will trigger a lookup
4612 failure and thereby defer the delivery if it tries to use (for example) a
4613 cached ldap connection that the parent has called unbind on.
4615 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4616 of $address_data from the recipient verification was clobbered by the
4617 sender verification.
4619 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4620 was its contents. (It was OK if the option was not defined at all.)
4622 11. A "Completed" log line is now written for messages that are removed from
4623 the spool by the -Mrm option.
4625 12. $host_address is now set to the target address during the checking of
4626 ignore_target_hosts.
4628 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4629 being passed; this would have caused $sender_host_name to have been used if
4630 matching the list had actually called for a host name (not very likely,
4631 since this list is usually IP addresses). A host name is now passed as
4634 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4635 code that specifies a non-restarting handler (typically sigaction() in
4636 modern systems) in an attempt to fix a rare and obscure crash bug.
4638 15. Narrowed the window for a race in the daemon that could cause it to ignore
4639 SIGCHLD signals. This is not a major problem, because they are used only to
4640 wake it up if nothing else does.
4642 16. A malformed maildirsize file could cause Exim to calculate negative values
4643 for the mailbox size or file count. Odd effects could occur as a result.
4644 The maildirsize information is now recalculated if the size or filecount
4647 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4648 support for a long time. Removed HAVE_SYS_VFS_H.
4650 18. Updated exipick to current release from John Jetmore.
4652 19. Allow an empty sender to be matched against a lookup in an address list.
4653 Previously the only cases considered were a regular expression, or an
4656 20. Exim went into a mad DNS lookup loop when doing a callout where the
4657 host was specified on the transport, if the DNS lookup yielded more than
4660 21. The RFC2047 encoding function was originally intended for short strings
4661 such as real names; it was not keeping to the 75-character limit for
4662 encoded words that the RFC imposes. It now respects the limit, and
4663 generates multiple encoded words if necessary. To be on the safe side, I
4664 have increased the buffer size for the ${rfc2047: expansion operator from
4667 22. Failure to deliver a bounce message always caused it to be frozen, even if
4668 there was an errors_to setting on the router. The errors_to setting is now
4671 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4672 canonical form (fully expanded) before being placed in
4673 $sender_host_address.
4675 24. Updated eximstats to version 1.33
4677 25. Include certificate and key file names in error message when GnuTLS fails
4678 to set them up, because the GnuTLS error message doesn't include the name
4679 of the failing file when there is a problem reading it.
4681 26. Expand error message when OpenSSL has problems setting up cert/key files.
4684 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4685 (this can affect the format of dates).
4687 28. exim_tidydb, when checking for the continued existence of a message for
4688 which it has found a message-specific retry record, was not finding
4689 messages that were in split spool directories. Consequently, it was
4690 deleting retry records that should have stayed in existence.
4692 29. eximstats updated to version 1.35
4693 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4694 1.35 - bugfix such that pie charts by volume are generated correctly
4696 30. The SPA authentication driver was not abandoning authentication and moving
4697 on to the next authenticator when an expansion was forced to fail,
4698 contradicting the general specification for all authenticators. Instead it
4699 was generating a temporary error. It now behaves as specified.
4701 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4702 (the order specifies the preference for clients). The order is now AES256,
4703 AES128, 3DES, ARCFOUR128.
4705 31. Small patch to Sieve code - explicitly set From: when generating an
4708 32. Exim crashed if a remote delivery caused a very long error message to be
4709 recorded - for instance if somebody sent an entire SpamAssassin report back
4710 as a large number of 550 error lines. This bug was coincidentally fixed by
4711 increasing the size of one of Exim's internal buffers (big_buffer) that
4712 happened as part of the Exiscan merge. However, to be on the safe side, I
4713 have made the code more robust (and fixed the comments that describe what
4716 33. Some experimental protocols are using DNS PTR records for new purposes. The
4717 keys for these records are domain names, not reversed IP addresses. The
4718 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4719 leaves it alone. Component reversal etc. now happens only for IP addresses.
4722 34. The host_aton() function is supposed to be passed a string that is known
4723 to be a valid IP address. However, in the case of IPv6 addresses, it was
4724 not checking this. This is a hostage to fortune. Exim now panics and dies
4725 if the condition is not met. A case was found where this could be provoked
4726 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4727 components; fortuitously, this particular loophole had already been fixed
4728 by change 4.50/55 or 4.44/33 above.
4730 If there are any other similar loopholes, the new check in host_aton()
4731 itself should stop them being exploited. The report I received stated that
4732 data on the command line could provoke the exploit when Exim was running as
4733 exim, but did not say which command line option was involved. All I could
4734 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4735 running as the user.
4738 35. There was a buffer overflow vulnerability in the SPA authentication code
4739 (which came originally from the Samba project). I have added a test to the
4740 spa_base64_to_bits() function which I hope fixes it.
4743 36. The daemon start-up calls getloadavg() while still root for those OS that
4744 need the first call to be done as root, but it missed one case: when
4745 deliver_queue_load_max is set with deliver_drop_privilege. This is
4746 necessary for the benefit of the queue runner, because there is no re-exec
4747 when deliver_drop_privilege is set.
4749 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4750 was in use, was not putting the data itself into the right store pool;
4751 consequently, it could be overwritten for a subsequent message in the same
4752 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4755 38. Sometimes the final signoff response after QUIT could fail to get
4756 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4757 < 0 before doing a fflush(). This bug looks as though it goes back to the
4758 introduction of TLS in release 3.20, but "sometimes" must have been rare
4759 because the tests only now provoked it.
4765 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4766 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4767 Now it returns an integer. A number of calls were still expecting a T/F
4768 return. Fortuitously, in all cases, the tests worked in OK situations,
4769 which is the norm. However, things would have gone wrong on any write
4770 failures on the smtp file descriptor. This function is used when sending
4771 messages over SMTP and also when doing verify callouts.
4773 2. When Exim is called to do synchronous delivery of a locally submitted
4774 message (the -odf or -odi options), it no longer closes stderr before doing
4777 3. Implemented the mua_wrapper option.
4779 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4781 5. Implemented the functions header_remove(), header_testname(),
4782 header_add_at_position(), and receive_remove_recipient(), and exported them
4785 6. If an ACL "warn" statement specified the addition of headers, Exim already
4786 inserted X-ACL-Warn: at the start if there was no header name. However, it
4787 was not making this test for the second and subsequent header lines if
4788 there were newlines in the string. This meant that an invalid header could
4789 be inserted if Exim was badly configured.
4791 7. Allow an ACL "warn" statement to add header lines at the start or after all
4792 the Received: headers, as well as at the end.
4794 8. Added the rcpt_4xx retry error code.
4796 9. Added postmaster_mailfrom=xxx to callout verification option.
4798 10. Added mailfrom=xxxx to the callout verification option, for verify=
4801 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4802 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4804 12. Inserted some casts to stop certain compilers warning when using pointer
4805 differences as field lengths or precisions in printf-type calls (mostly
4806 affecting debugging statements).
4808 13. Added optional readline() support for -be (dynamically loaded).
4810 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4811 same clock tick as a message's arrival, so that its received time was the
4812 same as the "first fail" time on the retry record, and that message
4813 remained on the queue past the ultimate address timeout, every queue runner
4814 would try a delivery (because it was past the ultimate address timeout) but
4815 after another failure, the ultimate address timeout, which should have then
4816 bounced the address, did not kick in. This was a "< instead of <=" error;
4817 in most cases the first failure would have been in the next clock tick
4818 after the received time, and all would be well.
4820 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4821 being recognized when the domain list was tested by the match_domain
4822 condition in an expansion string.
4824 16. Added the ${str2b64: operator.
4826 17. Exim was always calling setrlimit() to set a large limit for the number of
4827 processes, without checking whether the existing limit was already
4828 adequate. (It did check for the limit on file descriptors.) Furthermore,
4829 errors from getrlimit() and setrlimit() were being ignored. Now they are
4830 logged to the main and panic logs, but Exim does carry on, to try to do its
4831 job under whatever limits there are.
4833 18. Imported PCRE 5.0.
4835 19. Trivial typo in log message " temporarily refused connection" (the leading
4838 20. If the log selector return_path_on_delivery was set and an address was
4839 redirected to /dev/null, the delivery process crashed because it assumed
4840 that a return path would always be set for a "successful" delivery. In this
4841 case, the whole delivery is bypassed as an optimization, and therefore no
4844 21. Internal re-arrangement: the function for sending a challenge and reading
4845 a response while authentication was assuming a zero-terminated challenge
4846 string. It's now changed to take a pointer and a length, to allow for
4847 binary data in such strings.
4849 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4851 23. Exim was not respecting finduser_retries when seeking the login of the
4852 uid under which it was called; it was always trying 10 times. (The default
4853 setting of finduser_retries is zero.) Also, it was sleeping after the final
4854 failure, which is pointless.
4856 24. Implemented tls_on_connect_ports.
4858 25. Implemented acl_smtp_predata.
4860 26. If the domain in control=submission is set empty, Exim assumes that the
4861 authenticated id is a complete email address when it generates From: or
4862 Sender: header lines.
4864 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4865 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4866 chown and chgrp in /bin and hostname in /usr/bin.
4868 28. Exim was keeping the "process log" file open after each use, just as it
4869 does for the main log. This opens the possibility of it remaining open for
4870 long periods when the USR1 signal hits a daemon. Occasional processlog
4871 errors were reported, that could have been caused by this. Anyway, it seems
4872 much more sensible not to leave this file open at all, so that is what now
4875 29. The long-running daemon process does not normally write to the log once it
4876 has entered its main loop, and it closes the log before doing so. This is
4877 so that log files can straightforwardly be renamed and moved. However,
4878 there are a couple of unusual error situations where the daemon does write
4879 log entries, and I had neglected to close the log afterwards.
4881 30. The text of an SMTP error response that was received during a remote
4882 delivery was being truncated at 512 bytes. This is too short for some of
4883 the long messages that one sometimes sees. I've increased the limit to
4886 31. It is now possible to make retry rules that apply only when a message has a
4887 specific sender, in particular, an empty sender.
4889 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4890 it possible to be selective about when SMTP synchronization is enforced.
4892 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4894 32. Implemented hosts_connection_nolog.
4896 33. Added an ACL for QUIT.
4898 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4901 35. Added mailbox_size and mailbox_filecount to appendfile.
4903 36. Added control = no_multiline_responses to ACLs.
4905 37. There was a bug in the logic of the code that waits for the clock to tick
4906 in the case where the clock went backwards by a substantial amount such
4907 that the microsecond fraction of "now" was more than the microsecond
4908 fraction of "then" (but the whole seconds number was less).
4910 38. Added support for the libradius Radius client library this is found on
4911 FreeBSD (previously only the radiusclient library was supported).
4917 1. When certain lookups returned multiple values in the form name=value, the
4918 quoting of the values was not always being done properly. Specifically:
4919 (a) If the value started with a double quote, but contained no whitespace,
4921 (b) If the value contained whitespace other than a space character (i.e.
4922 tabs or newlines or carriage returns) it was not quoted.
4923 This fix has been applied to the mysql and pgsql lookups by writing a
4924 separate quoting function and calling it from the lookup code. The fix
4925 should probably also be applied to nisplus, ibase and oracle lookups, but
4926 since I cannot test any of those, I have not disturbed their existing code.
4928 2. A hit in the callout cache for a specific address caused a log line with no
4929 reason for rejecting RCPT. Now it says "Previous (cached) callout
4930 verification failure".
4932 3. There was an off-by-one bug in the queryprogram router. An over-long
4933 return line was truncated at 256 instead of 255 characters, thereby
4934 overflowing its buffer with the terminating zero. As well as fixing this, I
4935 have increased the buffer size to 1024 (and made a note to document this).
4937 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4938 when Exim is waiting for an SMTP response from a remote server, Exim
4939 restarts its select() call on the socket, thereby resetting its timeout.
4940 This is not a problem when such interrupts are rare. Somebody set up a cron
4941 job to run exiwhat every 2 minutes, which is less than the normal select()
4942 timeout (5 or 10 minutes). This meant that the select() timeout never
4943 kicked in because it was always reset. I have fixed this by comparing the
4944 time when an interrupt arrives with the time at the start of the first call
4945 to select(). If more time than the timeout has elapsed, the interrupt is
4946 treated as a timeout.
4948 5. Some internal re-factoring in preparation for the addition of Sieve
4949 extensions (by MH). In particular, the "personal" test is moved to a
4950 separate function, and given an option for scanning Cc: and Bcc: (which is
4951 not set for Exim filters).
4953 6. When Exim created an email address using the login of the caller as the
4954 local part (e.g. when creating a From: or Sender: header line), it was not
4955 quoting the local part when it contained special characters such as @.
4957 7. Installed new OpenBSD configuration files.
4959 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4960 try to make them clearer.
4962 9. Callout options, other than the timeout value, were being ignored when
4963 verifying sender addresses in header lines. For example, when using
4965 verify = header_sender/callout=no_cache
4967 the cache was (incorrectly) being used.
4969 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4970 only the Cygwin environment.
4972 11. When return_path_on_delivery was set as a log selector, if different remote
4973 addresses in the same message used different return paths and parallel
4974 remote delivery occurred, the wrong values would sometimes be logged.
4975 (Whenever a remote delivery process finished, the return path value from
4976 the most recently started remote delivery process was logged.)
4978 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4979 lines when AUTH and/or TLS are in use. This is the "received protocol"
4980 field. Exim used to use "asmtp" for authenticated SMTP, without any
4981 indication (in the protocol name) for TLS use. Now it follows the RFC and
4982 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4983 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4984 names appear in log lines as well as in Received: header lines.
4986 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4987 extensions, and comparison tests, and to fix some bugs.
4989 14. Changes to the "personal" filter test:
4991 (1) The test was buggy in that it was just doing the equivalent of
4992 "contains" tests on header lines. For example, if a user's address was
4993 anne@some.where, the "personal" test would incorrectly be true for
4995 To: susanne@some.where
4997 This test is now done by extracting each address from the header in turn,
4998 and checking the entire address. Other tests that are part of "personal"
4999 are now done using regular expressions (for example, to check local parts
5000 of addresses in From: header lines).
5002 (2) The list of non-personal local parts in From: addresses has been
5003 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5004 taken from the Sieve specification recommendations.
5006 (3) If the message contains any header line starting with "List-" it is
5007 treated as non-personal.
5009 (4) The test for "circular" in the Subject: header line has been removed
5010 because it now seems ill-conceived.
5012 15. Minor typos in src/EDITME comments corrected.
5014 16. Installed latest exipick from John Jetmore.
5016 17. If headers_add on a router specified a text string that was too long for
5017 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5018 of string_sprintf() is now avoided.
5020 18. $message_body_size was not set (it was always zero) when running the DATA
5021 ACL and the local_scan() function.
5023 19. For the "mail" command in an Exim filter, no default was being set for
5024 the once_repeat time, causing a random time value to be used if "once" was
5025 specified. (If the value happened to be <= 0, no repeat happened.) The
5026 default is now 0s, meaning "never repeat". The "vacation" command was OK
5027 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5028 (I found it when inspecting the code).
5030 20. There is now an overall timeout for performing a callout verification. It
5031 defaults to 4 times the callout timeout, which applies to individual SMTP
5032 commands during the callout. The overall timeout applies when there is more
5033 than one host that can be tried. The timeout is checked before trying the
5034 next host. This prevents very long delays if there are a large number of
5035 hosts and all are timing out (e.g. when the network connections are timing
5036 out). The value of the overall timeout can be changed by specifying an
5037 additional sub-option for "callout", called "maxwait". For example:
5039 verify = sender/callout=5s,maxwait=20s
5041 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5042 to the end before writing, but this should make it even safer).
5044 22. Exim was forgetting that it had advertised PIPELINING for the second and
5045 subsequent messages on an SMTP connection. It was also not resetting its
5046 memory on STARTTLS and an internal HELO.
5048 23. When Exim logs an SMTP synchronization error within a session, it now
5049 records whether PIPELINING has been advertised or not.
5051 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5052 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5053 rather than long int.
5055 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5057 26. Added the never_mail option to autoreply.
5063 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5064 crash if the getsockname() call failed; this can happen if a connection is
5065 closed very soon after it is established. The problem was simply in the
5066 order in which certain operations were done, causing Exim to try to write
5067 to the SMTP stream before it had set up the file descriptor. The bug has
5068 been fixed by making things happen in the correct order.
5074 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5075 before the connection was closed, thus losing the rejection response.
5077 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5078 some early Solaris releases, but causes trouble in current releases where
5079 socklen_t is defined.
5081 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5084 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5087 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5088 that releasing the top of it at the end releases what was used for sub-
5089 expansions (unless the block got too big). However, discard this block if
5090 the first thing is a variable or header, so that we can use its block when
5091 it is dynamic (useful for very large $message_headers, for example).
5093 6. Lookups now cache *every* query, not just the most recent. A new, separate
5094 store pool is used for this. It can be recovered when all lookup caches are
5095 flushed. Lookups now release memory at the end of their result strings.
5096 This has involved some general refactoring of the lookup sources.
5098 7. Some code has been added to the store_xxx() functions to reduce the amount
5099 of flapping under certain conditions.
5101 8. log_incoming_interface used to affect only the <= reception log lines. Now
5102 it causes the local interface and port to be added to several more SMTP log
5103 lines, for example "SMTP connection from", and rejection lines.
5105 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5107 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5109 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5110 could be overwritten at the end of the current message (or the start of a
5111 new message if it was set in a HELO ACL). The value is now preserved for
5112 the duration of the SMTP connection.
5114 12. If a transport had a headers_rewrite setting, and a matching header line
5115 contained an unqualified address, that address was qualified, even if it
5116 did not match any rewriting rules. The underlying bug was that the values
5117 of the flags that permit the existence of unqualified sender and recipient
5118 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5119 non-local messages, and by -bnq for local messages) were not being
5120 preserved with the message after it was received.
5122 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5123 "next input=" as part of the text comprising the host identity instead of
5124 the correct text. The code was using the same buffer for two different
5125 strings. However, depending on which order the printing function evaluated
5126 its arguments, the bug did not always show up. Under Linux, for example, my
5127 test suite worked just fine.
5129 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5130 doesn't work with some older versions of Perl. It has been changed to "my",
5131 which in any case is probably the better facility to use.
5133 15. A really picky compiler found some instances of statements for creating
5134 error messages that either had too many or two few arguments for the format
5137 16. The size of the buffer for calls to the DNS resolver has been increased
5138 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5139 for addresses that have a lot of PTR records. This alleviates a problem; it
5140 does not fully solve it.
5142 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5143 buffer now truncates the list and logs the incident, which is the same
5144 action as happens when Exim is looking up a host name and its aliases.
5145 Previously in this situation something unpredictable would happen;
5146 sometimes it was "internal error: store_reset failed".
5148 18. If a server dropped the connection unexpectedly when an Exim client was
5149 using GnuTLS and trying to read a response, the client delivery process
5150 crashed while trying to generate an error log message.
5152 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5153 string, for example:
5155 warn message = H1: something\nH2: something
5157 the text was added as a single header line from Exim's point of view
5158 though it ended up OK in the delivered message. However, searching for the
5159 second and subsequent header lines using $h_h2: did not work. This has been
5160 fixed. Similarly, if a system filter added multiple headers in this way,
5161 the routers could not see them.
5163 20. Expanded the error message when iplsearch is called with an invalid key to
5164 suggest using net-iplsearch in a host list.
5166 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5167 ACLs are no longer actually imposed (and a message to that effect is
5170 22. If a "gecos" field in a passwd entry contained escaped characters, in
5171 particular, if it contained a \" sequence, Exim got it wrong when building
5172 a From: or a Sender: header from that name. A second bug also caused
5173 incorrect handling when an unquoted " was present following a character
5174 that needed quoting.
5176 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5177 was not being matched caselessly.
5179 24. Arranged for all hyphens in the exim.8 source to be escaped with
5182 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5183 properly. Recipient callout cache records were still being keyed to include
5184 the sender, even when use_sender was set false. This led to far more
5185 callouts that were necessary. The sender is no longer included in the key
5186 when use_sender is false.
5188 26. Added "control = submission" modifier to ACLs.
5190 27. Added the ${base62d: operator to decode base 62 numbers.
5192 28. dnsdb lookups can now access SRV records.
5194 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5195 the configuration file.
5197 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5198 (-v) mode. This makes the output for a verbose queue run more intelligible.
5200 31. Added a use_postmaster feature to recipient callouts.
5202 32. Added the $body_zerocount variable, containing the number of binary zero
5203 bytes in the message body.
5205 33. The time of last modification of the "new" subdirectory is now used as the
5206 "mailbox time last read" when there is a quota error for a maildir
5209 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5211 35. Added +ignore_unknown as a special item in host lists.
5213 36. Code for decoding IPv6 addresses in host lists is now included, even if
5214 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5215 address was recognized as an IP address, but was then not correctly decoded
5216 into binary, causing unexpected and incorrect effects when compared with
5223 1. Very minor rewording of debugging text in manualroute to say "list of
5224 hosts" instead of "hostlist".
5226 2. If verify=header_syntax was set, and a header line with an unqualified
5227 address (no domain) and a large number of spaces between the end of the
5228 name and the colon was received, the reception process suffered a buffer
5229 overflow, and (when I tested it) crashed. This was caused by some obsolete
5230 code that should have been removed. The fix is to remove it!
5232 3. When running in the test harness, delay a bit after writing a bounce
5233 message to get a bit more predictability in the log output.
5235 4. Added a call to search_tidyup() just before forking a reception process. In
5236 theory, someone could use a lookup in the expansion of smtp_accept_max_
5237 per_host which, without the tidyup, could leave open a database connection.
5239 5. Added the variables $recipient_data and $sender_data which get set from a
5240 lookup success in an ACL "recipients" or "senders" condition, or a router
5241 "senders" option, similar to $domain_data and $local_part_data.
5243 6. Moved the writing of debug_print from before to after the "senders" test
5246 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5247 problems for message scanning, either using a data ACL, or using
5248 local_scan() because the Received: header was not generated till after they
5249 were called (in order to set the time as the time of reception completion).
5250 I have revised the way this works. The header is now generated after the
5251 body is received, but before the ACL or local_scan() are called. After they
5252 are run, the timestamp in the header is updated.
5258 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5259 before starting a queue runner without re-exec. This happened only when
5260 deliver_drop_privilege was set or when the Exim user was set to root. The
5261 effect of the bug was that timeouts during subsequent deliveries caused
5262 crashes instead of being properly handled. The handler is now left at its
5263 default (and expected) setting.
5265 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5266 message, again when deliver_drop_privilege is set or Exim is run as root.
5267 The bug described in (1) was not present in this case, but the tidying up
5268 of the other signals was missing. I have made the two cases consistent.
5270 3. The ignore_target_hosts setting on a manualroute router was being ignored
5271 for hosts that were looked up using the /MX notation.
5273 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5276 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5277 operated on the sender address. After changing the $sender_address to <>
5278 for the sender address verify, Exim was re-instated it as the original
5279 (before rewriting) address, but remembering that it had rewritten it, so it
5280 wasn't rewriting it again. This bug also had the effect of breaking the
5281 sender address verification caching when the sender address was rewritten.
5283 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5284 This has been changed so that if the ip literal address matches
5285 ignore_target_hosts, the router declines.
5287 7. Added expansion conditions match_domain, match_address, and match_local_
5288 part (NOT match_host).
5290 8. The placeholder for the Received: header didn't have a length field set.
5292 9. Added code to Exim itself and to exim_lock to test for a specific race
5293 condition that could lead to file corruption when using MBX delivery. The
5294 issue is with the lockfile that is created in /tmp. If this file is removed
5295 after a process has opened it but before that process has acquired a lock,
5296 there is the potential for a second process to recreate the file and also
5297 acquire a lock. This could lead to two Exim processes writing to the file
5298 at the same time. The added code performs the same test as UW imapd; it
5299 checks after acquiring the lock that its file descriptor still refers to
5300 the same named file.
5302 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5303 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5306 11. Added the smtp_active_hostname option. If used, this will typically be made
5307 to depend on the incoming interface address. Because $interface_address is
5308 not set up until the daemon has forked a reception process, error responses
5309 that can happen earlier (such as "too many connections") no longer contain
5312 12. If an expansion in a condition on a "warn" statement fails because a lookup
5313 defers, the "warn" statement is abandoned, and the next ACL statement is
5314 processed. Previously this caused the whole ACL to be aborted.
5316 13. Added the iplsearch lookup type.
5318 14. Added ident_timeout as a log selector.
5320 15. Added tls_certificate_verified as a log selector.
5322 16. Added a global option tls_require_ciphers (compare the smtp transport
5323 option of the same name). This controls incoming TLS connections.
5325 17. I finally figured out how to make tls_require_ciphers do a similar thing
5326 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5327 before starting the TLS session.
5329 18. Tabs are now shown as \t in -bP output.
5331 19. If the log selector return_path_on_delivery was set, Exim crashed when
5332 bouncing a message because it had too many Received: header lines.
5334 20. If two routers both had headers_remove settings, and the first one included
5335 a superfluous trailing colon, the final name in the first list and the
5336 first name in the second list were incorrectly joined into one item (with a
5337 colon in the middle).
5343 1. Added -C and -D options to the exinext utility, mainly to make it easier
5344 to include in the automated testing, but these could be helpful when
5345 multiple configurations are in use.
5347 2. The exinext utility was not formatting the output nicely when there was
5348 an alternate port involved in the retry record key, nor when there was a
5349 message id as well (for retries that were specific to a specific message
5350 and a specific host). It was also confused by IPv6 addresses, because of
5351 the additional colons they contain. I have fixed the IPv4 problem, and
5352 patched it up to do a reasonable job for IPv6.
5354 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5355 delivery, the log line now contains "pipelined" if PIPELINING was used.
5357 4. An SMTP transport process used to panic and die if the bind() call to set
5358 an explicit outgoing interface failed. This has been changed; it is now
5359 treated in the same way as a connect() failure.
5361 5. A reference to $sender_host_name in the part of a conditional expansion
5362 that was being skipped was still causing a DNS lookup. This no longer
5365 6. The def: expansion condition was not recognizing references to header lines
5366 that used bh_ and bheader_.
5368 7. Added the _cache feature to named lists.
5370 8. The code for checking quota_filecount in the appendfile transport was
5371 allowing one more file than it should have been.
5373 9. For compatibility with Sendmail, the command line option
5381 and sets the incoming protocol and host name (for trusted callers). The
5382 host name and its colon can be omitted when only the protocol is to be set.
5383 Note the Exim already has two private options, -pd and -ps, that refer to
5384 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5385 "s", but I don't think that's a major issue.
5387 10. A number of refactoring changes to the code, none of which should affect
5390 (a) The number of logging options was getting close to filling up the
5391 32-bit word that was used as a bit map. I have split them into two classes:
5392 those that are passed in the argument to log_write(), and those that are
5393 only ever tested independently outside of that function. These are now in
5394 separate 32-bit words, so there is plenty of room for expansion again.
5395 There is no change in the user interface or the logging behaviour.
5397 (b) When building, for example, log lines, the code previously used a
5398 macro that called string_cat() twice, in order to add two strings. This is
5399 not really sufficiently general. Furthermore, there was one instance where
5400 it was actually wrong because one of the argument was used twice, and in
5401 one call a function was used. (As it happened, calling the function twice
5402 did not affect the overall behaviour.) The macro has been replaced by a
5403 function that can join an arbitrary number of extra strings onto a growing
5406 (c) The code for expansion conditions now uses a table and a binary chop
5407 instead of a serial search (which was left over from when there were very
5408 few conditions). Also, it now recognizes conditions like "pam" even when
5409 the relevant support is not compiled in: a suitably worded error message is
5410 given if an attempt is made to use such a condition.
5412 11. Added ${time_interval:xxxxx}.
5414 12. A bug was causing one of the ddress fields not to be passed back correctly
5415 from remote delivery subprocesses. The field in question was not being
5416 subsequently used, so this caused to problems in practice.
5418 13. Added new log selectors queue_time and deliver_time.
5420 14. Might have fixed a bug in maildirsizefile handling that threw up
5421 "unexpected character" debug warnings, and recalculated the data
5422 unnecessarily. In any case, I expanded the warning message to give more
5425 15. Added the message "Restricted characters in address" to the statements in
5426 the default ACL that block characters like @ and % in local parts.
5428 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5429 Three changes have been made:
5431 (a) There was a serious bug; a negative response to MAIL caused the whole
5432 recipient domain to be cached as invalid, thereby blocking all messages
5433 to all local parts at the same domain, from all senders. This bug has
5434 been fixed. The domain is no longer cached after a negative response to
5435 MAIL if the sender used is not empty.
5437 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5440 (c) A new callout option, "use_sender" has been added for people who want
5441 the modified behaviour.
5447 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5450 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5451 indeed breaks things for older releases.
5453 3. Added additional logging to the case where there is a problem reading data
5454 from a filter that is running in a subprocess using a pipe, in order to
5455 try to track down a specific problem.
5457 4. Testing facility fudge: when running in the test harness and attempting
5458 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5459 getting "No route to host". Convert this to a timeout.
5461 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5464 6. Some OS don't have socklen_t but use size_t instead. This affects the
5465 fifth argument of getsockopt() amongst other things. This is now
5466 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5467 can be set for individual OS. I have set it for SunOS5, OSF1, and
5468 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5469 some earlier ones do not.
5471 7. Change 4.30/15 was not doing the test caselessly.
5473 8. The standard form for an IPv6 address literal was being rejected by address
5474 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5475 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5476 this, as well as the form without the "IPv6" on the front (but only when
5477 address literals are enabled, of course).
5479 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5481 10. Exim crashed if a message with an empty sender address specified by -f
5482 encountered a router with an errors_to setting. This could be provoked only
5483 by a command such as
5487 where an empty string was supplied; "<>" did not hit this bug.
5489 11. Installed PCRE release 4.5.
5491 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5492 remained set. It is now erased.
5494 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5495 times from message ids (which are base 36 rather than the normal 62).
5497 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5498 were being counted as actual protocol errors, and logged if the log
5499 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5500 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5501 and DATA following a set of rejected RCPTs do not count as protocol errors.
5502 In other words, Exim assumes they were pipelined, though this may not
5503 actually be the case. Of course, in all cases the client gets an
5504 appropriate error code.
5506 15. If a lookup fails in an ACL condition, a message about the failure may
5507 be available; it is used if testing the ACL cannot continue, because most
5508 such messages specify what the cause of the deferral is. However, some
5509 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5510 that caused an old message to be retained and used if a later statement
5511 caused a defer, replacing the real cause of the deferral.
5513 16. If an IP address had so many PTR records that the DNS lookup buffer
5514 was not large enough to hold them, Exim could crash while trying to process
5515 the truncated data. It now detects and logs this case.
5517 17. Further to 4.21/58, another change has been made: if (and only if) the
5518 first line of a message (the first header line) ends with CRLF, a bare LF
5519 in a subsequent header line has a space inserted after it, so as not to
5520 terminate the header.
5522 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5523 unnecessarily, used atoi() instead of strtol(), and didn't check the
5524 termination when getting file sizes from file names by regex.
5526 19. Completely re-implemented the support for maildirsize files, in the light
5527 of a number of problems with the previous contributed implementation
5528 (4.30/29). In particular:
5530 . If the quota is zero, the maildirsize file is maintained, but no quota is
5533 . If the maildir directory does not exist, it is created before any attempt
5534 to write a maildirsize file.
5536 . The quota value in the file is just a cache; if the quota is changed in
5537 the transport, the new value overrides.
5539 . A regular expression is available for excluding directories from the
5542 20. The autoreply transport checks the characters in options that define the
5543 message's headers; it allows continued headers, but it was checking with
5544 isspace() after an embedded newline instead of explicitly looking for a
5547 21. If all the "regular" hosts to which an address was routed had passed their
5548 expiry times, and had not reached their retry times, the address was
5549 bounced, even if fallback hosts were defined. Now Exim should go on to try
5552 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5553 equivalent code in the SMTP transport. Some hosts send humungous responses
5554 to HELO/EHLO, more than 1024 it seems.
5556 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5557 gives compiler warnings in some environments. I've now done it "properly",
5560 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5561 (because of problems with the built-in one) was declared to return uschar *
5562 instead of char *, causing compiler failure.
5564 25. Fixed a file descriptor leak when processing alias/forward files.
5566 26. Fixed a minor format string issue in dbfn.c.
5568 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5570 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5571 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5574 29. When a sender address is verified, it is cached, to save repeating the test
5575 when there is more than one recipient in a message. However, when the
5576 verification involves a callout, it is possible for different callout
5577 options to be set for different recipients. It is too complicated to keep
5578 track of this in the cache, so now Exim always runs a verification when a
5579 callout is required, relying on the callout cache for the optimization.
5580 The overhead is duplication of the address routing, but this should not be
5583 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5584 to be verified with callout=postmaster, and the main callout worked but the
5585 postmaster check failed, the verification correctly failed. However, if a
5586 subsequent RCPT command asked for sender verification *without* the
5587 postmaster check, incorrect caching caused this verification also to fail,
5590 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5591 it was not caching the DNS options (qualify_single, search_parents) that
5592 were used when the lookup failed. A subsequent lookup with different
5593 options therefore always gave the same answer, though there were cases
5594 where it should not have. (Example: a "domains = !$mx_any" option on a
5595 dnslookup router: the "domains" option is always processed without any
5596 widening, but the router might have qualify_single set.) Now Exim uses the
5597 cached value only when the same options are set.
5599 32. Added John Jetmore's "exipick" utility to the distribution.
5601 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5602 than a timeout (e.g. a certificate is required, and is not provided), an
5603 Exim server now closes the connection immediately. Previously it waited for
5604 the client to close - but if the client is SSL, it seems that they each
5605 wait for each other, leading to a delay before one of them times out.
5607 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5608 maintained 0.8.x compatibility because I don't think many are using it, and
5609 it is clearly obsolete.
5611 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5614 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5615 client certificate was expired. A simple patch fixes this, though I don't
5616 understand the full logic of why the verify callback is called multiple
5619 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5620 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5621 which causes problems with some clients (such as the Certicom SSL Plus
5622 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5623 disables the coutermeasure allowing Eudora to connect."
5625 38. Exim was not checking that a write() to a log file succeeded. This could
5626 lead to Bad Things if a log got too big, in particular if it hit a file
5627 size limit. Exim now panics and dies if it cannot write to a log file, just
5628 as it does if it cannot open a log file.
5630 39. Modified OS/Makefile-Linux so that it now contains
5632 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5634 The two -D definitions ensure that Exim is compiled with large file
5635 support, which makes it possible to handle log files that are bigger than
5638 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5639 instance) a domain was checked against a named list that involved a lookup,
5640 causing $domain_data to be set, then another domain was checked against the
5641 same list, then the first domain was re-checked, the value of $domain_data
5642 after the final check could be wrong. In particular, if the second check
5643 failed, it could be set empty. This bug probably also applied to
5646 41. The strip_trailing_dot option was not being applied to the address given
5647 with the -f command-line option.
5649 42. The code for reading a message's header from the spool was incrementing
5650 $received_count, but never initializing it. This meant that the value was
5651 incorrect (doubled) while delivering a message in the same process in which
5652 it was received. In the most common configuration of Exim, this never
5653 happens - a fresh exec is done - but it can happen when
5654 deliver_drop_privilege is set.
5656 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5657 it now includes up to 150 characters of the unexpected data in the log
5660 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5661 and building data strings. The size of both of these buffers was 10 000
5662 bytes - far larger than anybody would *ever* want, thought I. Needless to
5663 say, somebody hit the limit. I have increased the maximum line length to
5664 20 000 and the maximum data length of concatenated lines to 100 000. I have
5665 also fixed two bugs, because there was no checking on these buffers. Tsk,
5666 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5667 buffer is too small.
5669 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5670 lsearch lookups. Now it does.
5672 46. When parsing a route_list item in a manualroute router, a fixed-length
5673 buffer was used for the list of hosts. I made this 1024 bytes long,
5674 thinking that nobody would ever have a list of hosts that long. Wrong.
5675 Somebody had a whole pile of complicated expansion conditions, and the
5676 string was silently truncated, leading to an expansion error. It turns out
5677 that it is easier to change to an unlimited length (owing to other changes
5678 that have happened since this code was originally written) than to build
5679 structure for giving a limitation error. The length of the item that
5680 expands into the list of hosts is now unlimited.
5682 47. The lsearch lookup could not handle data where the length of text line was
5683 more than 4095 characters. Such lines were truncated, leading to shortened
5684 data being returned. It should now handle lines of any length.
5686 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5687 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5690 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5691 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5692 no longer happens in comment lines. A list of replacements is now placed
5693 at the head of all of the source files, except those whose only change is
5694 to replace PERL_COMMAND in the very first #! line.
5696 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5697 messages on the queue, with a bottom-up merge sort, using code contributed
5698 by Michael Haardt. This should make operations like -bp somewhat faster on
5699 large queues. It won't affect queue runners, except when queue_run_in_order
5702 51. Installed eximstats 1.31 in the distribution.
5704 52. Added support for SRV lookups to the dnslookup router.
5706 53. If an ACL referred to $message_body or $message_body_end, the value was not
5707 reset for any messages that followed in the same SMTP session.
5709 54. The store-handling optimization for building very long strings was not
5710 differentiating between the different store pools. I don't think this
5711 actually made any difference in practice, but I've tidied it.
5713 55. While running the routers to verify a sender address, $sender_address
5714 was still set to the sender address. This is wrong, because when routing to
5715 send a bounce to the sender, it would be empty. Therefore, I have changed
5716 it so that, while verifying a sender address, $sender_address is set to <>.
5717 (There is no change to what happens when verifying a recipient address.)
5719 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5720 target A or AAAA records (if not already returned) without resetting the
5721 qualify_single or search_parents options of the DNS resolver. These are
5722 inappropriate in this case because the targets of MX and SRV records must
5723 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5724 target that, when qualified, matched something in the local domain. These
5725 two options are now turned off when doing these lookups.
5727 57. It seems that at least some releases of Reiserfs (which does not have the
5728 concept of a fixed number of inodes) returns zero and not -1 for the
5729 number of available inodes. This interacted badly with check_spool_inodes,
5730 which assumed that -1 was the "no such thing" setting. What I have done is
5731 to check that the total number of inodes is greater than zero before doing
5732 the test of how many are available.
5734 58. When a "warn" ACL statement has a log_message modifier, the message is
5735 remembered, and not repeated. This is to avoid a lot of repetition when a
5736 message has many recipients that cause the same warning to be written.
5737 However, Exim was preserving the list of already written lines for an
5738 entire SMTP session, which doesn't seem right. The memory is now reset if a
5739 new message is started.
5741 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5742 debugging output unless log_rewrite was also set.
5744 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5745 of (int)(handle) when we know that handle contains (void *)(-1).
5747 61. The Exim daemon panic-logs an error return when it closes the incoming
5748 connection. However "connection reset by peer" seems to be common, and
5749 isn't really an error worthy of noting specially, so that particular error
5752 62. When Exim is trying to find all the local interfaces, it used to panic and
5753 die if the ioctl to get the interface flags failed. However, it seems that
5754 on at least one OS (Solaris 9) it is possible to have an interface that is
5755 included in the list of interfaces, but for which you get a failure error
5756 for this call. This happens when the interface is not "plumbed" into a
5757 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5758 failure of the "get flags" call assumes that the interface is down.
5760 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5761 makes life easier for people who are doing arithmetic on fields extracted
5762 from dates, where you often get leading zeros that should not be
5763 interpreted as octal.
5765 64. Added qualify_domain to the redirect router, to override the global
5768 65. If a pathologically long header line contained very many addresses (the
5769 report of this problem mentioned 10 000) and each of them was rewritten,
5770 Exim could use up a very large amount of memory. (It kept on making new
5771 copies of the header line as it rewrote, and never released the old ones.)
5772 At the expense of a bit more processing, the header rewriting function has
5773 been changed so that it no longer eats memory in this way.
5775 66. The generation of the Received: header has been moved from the time that a
5776 message starts to be received, to the time that it finishes. The timestamp
5777 in the Received: header should now be very close to that of the <= log
5778 line. There are two side-effects of this change:
5780 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5781 logged header lines no longer include the local Received: line, because
5782 it has not yet been created. The same applies to a copy of the message
5783 that is returned to a non-SMTP sender when a message is rejected.
5785 (b) When a filter file is tested using -bf, no additional Received: header
5786 is added to the test message. After some thought, I decided that this
5789 This change does not affect the value of $received_for. It is still set
5790 after address rewriting, but before local_scan() is called.
5792 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5794 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5795 gave an unhelpful panic error message, and a defer error. I have managed to
5796 change this behaviour so that it now rejects any supplied certificate,
5797 which seems right, as the list of acceptable certificates is empty.
5799 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5800 gave an unhelpful defer error. I have not managed to make this reject any
5801 supplied certificates, but the error message it gives is "no certificate
5802 supplied", which is not helpful.
5804 70. exigrep's output now also includes lines that are not associated with any
5805 message, but which match the given pattern. Implemented by a patch from
5806 Martin Sluka, which also tidied up the Perl a bit.
5808 71. Recipient callout verification, like sender verification, was using <> in
5809 the MAIL FROM command. This isn't really the right thing, since the actual
5810 sender may affect whether the remote host accepts the recipient or not. I
5811 have changed it to use the actual sender in the callout; this means that
5812 the cache record is now keyed on a recipient/sender pair, not just the
5813 recipient address. There doesn't seem to be a real danger of callout loops,
5814 since a callout by the remote host to check the sender would use <>.
5815 [SEE ABOVE: changed after hitting problems.]
5817 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5818 temporary errors. However, in the case of such a code being given after
5819 the end of a data transmission (i.e. after ".") Exim was failing to write
5820 a retry record for the message. (Yes, there was some broken host that was
5821 actually sending 8xx at this point.)
5823 73. An unknown lookup type in a host list could cause Exim to panic-die when
5824 the list was checked. (An example that provoked this was putting <; in the
5825 middle of a list instead of at the start.) If this happened during a DATA
5826 ACL check, a -D file could be left lying around. This kind of configuration
5827 error no longer causes Exim to die; instead it causes a defer error. The
5828 incident is still logged to the main and panic logs.
5830 74. Buglet left over from Exim 3 conversion. The message "too many messages
5831 in one connection" was written to the rejectlog but not the mainlog, except
5832 when address rewriting (yes!) was being logged.
5834 75. Added write_rejectlog option.
5836 76. When a system filter was run not as root (that is, when system_filter_user
5837 was set), the values of the $n variables were not being returned to the
5838 main process; thus, they were not subsequently available in the $sn
5841 77. Added +return_path_on_delivery log selector.
5843 78. A connection timeout was being treated differently from recipients deferred
5844 when testing hosts_max_try with a message that was older than the host's
5845 retry timeout. (The host should not be counted, thus allowing all hosts to
5846 be tried at least once before bouncing.) This may have been the cause of an
5847 occasionally reported bug whereby a message would remain on the queue
5848 longer than the retry timeout, but would be bounced if a delivery was
5849 forced. I say "may" because I never totally pinned down the problem;
5850 setting up timeout/retry tests is difficult. See also the next item.
5852 79. The ultimate address timeout was not being applied to errors that involved
5853 a combination of host plus message (for example, a timeout on a MAIL
5854 command). When an address resolved to a number of possible hosts, and they
5855 were not all tried for each delivery (e.g. because of hosts_max_try), a
5856 message could remain on the queue longer than the retry timeout.
5858 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5861 81. Fixed an obscure SMTP outgoing bug which required at least the following
5862 conditions: (a) there was another message waiting for the same server;
5863 (b) the server returned 5xx to all RCPT commands in the first message so
5864 that the message was not completed; (c) the server dropped the connection
5865 or gave a negative response to the RSET that Exim sends to abort the
5866 transaction. The observed case was a dropped connection after DATA that had
5867 been sent in pipelining mode. That is, the server had advertised PIPELINING
5868 but was not implementing it correctly. The effect of the bug was incorrect
5869 behaviour, such as trying another host, and this could lead to a crash.
5875 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5876 and daemon.c were passed as pointers to ints; they should have been
5877 pointers to socklen_t variables (which are typically unsigned ints).
5879 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5882 3. Fixed a really odd bug that affected only the testing scheme; patching a
5883 certain fixed string in the binary changed the value of another string that
5884 happened to be identical to the end of the original first string.
5886 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5887 name", it returns that address as the IP address. On some operating
5888 systems (e.g. Solaris), it also passes back the IP address string as the
5889 "host name". However, on others (e.g. Linux), it passes back an empty
5890 string. Exim wasn't checking for this, and was changing the host name to an
5891 empty string, assuming it had been canonicalized.
5893 5. Although rare, it is permitted to have more than one PTR record for a given
5894 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5895 all the names associated with an address, because they do in Solaris.
5896 However, it seems that they do not in Linux for data that comes from the
5897 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5898 I found this out when I moved to a new Linux workstation and tried to run
5899 the Exim test suite.
5901 To get round this problem I have changed the code so that it now does its
5902 own call to the DNS to look up PTR records when searching for a host name.
5903 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5904 addresses that are only in /etc/hosts are still found.
5906 This behaviour is, however, controlled by an option called host_lookup_
5907 order, which defaults to "bydns:byaddr". If people want to use the other
5908 order, or indeed, just use one or the other means of lookup, they can
5909 specify it in this variable.
5911 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5912 some operating systems, this comes back from gethostbyaddr() as an empty
5913 string, and this is what Exim used to test for. However, it seems that in
5914 other systems, "." is yielded. Exim now tests for this case too.
5916 7. The values of check_spool_space and check_log_space are now held internally
5917 as a number of kilobytes instead of an absolute number of bytes. If a
5918 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5919 kilobyte. This means that much larger values can be stored.
5921 8. Exim monitor: an attempt to get the action menu when not actually pointing
5922 at a message produces an empty menu entitled "No message selected". This
5923 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5924 no entries in it ("Shell widget menu has zero width and/or height"). So I
5925 have added a single, blank menu entry in this case.
5927 9. Added ${quote_local_part.
5929 10. MIME decoding is now applied to the contents of Subject: header lines when
5932 11. Now that a reference to $sender_host_address automatically causes a reverse
5933 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5934 host lookup before query-style lookups in lists that might use this
5935 variable. This has therefore been abolished, and the "net-" prefix is no
5936 longer necessary for query-style lookups.
5938 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5939 have been a typo for LFLAGS, so it has been changed.
5941 13. The install script calls Exim with "-C /dev/null" in order to find the
5942 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5943 to be output. However, since Exim outputs its version number before the
5944 error, it didn't break the script. It just looked ugly. I fixed this by
5945 always allowing "-C /dev/null" if the caller is root.
5947 14. Ignore overlarge ACL variable number when reading spool file - insurance
5948 against a later release with more variables having written the file.
5950 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5951 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5952 this, as well as the form without the "IPv6" on the front.
5954 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5955 OS/Makefile-Darwin file.
5957 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5958 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5960 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5961 variable. However, some OS use ints for inodes. I've added cast to long int
5962 to get rid of the compiler warning.
5964 19. I had forgotten to lock out "/../" in configuration file names when
5965 ALT_CONFIG_PREFIX was set.
5967 20. Routers used for verification do not need to specify transports. However,
5968 if such a router generated a host list, and callout was configured, Exim
5969 crashed, because it could not find a port number from the (non-existent)
5970 transport. It now assumes port 25 in this circumstance.
5972 21. Added the -t option to exigrep.
5974 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5975 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5976 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5977 them is set, LOOKUP_LSEARCH is forced.
5979 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5981 24. Added sender and host information to the "rejected by local_scan()" log
5982 line; previously there was no indication of these.
5984 25. Added .include_if_exists.
5986 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5987 Linux. It turns out that not all file systems support this. Apparently some
5988 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5989 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5990 not supported on the file descriptor, is now ignored when Exim is trying to
5991 sync a directory. This applies only to Linux.
5993 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5995 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5996 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5999 29. Added support for maildirsize files from supplied patch (modified a bit).
6001 30. The use of :fail: followed by an empty string could lead Exim to respond to
6002 sender verification failures with (e.g.):
6004 550 Verification failed for <xxx>
6005 550 Sender verify failed
6007 where the first response line was missing the '-' that indicates it is not
6008 the final line of the response.
6010 31. The loop for finding the name of the user that called Exim had a hardwired
6011 limit of 10; it now uses the value of finduser_retries, which is used for
6012 all other user lookups.
6014 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6017 33. Exim was neglecting to zero errno before one call of strtol() when
6018 expanding a string and expecting an integer value. On some systems this
6019 resulted in spurious "integer overflow" errors. Also, it was casting the
6020 result into an int without checking.
6022 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6023 did not work. The code looks as if it has *never* worked, though it appears
6024 to have been documented since at least release 1.62. I have made it work.
6026 35. The "timeout_DNS" error in retry rules, also documented since at least
6027 1.62, also never worked. As it isn't clear exactly what this means, and
6028 clearly it isn't a major issue, I have abolished the feature by treating it
6029 as "timeout", and writing a warning to the main and panic logs.
6031 36. The display of retry rules for -brt wasn't always showing the error code
6034 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6035 timeout_connect_A, timeout_connect_MX.
6037 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6038 to the empty sender.
6040 39. The daemon was not analysing the content of -oX till after it had closed
6041 stderr and disconnected from the controlling terminal. This meant that any
6042 syntax errors were only noted on the panic log, and the return code from
6043 the command was 0. By re-arranging the code a little, I've made the
6044 decoding happen first, so such errors now appear on stderr, and the return
6045 code is 1. However, the actual setting up of the sockets still happens in
6046 the disconnected process, so errors there are still only recorded on the
6049 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6050 connections (as happens on some IP stacks) was logged at start up time as
6051 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6052 it from "IPv6 and IPv4", which means that two separate sockets are being
6055 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6056 says whether AF_INET or AF_INET6 was passed as an argument.
6058 42. Exiwhat output was messed up when time zones were included in log
6061 43. Exiwhat now gives more information about the daemon's listening ports,
6062 and whether -tls-on-connect was used.
6064 44. The "port" option of the smtp transport is now expanded.
6066 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6067 silently ignored. Now an error message is written to the main and panic
6070 46. There's a new ACL modifier called "logwrite" which writes to a log file
6071 as soon as it is encountered.
6073 47. Added $local_user_uid and $local_user_gid at routing time.
6075 48. Exim crashed when trying to verify a sender address that was being
6078 49. Exim was recognizing only a space character after ".include". It now also
6079 recognizes a tab character.
6081 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6082 extracting the relevant information from the specification. The man page no
6083 longer contains scrambled data for the -d option, and I've added a section
6084 at the front about calling Exim under different names.
6086 51. Added "extra_headers" argument to the "mail" command in filter files.
6088 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6091 53. Installed eximstats 1.29.
6093 54. Added transport_filter_timeout as a generic transport option.
6095 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6096 Cc: header lines. This was required by RFC 822, but it not required by RFC
6099 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6100 incoming messages that did not have them. Now it does so only if the
6101 message originates locally, that is, if there is no associated remote host
6102 address. When Resent- header lines are present, this applies to the Resent-
6103 lines rather than the non-Resent- lines.
6105 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6106 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6108 58. Messages for configuration errors now include the name of the main
6109 configuration file - useful now that there may be more than one file in a
6110 list (.included file names were always shown).
6112 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6113 for those rare installations that do not start the daemon as root or run it
6114 setuid root. I've cut out the call to initgroups() if the daemon is not
6117 60. The Exim user and group can now be bound into the binary as text strings
6118 that are looked up at the start of Exim's processing.
6120 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6122 62. Added $mailstore_basename variable.
6124 63. Installed patch to sieve.c from Michael Haardt.
6126 64. When Exim failed to open the panic log after failing to open the main log,
6127 the original message it was trying to log was written to stderr and debug
6128 output, but if they were not available (the usual case in production), it
6129 was lost. Now it is written to syslog before the two lines that record the
6130 failures to open the logs.
6132 65. Users' Exim filters run in subprocesses under the user's uid. It is
6133 possible for a "deliver" command or an alias in a "personal" command to
6134 provoke an address rewrite. If logging of address rewriting is configured,
6135 this fails because the process is not running as root or exim. There may be
6136 a better way of dealing with this, but for the moment (because 4.30 needs
6137 to be released), I have disabled address rewrite logging when running a
6138 filter in a non-root, non-exim process.
6144 1. The buildconfig auxiliary program wasn't quoting the value set for
6145 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6146 not defined. This bug was masked in 4.22 by the effect that was fixed in
6149 2. Some messages that were rejected after a message id was allocated were
6150 shown as "incomplete" by exigrep. It no longer does this for messages that
6151 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6153 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6154 have allow_domain_literals set, the ID did not get logged in the <= line.
6155 Domain literals are now always recognized in Message-ID: header lines.
6157 4. The first argument for a ${extract expansion item is the key name or field
6158 number. Leading and trailing spaces in this item were not being ignored,
6159 causing some misleading effects.
6161 5. When deliver_drop_privilege was set, single queue runner processes started
6162 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6163 same command in the process it spins off) were not dropping privilege.
6165 6. When the daemon running as "exim" started a queue runner, it always
6166 re-executed Exim in the spun-off process. This is a waste of effort when
6167 deliver_drop_privilege is set. The new process now just calls the
6168 queue-runner function directly.
6174 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6177 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6178 ignored. Though the use of -oP was forcing the writing of a pid file, it
6179 was always written to the default place.
6181 3. If the message "no IP address found for host xxxx" is generated during
6182 incoming verification, it is now followed by identification of the incoming
6183 connection (so you can more easily find what provoked it).
6185 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6187 5. Added some features to "harden" Exim a bit more against certain attacks:
6189 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6190 be put in Local/Makefile. This is like the never_users runtime option,
6191 but it cannot be overridden. The default setting is "root".
6193 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6194 prefix string with which any file named in a -C command line option
6197 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6198 is retained for -C and -D only if the caller of Exim is root. Without
6199 it, the exim user may also use -C and -D and retain privilege.
6201 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6202 command line option is disabled.
6204 6. Macro names set by the -D option must start with an upper case letter, just
6205 like macro names defined in the configuration file.
6207 7. Added "dereference=" facility to LDAP.
6209 8. Two instances of the typo "uknown" in the source files are fixed.
6211 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6212 the Configure-Makefile script screwed up while processing it.
6214 10. Incorporated PCRE 4.4.
6216 11. The SMTP synchronization check was not operating right at the start of an
6217 SMTP session. For example, it could not catch a HELO sent before the client
6218 waited for the greeting. There is now a check for outstanding input at the
6219 point when the greeting is written. Because of the duplex, asynchronous
6220 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6221 way, but not yet received, when the check is performed.
6223 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6224 on TCP/IP sockets, because this apparently causes some broken clients to
6227 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6228 unchanged) from the Cygwin maintainer.
6230 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6231 was supported instead of testing for mbx. Effectively a typo.
6233 15. The spa authenticator server code was not checking that the input it
6234 received was valid base64.
6236 16. The debug output line for the "set" modifier in ACLs was not showing the
6237 name of the variable that was being set.
6239 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6241 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6242 lookup on its own. Something else was needed to trigger the lookup. For
6243 example, a match in host_lookup or the need for a host name in a host list.
6244 Now, if $sender_host_name is referenced and the host name has not yet been
6245 looked up, a lookup is performed. If the lookup fails, the variable remains
6246 empty, and $host_lookup_failed is set to "1".
6248 19. Added "eqi" as a case-independent comparison operator.
6250 20. The saslauthd authentication condition could segfault if neither service
6251 nor realm was specified.
6253 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6254 error message that was logged was misleading, and incoming SMTP
6255 connections were dropped. The message is now more accurate, and temporary
6256 errors are given to SMTP connections.
6258 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6259 (except RSET) with a 421 error, until QUIT is received. However, it was
6260 failing to send a response to QUIT.
6262 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6263 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6264 ACL is now run after the helo_try_verify_hosts code.
6266 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6267 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6268 case-independent, so other case variants are also recognized). Apparently
6269 some systems use these upper case variants.
6271 25. If more than two messages were waiting for the same host, and a transport
6272 filter was specified for the transport, Exim sent two messages over the
6273 same TCP/IP connection, and then failed with "socket operation on non-
6274 socket" when it tried to send the third.
6276 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6278 27. The extern definition of crypt16() in expand.c was not being excluded when
6279 the OS had its own crypt16() function.
6281 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6282 as a preferred synonym for return_size_limit, both as an option and as an
6285 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6287 30. Changed the default configuration ACL to relax the local part checking rule
6288 for addresses that are not in any local domains. For these addresses,
6289 slashes and pipe symbols are allowed within local parts, but the sequence
6290 /../ is explicitly forbidden.
6292 31. SPA server authentication was not clearing the challenge buffer before
6295 32. log_message in a "warn" ACL statement was writing to the reject log as
6296 well as to the main log, which contradicts the documentation and doesn't
6297 seem right (because no rejection is happening). So I have stopped it.
6299 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6300 However, I am unable to do any testing of this.
6302 34. Fixed an infelicity in the appendfile transport. When checking directories
6303 for a mailbox, to see if any needed to be created, it was accidentally
6304 using path names with one or more superfluous leading slashes; tracing
6305 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6307 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6308 contents are added to the log line that is written for every discarded
6309 recipient. (Previously a log_message setting was ignored.)
6311 36. The ${quote: operator now quotes the string if it is empty.
6313 37. The install script runs exim in order to find its version number. If for
6314 some reason other than non-existence or emptiness, which it checks, it
6315 could not run './exim', it was installing it with an empty version number,
6316 i.e. as "exim-". This error state is now caught, and the installation is
6319 38. An argument was missing from the function that creates an error message
6320 when Exim fails to connect to the socket for saslauthd authentication.
6321 This could cause Exim to crash, or give a corrupted message.
6323 39. Added isip, isip4, and isip6 to ${if conditions.
6325 40. The ACL variables $acl_xx are now saved with the message, and can be
6326 accessed later in routers, transports, and filters.
6328 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6329 strings in the file are not string-expanded.
6331 42. If a MAIL command specified a SIZE value that was too large to fit into an
6332 int variable, the check against message_size_limit failed. Such values are
6333 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6334 day this will have to be increased, but I don't think I want to be around
6335 when emails are that large.
6342 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6343 iconv() is not standard in FreeBSD.
6345 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6346 IPv6 enabled. The observed symptom was a segmentation fault on return from
6347 the function os_common_find_running_interfaces() in src/os.c.
6349 3. In the check_special_case() function in daemon.c I had used "errno" as an
6350 argument name, which causes warnings on some systems. This was basically a
6351 typo, since it was named "eno" in the comments!
6353 4. The code that waits for the clock to tick (at a resolution of some fraction
6354 of a second) so as to ensure message-id uniqueness was always waiting for
6355 at least one whole tick, when it could have waited for less. [This is
6356 almost certainly not relevant at current processor speeds, where it is
6357 unlikely to ever wait at all. But we try to future-proof.]
6359 5. The function that sleeps for a time interval that includes fractions of a
6360 second contained a race. It did not block SIGALRM between setting the
6361 timer, and suspending (a couple of lines later). If the interval was short
6362 and the sigsuspend() was delayed until after it had expired, the suspension
6363 never ended. On busy systems this could lead to processes getting stuck for
6366 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6367 process, before it forks any delivery processes. The open lookup caching
6368 mechanism meant that the open file or database connection was passed into
6369 the delivery process. The problem was that delivery processes always tidy
6370 up cached lookup data. This could cause a problem for the next delivery
6371 process started by the queue runner, because the external queue runner
6372 process does not know about the closure. So the next delivery process
6373 still has data in the lookup cache. In the case of a file lookup, there was
6374 no problem because closing a file descriptor in a subprocess doesn't affect
6375 the parent. However, if the lookup was caching a connection to a database,
6376 the connection was closed, and the second delivery process was likely to
6377 see errors such as "PGSQL: query failed: server closed the connection
6378 unexpectedly". The problem has been fixed by closing all cached lookups
6379 in a queue runner before running a delivery process.
6381 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6382 seem to have the "const" qualifier which it has on other OS. I've
6385 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6386 *of the same type* (client or server) with the same public name that an
6387 error should be diagnosed.
6389 9. When Exim looked up a host name for an IP address, but failed to find the
6390 original IP address when looking up the host name (a safety check), it
6391 output the message "<ip address> does not match any IP for NULL", which was
6392 confusing, to say the least. The bug was that the host name should have
6393 appeared instead of "NULL".
6395 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6396 user that is built into the binary, and the -C or -D options is used, root
6397 privilege is dropped before the configuration file is read. In addition,
6398 logging is switched to stderr instead of the normal log files. If the
6399 configuration then re-defines the Exim user, the unprivileged environment
6400 is probably not what is expected, so Exim logs a panic warning message (but
6403 However, if deliver_drop_privilege is set, the unprivileged state may well
6404 be exactly what is intended, so the warning has been cut out in that case,
6405 and Exim is allowed to try to write to its normal log files.
6411 1. smtp_return_error_details was not giving details for temporary sender
6412 or receiver verification errors.
6414 2. Diagnose a configuration error if two authenticators have the same public
6417 3. Exim used not to create the message log file for a message until the first
6418 delivery attempt. This could be confusing when incoming messages were held
6419 for policy or load reasons. The message log file is now created at the time
6420 the message is received, and an initial "Received" line is written to it.
6422 4. The automatically generated man page for command line options had a minor
6423 bug that caused no ill effects; however, a more serious problem was that
6424 the procedure for building the man page automatically didn't always
6425 operate. Consequently, release 4.20 contains an out-of-date version. This
6426 shouldn't happen again.
6428 5. When building Exim with embedded Perl support, the script that builds the
6429 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6430 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6432 6. The freeze_tell option was not being used for messages that were frozen on
6433 arrival, either by an ACL or by local_scan().
6435 7. Added the smtp_incomplete_transaction log selector.
6437 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6438 was accepting AUTH without a new EHLO.
6440 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6441 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6444 10. Logging of TCP/IP connections (when configured) now happens in the main
6445 daemon process instead of the child process, so that the TCP/IP connection
6446 count is more accurate (but it can never be perfect).
6448 11. The use of "drop" in a nested ACL was not being handled correctly in the
6449 outer ACL. Now, if condition failure induced by the nested "drop" causes
6450 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6451 or "require"), the connection is dropped.
6453 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6454 that yield "discard" can now be used with an "accept" or a "discard" verb,
6455 but an error is generated for any others (because I can't see a useful way
6456 to define what should happen).
6458 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6459 are now processed in the same way as lines in the Exim configuration file.
6460 In particular, continuation lines are supported.
6462 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6464 15. Added -ti meaning -t -i.
6466 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6467 domains, and warn by logging if others are found.
6469 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6470 returned from GIFCONF when Exim is trying to find the list of interfaces on
6471 a host. The code in os.c has been modified to copy each ifreq to an aligned
6472 structure in all cases.
6474 Also, in some cases, the returned ifreq's were being copied to a 'struct
6475 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6476 means the last couple of bytes of an IPv6 address could be chopped if the
6477 ifreq contained only a normal sockaddr (14 bytes storage).
6479 18. Named domain lists were not supported in the hosts_treat_as_local option.
6480 An entry such as +xxxx was not recognized, and was treated as a literal
6483 19. Ensure that header lines added by a DATA ACL are included in the reject log
6484 if the ACL subsequently rejects the message.
6486 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6487 MD5 (which is deprecated).
6489 21. When testing a filter file using -bf, Exim was writing a message when it
6490 took the sender from a "From " line in the message, but it was not doing so
6491 when it took $return_path from a Return-Path: header line. It now does.
6493 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6494 with a valid header line field name (a series of printing characters
6495 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6497 23. Changed "disc" in the source to "disk" to conform to the documentation and
6498 the book and for uniformity.
6500 24. Ignore Sendmail's -Ooption=value command line item.
6502 25. When execve() failed while trying to run a command in a pipe transport,
6503 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6504 could be confused with a return value of 69 from the command itself. This
6505 has been changed to 127, the value the shell returns if it is asked to run
6506 a non-existent command. The wording for the related log line suggests a
6507 non-existent command as the problem.
6509 26. If received_header_text expands to an empty string, do not add a Received:
6510 header line to the message. (Well, it adds a token one on the spool, but
6511 marks it "old" so that it doesn't get used or transmitted.)
6513 27. Installed eximstats 1.28 (addition of -nt option).
6515 28. There was no check for failure on the call to getsockname() in the daemon
6516 code. This can fail if there is a shortage of resources on the system, with
6517 ENOMEM, for example. A temporary error is now given on failure.
6519 29. Contrary to the C standard, it seems that in some environments, the
6520 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6521 program. Exim now does this explicitly; it affects the formatting of
6522 timestamps using strftime().
6524 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6525 complaints. I've now initialized all the variables, to avoid this.
6527 32. Header lines added by a system filter were not being "seen" during
6528 transport-time rewrites.
6530 33. The info_callback() function passed to OpenSSL is set up with type void
6531 (*)(SSL *, int, int), as described somewhere. However, when calling the
6532 function (actually a macro) that sets it up, the type void(*)() is
6533 expected. I've put in a cast to prevent warnings from picky compilers.
6535 34. If a DNS black list lookup found a CNAME record, but there were no A
6536 records associated with the domain it pointed at, Exim crashed.
6538 35. If a DNS black list lookup returned more than one A record, Exim ignored
6539 all but the first. It now scans all returned addresses if a particular IP
6540 value is being sought. In this situation, the contents of the
6541 $dnslist_value variable are a list of all the addresses, separated by a
6544 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6545 to accept a host name and all its aliases if the forward lookup for any of
6546 them yielded the IP address of the incoming connection. Now it accepts only
6547 those names whose forward lookup yields the correct IP address. Any other
6548 names are discarded. This closes a loophole whereby a rogue DNS
6549 administrator could create reverse DNS records to break through a
6550 wildcarded host restriction in an ACL.
6552 37. If a user filter or a system filter that ran in a subprocess used any of
6553 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6554 the wrong values were passed to the pipe command ($thisaddress had the
6555 value of $0, $0 had the value of $1, etc). This bug was introduced by
6556 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6558 38. Improved the line breaking for long SMTP error messages from ACLs.
6559 Previously, if there was no break point between 40 and 75 characters, Exim
6560 left the rest of the message alone. Two changes have been made: (a) I've
6561 reduced the minimum length to 35 characters; (b) if it can't find a break
6562 point between 35 and 75 characters, it looks ahead and uses the first one
6563 that it finds. This may give the occasional overlong line, but at least the
6564 remaining text gets split now.
6566 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6567 file descriptors might be low, and that setting 1000 would always raise it.
6568 It turns out that in some environments, the limit is already over 1000 and
6569 that lowering it causes trouble. So now Exim takes care not to decrease it.
6571 40. When delivering a message, the value of $return_path is set to $sender_
6572 address at the start of routing (routers may change the value). By an
6573 oversight, this default was not being set up when an address was tested by
6574 -bt or -bv, which affected the outcome if any router or filter referred to
6577 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6578 log only when "message" or "log_message" are set. However, if one of the
6579 conditions was an address verification, or a call to a nested ACL, the
6580 messages generated by the underlying test were being passed through. This
6581 no longer happens. The underlying message is available in $acl_verify_
6582 message for both "message" and "log_message" expansions, so it can be
6583 passed through if needed.
6585 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6586 new expansion $bh_ to give the encoded byte string without charset
6587 translation. Translation happens only if iconv() is available; HAVE_ICONV
6588 indicates this at build time. HEADERS_CHARSET gives the charset to
6589 translate to; headers_charset can change it in the configuration, and
6590 "headers charset" can change it in an individual filter file.
6592 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6593 that creates RFC 2047 encoded "words" labels them as that charset instead
6594 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6595 expansion operator; (ii) when Exim creates a From: line for a local
6596 message; (iii) when a header line is rewritten to include a "phrase" part.
6598 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6599 buggy, causing it to skip the first lines of messages whose message ID
6600 ended in 'D'. This would not have bitten before Exim release 4.14, because
6601 message IDs were unlikely to end in 'D' before then. The effect was to have
6602 incorrect size information for certain domains.
6604 45. #include "config.h" was missing at the start of the crypt16.c module. This
6605 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6608 46. If there was a timeout during a "random" callout check, Exim treated it as
6609 a failure of the random address, and carried on sending RSET and the real
6610 address. If the delay was just some slowness somewhere, the response to the
6611 original RCPT would be taken as a response to RSET and so on, causing
6612 mayhem of various kinds.
6614 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6615 when I implemented it. It didn't allow for the fact that some option values
6616 may legitimately be negative (e.g. size_addition), and it didn't even do
6617 the right test for positive values.
6619 48. Domain names in DNS records are case-independent. Exim always looks them up
6620 in lower case. Some resolvers return domain names in exactly the case they
6621 appear in the zone file, that is, they may contain uppercase letters. Not
6622 all resolvers do this - some return always lower case. Exim was treating a
6623 change of case by a resolver as a change of domain, similar to a widening
6624 of a domain abbreviation. This triggered its re-routing code and so it was
6625 trying to route what was effectively the same domain again. This normally
6626 caused routing to fail (because the router wouldn't handle the domain
6627 twice). Now Exim checks for this case specially, and just changes the
6628 casing of the domain that it ultimately uses when it transmits the message
6631 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6634 50. If a filter generated a file delivery with a non-absolute name (possible if
6635 no home directory exists for the router), the forbid_file option was not
6638 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6639 the existing equality matching.
6641 52. Exim was using ints instead of ino_t variables in some places where it was
6642 dealing with inode numbers.
6644 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6645 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6646 called TMPDIR, and if it finds it is different, it changes its value.
6648 54. The smtp_printf() function is now made available to local_scan() so
6649 additional output lines can be written before returning. There is also an
6650 smtp_fflush() function to enable the detection of a dropped connection.
6651 The variables smtp_input and smtp_batched_input are exported to
6654 55. Changed the default runtime configuration: the message "Unknown user"
6655 has been removed from the ACL, and instead placed on the localuser router,
6656 using the cannot_route_message feature. This means that any verification
6657 failures that generate their own messages won't get overridden. Similarly,
6658 the "Unrouteable address" message that was in the ACL for unverifiable
6659 relay addresses has also been removed.
6661 56. Added hosts_avoid_esmtp to the smtp transport.
6663 57. The exicyclog script was not checking for the esoteric option
6664 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6665 will work only if exicyclog is run under the appropriate euid.
6667 58. Following a discussion on the list, the rules by which Exim recognises line
6668 endings on incoming messages have been changed. The -dropcr and drop_cr
6669 options are now no-ops, retained only for backwards compatibility. The
6670 following line terminators are recognized: LF CRLF CR. However, special
6671 processing applies to CR:
6673 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6674 nor a local message in the state where . is a terminator.
6676 (ii) If a bare CR is encountered in a header line, an extra space is added
6677 after the line terminator so as not to end the header. The reasoning
6678 behind this is that bare CRs in header lines are most likely either
6679 to be mistakes, or people trying to play silly games.
6681 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6682 was being incorrectly given as 18 bytes larger than it should have been.
6683 This is a VOB (very old bug).
6685 60. This may never have affected anything current, but just in case it has:
6686 When the local host is found other than at the start of a list of hosts,
6687 the local host, those with the same MX, and any that follow, are discarded.
6688 When the list in question was part of a longer list of hosts, the following
6689 hosts (not currently being processed) were also being discarded. This no
6690 longer happens. I'm not sure if this situation could ever has previously
6693 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6696 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6697 header. This is something that is recommended in a new Internet Draft, and
6698 is something that is documented as being done by Sendmail. There are two
6699 possible values. For messages generated by the autoreply transport, Exim
6702 Auto-Submitted: auto-replied
6704 whereas for all other generated messages (e.g. bounces) it adds
6706 Auto-Submitted: auto-generated
6708 63. The "personal" condition in filters now includes a test for the
6709 Auto-Submitted: header. If it contains the string "auto-" the message it
6710 not considered personal.
6712 64. Added rcpt_include_affixes as a generic transport option.
6714 65. Added queue_only_override (default true).
6716 66. Added the syslog_duplication option.
6718 67. If what should have been the first header line of a message consisted of
6719 a space followed by a colon, Exim was mis-interpreting it as a header line.
6720 It isn't of course - it is syntactically invalid and should therefore be
6721 treated as the start of the message body. The misbehaviour could have
6722 caused a number of strange effects, including loss of data in subsequent
6723 header lines, and spool format errors.
6725 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6726 client host had authenticated. This control can now be exercised by an ACL
6727 for more flexibility.
6729 69. By default, callouts do not happen when testing with -bh. There is now a
6730 variant, -bhc, which does actually run the callout code, including
6731 consulting and updating the callout cache.
6733 70. Added support for saslauthd authentication, courtesy of Alexander
6736 71. If statvfs() failed on the spool or log directories while checking their
6737 size for availability, Exim confusingly gave the error "space shortage".
6738 Furthermore, in debugging mode it crashed with a floating point exception.
6739 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6740 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6741 serious problem, Exim now writes to the main and panic logs when this
6742 happens, with details of the failure. It then refuses to accept the
6743 incoming message, giving the message "spool directory problem" or "log
6744 directory problem" with a 421 code for SMTP messages.
6746 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6747 0, 1, and 2 exist, because some OS complain for execs without them (see
6748 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6749 descriptors. However, the code omitted to check that the open succeeded,
6750 causing mysterious errors if for some reason the permissions on /dev/null
6751 got screwed. Now Exim writes a message to the main and panic logs, and
6752 bombs out if it can't open /dev/null.
6754 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6755 interact so that it is all more flexible. It is supposed to remain
6756 backwards compatible. Also added extra_local_interfaces.
6758 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6759 to bomb out with an assertion failure - to the client this appears as a
6760 connection drop. This problem occurs in the part of the code that was taken
6761 from the Samba project. Fortunately, the assertion is in a very simple
6762 function, so I have fixed this by reproducing the function inline in the
6763 one place where it is called, and arranging for authentication to fail
6764 instead of killing the process with assert().
6766 75. The SPA client code was not working when the server requested OEM rather
6767 than Unicode encoding.
6769 76. Added code to make require_files with a specific uid setting more usable in
6770 the case where statting the file as root fails - usually a non-root-mounted
6771 NFS file system. When this happens and the failure is EACCES, Exim now
6772 forks a subprocess and does the per-uid checking as the relevant uid.
6774 77. Added process_log_path.
6776 78. If log_file_path was not explicitly set, a setting of check_log_space or
6777 check_log_inodes was ignored.
6779 79. If a space check for the spool or log partitions fails, the incident is now
6780 logged. Of course, in the latter case the data may get lost...
6782 80. Added the %p formatting code to string_format() so that it can be used to
6783 print addresses in debug_print(). Adjusted all the address printing in the
6784 debugging in store.c to use %p rather than %d.
6786 81. There was a concern that a line of code in smtp_in.c could overflow a
6787 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6788 initially expressed, the concern was not well-founded, because trailing
6789 spaces are removed early. However, if the trailing spaces were followed by
6790 a NULL, they did not get removed, so the overflow was possible. Two fixes
6793 (a) I re-wrote the offending code in a cleaner fashion.
6794 (b) If an incoming SMTP command contains a NULL character, it is rejected
6797 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6798 runs initgroups(), so that if the Exim user is in any additional groups,
6799 they will be used during message reception.
6805 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob