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
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
64 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
65 4.77 as the kernel support then wasn't solid, having issues in 64bit
66 mode. Now, it's been long enough.
68 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
69 case where the routing stage had gathered several addresses to send to
70 a host before calling the transport for the first, we previously failed
71 to close down TLS in the old transport process before passing the TCP
72 connection to the new process. The new one sent a STARTTLS command
73 which naturally failed, giving a failed delivery and bloating the retry
74 database. Investigation and fix prototype from Wolfgang Breyha.
76 JH/12 Fix check on SMTP command input synchronisation. Previously there were
77 false-negatives in the check that the sender had not preempted a response
78 or prompt from Exim (running as a server), due to that code's lack of
79 awareness of the SMTP input buferring.
85 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
86 than -2003 did; needs libidn2 in addition to libidn.
88 JH/02 The path option on a pipe transport is now expanded before use.
90 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
91 Patch provided by "Björn", documentation fix added too.
93 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
94 missing a wire-to-host endian conversion.
96 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
97 close after a BDAT command line could be taken as a following command,
98 giving a synch failure. Fix by only checking for synch immediately
99 before acknowledging the chunk.
101 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
102 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
103 macro. Patches provided by Josh Soref.
105 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
106 Previously we did not; the RFC seems ambiguous and VRFY is not listed
107 by IANA as a service extension. However, John Klensin suggests that we
110 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
111 the dkim code may be unix-mode line endings rather than smtp wire-format
112 CRLF, so prepend a CR to any bare LF.
114 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
115 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
117 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
118 the first were themselves being wrongly included in the feed into dkim
119 processing; with most chunk sizes in use this resulted in an incorrect
120 body hash calculated value.
122 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
123 DKIM signature block, for verification. Although advised against by
124 standards it is specifically not ruled illegal.
126 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
128 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
129 missing a body hash (the bh= tag).
131 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
132 It seems that HAProxy sends the Proxy Protocol information in clear and
133 only then does a TLS startup, so do the same.
135 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
136 TCP connections (such as for Spamd) unless the daemon successfully set
137 Fast Open mode on its listening sockets. This fixes breakage seen on
138 too-old kernels or those not configured for Fast Open, at the cost of
139 requiring both directions being enabled for TFO, and TFO never being used
140 by non-daemon-related Exim processes.
142 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
143 endings, at least on the first header line. Try to canonify any that get
144 past that check, despite the cost.
146 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
147 now limited to an arbitrary five deep, while parsing addresses with the
148 strip_excess_angle_brackets option enabled.
150 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
151 instead leave the unprompted TLS handshake in socket buffer for the
152 TLS library to consume.
154 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
156 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
158 JH/16 Drop variables when they go out of scope. Memory management drops a whole
159 region in one operation, for speed, and this leaves assigned pointers
160 dangling. Add checks run only under the testsuite which checks all
161 variables at a store-reset and panics on a dangling pointer; add code
162 explicitly nulling out all the variables discovered. Fixes one known
163 bug: a transport crash, where a dangling pointer for $sending_ip_address
164 originally assigned in a verify callout, is re-used.
166 PP/06 Drop '.' from @INC in various Perl scripts.
168 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
170 PP/08 Reduce a number of compilation warnings under clang; building with
171 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
172 should be warning-free.
174 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
176 HS/01 Fix portability problems introduced by PP/08 for platforms where
177 realloc(NULL) is not equivalent to malloc() [SunOS et al].
179 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
180 chunk. This allows us to accept broken chunked messages. We need a more
181 general solution here.
183 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
184 already-broken messages in the queue.
186 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
188 JH/19 Fix reference counting bug in routing-generated-address tracking.
194 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
195 supports it and a size is available (ie. the sending peer gave us one).
197 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
198 years of being deprecated). The replacements are the ACLs
199 acl_smtp_mime and acl_not_smtp_mime.
201 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
202 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
203 or even an in-clear connection were permitted. Now, if the host lookup
204 was dnssec and dane was requested then the host is only used if the
205 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
206 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
207 if one fails this test.
208 This means that a poorly-configured remote DNS will make it incommunicado;
209 but it protects against a DNS-interception attack on it.
211 JH/04 Bug 1810: make continued-use of an open smtp transport connection
212 non-noisy when a race steals the message being considered.
214 JH/05 If main configuration option tls_certificate is unset, generate a
215 self-signed certificate for inbound TLS connections.
217 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
218 in rewrites and routers.
220 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
221 and logged a warning sing 4.83; now they are a configuration file error.
223 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
224 (lacking @domain). Apply the same qualification processing as RCPT.
226 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
228 JH/10 Support ${sha256:} applied to a string (as well as the previous
231 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
232 a cutthrough deliver is pending, as we always want to make a connection.
233 This also avoids re-routing the message when later placing the cutthrough
234 connection after a verify cache hit.
235 Do not update it with the verify result either.
237 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
238 when routing results in more than one destination address.
240 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
241 signing (which inhibits the cutthrough capability). Previously only
242 the presence of an option was tested; now an expansion evaluating as
243 empty is permissible (obviously it should depend only on data available
244 when the cutthrough connection is made).
246 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
247 the relevant preceding SMTP command did not note the pipelining mode.
249 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
250 Previously they were not counted.
252 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
253 as one having no matching records. Previously we deferred the message
254 that needed the lookup.
256 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
257 distinguished as "(=".
259 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
260 for missing MX records. Previously it only worked for missing A records.
262 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
264 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
265 after the data-go-ahead and data-ack. Patch from Jason Betts.
267 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
268 even for a "none" policy. Patch from Tony Meyer.
270 JH/22 Fix continued use of a connection for further deliveries. If a port was
271 specified by a router, it must also match for the delivery to be
274 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
275 When one of the recipients of a message was unsuitable for the connection
276 (has no matching addresses), we lost track of needing to mark it
277 deferred. As a result mail would be lost.
279 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
281 JH/25 Decoding ACL controls is now done using a binary search; the source code
282 takes up less space and should be simpler to maintain. Merge the ACL
283 condition decode tables also, with similar effect.
285 JH/26 Fix problem with one_time used on a redirect router which returned the
286 parent address unchanged. A retry would see the parent address marked as
287 delivered, so not attempt the (identical) child. As a result mail would
290 JH/27 Fix a possible security hole, wherein a process operating with the Exim
291 UID can gain a root shell. Credit to http://www.halfdog.net/ for
292 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
295 JH/28 Enable {spool,log} filesystem space and inode checks as default.
296 Main config options check_{log,spool}_{inodes,space} are now
297 100 inodes, 10MB unless set otherwise in the configuration.
299 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
300 Previously it only applied to the main-section connection policy
303 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
305 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
306 by me. Added RFC7919 DH primes as an alternative.
308 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
309 are not in the system include path.
311 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
312 GnuTLS, when a session startup failed (eg because the client disconnected)
313 Exim did stdio operations after fclose. This was exposed by a recent
314 change which nulled out the file handle after the fclose.
316 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
317 signed directly by the cert-signing cert, rather than an intermediate
318 OCSP-signing cert. This is the model used by LetsEncrypt.
320 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
322 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
323 an incoming connection.
325 HS/02 Bug 1802: Do not half-close the connection after sending a request
328 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
329 fallback to "prime256v1".
331 JH/34 SECURITY: Use proper copy of DATA command in error message.
332 Could leak key material. Remotely exploitable. CVE-2016-9963.
338 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
339 and 3.4.4 - once the server is enabled to respond to an OCSP request
340 it does even when not requested, resulting in a stapling non-aware
341 client dropping the TLS connection.
343 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
344 support variable-length bit vectors. No functional change.
346 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
347 The I= interface field on outgoing lines is now after the H= remote
348 host field, same as incoming lines. There is a separate
349 outgoing_interface log selector which allows you to disable the
352 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
353 If not running log_selector +smtp_connection the mainlog would be held
354 open indefinitely after a "too many connections" event, including to a
355 deleted file after a log rotate. Leave the per net connection logging
356 leaving it open for efficiency as that will be quickly detected by the
357 check on the next write.
359 HS/01 Bug 1671: Fix post transport crash.
360 Processing the wait-<transport> messages could crash the delivery
361 process if the message IDs didn't exist for some reason. When
362 using 'split_spool_directory=yes' the construction of the spool
363 file name failed already, exposing the same netto behaviour.
365 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
366 mime_regex ACL conditions.
368 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
369 to DSN fail messages (bounces): remote IP, remote greeting, remote response
370 to HELO, local diagnostic string.
372 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
373 log line to debug. Even when configured with a tls authenticator many
374 client connections are expected to not authenticate in this way, so
375 an authenticate fail is not an error.
377 HS/02 Add the Exim version string to the process info. This way exiwhat
378 gives some more detail about the running daemon.
380 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
381 matter for fast-change records such as DNSBLs.
383 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
384 retry record, even if constant. There may be multiple transports with
385 different interface settings and the retry behaviour needs to be kept
388 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
390 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
392 JH/10 Bug 840: fix log_defer_output option of pipe transport
394 JH/11 Bug 830: use same host for all RCPTS of a message, even under
395 hosts_randomize. This matters a lot when combined with mua_wrapper.
397 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
398 ${quote_pgsql:<string>} operator.
400 JH/13 Bug 1708: avoid misaligned access in cached lookup.
402 JH/14 Change header file name for freeradius-client. Relevant if compiling
403 with Radius support; from the Gentoo tree and checked under Fedora.
405 JH/15 Bug 1712: Introduce $prdr_requested flag variable
407 JH/16 Bug 1714: Permit an empty string as expansion result for transport
408 option transport_filter, meaning no filtering.
410 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
412 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
413 defaults to "*" (all hosts). The variable is now available when not built
414 with TLS, default unset, mainly to enable keeping the testsuite sane.
415 If a server certificate is not supplied (via tls_certificate) an error is
416 logged, and clients will find TLS connections fail on startup. Presumably
417 they will retry in-clear.
418 Packagers of Exim are strongly encouraged to create a server certificate
419 at installation time.
421 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
422 with the $config_file variable.
424 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
425 in transport context, after the attempt, and per-recipient. The latter type
426 is per host attempted. The event data is the error message, and the errno
427 information encodes the lookup type (A vs. MX) used for the (first) host,
428 and the trailing two digits of the smtp 4xx response.
430 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
431 to write to mainlog (or rejectlog, paniclog) in the window between file
432 creation and permissions/ownership being changed. Particularly affects
433 installations where exicyclog is run as root, rather than exim user;
434 result is that the running daemon panics and dies.
436 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
438 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
439 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
440 "pri" and "weight". Note that the previous implicit priority given by the
441 list order is no longer honoured.
443 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
446 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
447 by defining SUPPORT_SOCKS.
449 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
450 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
451 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
452 variables are renamed to proxy_{local,external}_{address,port}.
454 JH/27 Move Internationalisation support from Experimental to mainline, enabled
455 for a build by defining SUPPORT_I18N
457 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
458 of the query string, and make ${quote_redis:} do that quoting.
460 JH/29 Move Events support from Experimental to mainline, enabled by default
461 and removable for a build by defining DISABLE_EVENT.
463 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
465 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
466 cached by the daemon.
468 JH/32 Move Redis support from Experimental to mainline, enabled for a build
469 by defining LOOKUP_REDIS. The libhiredis library is required.
471 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
472 keys are given for lookup.
474 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
475 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
476 only supported when built with TLS support. The PolarSSL SHA routines
477 are still used when the TLS library is too old for convenient support.
479 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
480 openssl_options), for security. OpenSSL forces this from version 1.1.0
481 server-side so match that on older versions.
483 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
484 allocation for $value could be released as the expansion processing
485 concluded, but leaving the global pointer active for it.
487 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
488 and to use the domains and local_parts ACL conditions.
490 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
491 incorrectly not doubled on cutthrough transmission, hence seen as a
492 body-termination at the receiving system - resulting in truncated mails.
493 Commonly the sender saw a TCP-level error, and retransmitted the message
494 via the normal store-and-forward channel. This could result in duplicates
495 received - but deduplicating mailstores were liable to retain only the
496 initial truncated version.
498 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
500 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
502 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
503 we're in there, support oversigning also; bug 1309.
505 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
507 HS/04 Add support for keep_environment and add_environment options.
509 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
510 either intentional arithmetic overflow during PRNG, or testing config-
513 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
514 delivery resulted in actual delivery. Cancel cutthrough before DATA
517 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
518 rejects a recipient: pass the reject to the originator.
520 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
521 Many were false-positives and ignorable, but it's worth fixing the
524 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
525 for the new environment-manipulation done at startup. Move the routines
526 from being local to tls.c to being global via the os.c file.
528 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
529 an extract embedded as result-arg for a map, the first arg for extract
530 is unavailable so we cannot tell if this is a numbered or keyed
531 extraction. Accept either.
537 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
540 JH/02 The smtp transport option "multi_domain" is now expanded.
542 JH/03 The smtp transport now requests PRDR by default, if the server offers
545 JH/04 Certificate name checking on server certificates, when exim is a client,
546 is now done by default. The transport option tls_verify_cert_hostnames
547 can be used to disable this per-host. The build option
548 EXPERIMENTAL_CERTNAMES is withdrawn.
550 JH/05 The value of the tls_verify_certificates smtp transport and main options
551 default to the word "system" to access the system default CA bundle.
552 For GnuTLS, only version 3.0.20 or later.
554 JH/06 Verification of the server certificate for a TLS connection is now tried
555 (but not required) by default. The verification status is now logged by
556 default, for both outbound TLS and client-certificate supplying inbound
559 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
562 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
563 Status Notification (bounce) messages are now MIME format per RFC 3464.
564 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
565 under the control of the dsn_advertise_hosts option, and routers may
566 have a dsn_lasthop option.
568 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
569 default, modifiable by a malware= option. The list separator for
570 the options can now be changed in the usual way. Bug 68.
572 JH/10 The smtp_receive_timeout main option is now expanded before use.
574 JH/11 The incoming_interface log option now also enables logging of the
575 local interface on delivery outgoing connections.
577 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
578 if the interface and destination host and port all match.
580 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
583 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
584 Patch from Andrew Lewis.
586 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
587 now supports optional time-restrictions, weighting, and priority
588 modifiers per server. Patch originally by <rommer@active.by>.
590 JH/16 The spamd_address main option now supports a mixed list of local
591 and remote servers. Remote servers can be IPv6 addresses, and
592 specify a port-range.
594 JH/17 Bug 68: The spamd_address main option now supports an optional
595 timeout value per server.
597 JH/18 Bug 1581: Router and transport options headers_add/remove can
598 now have the list separator specified.
600 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
603 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
606 JH/21 Support for the A6 type of dns record is withdrawn.
608 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
609 rather than the verbs used.
611 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
612 from 255 to 1024 chars.
614 JH/24 Verification callouts now attempt to use TLS by default.
616 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
617 are generic router options now. The defaults didn't change.
619 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
620 Original patch from Alexander Shikoff, worked over by JH.
622 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
623 files if ZCAT_COMMAND is not executable.
625 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
627 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
629 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
630 Normally benign, it bites when the pair was led to by a CNAME;
631 modern usage is to not canonicalize the domain to a CNAME target
632 (and we were inconsistent anyway for A-only vs AAAA+A).
634 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
636 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
637 when evaluating $sender_host_dnssec.
639 JH/31 Check the HELO verification lookup for DNSSEC, adding new
640 $sender_helo_dnssec variable.
642 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
644 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
646 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
648 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
649 documented as working, but never had. Support all but $spam_report.
651 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
652 added for tls authenticator.
654 HS/03 Add perl_taintmode main config option
660 TL/01 When running the test suite, the README says that variables such as
661 no_msglog_check are global and can be placed anywhere in a specific
662 test's script, however it was observed that placement needed to be near
663 the beginning for it to behave that way. Changed the runtest perl
664 script to read through the entire script once to detect and set these
665 variables, reset to the beginning of the script, and then run through
666 the script parsing/test process like normal.
668 TL/02 The BSD's have an arc4random API. One of the functions to induce
669 adding randomness was arc4random_stir(), but it has been removed in
670 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
671 function when detected.
673 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
674 cause callback expansion.
676 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
677 syntax errors in an expansion can be treated as a string instead of
678 logging or causing an error, due to the internal use of bool_lax
679 instead of bool when processing it.
681 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
682 server certificates when making smtp deliveries.
684 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
686 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
688 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
690 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
691 Merged patch from Sebastian Wiedenroth.
693 JH/05 Fix results-pipe from transport process. Several recipients, combined
694 with certificate use, exposed issues where response data items split
695 over buffer boundaries were not parsed properly. This eventually
696 resulted in duplicates being sent. This issue only became common enough
697 to notice due to the introduction of connection certificate information,
698 the item size being so much larger. Found and fixed by Wolfgang Breyha.
700 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
701 size buffer was used, resulting in syntax errors when an expansion
704 JH/07 Add support for directories of certificates when compiled with a GnuTLS
705 version 3.3.6 or later.
707 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
708 is EXPERIMENTAL_EVENT, the main-configuration and transport options
709 both become "event_action", the variables become $event_name, $event_data
710 and $event_defer_errno. There is a new variable $verify_mode, usable in
711 routers, transports and related events. The tls:cert event is now also
712 raised for inbound connections, if the main configuration event_action
715 TL/06 In test suite, disable OCSP for old versions of openssl which contained
716 early OCSP support, but no stapling (appears to be less than 1.0.0).
718 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
719 server certificate names available under the smtp transport option
720 "tls_verify_cert_hostname" now do not permit multi-component wildcard
723 JH/10 Time-related extraction expansions from certificates now use the main
724 option "timezone" setting for output formatting, and are consistent
725 between OpenSSL and GnuTLS compilations. Bug 1541.
727 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
728 encoded parameter in the incoming message. Bug 1558.
730 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
731 include certificate info, eximon was claiming there were spoolfile
734 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
736 JH/14 Log delivery-related information more consistently, using the sequence
737 "H=<name> [<ip>]" wherever possible.
739 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
740 are problematic for Debian distribution, omit them from the release
743 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
745 JH/16 Fix string representation of time values on 64bit time_t architectures.
748 JH/17 Fix a null-indirection in certextract expansions when a nondefault
749 output list separator was used.
754 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
755 checkers that were complaining about end of non-void function with no
758 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
759 This was a regression introduced in 4.83 by another bugfix.
761 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
763 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
764 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
770 TF/01 Correctly close the server side of TLS when forking for delivery.
772 When a message was received over SMTP with TLS, Exim failed to clear up
773 the incoming connection properly after forking off the child process to
774 deliver the message. In some situations the subsequent outgoing
775 delivery connection happened to have the same fd number as the incoming
776 connection previously had. Exim would try to use TLS and fail, logging
777 a "Bad file descriptor" error.
779 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
780 utilities have not been installed.
782 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
783 temporary space as the ACL may create new global variables.
785 TL/01 LDAP support uses per connection or global context settings, depending
786 upon the detected version of the libraries at build time.
788 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
789 to extract and use the src ip:port in logging and expansions as if it
790 were a direct connection from the outside internet. PPv2 support was
791 updated based on HAProxy spec change in May 2014.
793 JH/02 Add ${listextract {number}{list}{success}{fail}}.
795 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
796 Properly escape header and check for NULL return.
798 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
801 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
803 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
804 characters in header names, implemented as a verify condition.
805 Contributed by Michael Fischer v. Mollard.
807 TL/05 Rename SPF condition results err_perm and err_temp to standardized
808 results permerror and temperror. Previous values are deprecated but
809 still accepted. In a future release, err_perm and err_temp will be
810 completely removed, which will be a backward incompatibility if the
811 ACL tests for either of these two old results. Patch contributed by
812 user bes-internal on the mailing list.
814 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
816 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
817 selectors, in both main and reject logs.
819 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
822 JH/07 Add malware type "sock" for talking to simple daemon.
824 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
826 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
827 routers/transports under cutthrough routing.
829 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
830 numbers. Touch up "bool" conditional to keep the same definition.
832 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
834 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
836 JH/12 Expand items in router/transport headers_add or headers_remove lists
837 individually rather than the list as a whole. Bug 1452.
839 Required for reasonable handling of multiple headers_ options when
840 they may be empty; requires that headers_remove items with embedded
841 colons must have them doubled (or the list-separator changed).
843 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
844 view the policy declared in the DMARC record. Currently, $dmarc_status
845 is a combined value of both the record presence and the result of the
848 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
850 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
851 dnslookup router and the smtp transport (applying to the forward
854 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
855 of ldap servers used for a specific lookup. Patch provided by Heiko
858 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
859 New variable $lookup_dnssec_authenticated for observability.
861 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
862 Patch submitted by Lars Timman.
864 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
866 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
867 Requires trusted mode and valid format message id, aborts otherwise.
868 Patch contributed by Heiko Schlichting.
870 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
871 certextract with support for various fields. Bug 1358.
873 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
874 is requested by default, modifiable by smtp transport option
877 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
878 operate on certificate variables to give certificate fingerprints
879 Also new ${sha256:cert_variable}.
881 JH/23 The PRDR feature is moved from being Experimental into the mainline.
883 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
884 Christian Aistleitner.
886 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
888 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
889 file. Patch from Wolfgang Breyha.
891 JH/25 Expand the coverage of the delivery $host and $host_address to
892 client authenticators run in verify callout. Bug 1476.
894 JH/26 Port service names are now accepted for tls_on_connect_ports, to
895 align with daemon_smtp_ports. Bug 72.
897 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
898 support and error reporting did not work properly.
900 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
901 and is readable. Patch from Andrew Colin Kissa.
903 TL/14 Enhance documentation of ${run expansion and how it parses the
904 commandline after expansion, particularly in the case when an
905 unquoted variable expansion results in an empty value.
907 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
909 PP/02 Fix internal collision of T_APL on systems which support RFC3123
910 by renaming away from it. Addresses GH issue 15, reported by
913 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
915 TL/15 SECURITY: prevent double expansion in math comparison functions
916 (can expand unsanitized data). Not remotely exploitable.
923 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
925 PP/02 Make -n do something, by making it not do something.
926 When combined with -bP, the name of an option is not output.
928 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
931 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
932 $sender_host_name and config options to manage this, and basic check
935 PP/05 DSCP support for outbound connections and control modifier for inbound.
937 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
938 (Only plugin which currently uses this is kerberos4, which nobody should
939 be using, but we should make it available and other future plugins might
940 conceivably use it, even though it would break NAT; stuff *should* be
941 using channel bindings instead).
943 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
944 name; added for Sendmail compatibility; requires admin caller.
945 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
946 just ignore it); requires trusted caller.
947 Also parse but ignore: -Ac -Am -X<logfile>
950 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
952 TL/02 Add +smtp_confirmation as a default logging option.
954 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
955 Patch by Magnus Holmgren from 2007-02-20.
957 TL/04 Bugzilla 1281 - Spec typo.
958 Bugzilla 1283 - Spec typo.
959 Bugzilla 1290 - Spec grammar fixes.
961 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
963 TL/06 Add Experimental DMARC support using libopendmarc libraries.
965 TL/07 Fix an out of order global option causing a segfault. Reported to dev
966 mailing list by by Dmitry Isaikin.
968 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
970 JH/02 Support "G" suffix to numbers in ${if comparisons.
972 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
974 NM/01 Bugzilla 1197 - Spec typo
975 Bugzilla 1196 - Spec examples corrections
977 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
979 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
980 gnutls_enable_pkcs11, but renamed to more accurately indicate its
983 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
984 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
986 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
987 "acl {{name}{arg}...}", and optional args on acl condition
990 JH/05 Permit multiple router/transport headers_add/remove lines.
992 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
994 JH/07 Avoid using a waiting database for a single-message-only transport.
995 Performance patch from Paul Fisher. Bugzilla 1262.
997 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1000 JH/09 Add $headers_added variable, with content from use of ACL modifier
1001 add_header (but not yet added to the message). Bugzilla 199.
1003 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1004 Pulled from Bugzilla 817 by Wolfgang Breyha.
1006 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1008 (nb: this is the same fix as in Exim 4.80.1)
1010 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1013 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1016 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1018 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1019 advertise SMTP AUTH mechanism to us, instead of a generic
1020 protocol violation error. Also, make Exim more robust to bad
1021 data from the Dovecot auth socket.
1023 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1025 When a queue runner is handling a message, Exim first routes the
1026 recipient addresses, during which it prunes them based on the retry
1027 hints database. After that it attempts to deliver the message to
1028 any remaining recipients. It then updates the hints database using
1031 So if a recipient address works intermittently, it can get repeatedly
1032 deferred at routing time. The retry hints record remains fresh so the
1033 address never reaches the final cutoff time.
1035 This is a fairly common occurrence when a user is bumping up against
1036 their storage quota. Exim had some logic in its local delivery code
1037 to deal with this. However it did not apply to per-recipient defers
1038 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1040 This change adds a proper retry rule check during routing so that the
1041 final cutoff time is checked against the message's age. We only do
1042 this check if there is an address retry record and there is not a
1043 domain retry record; this implies that previous attempts to handle
1044 the address had the retry_use_local_parts option turned on. We use
1045 this as an approximation for the destination being like a local
1046 delivery, as in LMTP.
1048 I suspect this new check makes the old local delivery cutoff check
1049 redundant, but I have not verified this so I left the code in place.
1051 TF/02 Correct gecos expansion when From: is a prefix of the username.
1053 Test 0254 submits a message to Exim with the header
1057 When I ran the test suite under the user fanf2, Exim expanded
1058 the header to contain my full name, whereas it should have added
1059 a Resent-Sender: header. It erroneously treats any prefix of the
1060 username as equal to the username.
1062 This change corrects that bug.
1064 GF/01 DCC debug and logging tidyup
1065 Error conditions log to paniclog rather than rejectlog.
1066 Debug lines prefixed by "DCC: " to remove any ambiguity.
1068 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1070 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1071 Bug spotted by Jeremy Harris; was flawed since initial commit.
1072 Would have resulted in OCSP responses post-SNI triggering an Exim
1073 NULL dereference and crash.
1075 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1077 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1078 Bug detection, analysis and fix by Samuel Thibault.
1079 Bugzilla 1331, Debian bug #698092.
1081 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1083 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1084 Server implementation by Todd Lyons, client by JH.
1085 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1086 config variable "prdr_enable" controls whether the server
1087 advertises the facility. If the client requests PRDR a new
1088 acl_data_smtp_prdr ACL is called once for each recipient, after
1089 the body content is received and before the acl_smtp_data ACL.
1090 The client is controlled by both of: a hosts_try_prdr option
1091 on the smtp transport, and the server advertisement.
1092 Default client logging of deliveries and rejections involving
1093 PRDR are flagged with the string "PRDR".
1095 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1096 fclose(). Diagnosis by Todd Lyons.
1098 PP/17 Update configure.default to handle IPv6 localhost better.
1099 Patch by Alain Williams (plus minor tweaks).
1102 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1103 This is now consistent with GnuTLS, and is now documented: the
1104 previous undocumented portable approach to treating the option as
1105 unset was to force an expansion failure. That still works, and
1106 an empty string is now equivalent.
1108 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1109 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1110 not performing validation itself.
1112 PP/20 Added force_command boolean option to pipe transport.
1113 Patch from Nick Koston, of cPanel Inc.
1115 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1118 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1120 PP/21 Fix eximon continuous updating with timestamped log-files.
1121 Broken in a format-string cleanup in 4.80, missed when I repaired the
1122 other false fix of the same issue.
1123 Report and fix from Heiko Schlichting.
1126 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1127 Report from Prashanth Katuri.
1129 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1130 It's SecureTransport, so affects any MacOS clients which use the
1131 system-integrated TLS libraries, including email clients.
1133 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1134 using a MIME ACL for non-SMTP local injection.
1135 Report and assistance in diagnosis by Warren Baker.
1137 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1139 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1141 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1142 last id that failed so it may be referenced in subsequent ACL's.
1144 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1147 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1148 ldap library initialization, allowing self-signed CA's to be
1149 used. Also properly sets require_cert option later in code by
1150 using NULL (global ldap config) instead of ldap handle (per
1151 session). Bug diagnosis and testing by alxgomz.
1153 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1154 the src/util/ subdirectory.
1156 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1157 renamed to Transport Post Delivery Action by Jeremy Harris, as
1160 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1161 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1162 redis_servers = needs to be configured which will be used by the redis
1163 lookup. Patch from Warren Baker, of The Packet Hub.
1165 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1167 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1168 hostname or reverse DNS when processing a host list. Used suggestions
1169 from multiple comments on this bug.
1171 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1173 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1174 Missed a few lines, added it to make the runtest require no keyboard
1177 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1178 contains upper case chars. Make router use caseful_local_part.
1180 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1181 support when GnuTLS has been built with p11-kit.
1187 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1189 This, or similar/improved, will also be change PP/11 of 4.82.
1195 PP/01 Handle short writes when writing local log-files.
1196 In practice, only affects FreeBSD (8 onwards).
1197 Bugzilla 1053, with thanks to Dmitry Isaikin.
1199 NM/01 Bugzilla 949 - Documentation tweak
1201 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1204 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1206 PP/02 Implemented gsasl authenticator.
1208 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1210 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1211 `pkg-config foo` for cflags/libs.
1213 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1214 with rest of GSASL and with heimdal_gssapi.
1216 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1217 `pkg-config foo` for cflags/libs for the TLS implementation.
1219 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1220 properties get this fed in as external SSF. A number of robustness
1221 and debugging improvements to the cyrus_sasl authenticator.
1223 PP/08 cyrus_sasl server now expands the server_realm option.
1225 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1226 Patch by Jeremy Harris.
1228 PP/10 Added dbmjz lookup type.
1230 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1232 PP/12 MAIL args handles TAB as well as SP, for better interop with
1233 non-compliant senders.
1234 Analysis and variant patch by Todd Lyons.
1236 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1237 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1238 Patch from Dirk Mueller <dmueller@suse.com>
1240 PP/13 tls_peerdn now print-escaped for spool files.
1241 Observed some $tls_peerdn in wild which contained \n, which resulted
1242 in spool file corruption.
1244 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1245 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1246 or write after TLS renegotiation, which otherwise led to messages
1249 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1250 as a tracking header (ie: a signed header comes before the signature).
1251 Patch from Wolfgang Breyha.
1253 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1254 comma-sep list; embedded commas doubled.
1256 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1258 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1260 Report and patch from Dmitry Banschikov.
1262 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1263 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1264 failures. We appear to now support TLS1.1+ with Exim.
1266 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1267 lets Exim select keys and certificates based upon TLS SNI from client.
1268 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1269 before an outbound SMTP session. New log_selector, +tls_sni.
1271 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1272 NULL dereference. Report and patch from Alun Jones.
1274 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1275 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1276 Not seeing resolver debug output on NetBSD, but suspect this is a
1277 resolver implementation change.
1279 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1280 Left warnings. Added "eximon gdb" invocation mode.
1282 PP/21 Defaulting "accept_8bitmime" to true, not false.
1284 PP/22 Added -bw for inetd wait mode support.
1286 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1287 locate the relevant includes and libraries. Made this the default.
1289 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1290 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1292 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1293 This may cause build issues on older platforms.
1295 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1296 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1297 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1298 Added SNI support via GnuTLS too.
1299 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1301 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1303 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1304 Applied second patch from Janne, implementing suggestion to default
1305 multiple-strings-in-record handling to match SPF spec.
1307 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1309 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1310 read-only, out of scope).
1311 Patch from Wolfgang Breyha, report from Stuart Northfield.
1313 PP/29 Fix three issues highlighted by clang analyser static analysis.
1314 Only crash-plausible issue would require the Cambridge-specific
1315 iplookup router and a misconfiguration.
1316 Report from Marcin Mirosław.
1318 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1320 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1321 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1322 As part of this, removing so much warning spew let me fix some minor
1323 real issues in debug logging.
1325 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1326 assignment on my part. Fixed.
1328 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1329 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1330 Janne Snabb (who went above and beyond: thank you).
1332 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1333 string otherwise requires a connection and a bunch more work and it's
1334 relatively easy to get wrong. Should also expose TLS library linkage
1337 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1338 64-bit ${eval} (JH/03).
1340 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1341 GNU libc to support some of the 64-bit stuff, should not lead to
1342 conflicts. Defined before os.h is pulled in, so if a given platform
1343 needs to override this, it can.
1345 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1346 protection layer was required, which is not implemented.
1347 Bugzilla 1254, patch from Wolfgang Breyha.
1349 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1350 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1351 tls_dhparam take prime identifiers. Also unbreak combination of
1352 OpenSSL+DH_params+TLSSNI.
1354 PP/39 Disable SSLv2 by default in OpenSSL support.
1360 PP/01 Solaris build fix for Oracle's LDAP libraries.
1361 Bugzilla 1109, patch from Stephen Usher.
1363 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1365 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1368 TF/02 Fix a couple more cases where we did not log the error message
1369 when unlink() failed. See also change 4.74-TF/03.
1371 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1372 lock up or crash if it happened to be inside a call to libc when it
1373 got a SIGUSR1 from exiwhat.
1375 The SIGUSR1 handler appends the current process status to the process
1376 log which is later printed by exiwhat. It used to use the general
1377 purpose logging code to do this, but several functions it calls are
1378 not safe for signals.
1380 The new output code in the SIGUSR1 handler is specific to the process
1381 log, and simple enough that it's easy to inspect for signal safety.
1382 Removing some special cases also simplifies the general logging code.
1383 Removing the spurious timestamps from the process log simplifies
1386 TF/04 Improved ratelimit ACL condition.
1388 The /noupdate option has been deprecated in favour of /readonly which
1389 has clearer semantics. The /leaky, /strict, and /readonly update modes
1390 are mutually exclusive. The update mode is no longer included in the
1391 database key; it just determines when the database is updated. (This
1392 means that when you upgrade Exim will forget old rate measurements.)
1394 Exim now checks that the per_* options are used with an update mode that
1395 makes sense for the current ACL. For example, when Exim is processing a
1396 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1397 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1398 must specify per_mail/readonly. If you omit the update mode it defaults to
1399 /leaky where that makes sense (as before) or /readonly where required.
1401 The /noupdate option is now undocumented but still supported for
1402 backwards compatibility. It is equivalent to /readonly except that in
1403 ACLs where /readonly is required you may specify /leaky/noupdate or
1404 /strict/noupdate which are treated the same as /readonly.
1406 A useful new feature is the /count= option. This is a generalization
1407 of the per_byte option, so that you can measure the throughput of other
1408 aggregate values. For example, the per_byte option is now equivalent
1409 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1411 The per_rcpt option has been generalized using the /count= mechanism
1412 (though it's more complicated than the per_byte equivalence). When it is
1413 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1414 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1415 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1416 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1417 Note that using per_rcpt with a non-readonly update mode in more than
1418 one ACL will cause the recipients to be double-counted. (The per_mail
1419 and per_byte options don't have this problem.)
1421 The handling of very low rates has changed slightly. If the computed rate
1422 is less than the event's count (usually one) then this event is the first
1423 after a long gap. In this case the rate is set to the same as this event's
1424 count, so that the first message of a spam run is counted properly.
1426 The major new feature is a mechanism for counting the rate of unique
1427 events. The new per_addr option counts the number of different
1428 recipients that someone has sent messages to in the last time period. It
1429 behaves like per_rcpt if all the recipient addresses are different, but
1430 duplicate recipient addresses do not increase the measured rate. Like
1431 the /count= option this is a general mechanism, so the per_addr option
1432 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1433 example, measure the rate that a client uses different sender addresses
1434 with the options per_mail/unique=$sender_address. There are further
1435 details in the main documentation.
1437 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1439 TF/06 Removed a few PCRE remnants.
1441 TF/07 Automatically extract Exim's version number from tags in the git
1442 repository when doing development or release builds.
1444 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1445 Bugzilla 879. Patch from Paul Fisher.
1447 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1448 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1451 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1453 PP/05 Variable $av_failed, true if the AV scanner deferred.
1454 Bugzilla 1078. Patch from John Horne.
1456 PP/06 Stop make process more reliably on build failure.
1457 Bugzilla 1087. Patch from Heiko Schlittermann.
1459 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1460 Bugzilla 1089. Patch from Heiko Schlittermann.
1462 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1463 Bugzilla 1131. Patch from Holger Weiß.
1465 PP/09 Handle IPv6 addresses with SPF.
1466 Bugzilla 860. Patch from Wolfgang Breyha.
1468 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1470 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1473 PP/11 match_* no longer expand right-hand-side by default.
1474 New compile-time build option, EXPAND_LISTMATCH_RHS.
1475 New expansion conditions, "inlist", "inlisti".
1477 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1479 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1481 PP/14 fix log_write() format string regression from TF/03.
1482 Bugzilla 1152. Patch from Dmitry Isaikin.
1488 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1490 PP/02 Harmonised TLS library version reporting; only show if debugging.
1491 Layout now matches that introduced for other libraries in 4.74 PP/03.
1493 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1495 PP/04 New "dns_use_edns0" global option.
1497 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1500 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1501 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1503 TK/01 Updated PolarSSL code to 0.14.2.
1504 Bugzilla 1097. Patch from Andreas Metzler.
1506 PP/07 Catch divide-by-zero in ${eval:...}.
1507 Fixes bugzilla 1102.
1509 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1512 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1513 format-string attack -- SECURITY: remote arbitrary code execution.
1515 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1516 time unintentionally subject to list matching rules, letting the header
1517 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1518 arbitrary string expansion). This allowed for information disclosure.
1520 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1521 INT_MIN/-1 -- value coerced to INT_MAX.
1527 NM/01 Workaround for PCRE version dependency in version reporting
1530 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1531 This fixes portability to compilers other than gcc, notably
1532 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1534 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1535 makefiles for portability to HP-UX and POSIX correctness.
1537 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1538 Also via indented variable definition in the Makefile.
1539 (Debugging by Oliver Heesakkers).
1541 PP/02 Restore caching of spamd results with expanded spamd_address.
1542 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1544 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1545 Improves build reliability. Fix from: Frank Elsner
1547 NM/02 Fix wide character breakage in the rfc2047 coding
1548 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1550 NM/03 Allow underscore in dnslist lookups
1551 Fixes bug 1026. Patch from Graeme Fowler
1553 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1554 Code patches from Adam Ciarcinski of NetBSD.
1556 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1559 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1560 is logged, to avoid truncation. Patch from John Horne.
1562 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1563 Patch from Jakob Hirsch.
1565 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1566 SQL string expansion failure details.
1567 Patch from Andrey Oktyabrski.
1569 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1570 Patch from Simon Arlott.
1572 PP/09 New lookups functionality failed to compile on old gcc which rejects
1573 extern declarations in function scope.
1574 Patch from Oliver Fleischmann
1576 PP/10 Use sig_atomic_t for flags set from signal handlers.
1577 Check getgroups() return and improve debugging.
1578 Fixed developed for diagnosis in bug 927 (which turned out to be
1581 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1582 Patch from Mark Zealey.
1584 PP/12 Bugzilla 1056: Improved spamd server selection.
1585 Patch from Mark Zealey.
1587 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1588 Based on patch from Heiko Schlittermann.
1590 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1591 Patch from Uwe Doering, sign-off by Michael Haardt.
1593 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1594 variable declaration deep within a block. Bug and patch from
1597 PP/15 lookups-Makefile IRIX compatibility coercion.
1599 PP/16 Make DISABLE_DKIM build knob functional.
1601 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1602 Patch by Simon Arlott
1604 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1605 variable argument macros. Our copy now differs from upstream.
1611 TF/01 Failure to get a lock on a hints database can have serious
1612 consequences so log it to the panic log.
1614 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1615 controlled using the smtp_confirmation log selector.
1617 TF/03 Include the error message when we fail to unlink a spool file.
1619 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1620 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1621 for maintaining out-of-tree patches for some time.
1623 PP/01 Bugzilla 139: Documentation and portability issues.
1624 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1625 Handle per-OS dynamic-module compilation flags.
1627 PP/02 Let /dev/null have normal permissions.
1628 The 4.73 fixes were a little too stringent and complained about the
1629 permissions on /dev/null. Exempt it from some checks.
1630 Reported by Andreas M. Kirchwitz.
1632 PP/03 Report version information for many libraries, including
1633 Exim version information for dynamically loaded libraries. Created
1634 version.h, now support a version extension string for distributors
1635 who patch heavily. Dynamic module ABI change.
1637 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1638 privilege escalation vulnerability whereby the Exim run-time user
1639 can cause root to append content of the attacker's choosing to
1642 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1645 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1646 If dropping privileges for untrusted macros, we disabled normal logging
1647 on the basis that it would fail; for the Exim run-time user, this is not
1648 the case, and it resulted in successful deliveries going unlogged.
1649 Fixed. Reported by Andreas Metzler.
1655 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1656 only prepend for the Resent-* case. Fixes regression introduced in
1657 Exim 4.70 by NM/22 for Bugzilla 607.
1659 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1660 increasing numbers of administrators be bitten by this.
1662 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1664 PP/03 Bugzilla 994: added openssl_options main configuration option.
1666 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1668 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1670 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1672 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1673 without a peer certificate, leading to a segfault because of an
1674 assumption that peers always have certificates. Be a little more
1675 paranoid. Problem reported by Martin Tscholak.
1677 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1678 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1679 NB: ClamAV planning to remove STREAM in "middle of 2010".
1680 CL also introduces -bmalware, various -d+acl logging additions and
1681 more caution in buffer sizes.
1683 PP/09 Implemented reverse_ip expansion operator.
1685 PP/10 Bugzilla 937: provide a "debug" ACL control.
1687 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1689 PP/12 Bugzilla 973: Implement --version.
1691 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1693 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1695 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1697 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1698 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1699 ignore trailing whitespace.
1701 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1703 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1706 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1707 Notification from Dr Andrew Aitchison.
1709 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1710 ExtendedDetectionInfo response format.
1711 Notification from John Horne.
1713 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1716 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1717 XSL and documented dependency on system catalogs, with examples of how
1720 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1723 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1724 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1725 configuration file which is writeable by the Exim user or group.
1727 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1728 of configuration files to cover files specified with the -C option if
1729 they are going to be used with root privileges, not just the default
1732 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1733 option (effectively making it always true).
1735 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1736 files to be used while preserving root privileges.
1738 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1739 that rogue child processes cannot use them.
1741 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1742 run-time user, instead of root.
1744 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1745 Exim run-time user without dropping privileges.
1747 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1748 result string, instead of calling string_vformat() twice with the same
1751 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1752 for other users. Others should always drop root privileges if they use
1753 -C on the command line, even for a whitelisted configure file.
1755 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1757 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1763 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1764 $data_path, and $header_path variables; fixed documentation bugs and
1767 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1768 exipick to access non-standard spools, including the "frozen" queue
1771 NM/01 Bugzilla 965: Support mysql stored procedures.
1772 Patch from Alain Williams
1774 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1776 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1777 Patch from Andreas Metzler
1779 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1780 Patch from Kirill Miazine
1782 NM/05 Bugzilla 671: Added umask to procmail example.
1784 JJ/03 installed exipick 20100323.0, fixing doc bug
1786 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1787 directory. Notification and patch from Dan Rosenberg.
1789 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1791 TK/02 Improve log output when DKIM signing operation fails.
1793 MH/01 Treat the transport option dkim_domain as a colon separated
1794 list, not as a single string, and sign the message with each element,
1795 omitting multiple occurences of the same signer.
1797 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1798 Bugzilla 985, 986. Patch by Simon Arlott
1800 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1801 Patch by Simon Arlott
1803 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1804 MBX locking. Notification from Dan Rosenberg.
1810 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1812 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1814 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1816 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1818 NM/04 Bugzilla 915: Flag broken perl installation during build.
1824 TK/01 Added patch by Johannes Berg that expands the main option
1825 "spamd_address" if it starts with a dollar sign.
1827 TK/02 Write list of recipients to X-Envelope-Sender header when building
1828 the mbox-format spool file for content scanning (suggested by Jakob
1831 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1832 (http://www.dcc-servers.net/) support via dccifd. Activated by
1833 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1835 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1836 by Mark Daniel Reidel <mr@df.eu>.
1838 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1839 When building exim an external PCRE library is now needed -
1840 PCRE is a system library on the majority of modern systems.
1841 See entry on PCRE_LIBS in EDITME file.
1843 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1844 conversation. Added nologin parameter to request.
1845 Patch contributed by Kirill Miazine.
1847 TF/01 Do not log submission mode rewrites if they do not change the address.
1849 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1851 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1852 log files in place. Contributed by Roberto Lima.
1854 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1856 TF/03 Bugzilla 615: When checking the local_parts router precondition
1857 after a local_part_suffix or local_part_prefix option, Exim now
1858 does not use the address's named list lookup cache, since this
1859 contains cached lookups for the whole local part.
1861 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1862 Robert Millan. Documentation is in experimental-spec.txt.
1864 TF/04 Bugzilla 668: Fix parallel build (make -j).
1866 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1868 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1869 Patch provided by Jan Srzednicki.
1871 TF/05 Leading white space used to be stripped from $spam_report which
1872 wrecked the formatting. Now it is preserved.
1874 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1875 that they are available at delivery time.
1877 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1879 TF/08 TLS error reporting now respects the incoming_interface and
1880 incoming_port log selectors.
1882 TF/09 Produce a more useful error message if an SMTP transport's hosts
1883 setting expands to an empty string.
1885 NM/06 Bugzilla 744: EXPN did not work under TLS.
1886 Patch provided by Phil Pennock.
1888 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1889 Patch provided by Richard Godbee.
1891 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1892 acl_smtp_notquit, added index entry.
1894 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1895 Patch provided by Eugene Bujak.
1897 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1898 accept(). Patch provided by Maxim Dounin.
1900 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1901 Patch provided by Phil Pennock.
1903 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1905 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1906 Patch provided by Brad "anomie" Jorsch.
1908 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1909 Patch provided by Dean Brooks.
1911 TK/05 Add native DKIM support (does not depend on external libraries).
1913 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1914 Patch provided by Graeme Fowler.
1916 NM/16 Bugzilla 851: Documentation example syntax fix.
1918 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1920 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1923 NM/19 Bugzilla 745: TLS version reporting.
1924 Patch provided by Phil Pennock.
1926 NM/20 Bugzilla 167: bool: condition support.
1927 Patch provided by Phil Pennock.
1929 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1930 clients. Patch provided by Phil Pennock.
1932 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1933 Patch provided by Brad "anomie" Jorsch.
1935 NM/23 Bugzilla 687: Fix misparses in eximstats.
1936 Patch provided by Heiko Schlittermann.
1938 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1939 Patch provided by Heiko Schlittermann.
1941 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1942 plus update to original patch.
1944 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1946 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1947 Patch provided by David Brownlee.
1949 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1951 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1953 NM/30 Bugzilla 888: TLS documentation bugfixes.
1955 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1957 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1958 Unlike the original bugzilla I have changed all shell scripts in src tree.
1960 NM/33 Bugzilla 898: Transport filter timeout fix.
1961 Patch by Todd Rinaldo.
1963 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1964 Patch by Serge Demonchaux.
1966 NM/35 Bugzilla 39: Base64 decode bug fixes.
1967 Patch by Jakob Hirsch.
1969 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1971 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1973 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1975 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1981 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1982 ALT-N's libdkim that I have put here:
1983 http://duncanthrax.net/exim-experimental/
1985 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1986 were called 'true' and it seems that C99 defines that as a reserved
1987 keyword to be used with 'bool' variable types. That means you could
1988 not include C99-style headers which use bools without triggering
1989 build errors in sieve.c.
1991 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1992 as mailq or other aliases. Changed the --help handling significantly
1993 to do whats expected. exim_usage() emits usage/help information.
1995 SC/01 Added the -bylocaldomain option to eximstats.
1997 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1999 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2001 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2007 PH/01 Another patch from the Sieve maintainer.
2009 PH/02 When an IPv6 address is converted to a string for single-key lookup
2010 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2011 dots are used instead of colons so that keys in lsearch files need not
2012 contain colons. This was done some time before quoting was made available
2013 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2014 (notated using the quote facility) so as to distinguish them from IPv4
2015 keys. This meant that lookups for IP addresses in host lists did not work
2016 for iplsearch lookups.
2018 This has been fixed by arranging for IPv6 addresses to be expressed with
2019 colons if the lookup type is iplsearch. This is not incompatible, because
2020 previously such lookups could never work.
2022 The situation is now rather anomalous, since one *can* have colons in
2023 ordinary lsearch keys. However, making the change in all cases is
2024 incompatible and would probably break a number of configurations.
2026 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2029 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2030 conversion specification without a maximum field width, thereby enabling
2031 a rogue spamd server to cause a buffer overflow. While nobody in their
2032 right mind would setup Exim to query an untrusted spamd server, an
2033 attacker that gains access to a server running spamd could potentially
2034 exploit this vulnerability to run arbitrary code as the Exim user.
2036 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2037 $primary_hostname instead of what libspf2 thinks the hosts name is.
2039 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2040 a directory entry by the name of the lookup key. Previously, if a
2041 symlink pointed to a non-existing file or a file in a directory that
2042 Exim lacked permissions to read, a lookup for a key matching that
2043 symlink would fail. Now it is enough that a matching directory entry
2044 exists, symlink or not. (Bugzilla 503.)
2046 PH/03 The body_linecount and body_zerocount variables are now exported in the
2049 PH/04 Added the $dnslist_matched variable.
2051 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2052 This means they are set thereafter only if the connection becomes
2055 PH/06 Added the client_condition to authenticators so that some can be skipped
2056 by clients under certain conditions.
2058 PH/07 The error message for a badly-placed control=no_multiline_responses left
2059 "_responses" off the end of the name.
2061 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2063 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2064 (without spaces) instead of just copying the configuration text.
2066 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2068 PH/11 Added $max_received_linelength.
2070 PH/12 Added +ignore_defer and +include_defer to host lists.
2072 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2073 way in which PCRE > 7.0 is built.
2075 PH/14 Implemented queue_only_load_latch.
2077 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2078 MAIL command. The effect was to mangle the value on 64-bit systems.
2080 PH/16 Another patch from the Sieve maintainer.
2082 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2084 PH/18 If a system quota error occurred while trying to create the file for
2085 a maildir delivery, the message "Mailbox is full" was not appended to the
2086 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2087 only to a quota excession during the actual writing of the file.
2089 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2090 characters?) which causes problems in log lines. The DN values are now
2091 passed through string_printing() before being added to log lines.
2093 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2094 and InterBase are left for another time.)
2096 PH/21 Added message_body_newlines option.
2098 PH/22 Guard against possible overflow in moan_check_errorcopy().
2100 PH/23 POSIX allows open() to be a macro; guard against that.
2102 PH/24 If the recipient of an error message contained an @ in the local part
2103 (suitably quoted, of course), incorrect values were put in $domain and
2104 $local_part during the evaluation of errors_copy.
2110 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2111 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2114 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2115 issue a MAIL command.
2117 PH/02 In an ACL statement such as
2119 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2121 if a client was not listed at all, or was listed with a value other than
2122 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2123 the condition was not true (as it should be), so access was not denied.
2124 The bug was that the ! inversion was incorrectly passed on to the second
2125 item. This has been fixed.
2127 PH/03 Added additional dnslists conditions == and =& which are different from
2128 = and & when the dns lookup returns more than one IP address.
2130 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2131 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2133 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2134 FSYNC, which compiles an option called disable_fsync that allows for
2135 bypassing fsync(). The documentation is heavily laced with warnings.
2137 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2139 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2140 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2141 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2142 including adding "make clean"; (3) Added -fPIC when compiling the test
2143 dynamically loaded module, to get rid of a warning.
2145 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2146 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2147 = 0s. The bug is otherwise harmless.
2149 PH/07 There was a bug in the dovecot authenticator such that the value of
2150 $auth1 could be overwritten, and so not correctly preserved, after a
2151 successful authentication. This usually meant that the value preserved by
2152 the server_setid option was incorrect.
2154 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2156 PH/09 Installed PCRE release 7.0.
2158 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2159 run for batched SMTP input. It is now run at the start of every message
2160 in the batch. While fixing this I discovered that the process information
2161 (output by running exiwhat) was not always getting set for -bs and -bS
2162 input. This is fixed, and it now also says "batched" for BSMTP.
2164 PH/11 Added control=no_pipelining.
2166 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2167 patch, slightly modified), and move the expansion of helo_data till after
2168 the connection is made in the smtp transport (so it can use these
2171 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2173 PH/14 Added log_selector = +pid.
2175 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2177 PH/16 Add ${if forany and ${if forall.
2179 PH/17 Added dsn_from option to vary the From: line in DSNs.
2181 PH/18 Flush SMTP output before performing a callout, unless control =
2182 no_callout_flush is set.
2184 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2185 was true (the default) a successful delivery failed to delete the retry
2186 item, thus causing premature timeout of the address. The bug is now
2189 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2191 PH/21 Long custom messages for fakedefer and fakereject are now split up
2192 into multiline responses in the same way that messages for "deny" and
2193 other ACL rejections are.
2195 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2196 with slight modification.
2198 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2199 draft, changing the syntax and factoring some duplicate code".
2201 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2202 for deliveries of the second and subsequent messages over the same SMTP
2205 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2206 ${reduce, with only minor "tidies".
2208 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2210 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2211 expansion side effects.
2213 PH/27 When a message times out after an over-quota error from an Exim-imposed
2214 quota, the bounce message says "mailbox is full". This message was not
2215 being given when it was a system quota that was exceeded. It now should
2218 MH/03 Made $recipients available in local_scan(). local_scan() already has
2219 better access to the recipient list through recipients_list[], but
2220 $recipients can be useful in postmaster-provided expansion strings.
2222 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2223 in the case of a MAIL command with additional options following the
2224 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2225 were accidentally chopped off.
2227 PH/29 SMTP synchronization checks are implemented when a command is read -
2228 there is a check that no more input is waiting when there shouldn't be
2229 any. However, for some commands, a delay in an ACL can mean that it is
2230 some time before the response is written. In this time, more input might
2231 arrive, invalidly. So now there are extra checks after an ACL has run for
2232 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2233 pipelining has not been advertised.
2235 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2237 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2238 returned in the SMTP response when smtp_return_error_details was false.
2239 This has been fixed.
2241 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2242 instead of the C I/O that was originally supplied, because problems were
2243 reported on Solaris.
2245 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2246 Exim which did not show up earlier: it was assuming that a call to
2247 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2248 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2249 was a macro that became an assignment, so it seemed to work. This has
2250 changed to a proper function call with a void return, hence the compile
2251 error. Exim's code has been fixed.
2253 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2256 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2258 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2260 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2261 because it turns out to be too big or there is a timeout) there is no
2262 "Completed" line in the log. When some messages of this type were
2263 selected by exigrep, they were listed as "not completed". Others were
2264 picked up by some special patterns. I have improved the selection
2265 criteria to be more general.
2267 PH/38 The host_find_failed option in the manualroute router can now be set
2268 to "ignore", to completely ignore a host whose IP address cannot be
2269 found. If all hosts are ignored, the behaviour is controlled by the new
2270 host_all_ignored option.
2272 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2273 homing or because of multiple MX records with /mx) generated more than
2274 one IP address, and the following item turned out to be the local host,
2275 all the secondary addresses of the first item were incorrectly removed
2276 from the list, along with the local host and any following hosts (which
2277 is what is supposed to happen).
2279 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2280 whoever called Exim into the -H file. In the case of the daemon it was
2281 behaving confusingly. When first started, it used values for whoever
2282 started the daemon, but after a SIGHUP it used the Exim user (because it
2283 calls itself on a restart). I have changed the code so that it now always
2286 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2287 message are rejected with the same error (e.g. no authentication or bad
2288 sender address), and a DATA command is nevertheless sent (as can happen
2289 with PIPELINING or a stupid MUA), the error message that was given to the
2290 RCPT commands is included in the rejection of the DATA command. This is
2291 intended to be helpful for MUAs that show only the final error to their
2294 PH/42 Another patch from the Sieve maintainer.
2296 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2297 Eximstats - Fixed some broken HTML links and added missing column headers
2299 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2300 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2302 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2304 PH/43 Yet another patch from the Sieve maintainer.
2306 PH/44 I found a way to check for a TCP/IP connection going away before sending
2307 the response to the final '.' that terminates a message, but only in the
2308 case where the client has not sent further data following the '.'
2309 (unfortunately, this is allowed). However, in many cases there won't be
2310 any further data because there won't be any more messages to send. A call
2311 to select() can be used: if it shows that the input is "ready", there is
2312 either input waiting, or the socket has been closed. An attempt to read
2313 the next input character can distinguish the two cases. Previously, Exim
2314 would have sent an OK response which the client would never have see.
2315 This could lead to message repetition. This fix should cure that, at
2316 least in a lot of common cases.
2318 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2319 advertised in response to EHLO.
2325 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2326 fixed by 4.65/MH/01 (is this a record?) are fixed:
2328 (i) An empty string was always treated as zero by the numeric comparison
2329 operators. This behaviour has been restored.
2331 (ii) It is documented that the numeric comparison operators always treat
2332 their arguments as decimal numbers. This was broken in that numbers
2333 starting with 0 were being interpreted as octal.
2335 While fixing these problems I realized that there was another issue that
2336 hadn't been noticed. Values of message_size_limit (both the global option
2337 and the transport option) were treated as octal if they started with 0.
2338 The documentation was vague. These values are now always treated as
2339 decimal, and I will make that clear in the documentation.
2345 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2346 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2349 MH/01 Don't check that the operands of numeric comparison operators are
2350 integers when their expansion is in "skipping" mode (fixes bug
2351 introduced by 4.64-PH/07).
2353 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2354 child addresses, Exim now panics and dies. Previously, because the count
2355 is held in a short int, deliveries were likely to be lost. As such a
2356 large number of recipients for a single message is ridiculous
2357 (performance will be very, very poor), I have chosen to impose a limit
2358 rather than extend the field.
2364 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2365 leftover -K file (the existence of which was triggered by #402).
2366 While we were at it, introduced process PID as part of the -K
2367 filename. This should rule out race conditions when creating
2370 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2371 processing considerably. Previous code took too long for large mails,
2372 triggering a timeout which in turn triggers #401.
2374 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2375 in the DK code in transports.c. sendfile() is not really portable,
2376 hence the _LINUX specificness.
2378 TF/01 In the add_headers option to the mail command in an Exim filter,
2379 there was a bug that Exim would claim a syntax error in any
2380 header after the first one which had an odd number of characters
2383 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2384 callout verification, Exim cached a "reject" for the entire domain. This
2385 is correct for most verifications, but it is not correct for a recipient
2386 verification with use_sender or use_postmaster set, because in that case
2387 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2388 case of MAIL FROM:<> rejection from other early rejections (e.g.
2389 rejection of HELO). When verifying a recipient using a non-null MAIL
2390 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2391 Whatever the result of the callout, the value of the domain cache is
2392 left unchanged (for any other kind of callout, getting as far as trying
2393 RCPT means that the domain itself is ok).
2395 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2398 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2399 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2400 session, the daemon ignores ECONNECT errors and logs others; it now
2401 ignores EPIPE as well.
2403 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2404 (quoted-printable decoding).
2406 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2407 later the small subsequent patch to fix an introduced bug.
2409 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2411 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2413 PH/08 An error is now given if message_size_limit is specified negative.
2415 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2416 to be given (somewhat) arbitrary names.
2418 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2421 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2422 miscellaneous code fixes
2424 PH/10 Added the log_reject_target ACL modifier to specify where to log
2427 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2428 hostname. This is wrong, because it relates to the incoming message (and
2429 probably the interface on which it is arriving) and not to the outgoing
2430 callout (which could be using a different interface). This has been
2431 changed to use the value of the helo_data option from the smtp transport
2432 instead - this is what is used when a message is actually being sent. If
2433 there is no remote transport (possible with a router that sets up host
2434 addresses), $smtp_active_hostname is used.
2436 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2437 tweaks were necessary in order to get it to work (see also 21 below):
2438 (a) The code assumed that strncpy() returns a negative number on buffer
2439 overflow, which isn't the case. Replaced with Exim's string_format()
2441 (b) There were several signed/unsigned issues. I just did the minimum
2442 hacking in of casts. There is scope for a larger refactoring.
2443 (c) The code used strcasecmp() which is not a standard C function.
2444 Replaced with Exim's strcmpic() function.
2445 (d) The code set only $1; it now sets $auth1 as well.
2446 (e) A simple test gave the error "authentication client didn't specify
2447 service in request". It would seem that Dovecot has changed its
2448 interface. Fortunately there's a specification; I followed it and
2449 changed what the client sends and it appears to be working now.
2451 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2454 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2455 address A is aliased to B and C, where B exists and C does not. Without
2456 -v the output is "A verified" because verification stops after a
2457 successful redirection if more than one address is generated. However,
2458 with -v the child addresses are also verified. Exim was outputting "A
2459 failed to verify" and then showing the successful verification for C,
2460 with its parentage. It now outputs "B failed to verify", showing B's
2461 parentage before showing the successful verification of C.
2463 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2464 look up a TXT record in a specific list after matching in a combined
2467 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2468 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2469 they consult the DNS. I had assumed they would set it the way they
2470 wanted; and indeed my experiments on Linux seem to show that in some
2471 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2472 To be on the safe side, however, I have now made the interface to
2473 host_find_byname() similar to host_find_bydns(), with an argument
2474 containing the DNS resolver options. The host_find_byname() function now
2475 sets these options at its start, just as host_find_bydns() does. The smtp
2476 transport options dns_qualify_single and dns_search_parents are passed to
2477 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2478 of host_find_byname() use the default settings of RES_DEFNAMES
2479 (qualify_single) but not RES_DNSRCH (search_parents).
2481 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2482 spool_read_header() do less string testing, by means of a preliminary
2483 switch on the second character of optional "-foo" lines. (This is
2484 overdue, caused by the large number of possibilities that now exist.
2485 Originally there were few.) While I was there, I also converted the
2486 str(n)cmp tests so they don't re-test the leading "-" and the first
2487 character, in the hope this might squeeze out yet more improvement.
2489 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2490 flag allowing group syntax was set by the header_syntax check but not
2491 turned off, possible causing trouble later; (2) The flag was not being
2492 set at all for the header_verify test, causing "group"-style headers to
2493 be rejected. I have now set it in this case, and also caused header_
2494 verify to ignore an empty address taken from a group. While doing this, I
2495 came across some other cases where the code for allowing group syntax
2496 while scanning a header line wasn't quite right (mostly, not resetting
2497 the flag correctly in the right place). These bugs could have caused
2498 trouble for malformed header lines. I hope it is now all correct.
2500 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2501 with the "reply" argument non-NULL. The code, however (which originally
2502 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2503 but it didn't always do it. This confused somebody who was copying the
2504 code for some other use. I have removed all the tests.
2506 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2507 feature that was used to support insecure browsers during the U.S. crypto
2508 embargo. It requires special client support, and Exim is probably the
2509 only MTA that supported it -- and would never use it because real RSA is
2510 always available. This code has been removed, because it had the bad
2511 effect of slowing Exim down by computing (never used) parameters for the
2512 RSA_EXPORT functionality.
2514 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2515 authenticator to fail if there's a tab character in the incoming data
2516 (there should never be unless someone is messing about, as it's supposed
2517 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2518 if the connection is using TLS or if the remote IP is the same as the
2519 local IP, and the "valid-client-cert option" if a client certificate has
2522 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2523 authenticators. This can be used for authorization after authentication
2524 succeeds. (In the case of plaintext, it servers for both authentication
2527 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2528 if any retry times were supplied.
2530 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2531 connection, where there is no client IP address to check. In this
2532 situation, the verify now always succeeds.
2534 PH/25 Applied John Jetmore's -Mset patch.
2536 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2538 PH/27 In a string expansion for a processed (not raw) header when multiple
2539 headers of the same name were present, leading whitespace was being
2540 removed from all of them, but trailing whitespace was being removed only
2541 from the last one. Now trailing whitespace is removed from each header
2542 before concatenation. Completely empty headers in a concatenation (as
2543 before) are ignored.
2545 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2546 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2548 PH/29 [Removed. This was a change that I later backed out, and forgot to
2549 correct the ChangeLog entry (that I had efficiently created) before
2550 committing the later change.]
2552 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2553 address errors (4xx response to RCPT) over the same connection as other
2554 messages routed to the same hosts. Such deliveries are always "forced",
2555 so retry times are not inspected. This resulted in far too many retries
2556 for the affected addresses. The effect occurred only when there were more
2557 hosts than the hosts_max_try setting in the smtp transport when it had
2558 the 4xx errors. Those hosts that it had tried were not added to the list
2559 of hosts for which the message was waiting, so if all were tried, there
2560 was no problem. Two fixes have been applied:
2562 (i) If there are any address or message errors in an SMTP delivery, none
2563 of the hosts (tried or untried) are now added to the list of hosts
2564 for which the message is waiting, so the message should not be a
2565 candidate for sending over the same connection that was used for a
2566 successful delivery of some other message. This seems entirely
2567 reasonable: after all the message is NOT "waiting for some host".
2568 This is so "obvious" that I'm not sure why it wasn't done
2569 previously. Hope I haven't missed anything, but it can't do any
2570 harm, as the worst effect is to miss an optimization.
2572 (ii) If, despite (i), such a delivery is accidentally attempted, the
2573 routing retry time is respected, so at least it doesn't keep
2574 hammering the server.
2576 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2577 in ${readsocket because some servers need this prod.
2579 PH/32 Added some extra debug output when updating a wait-xxx database.
2581 PH/33 The hint "could be header name not terminated by colon", which has been
2582 given for certain expansion errors for a long time, was not being given
2583 for the ${if def:h_colon_omitted{... case.
2585 PH/34 The spec says: "With one important exception, whenever a domain list is
2586 being scanned, $domain contains the subject domain." There was at least
2587 one case where this was not true.
2589 PH/35 The error "getsockname() failed: connection reset by peer" was being
2590 written to the panic log as well as the main log, but it isn't really
2591 panic-worthy as it just means the connection died rather early on. I have
2592 removed the panic log writing for the ECONNRESET error when getsockname()
2595 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2596 runs only) independently of the message's sender address. This meant
2597 that, if the 4xx error was in fact related to the sender, a different
2598 message to the same recipient with a different sender could confuse
2599 things. In particular, this can happen when sending to a greylisting
2600 server, but other circumstances could also provoke similar problems.
2601 I have changed the default so that the retry time for these errors is now
2602 based a combination of the sender and recipient addresses. This change
2603 can be overridden by setting address_retry_include_sender=false in the
2606 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2607 remote server are returned as part of bounce messages. This was not
2608 happening for LMTP over a pipe (the lmtp transport), but now it is the
2609 same for both kinds of LMTP.
2611 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2612 in header lines that were in fact CNAMEs. This is no longer the case.
2614 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2615 and queue runs started by the daemon processed all messages. This has
2616 been fixed so that -R and -S can now usefully be given with -q<time>.
2618 PH/40 Import PCRE release 6.7 (fixes some bugs).
2620 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2622 PH/42 Give an error if -q is specified more than once.
2624 PH/43 Renamed the variables $interface_address and $interface_port as
2625 $received_ip_address and $received_port, to make it clear that these
2626 values apply to message reception, and not to the outgoing interface when
2627 a message is delivered. (The old names remain recognized, of course.)
2629 PH/44 There was no timeout on the connect() call when using a Unix domain
2630 socket in the ${readsocket expansion. There now is.
2632 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2633 be meaningful with "accept".
2635 SC/01 Eximstats V1.43
2636 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2638 SC/02 Eximstats V1.44
2639 Use a glob alias rather than an array ref in the generated
2640 parser. This improves both readability and performance.
2642 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2643 Collect SpamAssassin and rejection statistics.
2644 Don't display local sender or destination tables unless
2645 there is data to show.
2646 Added average volumes into the top table text output.
2648 SC/04 Eximstats V1.46
2649 Collect data on the number of addresses (recipients)
2650 as well as the number of messages.
2652 SC/05 Eximstats V1.47
2653 Added 'Message too big' to the list of mail rejection
2654 reasons (thanks to Marco Gaiarin).
2656 SC/06 Eximstats V1.48
2657 Mainlog lines which have GMT offsets and are too short to
2658 have a flag are now skipped.
2660 SC/07 Eximstats V1.49 (Alain Williams)
2661 Added the -emptyok flag.
2663 SC/08 Eximstats V1.50
2664 Fixes for obtaining the IP address from reject messages.
2666 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2667 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2668 whitespace changes from 4.64-PH/27
2670 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2673 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2674 are found, allow negative numbers in numeric criteria)
2676 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2678 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2679 to match changes made in 4.64-PH/43
2681 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2683 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2684 "domain" is actually an IP address, and force a failure. This locks out
2685 those revolvers/nameservers that support "A-for-A" lookups, in
2686 contravention of the specifications.
2688 PH/48 When a host name was looked up from an IP address, and the subsequent
2689 forward lookup of the name timed out, the host name was left in
2690 $sender_host_name, contrary to the specification.
2692 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2693 restricted to single-key lookups, Exim was not diagnosing an error if
2694 * or *@ was used with a query-style lookup.
2696 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2698 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2699 long ago, but noone interested enough thought of it. Let's just say that
2700 the "1.1" means that there are some new functions that weren't there at
2701 some point in the past.
2703 PH/51 Error processing for expansion failure of helo_data from an smtp
2704 transport during callout processing was broken.
2706 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2707 tested/used via the -bh/-bhc/-bs options.
2709 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2710 bug, fixed in subsequent PCRE releases).
2712 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2713 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2715 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2721 SC/01 Use a glob alias rather than an array ref in eximstats generated
2722 parser. This improves both readability and performance.
2724 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2725 Don't display local sender or destination tables in eximstats unless
2726 there is data to show.
2727 Added average volumes into the eximstats top table text output.
2729 SC/03 Collect data on the number of addresses (recipients) as well
2730 as the number of messages in eximstats.
2732 TF/01 Correct an error in the documentation for the redirect router. Exim
2733 does (usually) call initgroups() when daemonizing.
2735 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2736 with consistent privilege compared to when running as a daemon.
2738 TF/03 Note in the spec that $authenticated_id is not set for local
2739 submissions from trusted users.
2741 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2742 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2744 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2745 by adding some example configuration directives to the default
2746 configuration file. A little bit of work is required to uncomment the
2747 directives and define how usernames and passwords are checked, but
2748 there is now a framework to start from.
2750 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2751 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2752 without this. I don't know how relevant this is to other LDAP libraries.
2754 PH/02 Add the verb name to the "unknown ACL verb" error.
2756 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2758 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2760 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2761 directory not expanded when it should be if an expanded home directory
2762 was set for the address (which is overridden by the transport).
2764 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2767 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2768 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2769 because it is too late at that time, and has no effect.
2771 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2772 security issue with \' (bugzilla #107). I could not use the
2773 PQescapeStringConn() function, because it needs a PGconn value as one of
2776 PH/08 When testing addresses using -bt, indicate those final addresses that
2777 are duplicates that would not cause an additional delivery. At least one
2778 person was confused, thinking that -bt output corresponded to deliveries.
2779 (Suppressing duplicates isn't a good idea as you lose the information
2780 about possibly different redirections that led to the duplicates.)
2782 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2783 systems where poll() doesn't work, in particular OS X.
2785 PH/10 Added more information to debugging output for retry time not reached.
2787 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2788 operations in malware.c.
2790 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2793 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2794 syslog_duplication set false, log lines that would normally be written
2795 both the the main log and to the reject log were not written to syslog at
2798 PH/14 In the default configuration, change the use of "message" in ACL warn
2799 statements to "add_header".
2801 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2802 not followed by a command (e.g. "seen endif").
2804 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2805 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2808 PH/17 Added extra conditions to the default value of delay_warning_condition
2812 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2813 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2814 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2817 The Auto-Submitted: and various List- headers are standardised, whereas I
2818 don't think Precedence: ever was.
2820 PH/18 Refactored debugging code in route_finduser() to show more information,
2821 in particular, the error code if getpwnam() issues one.
2823 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2824 This is apparently needed in addition to the PH/07 change above to avoid
2825 any possible encoding problems.
2827 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2828 but not after initializing Perl.
2830 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2831 output them only if debugging. By default they are written stderr,
2832 apparently, which is not desirable.
2834 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2837 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2840 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2842 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2843 authenticated or an ident call has been made. Suppress the default
2844 values for $authenticated_id and $authenticated_sender (but permit -oMai
2845 and -oMas) when testing with -bh.
2847 PH/24 Re-jigged the order of the tests in the default configuration so that the
2848 tests for valid domains and recipients precede the DNS black list and CSA
2849 tests, on the grounds that those ones are more expensive.
2851 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2852 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2853 This bug exists in every version of Exim that I still have, right back to
2856 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2857 However, an attempt to turn on case-sensitivity in a regex key by
2858 including (?-i) didn't work because the subject string was already
2859 lowercased, and the effects were non-intuitive. It turns out that a
2860 one-line patch can be used to allow (?-i) to work as expected.
2866 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2867 other effects) broke the use of negated acl sub-conditions.
2869 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2872 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2873 "Deny" causes Exim to reject the incoming connection with a 554 error.
2874 Unfortunately, if there is a major crisis, such as a disk failure,
2875 tcp-wrappers gives "deny", whereas what one would like would be some
2876 kind of temporary error. A kludge has been added to help with this.
2877 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2878 554 error is used if errno is still zero or contains ENOENT (which occurs
2879 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2882 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2884 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2885 errors. Otherwise a message that provokes a temporary error (when other
2886 messages do not) can cause a whole host to time out.
2888 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2889 addresses were routed directly to files or pipes from a redirect router.
2890 File deliveries just didn't batch; pipe deliveries might have suffered
2893 PH/06 A failure to get a lock for a hints database would erroneously always say
2894 "Failed to get write lock", even when it was really a read lock.
2896 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2897 of 0600. This has been changed to use the value of the lockfile_mode
2898 option (which defaults to 0600).
2900 PH/08 Applied small patch from the Sieve maintainer.
2902 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2903 folder from quota calculations, a direct delivery into this folder messed
2904 up the contents of the maildirsize file. This was because the regex was
2905 used only to exclude .Trash (or whatever) when the size of the mailbox
2906 was calculated. There was no check that a delivery was happening into an
2907 excluded directory. This bug has been fixed by ignoring all quota
2908 processing for deliveries into excluded directories.
2910 PH/10 Added the maildirfolder_create_regex option to appendfile.
2916 PH/01 The code for finding all the local interface addresses on a FreeBSD
2917 system running IPv6 was broken. This may well have applied to all BSD
2918 systems, as well as to others that have similar system calls. The broken
2919 code found IPv4 interfaces correctly, but gave incorrect values for the
2920 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2921 that it would not match correctly against @[] and not recognize the IPv6
2924 PH/02 The ipliteral router was not recognizing addresses of the form user@
2925 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2927 PH/03 Added disable_ipv6.
2929 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2930 decoded form, because it is most often used to construct To: headers
2931 lines in autoreplies, and the decoded form may well be syntactically
2932 invalid. However, $reply_address has leading white space removed, and all
2933 newlines turned into spaces so that the autoreply transport does not
2936 PH/05 If group was specified without a user on a router, and no group or user
2937 was specified on a transport, the group from the router was ignored.
2939 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2940 for visible compile-time settings that can be used to change these
2941 numbers, for those that want even more. Backwards compatibility with old
2942 spool files has been maintained. However, going back to a previous Exim
2943 release will lost any variables that are in spool files.
2945 PH/07 Two small changes when running in the test harness: increase delay when
2946 passing a TCP/IP connection to a new process, in case the original
2947 process has to generate a bounce, and remove special handling of
2948 127.0.0.2 (sic), which is no longer necessary.
2950 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2951 be the same on different OS.
2953 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2956 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2957 whether --show-vars was specified or not
2959 JJ/02 exipick: Added support for new ACL variable spool format introduced
2962 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2963 syntactically invalid From: or Reply-to: line, and a filter used this to
2964 generate an autoreply, and therefore failed to obtain an address for the
2965 autoreply, Exim could try to deliver to a non-existent relative file
2966 name, causing unrelated and misleading errors. What now happens is that
2967 it logs this as a hard delivery error, but does not attempt to create a
2970 PH/11 The exinext utility has a -C option for testing purposes, but although
2971 the given file was scanned by exinext itself; it wasn't being passed on
2972 when Exim was called.
2974 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2975 an end-of-file indication when reading a command response.
2977 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2978 compiled. In many other places in Exim, IPv6 addresses are always
2979 recognized, so I have changed this. It also means that IPv4 domain
2980 literals of the form [IPV4:n.n.n.n] are now always recognized.
2982 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2983 used if the router is not running as root, for example, when verifying at
2984 ACL time, or when using -bh. The debugging output from this situation was
2985 non-existent - all you got was a failure to exec. I have made two
2988 (a) Failures to set uid/gid, the current directory, or a process leader
2989 in a subprocess such as that created by queryprogram now generate
2990 suitable debugging output when -d is set.
2992 (b) The queryprogram router detects when it is not running as root,
2993 outputs suitable debugging information if -d is set, and then runs
2994 the subprocess without attempting to change uid/gid.
2996 PH/15 Minor change to Makefile for building test_host (undocumented testing
2999 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3000 additional section of a DNS packet that returns MX or SRV records.
3001 Instead, it always explicitly searches for A/AAAA records. This avoids
3002 major problems that occur when a DNS server includes only records of one
3003 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3004 fixed another bug: if SRV records were looked up and the corresponding
3005 address records were *not* found in the additional section, the port
3006 values from the SRV records were lost.
3008 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3009 using the correct key (the original address) when searching the retry
3010 rules in order to find which one to use for generating the retry hint.
3012 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3013 adding the default one. Similarly, if it contains a Reply-To: header, the
3014 errors_reply_to option, if set, is not used.
3016 PH/19 When calculating a retry time, Exim used to measure the "time since
3017 failure" by looking at the "first failed" field in the retry record. Now
3018 it does not use this if it is later than than the arrival time of the
3019 message. Instead it uses the arrival time. This makes for better
3020 behaviour in cases where some deliveries succeed, thus re-setting the
3021 "first failed" field. An example is a quota failure for a huge message
3022 when small messages continue to be delivered. Without this change, the
3023 "time since failure" will always be short, possible causing more frequent
3024 delivery attempts for the huge message than are intended.
3025 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3027 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3028 $1, $2, $3) because the numerical variables can be reset during some
3029 expansion items (e.g. "match"), thereby losing the authentication data.
3031 PH/21 Make -bV show the size of off_t variables so that the test suite can
3032 decide whether to run tests for quotas > 2G.
3034 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3035 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3036 filecount value is greater than 2G or if a quota value is greater than 2G
3037 on a system where the size of off_t is not greater than 4, a panic error
3040 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3041 never match. The debug and -bh output now contains an explicit error
3042 message indicating a malformed IPv4 address or mask.
3044 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3045 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3046 PH/23 above applies.
3048 PH/25 Do not write to syslog when running in the test harness. The only
3049 occasion when this arises is a failure to open the main or panic logs
3050 (for which there is an explicit test).
3052 PH/26 Added the /no_tell option to "control=freeze".
3054 PH/27 If a host name lookup failed very early in a connection, for example, if
3055 the IP address matched host_lookup and the reverse lookup yielded a name
3056 that did not have a forward lookup, an error message of the form "no IP
3057 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3058 could be logged. Now it outputs the IP address instead of "NULL".
3060 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3061 allows the sender and the authenticated sender to be set when
3062 submitting a message from within Exim. Since child_open_exim() is
3063 documented for local_scan(), the new function should be too.
3065 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3066 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3067 results in an empty string is now treated as unset.
3069 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3071 PH/31 Added sender_verify_fail logging option.
3073 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3074 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3075 tidied the source and removed it altogether.
3077 PH/33 When a queue run was abandoned because the load average was too high, a
3078 log line was always written; now it is written only if the queue_run log
3079 selector is set. In addition, the log line for abandonment now contains
3080 information about the queue run such as the pid. This is always present
3081 in "start" and "stop" lines but was omitted from the "abandon" line.
3083 PH/34 Omit spaces between a header name and the colon in the error message that
3084 is given when verify = headers_syntax fails (if there are lots of them,
3085 the message gets confusing).
3087 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3088 names, as there are now some PTR records that contain slashes. This check
3089 is only to protect against broken name servers that fall over on strange
3090 characters, so the fact that it applies to all lookups doesn't matter.
3092 PH/36 Now that the new test suite is complete, we can remove some of the
3093 special code in Exim that was needed for the old test suite. For example,
3094 sorting DNS records because real resolvers return them in an arbitrary
3095 order. The new test suite's fake resolver always returns records in the
3098 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3099 bounces) except when queue_only is set, to avoid logging races between
3100 the different processes.
3102 PH/38 Panic-die if .include specifies a non-absolute path.
3104 PH/39 A tweak to the "H" retry rule from its user.
3106 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3107 a label. They prevented compilation on older perls.
3109 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3110 a warning to be raised on newish perls.
3112 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3113 on queue. Changes to match documented behaviour of showing count of
3114 messages matching specified criteria.
3116 PH/40 Changed the default ident timeout from 30s to 5s.
3118 PH/41 Added support for the use of login_cap features, on those BSD systems
3119 that have them, for controlling the resources used by pipe deliveries.
3121 PH/42 The content-scanning code uses fopen() to create files in which to put
3122 message data. Previously it was not paying any attention to the mode of
3123 the files. Exim runs with umask(0) because the rest of the code creates
3124 files with open(), and sets the required mode explicitly. Thus, these
3125 files were ending up world-writeable. This was not a big issue, because,
3126 being within the spool directory, they were not world-accessible. I have
3127 created a function called modefopen, which takes an additional mode
3128 argument. It sets umask(777), creates the file, chmods it to the required
3129 mode, then resets the umask. All the relevant calls to fopen() in the
3130 content scanning code have been changed to use this function.
3132 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3133 to 24 hours. This avoids potential overflow problems when processing G
3134 and H retry rules. I suspect nobody ever tinkers with this value.
3136 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3138 PH/45 When the plaintext authenticator is running as a client, the server's
3139 challenges are checked to ensure they are valid base64 strings. By
3140 default, the authentication attempt is cancelled if an invalid string is
3141 received. Setting client_ignore_invalid_base64 true ignores these errors.
3142 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3143 they are received. Thus, the responses can be made to depend on the
3144 challenges. If an invalid string is ignored, an empty string is placed in
3147 PH/46 Messages that are created by the autoreply transport now contains a
3148 References: header, in accordance with RFCs 2822 and 3834.
3150 PH/47 Added authenticated_sender_force to the smtp transport.
3152 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3154 PH/49 Installed latest patch from the Sieve maintainer.
3156 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3157 was also set, the appendfile transport was unnecessarily scanning a
3158 directory of message files (e.g. for maildir delivery) to find the count
3159 of files (along with the size), even though it did not need this
3160 information. It now does the scan only if it needs to find either the
3161 size of the count of files.
3163 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3165 PH/52 Two bugs concerned with error handling when the smtp transport is
3168 (i) Exim was not creating retry information for temporary errors given
3169 for individual recipients after the DATA command when the smtp transport
3170 was used in LMTP mode. This meant that they could be retried too
3171 frequently, and not timed out correctly.
3173 (ii) Exim was setting the flag that allows error details to be returned
3174 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3175 recipients that were returned after the DATA command.
3177 PH/53 This is related to PH/52, but is more general: for any failing address,
3178 when detailed error information was permitted to be returned to the
3179 sender, but the error was temporary, then after the final timeout, only
3180 "retry timeout exceeded" was returned. Now it returns the full error as
3181 well as "retry timeout exceeded".
3183 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3184 do this, and (what is worse) MTAs that accept it.
3186 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3187 will now be deprecated.
3189 PH/56 New os.c-cygwin from the Cygwin maintainer.
3191 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3192 formats (previously only available in exim formats via -bpr, -bpru,
3193 and -bpra. Now also available in native and exiqgrep formats)
3195 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3196 with very large, slow to parse queues
3198 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3200 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3202 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3203 responses to authentication challenges, though it was showing the
3204 challenges; (ii) I've removed the CR characters from the debug output for
3207 PH/58 Allow for the insertion of a newline as well as a space when a string
3208 is turned into more than one encoded-word during RFC 2047 encoding. The
3209 Sieve code now uses this.
3211 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3212 data_4xx, lost_connection, tls_required.
3214 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3215 message was being sent as an SMTP response.
3217 PH/61 Add -l and -k options to exicyclog.
3219 PH/62 When verifying, if an address was redirected to one new address, so that
3220 verification continued, and the new address failed or deferred after
3221 having set something in $address_data, the value of $address_data was not
3222 passed back to the ACL. This was different to the case when no
3223 redirection occurred. The value is now passed back in both cases.
3225 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3226 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3227 use login_cap.h, so on its own it isn't the distinguishing feature. The
3228 new name refers directly to the setclassresources() function.
3230 PH/65 Added configuration files for NetBSD3.
3232 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3234 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3235 is preferred over IPv4.
3237 PH/68 The bounce_return_message and bounce_return_body options were not being
3238 honoured for bounces generated during the reception of non-SMTP messages.
3239 In particular, this applied to messages rejected by the ACL. This bug has
3240 been fixed. However, if bounce_return_message is true and bounce_return_
3241 body is false, the headers that are returned for a non-SMTP message
3242 include only those that have been read before the error was detected.
3243 (In the case of an ACL rejection, they have all been read.)
3245 PH/69 The HTML version of the specification is now built in a directory called
3246 spec_html instead of spec.html, because the latter looks like a path with
3247 a MIME-type, and this confuses some software.
3249 PH/70 Catch two compiler warnings in sieve.c.
3251 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3252 function verify_get_ident() calls ip_connect() to connect a socket, but
3253 if the "connect()" function timed out, ip_connect() used to close the
3254 socket. However, verify_get_ident() also closes the socket later, and in
3255 between Exim writes to the log, which may get opened at this point. When
3256 the socket was closed in ip_connect(), the log could get the same file
3257 descriptor number as the socket. This naturally causes chaos. The fix is
3258 not to close the socket in ip_connect(); the socket should be closed by
3259 the function that creates it. There was only one place in the code where
3260 this was missing, in the iplookup router, which I don't think anybody now
3261 uses, but I've fixed it anyway.
3263 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3264 well as to direct DNS lookups. Otherwise the handling of names in host
3265 lists is inconsistent and therefore confusing.
3271 PH/01 Two changes to the default runtime configuration:
3273 (1) Move the checks for relay_from_hosts and authenticated clients from
3274 after to before the (commented out) DNS black list checks.
3276 (2) Add control=submission to the relay_from_hosts and authenticated
3277 clients checks, on the grounds that messages accepted by these
3278 statements are most likely to be submissions.
3280 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3282 (1) Generate an error if the third argument for the ${prvs expansion is
3285 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3288 (3) Reset the variables that are obtained from the first argument of
3289 ${prvscheck and used in the second argument before leaving the code,
3290 because their memory is reclaimed, so using them afterwards may do
3293 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3294 one (it's much easier than Tom thought :-).
3296 (5) Because of (4), we can now allow for the use of $prvscheck_result
3297 inside the third argument.
3299 PH/03 For some reason, the default setting of PATH when running a command from
3300 a pipe transport was just "/usr/bin". I have changed it to
3303 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3304 anything to be listed in the output from -bV.
3306 PH/05 When a filter generated an autoreply, the entire To: header line was
3307 quoted in the delivery log line, like this:
3309 => >A.N.Other <ano@some.domain> <original@ddress> ...
3311 This has been changed so that it extracts the operative address. There
3312 may be more than one such address. If so, they are comma-separated, like
3315 => >ano@some.domain,ona@other.domain <original@ddress> ...
3317 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3318 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3319 not being looked up in the rDNS to get a host name, Exim was showing the
3320 IP address twice in Received: lines, even though the IP addresses were
3321 identical. For example:
3323 Received: from [1.2.3.4] (helo=[1.2.3.4])
3325 However, if the real host name was known, it was omitting the HELO data
3326 if it matched the actual IP address. This has been tidied up so that it
3327 doesn't show the same IP address twice.
3329 PH/07 When both +timestamp and +memory debugging was on, the value given by
3330 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3331 called by the debug printing, thereby overwriting the timestamp buffer.
3332 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3334 PH/08 When the original message was included in an autoreply transport, it
3335 always said "this is a copy of the message, including all the headers",
3336 even if body_only or headers_only was set. It now gives an appropriate
3339 PH/09 Applied a patch from the Sieve maintainer which:
3341 o fixes some comments
3342 o adds the (disabled) notify extension core
3343 o adds some debug output for the result of if/elsif tests
3344 o points to the current vacation draft in the documentation
3345 and documents the missing references header update
3349 o fixes a bug in processing the envelope test (when testing
3350 multiple envelope elements, the last element determined the
3353 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3354 Electronic Mail") by including:
3356 Auto-submitted: auto-generated
3358 in the messages that it generates (bounce messages and others, such as
3359 warnings). In the case of bounce messages for non-SMTP messages, there was
3360 also a typo: it was using "Auto_submitted" (underscore instead of
3361 hyphen). Since every message generated by Exim is necessarily in response
3362 to another message, thes have all been changed to:
3364 Auto-Submitted: auto-replied
3366 in accordance with these statements in the RFC:
3368 The auto-replied keyword:
3370 - SHOULD be used on messages sent in direct response to another
3371 message by an automatic process,
3373 - MUST NOT be used on manually-generated messages,
3375 - MAY be used on Delivery Status Notifications (DSNs) and Message
3376 Disposition Notifications (MDNs),
3378 - MUST NOT be used on messages generated by automatic or periodic
3379 processes, except for messages which are automatic responses to
3382 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3383 to the default Received: header definition.
3385 PH/12 Added log selector acl_warn_skipped (default on).
3387 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3388 variables because (a) they are in the wrong storage pool and (b) even if
3389 they were copied, it wouldn't work properly because of the caching.
3391 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3392 checking when decoding. Apparently there are clients that generate
3393 overlong encoded strings. Why am I not surprised?
3395 PH/15 If the first argument of "${if match_address" was not empty, but did not
3396 contain an "@" character, Exim crashed. Now it writes a panic log message
3397 and treats the condition as false.
3399 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3401 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3402 extension "envelope-auth". The code is finished and in agreement with
3403 other implementations, but there is no documentation so far and in fact,
3404 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3405 not changing the active code.
3407 Print executed "if" and "elsif" statements when debugging is used. This
3408 helps a great deal to understand what a filter does.
3410 Document more things not specified clearly in RFC3028. I had all this
3411 sorted out, when out of a sudden new issues came to my mind. Oops."
3413 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3416 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3417 canonical form (as documented). However, after a host name lookup from
3418 the IP address, check_host() was doing a simple string comparison with
3419 addresses acquired from the DNS when checking that the found name did
3420 have the original IP as one of its addresses. Since any found IPv6
3421 addresses are likely to be in abbreviated form, the comparison could
3422 fail. Luckily, there already exists a function for doing the comparison
3423 by converting both addresses to binary, so now that is used instead of
3424 the text comparison.
3426 PH/20 There was another similar case to PH/19, when a complete host name was
3427 given in a host list; looking up its IP address could give an abbreviated
3428 form, whereas the current host's name might or might not be abbreviated.
3429 The same fix has been applied.
3435 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3436 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3439 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3440 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3442 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3444 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3445 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3446 Also, for "Auto-Submitted", treat anything other than "no" as
3447 non-personal, in accordance with RFC 3834. (Previously it treated
3448 anything starting "auto-" as non-personal.)
3450 TF/01 The control=submission/name=... option had a problem with syntax
3451 errors if the name included a slash character. The /name= option
3452 now slurps the rest of the string, so it can include any characters
3453 but it must come last in the list of options (after /sender_retain
3456 PH/05 Some modifications to the interface to the fake nameserver for the new
3464 TK/01 Added the "success_on_redirect" address verification option. See
3465 NewStuff for rationale and an example.
3467 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3469 PH/02 Patch to exigrep to allow it to work on syslog lines.
3471 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3472 fread() to skip over the body file's header line, because in Cygwin the
3473 header line is locked and is inaccessible.
3475 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3476 co-exist for some time) to make it clear that it is the Exim ID that is
3477 referenced, not the Message-ID: header line.
3479 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3480 string_format() function, because snprintf() does not exist on all
3483 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3484 ${stat: expansion item.
3486 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3487 protocol synchronization error", to keep the pedants happy.
3489 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3490 well as for IRIX systems, when gcc is being used. See the host.c source
3493 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3495 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3498 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3499 transport and to the smtp transport in LMTP mode.
3501 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3503 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3504 run a filter in a subprocess. This could lead to confusion in subsequent
3505 lookups in the parent process. There should also be a search_tidyup() at
3506 the end of the subprocess.
3508 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3509 only if the host matched helo_try_verify_hosts, which caused the
3510 verification to occur when the EHLO/HELO command was issued. The ACL just
3511 tested the remembered result. Now, if a previous verification attempt has
3512 not happened, "verify = helo" does it there and then.
3514 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3516 TK/03 Fix log output including CR from clamd.
3518 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3519 exist provoked a memory error which could cause a segfault.
3521 PH/15 Installed PCRE 6.2
3523 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3525 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3526 of the problem. Specifically, suggested +O2 rather than +O1 for the
3529 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3531 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3532 record, the port was not correctly logged when the outgoing_port log
3533 selector was set (it logged the transort's default port).
3535 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3536 fallback_hosts, and "hosts" in the smtp transport.
3538 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3539 host errors such as "Connection refused".
3541 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3542 authentication with radiusclient 0.4.9:
3544 - Error returned from rc_read_config was caught wrongly
3545 - Username/password not passed on to radius server due to wrong length.
3547 The presumption is that some radiusclient API changes for 4.51/PH/17
3548 were not taken care of correctly. The code is still untested by me (my
3549 Linux distribution still has 0.3.2 of radiusclient), but it was
3550 contributed by a Radius user.
3552 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3553 expanding the "port" option of the smtp transport.
3555 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3556 while reading a MIME header. Thanks to Tom Hughes for a patch.
3558 PH/24 Include config.h inside local_scan.h so that configuration settings are
3561 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3562 that in an ACL for RCPT (for example), you can examine exactly what was
3565 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3566 commands, but it was not correctly comparing the address with the actual
3567 client host address. Thus, it would show the EHLO address in Received:
3568 header lines when this was not necessary.
3570 PH/27 Added the % operator to ${eval:}.
3572 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3573 it should be ignoring failures (because with -C, for example, it has lost
3574 privilege). It wasn't ignoring creation failures other than "already
3577 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3580 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3581 because an input line was too long, either on its own, or by virtue of
3582 too many continuations, the temporary file was not being removed, and the
3583 return code was incorrect.
3585 PH/31 Missing "BOOL" in function definition in filtertest.c.
3587 PH/32 Applied Sieve patches from the maintainer.
3589 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3591 PH/33 Added "verify = not_blind".
3593 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3594 Local/Makefile (with some defaults set). These are used in built scripts
3595 such as exicyclog, but they have never been used in the exim_install
3596 script (though there are many overriding facilities there). I have
3597 arranged that the exim_install script now takes note of these two
3600 PH/35 Installed configuration files for Dragonfly.
3602 PH/36 When a locally submitted message by a trusted user did not contain a
3603 From: header, and the sender address was obtained from -f or from an SMTP
3604 MAIL command, and the trusted user did not use -F to supply a sender
3605 name, $originator_name was incorrectly used when constructing a From:
3606 header. Furthermore, $originator_name was used for submission mode
3607 messages from external hosts without From: headers in a similar way,
3608 which is clearly wrong.
3610 PH/37 Added control=suppress_local_fixups.
3612 PH/38 When log_selector = +received_sender was set, and the addition of the
3613 sender made the log line's construction buffer exactly full, or one byte
3614 less than full, an overflow happened when the terminating "\n" was
3617 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3618 when the result of a list match is failure because a DNS lookup failed.
3620 PH/40 RM_COMMAND is now used in the building process.
3622 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3623 the "build-* directories that it finds.
3625 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3626 address in a domain literal was a prefix of an interface address.
3628 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3629 when verifying a sender address, unless rewrite_headers is false.
3631 PH/44 Wrote a long comment about why errors_to addresses are verified as
3632 recipients, not senders.
3634 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3635 the ratelimit ACL was added.
3637 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3639 PH/46 Added extra information about PostgreSQL errors to the error string.
3641 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3642 avoiding the need to install special zones in a real server. This is
3643 backwards compatible; if it can't find the fake resolver, it drops back.
3644 Thus, both old and new test suites can be run.
3646 TF/02 Added util/ratelimit.pl
3648 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3649 clock is set back in time.
3651 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3652 Candler <B.Candler@pobox.com>.
3654 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3655 OK for addresses that are the result of redirections.
3657 PH/48 A number of further additions for the benefit of the new test suite,
3658 including a fake gethostbyname() that interfaces to the fake DNS resolver
3661 TF/06 The fix for widen_domains has also been applied to qualify_single and
3662 search_parents which are the other dnslookup options that can cause
3665 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3668 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3670 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3671 service" if it encountered an MX record with an empty target hostname.
3672 The message is now "an MX or SRV record indicated no SMTP service".
3674 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3675 if the DNS of the sending site is misconfigured. This is quite a
3676 common situation. This change restores the behaviour of treating a
3677 helo verification defer as a failure.
3679 PH/51 If self=fail was set on a router, the bounce message did not include the
3680 actual error message.
3686 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3688 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3689 command itself ended in error, the underlying message about the transport
3690 filter timeout was being overwritten with the pipe command error. Now the
3691 underlying error message should be appended to the second error message.
3693 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3695 PH/02 Reduce the amount of output that "make" produces by default. Full output
3696 can still be requested.
3698 PH/03 The warning log line about a condition test deferring for a "warn" verb
3699 was being output only once per connection, rather than after each
3700 occurrence (because it was using the same function as for successful
3701 "warn" verbs). This seems wrong, so I have changed it.
3703 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3704 it should not have, which might have caused a crash in the right
3705 circumstances, but probably never did.
3707 PH/04 Installed a modified version of Tony Finch's patch to make submission
3708 mode fix the return path as well as the Sender: header line, and to
3709 add a /name= option so that you can make the user's friendly name appear
3712 TF/03 Added the control = fakedefer ACL modifier.
3714 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3715 Mark Lowes for thorough testing.
3717 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3719 TK/03 Merged latest SRS patch from Miles Wilton.
3721 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3722 with the definition in sysexits.h (which is #included earlier).
3723 Fortunately, Exim does not actually use EX_OK. The code used to try to
3724 preserve the sysexits.h value, by assuming that macro definitions were
3725 scanned for macro replacements. I have been disabused of this notion,
3726 so now the code just undefines EX_OK before #including unistd.h.
3728 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3729 in the smtp transport. When a block could not be written in a single
3730 write() function, the timeout was being re-applied to each part-write.
3731 This seems wrong - if the receiver was accepting one byte at a time it
3732 would take for ever. The timeout is now adjusted when this happens. It
3733 doesn't have to be particularly precise.
3735 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3736 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3738 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3739 without a domain if the check to <postmaster@domain> fails.
3741 SC/01 Eximstats: added -xls and the ability to specify output files
3742 (patch written by Frank Heydlauf).
3744 SC/02 Eximstats: use FileHandles for outputting results.
3746 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3748 SC/04 Eximstats: fixed display of large numbers with -nvr option
3750 SC/05 Eximstats: fixed merging of reports with empty tables.
3752 SC/06 Eximstats: added the -include_original_destination flag
3754 SC/07 Eximstats: removed tabs and trailing whitespace.
3756 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3758 TK/06 MBOX spool code: Add real "From " MBOX separator line
3759 so the .eml file is really in mbox format (even though
3760 most programs do not really care). Patch from Alex Miller.
3762 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3763 The latter is generated from $received_to and is only set if the
3764 message has one envelope recipient. SA can use these headers,
3765 obviously out-of-the-box. Patch from Alex Miller.
3767 PH/08 The ${def test on a variable was returning false if the variable's
3768 value was "0", contrary to what the specification has always said!
3769 The result should be true unless the variable is empty.
3771 PH/09 The syntax error of a character other than { following "${if
3772 def:variable_name" (after optional whitespace) was not being diagnosed.
3773 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3774 accidental colon was present, for example, could give incorrect results.
3776 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3777 result is used (not including appendfile, where other changes are about
3780 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3781 This involved changing a lot of size variables from int to off_t. It
3782 should work with maildirs and everything.
3784 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3785 spamd dying while we are connected to it.
3787 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3790 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3791 function for BDB 4.3.
3793 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3795 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3796 the sources for PCRE have been re-arranged and more files are now
3799 PH/15 The code I had for printing potentially long long variables in PH/11
3800 above was not the best (it lost precision). The length of off_t variables
3801 is now inspected at build time, and an appropriate printing format (%ld
3802 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3803 to be "long long int" or "long int". This is needed for the internal
3804 formatting function string_vformat().
3806 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3807 the configuration file to be ":syslog", then the script "guesses" where
3808 the logs files are, rather than using the compiled in default. In our
3809 case the guess is not the same as the compiled default, so the script
3810 suddenly stopped working when I started to use syslog. The patch checks
3811 to see if log_file_path is "". If so, it attempts to read it from exim
3812 with no configuration file to get the compiled in version, before it
3813 falls back to the previous guessing code."
3815 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3816 implementing BATV in an Exim configuration. See NewStuff for the gory
3819 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3820 Makefile that are specific to HP-UX.
3822 PH/18 If the "use_postmaster" option was set for a recipient callout together
3823 with the "random" option, the postmaster address was used as the MAIL
3824 FROM address for the random test, but not for the subsequent recipient
3825 test. It is now used for both.
3827 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3828 patch removes a few documentation additions to RFC 3028, because the
3829 latest draft now contains them. It adds the new en;ascii-case comparator
3830 and a new error check for 8bit text in MIME parts. Comparator and
3831 require names are now matched exactly. I enabled the subaddress
3832 extension, but it is not well tested yet (read: it works for me)."
3834 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3835 rework some of the code of TK/09 above to avoid the hardwired use of
3836 "%lld" and "long long". Replaced the call to snprintf() with a call to
3839 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3840 records point to non-existent hosts", "retry timeout exceeded", and
3841 "retry time not reached for any host after a long failure period".
3843 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3844 experimental DomainKeys support:
3846 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3847 (2) On an error such as an illegally used "control", the wrong name for
3848 the control was given.
3850 These problems did NOT occur unless DomainKeys support was compiled.
3852 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3854 PH/24 Added ${if match_ip condition.
3856 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3857 databases so that it will be absolutely obvious if a crash occurs in the
3858 DB library. This is a regular occurrence (often caused by mis-matched
3861 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3862 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3863 picked up on a user's system that detects such things. There doesn't seem
3864 to be a gcc warning option for this - only an attribute that has to be
3865 put on the function's prototype. It seems that in Fedora Core 4 they have
3866 set this on a number of new functions. No doubt there will be more in due
3869 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3870 specify a transport. However, if an address that was verified by such a
3871 router was the subject of a callout, Exim crashed because it tried to
3872 read the rcpt_include_affixes from the non-existent transport. Now it
3873 just assumes that the setting of that option is false. This bug was
3874 introduced by 4.51/PH/31.
3876 PH/28 Changed -d+all to exclude +memory, because that information is very
3877 rarely of interest, but it makes the output a lot bigger. People tend to
3878 do -d+all out of habit.
3880 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3881 code in os-type was giving problems when libc.so lives in lib64, like on
3884 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3885 aren't the modern standard, and it seems that some systems' include files
3886 don't always have them. Exim was already checking for some of the newer
3887 ones like T_AAAA, and defining it itself. I've added checks for all the
3888 record types that Exim uses.
3890 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3891 not automatically generating a new one, as it is supposed to. This
3892 prevented TLS from working. If the file did exist, but contained invalid
3893 data, a new version was generated, as expected. It was only the case of a
3894 non-existent file that was broken.
3896 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3897 with a change in libdomainkeys > 0.64.
3899 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3900 from DNS. If the selector record carries the flag, it now has
3901 precedence over the domain-wide flag.
3903 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3905 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3906 the use of a port name that isn't defined in /etc/services) occurred, the
3907 message was deferred as in a normal delivery, and thus remained on the
3908 spool, instead of being failed because of the mua_wrapper setting. This
3909 is now fixed, and I tidied up some of the mua_wrapper messages at the
3912 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3913 the messages in a hash of arrays rather than using individual hashes.
3914 This is a bit cleaner and results in dramatic memory savings, albeit
3915 at a slight CPU cost.
3917 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3918 as requested by Marc Sherman.
3920 SC/10 Eximstats: added histograms for user specified patterns as requested
3923 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3925 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3926 fopen() in the content-scanning modules that did not already have it.
3932 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3933 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3935 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3937 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3939 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3940 to test Sieve filters that use "vacation".
3942 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3943 that changes the way the GnuTLS parameters are stored in the cache file.
3944 The new format can be generated externally. For backward compatibility,
3945 if the data in the cache doesn't make sense, Exim assumes it has read an
3946 old-format file, and it generates new data and writes a new file. This
3947 means that you can't go back to an older release without removing the
3950 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3951 work if there are any delivery delays because "one_time" forces the
3952 parent to be marked "delivered", so its unseen clone is never tried
3953 again. For this reason, Exim now forbids the simultaneous setting of
3956 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3957 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3958 entry if you want to know the details. The fix, however, neglected to
3959 consider the case when local delivery batching is involved. The test for
3960 "previously delivered" was not happening when checking to see if an
3961 address could be batched with a previous (undelivered) one; under
3962 certain circumstances this could lead to multiple deliveries to the same
3965 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3966 in its include files, and this causes problems building Exim.
3968 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3969 header_syntax) but Exim was just ignoring anything given after a slash.
3970 In particular, this caused confusion with an attempt to use "verify =
3971 reverse_host_lookup/defer_ok". An error is now given when options are
3972 supplied for verify items that do not have them. (Maybe reverse_host_
3973 lookup should have a defer_ok option, but that's a different point.)
3975 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3976 defined by RFC 821) to 2048, because there were problems with some AUTH
3977 commands, and RFC 1869 says the size should be increased for extended
3978 SMTP commands that take arguments.
3980 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3983 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3984 "unknown" error; now it says that the functionality isn't in the binary.
3986 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3987 an address' error message when a string expansion fails (syntax or
3988 whatever). Otherwise the password may appear in the log. Following change
3989 PH/42 below, there is no longer a chance of it appearing in a bounce
3992 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3994 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3995 of its addresses was ever tried. (Bugzilla bug #2.)
3997 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3998 the result incorrectly in the debug output. (It correctly added a newline
3999 to what was transported.)
4001 TF/01 Added $received_time.
4003 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4004 commented out examples of how to interface to a virus scanner and to
4005 SpamAssassin. Also added commented examples of av_scanner and
4006 spamd_address settings.
4008 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4009 and controls are allowed in which ACLs. There were a couple of minor
4010 errors. Some of the entries in the conditions table (which is a table of
4011 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4012 negation of where the condition IS allowed.
4014 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4016 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4017 header file does not have a version number, so I've had to invent a new
4018 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4019 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4020 radiusclient), but it was contributed by a Radius user.
4022 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4023 files or directories, for OpenSSL.
4025 PH/19 When an Exim process that is running as root has to create an Exim log
4026 file, it does so in a subprocess that runs as exim:exim so as to get the
4027 ownership right at creation (otherwise, other Exim processes might see
4028 the file with the wrong ownership). There was no test for failure of this
4029 fork() call, which would lead to the process getting stuck as it waited
4030 for a non-existent subprocess. Forks do occasionally fail when resources
4031 run out. I reviewed all the other calls to fork(); they all seem to check
4034 PH/20 When checking for unexpected SMTP input at connect time (before writing
4035 the banner), Exim was not dealing correctly with a non-positive return
4036 from the read() function. If the client had disconnected by this time,
4037 the result was a log entry for a synchronization error with an empty
4038 string after "input=" when read() returned zero. If read() returned -1
4039 (an event I could not check), uninitialized data bytes were printed.
4040 There were reports of junk text (parts of files, etc) appearing after
4043 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4045 PH/22 Added support for macro redefinition, and (re)definition in between
4046 driver and ACL definitions.
4048 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4049 forgetting to use the resulting value; it was using the unexpanded value.
4051 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4052 hadn't been configured. The fix is from Juergen Kreileder, who
4053 understands it better than I do:
4055 "Here's what I see happening with three configured cyrus_sasl
4056 authenticators configured (plain, login, cram-md5):
4058 On startup auth_cyrus_sasl_init() gets called for each of these.
4059 This means three calls to sasl_listmech() without a specified mech_list.
4060 => SASL tests which mechs of all available mechs actually work
4061 => three warnings about OTP not working
4062 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4064 With the patch, sasl_listmech() also gets called three times. But now
4065 SASL's mech_list option is set to the server_mech specified in the the
4066 authenticator. Or in other words, the answer from sasl_listmech()
4067 gets limited to just the mech you're testing for (which is different
4069 => the return list contains just 'plain' or 'login', 'cram-md5' or
4070 nothing depending on the value of ob->server_mech.
4072 I've just tested the patch: Authentication still works fine,
4073 unavailable mechs specified in the exim configuration are still
4074 caught, and the auth.log warnings about OTP are gone."
4076 PH/25 When debugging is enabled, the contents of the command line are added
4077 to the debugging output, even when log_selector=+arguments is not
4080 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4081 answer is "GNU", and only if the return is "GNU/something" is the answer
4084 PH/27 $acl_verify_message is now set immediately after the failure of a
4085 verification in an ACL, and so is available in subsequent modifiers. In
4086 particular, the message can be preserved by coding like this:
4088 warn !verify = sender
4089 set acl_m0 = $acl_verify_message
4091 Previously, $acl_verify_message was set only while expanding "message"
4092 and "log_message" when a very denied access.
4094 PH/28 Modified OS/os.c-Linux with
4096 -#ifndef OS_LOAD_AVERAGE
4097 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4099 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4100 nomenclature these days.)
4102 PH/29 Installed patch from the Sieve maintainer that adds the options
4103 sieve_useraddress and sieve_subaddress to the redirect router.
4105 PH/30 In these circumstances:
4106 . Two addresses routed to the same list of hosts;
4107 . First host does not offer TLS;
4108 . First host accepts first address;
4109 . First host gives temporary error to second address;
4110 . Second host offers TLS and a TLS session is established;
4111 . Second host accepts second address.
4112 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4113 and peerdn, if requested) that were in fact used only for the second
4116 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4117 attention to any local part prefix or suffix that was matched by the
4118 router that accepted the address. It now behaves in the same way as it
4119 does for delivery: the affixes are removed from the local part unless
4120 rcpt_include_affixes is set on the transport.
4122 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4123 timeout during the DATA phase of an incoming message.
4125 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4126 applied a patch sanctioned by the Sieve maintainer.
4128 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4129 the uid or gid is negative. A case of a negative gid caused this to be
4130 noticed. The fix allows for either to be negative.
4132 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4133 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4136 PH/36 The change PH/12 above was broken. Fixed it.
4138 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4139 the grounds that routing the same address twice would always produce the
4140 same answer. This might have been true once, but it is certainly no
4141 longer true now. Routing a child address may depend on the previous
4142 routing that produced that child. Some complicated redirection strategies
4143 went wrong when messages had multiple recipients, and made Exim's
4144 behaviour dependent on the order in which the addresses were given.
4146 I have moved the duplicate checking until after the routing is complete.
4147 Exim scans the addresses that are assigned to local and remote
4148 transports, and removes any duplicates. This means that more work will be
4149 done, as duplicates will always all be routed, but duplicates are
4150 presumably rare, so I don't expect this is of any significance.
4152 For deliveries to pipes, files, and autoreplies, the duplicate checking
4153 still happens during the routing process, since they are not going to be
4156 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4157 It corrects a timeout issue with spamd. This is Ian's comment: "The
4158 background is that sometimes spamd either never reads data from a
4159 connection it has accepted, or it never writes response data. The exiscan
4160 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4161 blindly assumes that writes won't block so it may never time out."
4163 PH/39 Allow G after quota size as well as K and M.
4165 PH/40 The value set for $authenticated_id in an authenticator may not contain
4166 binary zeroes or newlines because the value is written to log lines and
4167 to spool files. There was no check on this. Now the value is run through
4168 the string_printing() function so that such characters are converted to
4169 printable escape sequences.
4171 PH/41 $message_linecount is a new variable that contains the total number of
4172 lines in the message. Compare $body_linecount, which is the count for the
4175 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4176 bounce and delay warning messages, except in certain special cases, which
4179 (a) An SMTP error message from a remote host;
4180 (b) A message specified in a :fail: redirection;
4181 (c) A message specified in a "fail" command in a system filter;
4182 (d) A message specified in a FAIL return from the queryprogram router;
4183 (e) A message specified by the cannot_route_message router option.
4185 In these cases only, Exim does include the error details in bounce and
4186 warning messages. There are also a few cases where bland messages such
4187 as "unrouteable address" or "local delivery error" are given.
4189 PH/43 $value is now also set for the "else" part of a ${run expansion.
4191 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4192 being worked on, but at least Exim now implements the latest version to
4195 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4196 process to complete was treated as a delivery failure, a timeout while
4197 writing the message to the pipe was logged, but erroneously treated as a
4198 successful delivery. Such timeouts include transport filter timeouts. For
4199 consistency with the overall process timeout, these timeouts are now
4200 treated as errors, giving rise to delivery failures by default. However,
4201 there is now a new Boolean option for the pipe transport called
4202 timeout_defer, which, if set TRUE, converts the failures into defers for
4203 both kinds of timeout. A transport filter timeout is now identified in
4206 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4207 systems where "make" and "gmake" are different, calling "gmake" at top
4208 level broke things. I've arranged for the value of $(MAKE) to be passed
4209 from the Makefile to this script so that it can call the same version of
4213 A note about Exim versions 4.44 and 4.50
4214 ----------------------------------------
4216 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4217 changes of various kinds. As a consequence, a big documentation update was
4218 needed. This delayed the release for rather longer than seemed good, especially
4219 in the light of a couple of (minor) security issues. Therefore, the changes
4220 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4221 release. So 4.44 and 4.50 are in effect two different branches that both start
4224 I have left the 4.50 change log unchanged; it contains all the changes since
4225 4.43. The change log for 4.44 is below; many of its items are identical to
4226 those for 4.50. This seems to be the most sensible way to preserve the
4227 historical information.
4233 1. Minor wording change to the doc/README.SIEVE file.
4235 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4236 computation of the current number of files was incorrect.
4238 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4239 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4242 4. Give more explanation in the error message when the command for a transport
4243 filter fails to execute.
4245 5. There are several places where Exim runs a non-Exim command in a
4246 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4247 was being done only for the command run by the queryprogram router. It is
4248 now done for all such subprocesses. The other cases are: ${run, transport
4249 filters, and the commands run by the lmtp and pipe transports.
4251 6. Added CONFIGURE_GROUP build-time option.
4253 7. Some older OS have a limit of 256 on the maximum number of file
4254 descriptors. Exim was using setrlimit() to set 1000 as a large value
4255 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4256 systems. I've change it so that if it can't get 1000, it tries for 256.
4258 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4259 was an oversight, and furthermore, ever since the addition of extra
4260 controls (e.g. 4.43/32), the checks on when to allow different forms of
4261 "control" were broken. There should now be diagnostics for all cases when a
4262 control that does not make sense is encountered.
4264 9. Added the /retain_sender option to "control=submission".
4266 10. $recipients is now available in the predata ACL (oversight).
4268 11. Tidy the search cache before the fork to do a delivery from a message
4269 received from the command line. Otherwise the child will trigger a lookup
4270 failure and thereby defer the delivery if it tries to use (for example) a
4271 cached ldap connection that the parent has called unbind on.
4273 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4274 of $address_data from the recipient verification was clobbered by the
4275 sender verification.
4277 13. The value of address_data from a sender verification is now available in
4278 $sender_address_data in subsequent conditions in the ACL statement.
4280 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4282 15. Added a new option "connect=<time>" to callout options, to set a different
4285 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4286 was its contents. (It was OK if the option was not defined at all.)
4288 17. A "Completed" log line is now written for messages that are removed from
4289 the spool by the -Mrm option.
4291 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4292 information about exactly what failed.
4294 19. Added -dd to debug only the daemon process.
4296 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4297 handles timeouts, both on the server side and network timeouts. Renamed the
4298 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4300 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4301 It is now set to "smtps".
4303 22. $host_address is now set to the target address during the checking of
4304 ignore_target_hosts.
4306 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4307 being passed; this would have caused $sender_host_name to have been used if
4308 matching the list had actually called for a host name (not very likely,
4309 since this list is usually IP addresses). A host name is now passed as
4312 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4313 code that specifies a non-restarting handler (typically sigaction() in
4314 modern systems) in an attempt to fix a rare and obscure crash bug.
4316 25. Narrowed the window for a race in the daemon that could cause it to ignore
4317 SIGCHLD signals. This is not a major problem, because they are used only to
4318 wake it up if nothing else does.
4320 26. A malformed maildirsize file could cause Exim to calculate negative values
4321 for the mailbox size or file count. Odd effects could occur as a result.
4322 The maildirsize information is now recalculated if the size or filecount
4325 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4326 support for a long time. Removed HAVE_SYS_VFS_H.
4328 28. Installed the latest version of exipick from John Jetmore.
4330 29. In an address list, if the pattern was not a regular expression, an empty
4331 subject address (from a bounce message) matched only if the pattern was an
4332 empty string. Non-empty patterns were not even tested. This was the wrong
4333 because it is perfectly reasonable to use an empty address as part of a
4334 database query. An empty address is now tested by patterns that are
4335 lookups. However, all the other forms of pattern expect the subject to
4336 contain a local part and a domain, and therefore, for them, an empty
4337 address still always fails if the pattern is not itself empty.
4339 30. Exim went into a mad DNS loop when attempting to do a callout where the
4340 host was specified on an smtp transport, and looking it up yielded more
4341 than one IP address.
4343 31. Re-factored the code for checking spool and log partition space into a
4344 function that finds that data and another that does the check. The former
4345 is then used to implement four new variables: $spool_space, $log_space,
4346 $spool_inodes, and $log_inodes.
4348 32. The RFC2047 encoding function was originally intended for short strings
4349 such as real names; it was not keeping to the 75-character limit for
4350 encoded words that the RFC imposes. It now respects the limit, and
4351 generates multiple encoded words if necessary. To be on the safe side, I
4352 have increased the buffer size for the ${rfc2047: expansion operator from
4355 33. It is now permitted to omit both strings after an "if" condition; if the
4356 condition is true, the result is "true". As before, when the second string
4357 is omitted, a false condition yields an empty string. This makes it less
4358 cumbersome to write custom ACL and router conditions.
4360 34. Failure to deliver a bounce message always caused it to be frozen, even if
4361 there was an errors_to setting on the router. The errors_to setting is now
4364 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4365 canonical form (fully expanded) before being placed in
4366 $sender_host_address.
4368 36. The table in the code that translates DNS record types into text (T_A to
4369 "A" for instance) was missing entries for NS and CNAME. It is just possible
4370 that this could have caused confusion if both these types were looked up
4371 for the same domain, because the text type is used as part of Exim's
4372 per-process caching. But the chance of anyone hitting this buglet seems
4375 37. The dnsdb lookup has been extended in a number of ways.
4377 (1) There is a new type, "zns", which walks up the domain tree until it
4378 finds some nameserver records. It should be used with care.
4380 (2) There is a new type, "mxh", which is like "mx" except that it returns
4381 just the host names, not the priorities.
4383 (3) It is now possible to give a list of domains (or IP addresses) to be
4384 looked up. The behaviour when one of the lookups defers can be
4385 controlled by a keyword.
4387 (4) It is now possible to specify the separator character for use when
4388 multiple records are returned.
4390 38. The dnslists ACL condition has been extended: it is now possible to supply
4391 a list of IP addresses and/or domains to be looked up in a particular DNS
4394 39. Added log_selector=+queue_time_overall.
4396 40. When running the queue in the test harness, wait just a tad after forking a
4397 delivery process, to get repeatability of debugging output.
4399 41. Include certificate and key file names in error message when GnuTLS fails
4400 to set them up, because the GnuTLS error message doesn't include the name
4401 of the failing file when there is a problem reading it.
4403 42. Allow both -bf and -bF in the same test run.
4405 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4407 44. The "Exiscan patch" is now merged into the mainline Exim source.
4409 45. Sometimes the final signoff response after QUIT could fail to get
4410 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4411 < 0 before doing a fflush(). This bug looks as though it goes back to the
4412 introduction of TLS in release 3.20, but "sometimes" must have been rare
4413 because the tests only now provoked it.
4415 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4416 (this can affect the format of dates).
4418 47. exim_tidydb, when checking for the continued existence of a message for
4419 which it has found a message-specific retry record, was not finding
4420 messages that were in split spool directories. Consequently, it was
4421 deleting retry records that should have stayed in existence.
4423 48. Steve fixed some bugs in eximstats.
4425 49. The SPA authentication driver was not abandoning authentication and moving
4426 on to the next authenticator when an expansion was forced to fail,
4427 contradicting the general specification for all authenticators. Instead it
4428 was generating a temporary error. It now behaves as specified.
4430 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4431 (the order specifies the preference for clients). The order is now AES256,
4432 AES128, 3DES, ARCFOUR128.
4434 51. Small patch to Sieve code - explicitly set From: when generating an
4437 52. Exim crashed if a remote delivery caused a very long error message to be
4438 recorded - for instance if somebody sent an entire SpamAssassin report back
4439 as a large number of 550 error lines. This bug was coincidentally fixed by
4440 increasing the size of one of Exim's internal buffers (big_buffer) that
4441 happened as part of the Exiscan merge. However, to be on the safe side, I
4442 have made the code more robust (and fixed the comments that describe what
4445 53. Now that there can be additional text after "Completed" in log lines (if
4446 the queue_time_overall log selector is set), a one-byte patch to exigrep
4447 was needed to allow it to recognize "Completed" as not the last thing in
4450 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4451 patch that reportedly fixes this has been added. I am not expert enough to
4452 create a test for it. This is what the patch creator wrote:
4454 "I found a little strange behaviour of ldap code when working with
4455 Windows 2003 AD Domain, where users was placed in more than one
4456 Organization Units. When I tried to give exim partial DN, the exit code
4457 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4458 But simultaneously result of request was absolutely normal ldap result,
4459 so I produce this patch..."
4461 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4462 REFERENCE, so I have modified the code to exclude the patch when that macro
4465 55. Some experimental protocols are using DNS PTR records for new purposes. The
4466 keys for these records are domain names, not reversed IP addresses. The
4467 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4468 leaves it alone. Component reversal etc. now happens only for IP addresses.
4471 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4473 57. Double the size of the debug message buffer (to 2048) so that more of very
4474 long debug lines gets shown.
4476 58. The exicyclog utility now does better if the number of log files to keep
4477 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4479 59. Two changes related to the smtp_active_hostname option:
4481 (1) $smtp_active_hostname is now available as a variable.
4482 (2) The default for smtp_banner uses $smtp_active_hostname instead
4483 of $primary_hostname.
4485 60. The host_aton() function is supposed to be passed a string that is known
4486 to be a valid IP address. However, in the case of IPv6 addresses, it was
4487 not checking this. This is a hostage to fortune. Exim now panics and dies
4488 if the condition is not met. A case was found where this could be provoked
4489 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4490 components; fortuitously, this particular loophole had already been fixed
4491 by change 4.50/55 above.
4493 If there are any other similar loopholes, the new check in host_aton()
4494 itself should stop them being exploited. The report I received stated that
4495 data on the command line could provoke the exploit when Exim was running as
4496 exim, but did not say which command line option was involved. All I could
4497 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4498 running as the user.
4501 61. There was a buffer overflow vulnerability in the SPA authentication code
4502 (which came originally from the Samba project). I have added a test to the
4503 spa_base64_to_bits() function which I hope fixes it.
4506 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4507 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4509 63. The daemon start-up calls getloadavg() while still root for those OS that
4510 need the first call to be done as root, but it missed one case: when
4511 deliver_queue_load_max is set with deliver_drop_privilege. This is
4512 necessary for the benefit of the queue runner, because there is no re-exec
4513 when deliver_drop_privilege is set.
4515 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4516 This has been fixed.
4518 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4519 was in use, was not putting the data itself into the right store pool;
4520 consequently, it could be overwritten for a subsequent message in the same
4521 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4524 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4526 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4527 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4528 the function were treating the return as a boolean value, which happened to
4529 work because 0=false and not-0=true, but is not correct code.
4531 68. The host_aton() function was not handling scoped IPv6 addresses (those
4532 with, for example, "%eth0" on the end) correctly.
4534 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4535 negated items (that is, ~something) in unsigned ints. Some compilers
4536 apparently mutter when there is no cast.
4538 70. If an address verification called from an ACL failed, and did not produce a
4539 user-specific message (i.e. there was only a "system" message), nothing was
4540 put in $acl_verify_message. In this situation, it now puts the system
4543 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4544 session; change 4.31/43 added the unwanted input to the log line - except
4545 that it did not do this in the start of session case. It now does.
4547 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4548 This is wrong and can cause the other end to generate a synchronization
4549 error if it is another Exim or anything else that does the synchronization
4550 check. A QUIT command is no longer sent after a timeout.
4552 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4553 during host lookups.
4555 74. The defer_ok option of callout verification was not working if it was used
4556 when verifying addresses in header lines, that is, for this case:
4558 verify = header_sender/callout=defer_ok
4560 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4561 those file descriptors could be used for SMTP connections. If anything
4562 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4563 could be sent to the SMTP client, causing chaos. The daemon now opens
4564 stdin, stdout, and stderr to /dev/null when it puts itself into the
4567 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4568 log by default. The user can override this with suitable Perl magic.
4570 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4571 the log message when discard triggers, was not working for the DATA ACL or
4572 for the non-SMTP ACL.
4574 78. Error message wording change in sieve.c.
4576 79. If smtp_accept_max_per_host was set, the number of connections could be
4577 restricted to fewer than expected, because the daemon was trying to set up
4578 a new connection before checking whether the processes handling previous
4579 connections had finished. The check for completed processes is now done
4580 earlier. On busy systems, this bug wouldn't be noticed because something
4581 else would have woken the daemon, and it would have reaped the completed
4584 80. If a message was submitted locally by a user whose login name contained one
4585 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4586 It caused a spool format error. I have fixed the spool reading code. A
4587 related problem was that the "from" clause in the Received: line became
4588 illegal because of the space(s). It is now covered by ${quote_local_part.
4590 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4593 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4595 83. Patch from Sieve maintainer to fix unterminated string problem in
4596 "vacation" handling.
4598 84. Some minor changes to the Linux configuration files to help with other
4599 OS variants using glibc.
4601 85. One more patch for Sieve to update vacation handling to latest spec.
4604 ----------------------------------------------------
4605 See the note above about the 4.44 and 4.50 releases.
4606 ----------------------------------------------------
4612 1. Change 4.43/35 introduced a bug that caused file counts to be
4613 incorrectly computed when quota_filecount was set in an appendfile
4616 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4617 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4620 3. Give more explanation in the error message when the command for a transport
4621 filter fails to execute.
4623 4. There are several places where Exim runs a non-Exim command in a
4624 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4625 was being done only for the command run by the queryprogram router. It is
4626 now done for all such subprocesses. The other cases are: ${run, transport
4627 filters, and the commands run by the lmtp and pipe transports.
4629 5. Some older OS have a limit of 256 on the maximum number of file
4630 descriptors. Exim was using setrlimit() to set 1000 as a large value
4631 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4632 systems. I've change it so that if it can't get 1000, it tries for 256.
4634 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4635 was an oversight, and furthermore, ever since the addition of extra
4636 controls (e.g. 4.43/32), the checks on when to allow different forms of
4637 "control" were broken. There should now be diagnostics for all cases when a
4638 control that does not make sense is encountered.
4640 7. $recipients is now available in the predata ACL (oversight).
4642 8. Tidy the search cache before the fork to do a delivery from a message
4643 received from the command line. Otherwise the child will trigger a lookup
4644 failure and thereby defer the delivery if it tries to use (for example) a
4645 cached ldap connection that the parent has called unbind on.
4647 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4648 of $address_data from the recipient verification was clobbered by the
4649 sender verification.
4651 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4652 was its contents. (It was OK if the option was not defined at all.)
4654 11. A "Completed" log line is now written for messages that are removed from
4655 the spool by the -Mrm option.
4657 12. $host_address is now set to the target address during the checking of
4658 ignore_target_hosts.
4660 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4661 being passed; this would have caused $sender_host_name to have been used if
4662 matching the list had actually called for a host name (not very likely,
4663 since this list is usually IP addresses). A host name is now passed as
4666 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4667 code that specifies a non-restarting handler (typically sigaction() in
4668 modern systems) in an attempt to fix a rare and obscure crash bug.
4670 15. Narrowed the window for a race in the daemon that could cause it to ignore
4671 SIGCHLD signals. This is not a major problem, because they are used only to
4672 wake it up if nothing else does.
4674 16. A malformed maildirsize file could cause Exim to calculate negative values
4675 for the mailbox size or file count. Odd effects could occur as a result.
4676 The maildirsize information is now recalculated if the size or filecount
4679 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4680 support for a long time. Removed HAVE_SYS_VFS_H.
4682 18. Updated exipick to current release from John Jetmore.
4684 19. Allow an empty sender to be matched against a lookup in an address list.
4685 Previously the only cases considered were a regular expression, or an
4688 20. Exim went into a mad DNS lookup loop when doing a callout where the
4689 host was specified on the transport, if the DNS lookup yielded more than
4692 21. The RFC2047 encoding function was originally intended for short strings
4693 such as real names; it was not keeping to the 75-character limit for
4694 encoded words that the RFC imposes. It now respects the limit, and
4695 generates multiple encoded words if necessary. To be on the safe side, I
4696 have increased the buffer size for the ${rfc2047: expansion operator from
4699 22. Failure to deliver a bounce message always caused it to be frozen, even if
4700 there was an errors_to setting on the router. The errors_to setting is now
4703 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4704 canonical form (fully expanded) before being placed in
4705 $sender_host_address.
4707 24. Updated eximstats to version 1.33
4709 25. Include certificate and key file names in error message when GnuTLS fails
4710 to set them up, because the GnuTLS error message doesn't include the name
4711 of the failing file when there is a problem reading it.
4713 26. Expand error message when OpenSSL has problems setting up cert/key files.
4716 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4717 (this can affect the format of dates).
4719 28. exim_tidydb, when checking for the continued existence of a message for
4720 which it has found a message-specific retry record, was not finding
4721 messages that were in split spool directories. Consequently, it was
4722 deleting retry records that should have stayed in existence.
4724 29. eximstats updated to version 1.35
4725 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4726 1.35 - bugfix such that pie charts by volume are generated correctly
4728 30. The SPA authentication driver was not abandoning authentication and moving
4729 on to the next authenticator when an expansion was forced to fail,
4730 contradicting the general specification for all authenticators. Instead it
4731 was generating a temporary error. It now behaves as specified.
4733 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4734 (the order specifies the preference for clients). The order is now AES256,
4735 AES128, 3DES, ARCFOUR128.
4737 31. Small patch to Sieve code - explicitly set From: when generating an
4740 32. Exim crashed if a remote delivery caused a very long error message to be
4741 recorded - for instance if somebody sent an entire SpamAssassin report back
4742 as a large number of 550 error lines. This bug was coincidentally fixed by
4743 increasing the size of one of Exim's internal buffers (big_buffer) that
4744 happened as part of the Exiscan merge. However, to be on the safe side, I
4745 have made the code more robust (and fixed the comments that describe what
4748 33. Some experimental protocols are using DNS PTR records for new purposes. The
4749 keys for these records are domain names, not reversed IP addresses. The
4750 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4751 leaves it alone. Component reversal etc. now happens only for IP addresses.
4754 34. The host_aton() function is supposed to be passed a string that is known
4755 to be a valid IP address. However, in the case of IPv6 addresses, it was
4756 not checking this. This is a hostage to fortune. Exim now panics and dies
4757 if the condition is not met. A case was found where this could be provoked
4758 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4759 components; fortuitously, this particular loophole had already been fixed
4760 by change 4.50/55 or 4.44/33 above.
4762 If there are any other similar loopholes, the new check in host_aton()
4763 itself should stop them being exploited. The report I received stated that
4764 data on the command line could provoke the exploit when Exim was running as
4765 exim, but did not say which command line option was involved. All I could
4766 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4767 running as the user.
4770 35. There was a buffer overflow vulnerability in the SPA authentication code
4771 (which came originally from the Samba project). I have added a test to the
4772 spa_base64_to_bits() function which I hope fixes it.
4775 36. The daemon start-up calls getloadavg() while still root for those OS that
4776 need the first call to be done as root, but it missed one case: when
4777 deliver_queue_load_max is set with deliver_drop_privilege. This is
4778 necessary for the benefit of the queue runner, because there is no re-exec
4779 when deliver_drop_privilege is set.
4781 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4782 was in use, was not putting the data itself into the right store pool;
4783 consequently, it could be overwritten for a subsequent message in the same
4784 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4787 38. Sometimes the final signoff response after QUIT could fail to get
4788 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4789 < 0 before doing a fflush(). This bug looks as though it goes back to the
4790 introduction of TLS in release 3.20, but "sometimes" must have been rare
4791 because the tests only now provoked it.
4797 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4798 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4799 Now it returns an integer. A number of calls were still expecting a T/F
4800 return. Fortuitously, in all cases, the tests worked in OK situations,
4801 which is the norm. However, things would have gone wrong on any write
4802 failures on the smtp file descriptor. This function is used when sending
4803 messages over SMTP and also when doing verify callouts.
4805 2. When Exim is called to do synchronous delivery of a locally submitted
4806 message (the -odf or -odi options), it no longer closes stderr before doing
4809 3. Implemented the mua_wrapper option.
4811 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4813 5. Implemented the functions header_remove(), header_testname(),
4814 header_add_at_position(), and receive_remove_recipient(), and exported them
4817 6. If an ACL "warn" statement specified the addition of headers, Exim already
4818 inserted X-ACL-Warn: at the start if there was no header name. However, it
4819 was not making this test for the second and subsequent header lines if
4820 there were newlines in the string. This meant that an invalid header could
4821 be inserted if Exim was badly configured.
4823 7. Allow an ACL "warn" statement to add header lines at the start or after all
4824 the Received: headers, as well as at the end.
4826 8. Added the rcpt_4xx retry error code.
4828 9. Added postmaster_mailfrom=xxx to callout verification option.
4830 10. Added mailfrom=xxxx to the callout verification option, for verify=
4833 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4834 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4836 12. Inserted some casts to stop certain compilers warning when using pointer
4837 differences as field lengths or precisions in printf-type calls (mostly
4838 affecting debugging statements).
4840 13. Added optional readline() support for -be (dynamically loaded).
4842 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4843 same clock tick as a message's arrival, so that its received time was the
4844 same as the "first fail" time on the retry record, and that message
4845 remained on the queue past the ultimate address timeout, every queue runner
4846 would try a delivery (because it was past the ultimate address timeout) but
4847 after another failure, the ultimate address timeout, which should have then
4848 bounced the address, did not kick in. This was a "< instead of <=" error;
4849 in most cases the first failure would have been in the next clock tick
4850 after the received time, and all would be well.
4852 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4853 being recognized when the domain list was tested by the match_domain
4854 condition in an expansion string.
4856 16. Added the ${str2b64: operator.
4858 17. Exim was always calling setrlimit() to set a large limit for the number of
4859 processes, without checking whether the existing limit was already
4860 adequate. (It did check for the limit on file descriptors.) Furthermore,
4861 errors from getrlimit() and setrlimit() were being ignored. Now they are
4862 logged to the main and panic logs, but Exim does carry on, to try to do its
4863 job under whatever limits there are.
4865 18. Imported PCRE 5.0.
4867 19. Trivial typo in log message " temporarily refused connection" (the leading
4870 20. If the log selector return_path_on_delivery was set and an address was
4871 redirected to /dev/null, the delivery process crashed because it assumed
4872 that a return path would always be set for a "successful" delivery. In this
4873 case, the whole delivery is bypassed as an optimization, and therefore no
4876 21. Internal re-arrangement: the function for sending a challenge and reading
4877 a response while authentication was assuming a zero-terminated challenge
4878 string. It's now changed to take a pointer and a length, to allow for
4879 binary data in such strings.
4881 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4883 23. Exim was not respecting finduser_retries when seeking the login of the
4884 uid under which it was called; it was always trying 10 times. (The default
4885 setting of finduser_retries is zero.) Also, it was sleeping after the final
4886 failure, which is pointless.
4888 24. Implemented tls_on_connect_ports.
4890 25. Implemented acl_smtp_predata.
4892 26. If the domain in control=submission is set empty, Exim assumes that the
4893 authenticated id is a complete email address when it generates From: or
4894 Sender: header lines.
4896 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4897 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4898 chown and chgrp in /bin and hostname in /usr/bin.
4900 28. Exim was keeping the "process log" file open after each use, just as it
4901 does for the main log. This opens the possibility of it remaining open for
4902 long periods when the USR1 signal hits a daemon. Occasional processlog
4903 errors were reported, that could have been caused by this. Anyway, it seems
4904 much more sensible not to leave this file open at all, so that is what now
4907 29. The long-running daemon process does not normally write to the log once it
4908 has entered its main loop, and it closes the log before doing so. This is
4909 so that log files can straightforwardly be renamed and moved. However,
4910 there are a couple of unusual error situations where the daemon does write
4911 log entries, and I had neglected to close the log afterwards.
4913 30. The text of an SMTP error response that was received during a remote
4914 delivery was being truncated at 512 bytes. This is too short for some of
4915 the long messages that one sometimes sees. I've increased the limit to
4918 31. It is now possible to make retry rules that apply only when a message has a
4919 specific sender, in particular, an empty sender.
4921 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4922 it possible to be selective about when SMTP synchronization is enforced.
4924 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4926 32. Implemented hosts_connection_nolog.
4928 33. Added an ACL for QUIT.
4930 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4933 35. Added mailbox_size and mailbox_filecount to appendfile.
4935 36. Added control = no_multiline_responses to ACLs.
4937 37. There was a bug in the logic of the code that waits for the clock to tick
4938 in the case where the clock went backwards by a substantial amount such
4939 that the microsecond fraction of "now" was more than the microsecond
4940 fraction of "then" (but the whole seconds number was less).
4942 38. Added support for the libradius Radius client library this is found on
4943 FreeBSD (previously only the radiusclient library was supported).
4949 1. When certain lookups returned multiple values in the form name=value, the
4950 quoting of the values was not always being done properly. Specifically:
4951 (a) If the value started with a double quote, but contained no whitespace,
4953 (b) If the value contained whitespace other than a space character (i.e.
4954 tabs or newlines or carriage returns) it was not quoted.
4955 This fix has been applied to the mysql and pgsql lookups by writing a
4956 separate quoting function and calling it from the lookup code. The fix
4957 should probably also be applied to nisplus, ibase and oracle lookups, but
4958 since I cannot test any of those, I have not disturbed their existing code.
4960 2. A hit in the callout cache for a specific address caused a log line with no
4961 reason for rejecting RCPT. Now it says "Previous (cached) callout
4962 verification failure".
4964 3. There was an off-by-one bug in the queryprogram router. An over-long
4965 return line was truncated at 256 instead of 255 characters, thereby
4966 overflowing its buffer with the terminating zero. As well as fixing this, I
4967 have increased the buffer size to 1024 (and made a note to document this).
4969 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4970 when Exim is waiting for an SMTP response from a remote server, Exim
4971 restarts its select() call on the socket, thereby resetting its timeout.
4972 This is not a problem when such interrupts are rare. Somebody set up a cron
4973 job to run exiwhat every 2 minutes, which is less than the normal select()
4974 timeout (5 or 10 minutes). This meant that the select() timeout never
4975 kicked in because it was always reset. I have fixed this by comparing the
4976 time when an interrupt arrives with the time at the start of the first call
4977 to select(). If more time than the timeout has elapsed, the interrupt is
4978 treated as a timeout.
4980 5. Some internal re-factoring in preparation for the addition of Sieve
4981 extensions (by MH). In particular, the "personal" test is moved to a
4982 separate function, and given an option for scanning Cc: and Bcc: (which is
4983 not set for Exim filters).
4985 6. When Exim created an email address using the login of the caller as the
4986 local part (e.g. when creating a From: or Sender: header line), it was not
4987 quoting the local part when it contained special characters such as @.
4989 7. Installed new OpenBSD configuration files.
4991 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4992 try to make them clearer.
4994 9. Callout options, other than the timeout value, were being ignored when
4995 verifying sender addresses in header lines. For example, when using
4997 verify = header_sender/callout=no_cache
4999 the cache was (incorrectly) being used.
5001 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5002 only the Cygwin environment.
5004 11. When return_path_on_delivery was set as a log selector, if different remote
5005 addresses in the same message used different return paths and parallel
5006 remote delivery occurred, the wrong values would sometimes be logged.
5007 (Whenever a remote delivery process finished, the return path value from
5008 the most recently started remote delivery process was logged.)
5010 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5011 lines when AUTH and/or TLS are in use. This is the "received protocol"
5012 field. Exim used to use "asmtp" for authenticated SMTP, without any
5013 indication (in the protocol name) for TLS use. Now it follows the RFC and
5014 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5015 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5016 names appear in log lines as well as in Received: header lines.
5018 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5019 extensions, and comparison tests, and to fix some bugs.
5021 14. Changes to the "personal" filter test:
5023 (1) The test was buggy in that it was just doing the equivalent of
5024 "contains" tests on header lines. For example, if a user's address was
5025 anne@some.where, the "personal" test would incorrectly be true for
5027 To: susanne@some.where
5029 This test is now done by extracting each address from the header in turn,
5030 and checking the entire address. Other tests that are part of "personal"
5031 are now done using regular expressions (for example, to check local parts
5032 of addresses in From: header lines).
5034 (2) The list of non-personal local parts in From: addresses has been
5035 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5036 taken from the Sieve specification recommendations.
5038 (3) If the message contains any header line starting with "List-" it is
5039 treated as non-personal.
5041 (4) The test for "circular" in the Subject: header line has been removed
5042 because it now seems ill-conceived.
5044 15. Minor typos in src/EDITME comments corrected.
5046 16. Installed latest exipick from John Jetmore.
5048 17. If headers_add on a router specified a text string that was too long for
5049 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5050 of string_sprintf() is now avoided.
5052 18. $message_body_size was not set (it was always zero) when running the DATA
5053 ACL and the local_scan() function.
5055 19. For the "mail" command in an Exim filter, no default was being set for
5056 the once_repeat time, causing a random time value to be used if "once" was
5057 specified. (If the value happened to be <= 0, no repeat happened.) The
5058 default is now 0s, meaning "never repeat". The "vacation" command was OK
5059 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5060 (I found it when inspecting the code).
5062 20. There is now an overall timeout for performing a callout verification. It
5063 defaults to 4 times the callout timeout, which applies to individual SMTP
5064 commands during the callout. The overall timeout applies when there is more
5065 than one host that can be tried. The timeout is checked before trying the
5066 next host. This prevents very long delays if there are a large number of
5067 hosts and all are timing out (e.g. when the network connections are timing
5068 out). The value of the overall timeout can be changed by specifying an
5069 additional sub-option for "callout", called "maxwait". For example:
5071 verify = sender/callout=5s,maxwait=20s
5073 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5074 to the end before writing, but this should make it even safer).
5076 22. Exim was forgetting that it had advertised PIPELINING for the second and
5077 subsequent messages on an SMTP connection. It was also not resetting its
5078 memory on STARTTLS and an internal HELO.
5080 23. When Exim logs an SMTP synchronization error within a session, it now
5081 records whether PIPELINING has been advertised or not.
5083 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5084 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5085 rather than long int.
5087 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5089 26. Added the never_mail option to autoreply.
5095 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5096 crash if the getsockname() call failed; this can happen if a connection is
5097 closed very soon after it is established. The problem was simply in the
5098 order in which certain operations were done, causing Exim to try to write
5099 to the SMTP stream before it had set up the file descriptor. The bug has
5100 been fixed by making things happen in the correct order.
5106 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5107 before the connection was closed, thus losing the rejection response.
5109 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5110 some early Solaris releases, but causes trouble in current releases where
5111 socklen_t is defined.
5113 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5116 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5119 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5120 that releasing the top of it at the end releases what was used for sub-
5121 expansions (unless the block got too big). However, discard this block if
5122 the first thing is a variable or header, so that we can use its block when
5123 it is dynamic (useful for very large $message_headers, for example).
5125 6. Lookups now cache *every* query, not just the most recent. A new, separate
5126 store pool is used for this. It can be recovered when all lookup caches are
5127 flushed. Lookups now release memory at the end of their result strings.
5128 This has involved some general refactoring of the lookup sources.
5130 7. Some code has been added to the store_xxx() functions to reduce the amount
5131 of flapping under certain conditions.
5133 8. log_incoming_interface used to affect only the <= reception log lines. Now
5134 it causes the local interface and port to be added to several more SMTP log
5135 lines, for example "SMTP connection from", and rejection lines.
5137 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5139 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5141 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5142 could be overwritten at the end of the current message (or the start of a
5143 new message if it was set in a HELO ACL). The value is now preserved for
5144 the duration of the SMTP connection.
5146 12. If a transport had a headers_rewrite setting, and a matching header line
5147 contained an unqualified address, that address was qualified, even if it
5148 did not match any rewriting rules. The underlying bug was that the values
5149 of the flags that permit the existence of unqualified sender and recipient
5150 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5151 non-local messages, and by -bnq for local messages) were not being
5152 preserved with the message after it was received.
5154 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5155 "next input=" as part of the text comprising the host identity instead of
5156 the correct text. The code was using the same buffer for two different
5157 strings. However, depending on which order the printing function evaluated
5158 its arguments, the bug did not always show up. Under Linux, for example, my
5159 test suite worked just fine.
5161 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5162 doesn't work with some older versions of Perl. It has been changed to "my",
5163 which in any case is probably the better facility to use.
5165 15. A really picky compiler found some instances of statements for creating
5166 error messages that either had too many or two few arguments for the format
5169 16. The size of the buffer for calls to the DNS resolver has been increased
5170 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5171 for addresses that have a lot of PTR records. This alleviates a problem; it
5172 does not fully solve it.
5174 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5175 buffer now truncates the list and logs the incident, which is the same
5176 action as happens when Exim is looking up a host name and its aliases.
5177 Previously in this situation something unpredictable would happen;
5178 sometimes it was "internal error: store_reset failed".
5180 18. If a server dropped the connection unexpectedly when an Exim client was
5181 using GnuTLS and trying to read a response, the client delivery process
5182 crashed while trying to generate an error log message.
5184 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5185 string, for example:
5187 warn message = H1: something\nH2: something
5189 the text was added as a single header line from Exim's point of view
5190 though it ended up OK in the delivered message. However, searching for the
5191 second and subsequent header lines using $h_h2: did not work. This has been
5192 fixed. Similarly, if a system filter added multiple headers in this way,
5193 the routers could not see them.
5195 20. Expanded the error message when iplsearch is called with an invalid key to
5196 suggest using net-iplsearch in a host list.
5198 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5199 ACLs are no longer actually imposed (and a message to that effect is
5202 22. If a "gecos" field in a passwd entry contained escaped characters, in
5203 particular, if it contained a \" sequence, Exim got it wrong when building
5204 a From: or a Sender: header from that name. A second bug also caused
5205 incorrect handling when an unquoted " was present following a character
5206 that needed quoting.
5208 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5209 was not being matched caselessly.
5211 24. Arranged for all hyphens in the exim.8 source to be escaped with
5214 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5215 properly. Recipient callout cache records were still being keyed to include
5216 the sender, even when use_sender was set false. This led to far more
5217 callouts that were necessary. The sender is no longer included in the key
5218 when use_sender is false.
5220 26. Added "control = submission" modifier to ACLs.
5222 27. Added the ${base62d: operator to decode base 62 numbers.
5224 28. dnsdb lookups can now access SRV records.
5226 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5227 the configuration file.
5229 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5230 (-v) mode. This makes the output for a verbose queue run more intelligible.
5232 31. Added a use_postmaster feature to recipient callouts.
5234 32. Added the $body_zerocount variable, containing the number of binary zero
5235 bytes in the message body.
5237 33. The time of last modification of the "new" subdirectory is now used as the
5238 "mailbox time last read" when there is a quota error for a maildir
5241 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5243 35. Added +ignore_unknown as a special item in host lists.
5245 36. Code for decoding IPv6 addresses in host lists is now included, even if
5246 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5247 address was recognized as an IP address, but was then not correctly decoded
5248 into binary, causing unexpected and incorrect effects when compared with
5255 1. Very minor rewording of debugging text in manualroute to say "list of
5256 hosts" instead of "hostlist".
5258 2. If verify=header_syntax was set, and a header line with an unqualified
5259 address (no domain) and a large number of spaces between the end of the
5260 name and the colon was received, the reception process suffered a buffer
5261 overflow, and (when I tested it) crashed. This was caused by some obsolete
5262 code that should have been removed. The fix is to remove it!
5264 3. When running in the test harness, delay a bit after writing a bounce
5265 message to get a bit more predictability in the log output.
5267 4. Added a call to search_tidyup() just before forking a reception process. In
5268 theory, someone could use a lookup in the expansion of smtp_accept_max_
5269 per_host which, without the tidyup, could leave open a database connection.
5271 5. Added the variables $recipient_data and $sender_data which get set from a
5272 lookup success in an ACL "recipients" or "senders" condition, or a router
5273 "senders" option, similar to $domain_data and $local_part_data.
5275 6. Moved the writing of debug_print from before to after the "senders" test
5278 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5279 problems for message scanning, either using a data ACL, or using
5280 local_scan() because the Received: header was not generated till after they
5281 were called (in order to set the time as the time of reception completion).
5282 I have revised the way this works. The header is now generated after the
5283 body is received, but before the ACL or local_scan() are called. After they
5284 are run, the timestamp in the header is updated.
5290 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5291 before starting a queue runner without re-exec. This happened only when
5292 deliver_drop_privilege was set or when the Exim user was set to root. The
5293 effect of the bug was that timeouts during subsequent deliveries caused
5294 crashes instead of being properly handled. The handler is now left at its
5295 default (and expected) setting.
5297 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5298 message, again when deliver_drop_privilege is set or Exim is run as root.
5299 The bug described in (1) was not present in this case, but the tidying up
5300 of the other signals was missing. I have made the two cases consistent.
5302 3. The ignore_target_hosts setting on a manualroute router was being ignored
5303 for hosts that were looked up using the /MX notation.
5305 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5308 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5309 operated on the sender address. After changing the $sender_address to <>
5310 for the sender address verify, Exim was re-instated it as the original
5311 (before rewriting) address, but remembering that it had rewritten it, so it
5312 wasn't rewriting it again. This bug also had the effect of breaking the
5313 sender address verification caching when the sender address was rewritten.
5315 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5316 This has been changed so that if the ip literal address matches
5317 ignore_target_hosts, the router declines.
5319 7. Added expansion conditions match_domain, match_address, and match_local_
5320 part (NOT match_host).
5322 8. The placeholder for the Received: header didn't have a length field set.
5324 9. Added code to Exim itself and to exim_lock to test for a specific race
5325 condition that could lead to file corruption when using MBX delivery. The
5326 issue is with the lockfile that is created in /tmp. If this file is removed
5327 after a process has opened it but before that process has acquired a lock,
5328 there is the potential for a second process to recreate the file and also
5329 acquire a lock. This could lead to two Exim processes writing to the file
5330 at the same time. The added code performs the same test as UW imapd; it
5331 checks after acquiring the lock that its file descriptor still refers to
5332 the same named file.
5334 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5335 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5338 11. Added the smtp_active_hostname option. If used, this will typically be made
5339 to depend on the incoming interface address. Because $interface_address is
5340 not set up until the daemon has forked a reception process, error responses
5341 that can happen earlier (such as "too many connections") no longer contain
5344 12. If an expansion in a condition on a "warn" statement fails because a lookup
5345 defers, the "warn" statement is abandoned, and the next ACL statement is
5346 processed. Previously this caused the whole ACL to be aborted.
5348 13. Added the iplsearch lookup type.
5350 14. Added ident_timeout as a log selector.
5352 15. Added tls_certificate_verified as a log selector.
5354 16. Added a global option tls_require_ciphers (compare the smtp transport
5355 option of the same name). This controls incoming TLS connections.
5357 17. I finally figured out how to make tls_require_ciphers do a similar thing
5358 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5359 before starting the TLS session.
5361 18. Tabs are now shown as \t in -bP output.
5363 19. If the log selector return_path_on_delivery was set, Exim crashed when
5364 bouncing a message because it had too many Received: header lines.
5366 20. If two routers both had headers_remove settings, and the first one included
5367 a superfluous trailing colon, the final name in the first list and the
5368 first name in the second list were incorrectly joined into one item (with a
5369 colon in the middle).
5375 1. Added -C and -D options to the exinext utility, mainly to make it easier
5376 to include in the automated testing, but these could be helpful when
5377 multiple configurations are in use.
5379 2. The exinext utility was not formatting the output nicely when there was
5380 an alternate port involved in the retry record key, nor when there was a
5381 message id as well (for retries that were specific to a specific message
5382 and a specific host). It was also confused by IPv6 addresses, because of
5383 the additional colons they contain. I have fixed the IPv4 problem, and
5384 patched it up to do a reasonable job for IPv6.
5386 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5387 delivery, the log line now contains "pipelined" if PIPELINING was used.
5389 4. An SMTP transport process used to panic and die if the bind() call to set
5390 an explicit outgoing interface failed. This has been changed; it is now
5391 treated in the same way as a connect() failure.
5393 5. A reference to $sender_host_name in the part of a conditional expansion
5394 that was being skipped was still causing a DNS lookup. This no longer
5397 6. The def: expansion condition was not recognizing references to header lines
5398 that used bh_ and bheader_.
5400 7. Added the _cache feature to named lists.
5402 8. The code for checking quota_filecount in the appendfile transport was
5403 allowing one more file than it should have been.
5405 9. For compatibility with Sendmail, the command line option
5413 and sets the incoming protocol and host name (for trusted callers). The
5414 host name and its colon can be omitted when only the protocol is to be set.
5415 Note the Exim already has two private options, -pd and -ps, that refer to
5416 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5417 "s", but I don't think that's a major issue.
5419 10. A number of refactoring changes to the code, none of which should affect
5422 (a) The number of logging options was getting close to filling up the
5423 32-bit word that was used as a bit map. I have split them into two classes:
5424 those that are passed in the argument to log_write(), and those that are
5425 only ever tested independently outside of that function. These are now in
5426 separate 32-bit words, so there is plenty of room for expansion again.
5427 There is no change in the user interface or the logging behaviour.
5429 (b) When building, for example, log lines, the code previously used a
5430 macro that called string_cat() twice, in order to add two strings. This is
5431 not really sufficiently general. Furthermore, there was one instance where
5432 it was actually wrong because one of the argument was used twice, and in
5433 one call a function was used. (As it happened, calling the function twice
5434 did not affect the overall behaviour.) The macro has been replaced by a
5435 function that can join an arbitrary number of extra strings onto a growing
5438 (c) The code for expansion conditions now uses a table and a binary chop
5439 instead of a serial search (which was left over from when there were very
5440 few conditions). Also, it now recognizes conditions like "pam" even when
5441 the relevant support is not compiled in: a suitably worded error message is
5442 given if an attempt is made to use such a condition.
5444 11. Added ${time_interval:xxxxx}.
5446 12. A bug was causing one of the ddress fields not to be passed back correctly
5447 from remote delivery subprocesses. The field in question was not being
5448 subsequently used, so this caused to problems in practice.
5450 13. Added new log selectors queue_time and deliver_time.
5452 14. Might have fixed a bug in maildirsizefile handling that threw up
5453 "unexpected character" debug warnings, and recalculated the data
5454 unnecessarily. In any case, I expanded the warning message to give more
5457 15. Added the message "Restricted characters in address" to the statements in
5458 the default ACL that block characters like @ and % in local parts.
5460 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5461 Three changes have been made:
5463 (a) There was a serious bug; a negative response to MAIL caused the whole
5464 recipient domain to be cached as invalid, thereby blocking all messages
5465 to all local parts at the same domain, from all senders. This bug has
5466 been fixed. The domain is no longer cached after a negative response to
5467 MAIL if the sender used is not empty.
5469 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5472 (c) A new callout option, "use_sender" has been added for people who want
5473 the modified behaviour.
5479 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5482 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5483 indeed breaks things for older releases.
5485 3. Added additional logging to the case where there is a problem reading data
5486 from a filter that is running in a subprocess using a pipe, in order to
5487 try to track down a specific problem.
5489 4. Testing facility fudge: when running in the test harness and attempting
5490 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5491 getting "No route to host". Convert this to a timeout.
5493 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5496 6. Some OS don't have socklen_t but use size_t instead. This affects the
5497 fifth argument of getsockopt() amongst other things. This is now
5498 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5499 can be set for individual OS. I have set it for SunOS5, OSF1, and
5500 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5501 some earlier ones do not.
5503 7. Change 4.30/15 was not doing the test caselessly.
5505 8. The standard form for an IPv6 address literal was being rejected by address
5506 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5507 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5508 this, as well as the form without the "IPv6" on the front (but only when
5509 address literals are enabled, of course).
5511 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5513 10. Exim crashed if a message with an empty sender address specified by -f
5514 encountered a router with an errors_to setting. This could be provoked only
5515 by a command such as
5519 where an empty string was supplied; "<>" did not hit this bug.
5521 11. Installed PCRE release 4.5.
5523 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5524 remained set. It is now erased.
5526 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5527 times from message ids (which are base 36 rather than the normal 62).
5529 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5530 were being counted as actual protocol errors, and logged if the log
5531 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5532 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5533 and DATA following a set of rejected RCPTs do not count as protocol errors.
5534 In other words, Exim assumes they were pipelined, though this may not
5535 actually be the case. Of course, in all cases the client gets an
5536 appropriate error code.
5538 15. If a lookup fails in an ACL condition, a message about the failure may
5539 be available; it is used if testing the ACL cannot continue, because most
5540 such messages specify what the cause of the deferral is. However, some
5541 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5542 that caused an old message to be retained and used if a later statement
5543 caused a defer, replacing the real cause of the deferral.
5545 16. If an IP address had so many PTR records that the DNS lookup buffer
5546 was not large enough to hold them, Exim could crash while trying to process
5547 the truncated data. It now detects and logs this case.
5549 17. Further to 4.21/58, another change has been made: if (and only if) the
5550 first line of a message (the first header line) ends with CRLF, a bare LF
5551 in a subsequent header line has a space inserted after it, so as not to
5552 terminate the header.
5554 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5555 unnecessarily, used atoi() instead of strtol(), and didn't check the
5556 termination when getting file sizes from file names by regex.
5558 19. Completely re-implemented the support for maildirsize files, in the light
5559 of a number of problems with the previous contributed implementation
5560 (4.30/29). In particular:
5562 . If the quota is zero, the maildirsize file is maintained, but no quota is
5565 . If the maildir directory does not exist, it is created before any attempt
5566 to write a maildirsize file.
5568 . The quota value in the file is just a cache; if the quota is changed in
5569 the transport, the new value overrides.
5571 . A regular expression is available for excluding directories from the
5574 20. The autoreply transport checks the characters in options that define the
5575 message's headers; it allows continued headers, but it was checking with
5576 isspace() after an embedded newline instead of explicitly looking for a
5579 21. If all the "regular" hosts to which an address was routed had passed their
5580 expiry times, and had not reached their retry times, the address was
5581 bounced, even if fallback hosts were defined. Now Exim should go on to try
5584 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5585 equivalent code in the SMTP transport. Some hosts send humungous responses
5586 to HELO/EHLO, more than 1024 it seems.
5588 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5589 gives compiler warnings in some environments. I've now done it "properly",
5592 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5593 (because of problems with the built-in one) was declared to return uschar *
5594 instead of char *, causing compiler failure.
5596 25. Fixed a file descriptor leak when processing alias/forward files.
5598 26. Fixed a minor format string issue in dbfn.c.
5600 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5602 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5603 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5606 29. When a sender address is verified, it is cached, to save repeating the test
5607 when there is more than one recipient in a message. However, when the
5608 verification involves a callout, it is possible for different callout
5609 options to be set for different recipients. It is too complicated to keep
5610 track of this in the cache, so now Exim always runs a verification when a
5611 callout is required, relying on the callout cache for the optimization.
5612 The overhead is duplication of the address routing, but this should not be
5615 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5616 to be verified with callout=postmaster, and the main callout worked but the
5617 postmaster check failed, the verification correctly failed. However, if a
5618 subsequent RCPT command asked for sender verification *without* the
5619 postmaster check, incorrect caching caused this verification also to fail,
5622 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5623 it was not caching the DNS options (qualify_single, search_parents) that
5624 were used when the lookup failed. A subsequent lookup with different
5625 options therefore always gave the same answer, though there were cases
5626 where it should not have. (Example: a "domains = !$mx_any" option on a
5627 dnslookup router: the "domains" option is always processed without any
5628 widening, but the router might have qualify_single set.) Now Exim uses the
5629 cached value only when the same options are set.
5631 32. Added John Jetmore's "exipick" utility to the distribution.
5633 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5634 than a timeout (e.g. a certificate is required, and is not provided), an
5635 Exim server now closes the connection immediately. Previously it waited for
5636 the client to close - but if the client is SSL, it seems that they each
5637 wait for each other, leading to a delay before one of them times out.
5639 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5640 maintained 0.8.x compatibility because I don't think many are using it, and
5641 it is clearly obsolete.
5643 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5646 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5647 client certificate was expired. A simple patch fixes this, though I don't
5648 understand the full logic of why the verify callback is called multiple
5651 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5652 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5653 which causes problems with some clients (such as the Certicom SSL Plus
5654 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5655 disables the coutermeasure allowing Eudora to connect."
5657 38. Exim was not checking that a write() to a log file succeeded. This could
5658 lead to Bad Things if a log got too big, in particular if it hit a file
5659 size limit. Exim now panics and dies if it cannot write to a log file, just
5660 as it does if it cannot open a log file.
5662 39. Modified OS/Makefile-Linux so that it now contains
5664 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5666 The two -D definitions ensure that Exim is compiled with large file
5667 support, which makes it possible to handle log files that are bigger than
5670 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5671 instance) a domain was checked against a named list that involved a lookup,
5672 causing $domain_data to be set, then another domain was checked against the
5673 same list, then the first domain was re-checked, the value of $domain_data
5674 after the final check could be wrong. In particular, if the second check
5675 failed, it could be set empty. This bug probably also applied to
5678 41. The strip_trailing_dot option was not being applied to the address given
5679 with the -f command-line option.
5681 42. The code for reading a message's header from the spool was incrementing
5682 $received_count, but never initializing it. This meant that the value was
5683 incorrect (doubled) while delivering a message in the same process in which
5684 it was received. In the most common configuration of Exim, this never
5685 happens - a fresh exec is done - but it can happen when
5686 deliver_drop_privilege is set.
5688 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5689 it now includes up to 150 characters of the unexpected data in the log
5692 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5693 and building data strings. The size of both of these buffers was 10 000
5694 bytes - far larger than anybody would *ever* want, thought I. Needless to
5695 say, somebody hit the limit. I have increased the maximum line length to
5696 20 000 and the maximum data length of concatenated lines to 100 000. I have
5697 also fixed two bugs, because there was no checking on these buffers. Tsk,
5698 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5699 buffer is too small.
5701 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5702 lsearch lookups. Now it does.
5704 46. When parsing a route_list item in a manualroute router, a fixed-length
5705 buffer was used for the list of hosts. I made this 1024 bytes long,
5706 thinking that nobody would ever have a list of hosts that long. Wrong.
5707 Somebody had a whole pile of complicated expansion conditions, and the
5708 string was silently truncated, leading to an expansion error. It turns out
5709 that it is easier to change to an unlimited length (owing to other changes
5710 that have happened since this code was originally written) than to build
5711 structure for giving a limitation error. The length of the item that
5712 expands into the list of hosts is now unlimited.
5714 47. The lsearch lookup could not handle data where the length of text line was
5715 more than 4095 characters. Such lines were truncated, leading to shortened
5716 data being returned. It should now handle lines of any length.
5718 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5719 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5722 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5723 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5724 no longer happens in comment lines. A list of replacements is now placed
5725 at the head of all of the source files, except those whose only change is
5726 to replace PERL_COMMAND in the very first #! line.
5728 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5729 messages on the queue, with a bottom-up merge sort, using code contributed
5730 by Michael Haardt. This should make operations like -bp somewhat faster on
5731 large queues. It won't affect queue runners, except when queue_run_in_order
5734 51. Installed eximstats 1.31 in the distribution.
5736 52. Added support for SRV lookups to the dnslookup router.
5738 53. If an ACL referred to $message_body or $message_body_end, the value was not
5739 reset for any messages that followed in the same SMTP session.
5741 54. The store-handling optimization for building very long strings was not
5742 differentiating between the different store pools. I don't think this
5743 actually made any difference in practice, but I've tidied it.
5745 55. While running the routers to verify a sender address, $sender_address
5746 was still set to the sender address. This is wrong, because when routing to
5747 send a bounce to the sender, it would be empty. Therefore, I have changed
5748 it so that, while verifying a sender address, $sender_address is set to <>.
5749 (There is no change to what happens when verifying a recipient address.)
5751 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5752 target A or AAAA records (if not already returned) without resetting the
5753 qualify_single or search_parents options of the DNS resolver. These are
5754 inappropriate in this case because the targets of MX and SRV records must
5755 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5756 target that, when qualified, matched something in the local domain. These
5757 two options are now turned off when doing these lookups.
5759 57. It seems that at least some releases of Reiserfs (which does not have the
5760 concept of a fixed number of inodes) returns zero and not -1 for the
5761 number of available inodes. This interacted badly with check_spool_inodes,
5762 which assumed that -1 was the "no such thing" setting. What I have done is
5763 to check that the total number of inodes is greater than zero before doing
5764 the test of how many are available.
5766 58. When a "warn" ACL statement has a log_message modifier, the message is
5767 remembered, and not repeated. This is to avoid a lot of repetition when a
5768 message has many recipients that cause the same warning to be written.
5769 However, Exim was preserving the list of already written lines for an
5770 entire SMTP session, which doesn't seem right. The memory is now reset if a
5771 new message is started.
5773 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5774 debugging output unless log_rewrite was also set.
5776 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5777 of (int)(handle) when we know that handle contains (void *)(-1).
5779 61. The Exim daemon panic-logs an error return when it closes the incoming
5780 connection. However "connection reset by peer" seems to be common, and
5781 isn't really an error worthy of noting specially, so that particular error
5784 62. When Exim is trying to find all the local interfaces, it used to panic and
5785 die if the ioctl to get the interface flags failed. However, it seems that
5786 on at least one OS (Solaris 9) it is possible to have an interface that is
5787 included in the list of interfaces, but for which you get a failure error
5788 for this call. This happens when the interface is not "plumbed" into a
5789 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5790 failure of the "get flags" call assumes that the interface is down.
5792 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5793 makes life easier for people who are doing arithmetic on fields extracted
5794 from dates, where you often get leading zeros that should not be
5795 interpreted as octal.
5797 64. Added qualify_domain to the redirect router, to override the global
5800 65. If a pathologically long header line contained very many addresses (the
5801 report of this problem mentioned 10 000) and each of them was rewritten,
5802 Exim could use up a very large amount of memory. (It kept on making new
5803 copies of the header line as it rewrote, and never released the old ones.)
5804 At the expense of a bit more processing, the header rewriting function has
5805 been changed so that it no longer eats memory in this way.
5807 66. The generation of the Received: header has been moved from the time that a
5808 message starts to be received, to the time that it finishes. The timestamp
5809 in the Received: header should now be very close to that of the <= log
5810 line. There are two side-effects of this change:
5812 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5813 logged header lines no longer include the local Received: line, because
5814 it has not yet been created. The same applies to a copy of the message
5815 that is returned to a non-SMTP sender when a message is rejected.
5817 (b) When a filter file is tested using -bf, no additional Received: header
5818 is added to the test message. After some thought, I decided that this
5821 This change does not affect the value of $received_for. It is still set
5822 after address rewriting, but before local_scan() is called.
5824 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5826 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5827 gave an unhelpful panic error message, and a defer error. I have managed to
5828 change this behaviour so that it now rejects any supplied certificate,
5829 which seems right, as the list of acceptable certificates is empty.
5831 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5832 gave an unhelpful defer error. I have not managed to make this reject any
5833 supplied certificates, but the error message it gives is "no certificate
5834 supplied", which is not helpful.
5836 70. exigrep's output now also includes lines that are not associated with any
5837 message, but which match the given pattern. Implemented by a patch from
5838 Martin Sluka, which also tidied up the Perl a bit.
5840 71. Recipient callout verification, like sender verification, was using <> in
5841 the MAIL FROM command. This isn't really the right thing, since the actual
5842 sender may affect whether the remote host accepts the recipient or not. I
5843 have changed it to use the actual sender in the callout; this means that
5844 the cache record is now keyed on a recipient/sender pair, not just the
5845 recipient address. There doesn't seem to be a real danger of callout loops,
5846 since a callout by the remote host to check the sender would use <>.
5847 [SEE ABOVE: changed after hitting problems.]
5849 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5850 temporary errors. However, in the case of such a code being given after
5851 the end of a data transmission (i.e. after ".") Exim was failing to write
5852 a retry record for the message. (Yes, there was some broken host that was
5853 actually sending 8xx at this point.)
5855 73. An unknown lookup type in a host list could cause Exim to panic-die when
5856 the list was checked. (An example that provoked this was putting <; in the
5857 middle of a list instead of at the start.) If this happened during a DATA
5858 ACL check, a -D file could be left lying around. This kind of configuration
5859 error no longer causes Exim to die; instead it causes a defer error. The
5860 incident is still logged to the main and panic logs.
5862 74. Buglet left over from Exim 3 conversion. The message "too many messages
5863 in one connection" was written to the rejectlog but not the mainlog, except
5864 when address rewriting (yes!) was being logged.
5866 75. Added write_rejectlog option.
5868 76. When a system filter was run not as root (that is, when system_filter_user
5869 was set), the values of the $n variables were not being returned to the
5870 main process; thus, they were not subsequently available in the $sn
5873 77. Added +return_path_on_delivery log selector.
5875 78. A connection timeout was being treated differently from recipients deferred
5876 when testing hosts_max_try with a message that was older than the host's
5877 retry timeout. (The host should not be counted, thus allowing all hosts to
5878 be tried at least once before bouncing.) This may have been the cause of an
5879 occasionally reported bug whereby a message would remain on the queue
5880 longer than the retry timeout, but would be bounced if a delivery was
5881 forced. I say "may" because I never totally pinned down the problem;
5882 setting up timeout/retry tests is difficult. See also the next item.
5884 79. The ultimate address timeout was not being applied to errors that involved
5885 a combination of host plus message (for example, a timeout on a MAIL
5886 command). When an address resolved to a number of possible hosts, and they
5887 were not all tried for each delivery (e.g. because of hosts_max_try), a
5888 message could remain on the queue longer than the retry timeout.
5890 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5893 81. Fixed an obscure SMTP outgoing bug which required at least the following
5894 conditions: (a) there was another message waiting for the same server;
5895 (b) the server returned 5xx to all RCPT commands in the first message so
5896 that the message was not completed; (c) the server dropped the connection
5897 or gave a negative response to the RSET that Exim sends to abort the
5898 transaction. The observed case was a dropped connection after DATA that had
5899 been sent in pipelining mode. That is, the server had advertised PIPELINING
5900 but was not implementing it correctly. The effect of the bug was incorrect
5901 behaviour, such as trying another host, and this could lead to a crash.
5907 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5908 and daemon.c were passed as pointers to ints; they should have been
5909 pointers to socklen_t variables (which are typically unsigned ints).
5911 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5914 3. Fixed a really odd bug that affected only the testing scheme; patching a
5915 certain fixed string in the binary changed the value of another string that
5916 happened to be identical to the end of the original first string.
5918 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5919 name", it returns that address as the IP address. On some operating
5920 systems (e.g. Solaris), it also passes back the IP address string as the
5921 "host name". However, on others (e.g. Linux), it passes back an empty
5922 string. Exim wasn't checking for this, and was changing the host name to an
5923 empty string, assuming it had been canonicalized.
5925 5. Although rare, it is permitted to have more than one PTR record for a given
5926 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5927 all the names associated with an address, because they do in Solaris.
5928 However, it seems that they do not in Linux for data that comes from the
5929 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5930 I found this out when I moved to a new Linux workstation and tried to run
5931 the Exim test suite.
5933 To get round this problem I have changed the code so that it now does its
5934 own call to the DNS to look up PTR records when searching for a host name.
5935 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5936 addresses that are only in /etc/hosts are still found.
5938 This behaviour is, however, controlled by an option called host_lookup_
5939 order, which defaults to "bydns:byaddr". If people want to use the other
5940 order, or indeed, just use one or the other means of lookup, they can
5941 specify it in this variable.
5943 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5944 some operating systems, this comes back from gethostbyaddr() as an empty
5945 string, and this is what Exim used to test for. However, it seems that in
5946 other systems, "." is yielded. Exim now tests for this case too.
5948 7. The values of check_spool_space and check_log_space are now held internally
5949 as a number of kilobytes instead of an absolute number of bytes. If a
5950 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5951 kilobyte. This means that much larger values can be stored.
5953 8. Exim monitor: an attempt to get the action menu when not actually pointing
5954 at a message produces an empty menu entitled "No message selected". This
5955 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5956 no entries in it ("Shell widget menu has zero width and/or height"). So I
5957 have added a single, blank menu entry in this case.
5959 9. Added ${quote_local_part.
5961 10. MIME decoding is now applied to the contents of Subject: header lines when
5964 11. Now that a reference to $sender_host_address automatically causes a reverse
5965 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5966 host lookup before query-style lookups in lists that might use this
5967 variable. This has therefore been abolished, and the "net-" prefix is no
5968 longer necessary for query-style lookups.
5970 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5971 have been a typo for LFLAGS, so it has been changed.
5973 13. The install script calls Exim with "-C /dev/null" in order to find the
5974 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5975 to be output. However, since Exim outputs its version number before the
5976 error, it didn't break the script. It just looked ugly. I fixed this by
5977 always allowing "-C /dev/null" if the caller is root.
5979 14. Ignore overlarge ACL variable number when reading spool file - insurance
5980 against a later release with more variables having written the file.
5982 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5983 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5984 this, as well as the form without the "IPv6" on the front.
5986 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5987 OS/Makefile-Darwin file.
5989 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5990 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5992 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5993 variable. However, some OS use ints for inodes. I've added cast to long int
5994 to get rid of the compiler warning.
5996 19. I had forgotten to lock out "/../" in configuration file names when
5997 ALT_CONFIG_PREFIX was set.
5999 20. Routers used for verification do not need to specify transports. However,
6000 if such a router generated a host list, and callout was configured, Exim
6001 crashed, because it could not find a port number from the (non-existent)
6002 transport. It now assumes port 25 in this circumstance.
6004 21. Added the -t option to exigrep.
6006 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6007 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6008 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6009 them is set, LOOKUP_LSEARCH is forced.
6011 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6013 24. Added sender and host information to the "rejected by local_scan()" log
6014 line; previously there was no indication of these.
6016 25. Added .include_if_exists.
6018 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6019 Linux. It turns out that not all file systems support this. Apparently some
6020 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6021 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6022 not supported on the file descriptor, is now ignored when Exim is trying to
6023 sync a directory. This applies only to Linux.
6025 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6027 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6028 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6031 29. Added support for maildirsize files from supplied patch (modified a bit).
6033 30. The use of :fail: followed by an empty string could lead Exim to respond to
6034 sender verification failures with (e.g.):
6036 550 Verification failed for <xxx>
6037 550 Sender verify failed
6039 where the first response line was missing the '-' that indicates it is not
6040 the final line of the response.
6042 31. The loop for finding the name of the user that called Exim had a hardwired
6043 limit of 10; it now uses the value of finduser_retries, which is used for
6044 all other user lookups.
6046 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6049 33. Exim was neglecting to zero errno before one call of strtol() when
6050 expanding a string and expecting an integer value. On some systems this
6051 resulted in spurious "integer overflow" errors. Also, it was casting the
6052 result into an int without checking.
6054 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6055 did not work. The code looks as if it has *never* worked, though it appears
6056 to have been documented since at least release 1.62. I have made it work.
6058 35. The "timeout_DNS" error in retry rules, also documented since at least
6059 1.62, also never worked. As it isn't clear exactly what this means, and
6060 clearly it isn't a major issue, I have abolished the feature by treating it
6061 as "timeout", and writing a warning to the main and panic logs.
6063 36. The display of retry rules for -brt wasn't always showing the error code
6066 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6067 timeout_connect_A, timeout_connect_MX.
6069 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6070 to the empty sender.
6072 39. The daemon was not analysing the content of -oX till after it had closed
6073 stderr and disconnected from the controlling terminal. This meant that any
6074 syntax errors were only noted on the panic log, and the return code from
6075 the command was 0. By re-arranging the code a little, I've made the
6076 decoding happen first, so such errors now appear on stderr, and the return
6077 code is 1. However, the actual setting up of the sockets still happens in
6078 the disconnected process, so errors there are still only recorded on the
6081 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6082 connections (as happens on some IP stacks) was logged at start up time as
6083 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6084 it from "IPv6 and IPv4", which means that two separate sockets are being
6087 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6088 says whether AF_INET or AF_INET6 was passed as an argument.
6090 42. Exiwhat output was messed up when time zones were included in log
6093 43. Exiwhat now gives more information about the daemon's listening ports,
6094 and whether -tls-on-connect was used.
6096 44. The "port" option of the smtp transport is now expanded.
6098 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6099 silently ignored. Now an error message is written to the main and panic
6102 46. There's a new ACL modifier called "logwrite" which writes to a log file
6103 as soon as it is encountered.
6105 47. Added $local_user_uid and $local_user_gid at routing time.
6107 48. Exim crashed when trying to verify a sender address that was being
6110 49. Exim was recognizing only a space character after ".include". It now also
6111 recognizes a tab character.
6113 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6114 extracting the relevant information from the specification. The man page no
6115 longer contains scrambled data for the -d option, and I've added a section
6116 at the front about calling Exim under different names.
6118 51. Added "extra_headers" argument to the "mail" command in filter files.
6120 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6123 53. Installed eximstats 1.29.
6125 54. Added transport_filter_timeout as a generic transport option.
6127 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6128 Cc: header lines. This was required by RFC 822, but it not required by RFC
6131 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6132 incoming messages that did not have them. Now it does so only if the
6133 message originates locally, that is, if there is no associated remote host
6134 address. When Resent- header lines are present, this applies to the Resent-
6135 lines rather than the non-Resent- lines.
6137 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6138 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6140 58. Messages for configuration errors now include the name of the main
6141 configuration file - useful now that there may be more than one file in a
6142 list (.included file names were always shown).
6144 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6145 for those rare installations that do not start the daemon as root or run it
6146 setuid root. I've cut out the call to initgroups() if the daemon is not
6149 60. The Exim user and group can now be bound into the binary as text strings
6150 that are looked up at the start of Exim's processing.
6152 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6154 62. Added $mailstore_basename variable.
6156 63. Installed patch to sieve.c from Michael Haardt.
6158 64. When Exim failed to open the panic log after failing to open the main log,
6159 the original message it was trying to log was written to stderr and debug
6160 output, but if they were not available (the usual case in production), it
6161 was lost. Now it is written to syslog before the two lines that record the
6162 failures to open the logs.
6164 65. Users' Exim filters run in subprocesses under the user's uid. It is
6165 possible for a "deliver" command or an alias in a "personal" command to
6166 provoke an address rewrite. If logging of address rewriting is configured,
6167 this fails because the process is not running as root or exim. There may be
6168 a better way of dealing with this, but for the moment (because 4.30 needs
6169 to be released), I have disabled address rewrite logging when running a
6170 filter in a non-root, non-exim process.
6176 1. The buildconfig auxiliary program wasn't quoting the value set for
6177 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6178 not defined. This bug was masked in 4.22 by the effect that was fixed in
6181 2. Some messages that were rejected after a message id was allocated were
6182 shown as "incomplete" by exigrep. It no longer does this for messages that
6183 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6185 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6186 have allow_domain_literals set, the ID did not get logged in the <= line.
6187 Domain literals are now always recognized in Message-ID: header lines.
6189 4. The first argument for a ${extract expansion item is the key name or field
6190 number. Leading and trailing spaces in this item were not being ignored,
6191 causing some misleading effects.
6193 5. When deliver_drop_privilege was set, single queue runner processes started
6194 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6195 same command in the process it spins off) were not dropping privilege.
6197 6. When the daemon running as "exim" started a queue runner, it always
6198 re-executed Exim in the spun-off process. This is a waste of effort when
6199 deliver_drop_privilege is set. The new process now just calls the
6200 queue-runner function directly.
6206 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6209 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6210 ignored. Though the use of -oP was forcing the writing of a pid file, it
6211 was always written to the default place.
6213 3. If the message "no IP address found for host xxxx" is generated during
6214 incoming verification, it is now followed by identification of the incoming
6215 connection (so you can more easily find what provoked it).
6217 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6219 5. Added some features to "harden" Exim a bit more against certain attacks:
6221 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6222 be put in Local/Makefile. This is like the never_users runtime option,
6223 but it cannot be overridden. The default setting is "root".
6225 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6226 prefix string with which any file named in a -C command line option
6229 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6230 is retained for -C and -D only if the caller of Exim is root. Without
6231 it, the exim user may also use -C and -D and retain privilege.
6233 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6234 command line option is disabled.
6236 6. Macro names set by the -D option must start with an upper case letter, just
6237 like macro names defined in the configuration file.
6239 7. Added "dereference=" facility to LDAP.
6241 8. Two instances of the typo "uknown" in the source files are fixed.
6243 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6244 the Configure-Makefile script screwed up while processing it.
6246 10. Incorporated PCRE 4.4.
6248 11. The SMTP synchronization check was not operating right at the start of an
6249 SMTP session. For example, it could not catch a HELO sent before the client
6250 waited for the greeting. There is now a check for outstanding input at the
6251 point when the greeting is written. Because of the duplex, asynchronous
6252 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6253 way, but not yet received, when the check is performed.
6255 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6256 on TCP/IP sockets, because this apparently causes some broken clients to
6259 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6260 unchanged) from the Cygwin maintainer.
6262 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6263 was supported instead of testing for mbx. Effectively a typo.
6265 15. The spa authenticator server code was not checking that the input it
6266 received was valid base64.
6268 16. The debug output line for the "set" modifier in ACLs was not showing the
6269 name of the variable that was being set.
6271 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6273 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6274 lookup on its own. Something else was needed to trigger the lookup. For
6275 example, a match in host_lookup or the need for a host name in a host list.
6276 Now, if $sender_host_name is referenced and the host name has not yet been
6277 looked up, a lookup is performed. If the lookup fails, the variable remains
6278 empty, and $host_lookup_failed is set to "1".
6280 19. Added "eqi" as a case-independent comparison operator.
6282 20. The saslauthd authentication condition could segfault if neither service
6283 nor realm was specified.
6285 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6286 error message that was logged was misleading, and incoming SMTP
6287 connections were dropped. The message is now more accurate, and temporary
6288 errors are given to SMTP connections.
6290 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6291 (except RSET) with a 421 error, until QUIT is received. However, it was
6292 failing to send a response to QUIT.
6294 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6295 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6296 ACL is now run after the helo_try_verify_hosts code.
6298 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6299 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6300 case-independent, so other case variants are also recognized). Apparently
6301 some systems use these upper case variants.
6303 25. If more than two messages were waiting for the same host, and a transport
6304 filter was specified for the transport, Exim sent two messages over the
6305 same TCP/IP connection, and then failed with "socket operation on non-
6306 socket" when it tried to send the third.
6308 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6310 27. The extern definition of crypt16() in expand.c was not being excluded when
6311 the OS had its own crypt16() function.
6313 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6314 as a preferred synonym for return_size_limit, both as an option and as an
6317 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6319 30. Changed the default configuration ACL to relax the local part checking rule
6320 for addresses that are not in any local domains. For these addresses,
6321 slashes and pipe symbols are allowed within local parts, but the sequence
6322 /../ is explicitly forbidden.
6324 31. SPA server authentication was not clearing the challenge buffer before
6327 32. log_message in a "warn" ACL statement was writing to the reject log as
6328 well as to the main log, which contradicts the documentation and doesn't
6329 seem right (because no rejection is happening). So I have stopped it.
6331 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6332 However, I am unable to do any testing of this.
6334 34. Fixed an infelicity in the appendfile transport. When checking directories
6335 for a mailbox, to see if any needed to be created, it was accidentally
6336 using path names with one or more superfluous leading slashes; tracing
6337 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6339 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6340 contents are added to the log line that is written for every discarded
6341 recipient. (Previously a log_message setting was ignored.)
6343 36. The ${quote: operator now quotes the string if it is empty.
6345 37. The install script runs exim in order to find its version number. If for
6346 some reason other than non-existence or emptiness, which it checks, it
6347 could not run './exim', it was installing it with an empty version number,
6348 i.e. as "exim-". This error state is now caught, and the installation is
6351 38. An argument was missing from the function that creates an error message
6352 when Exim fails to connect to the socket for saslauthd authentication.
6353 This could cause Exim to crash, or give a corrupted message.
6355 39. Added isip, isip4, and isip6 to ${if conditions.
6357 40. The ACL variables $acl_xx are now saved with the message, and can be
6358 accessed later in routers, transports, and filters.
6360 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6361 strings in the file are not string-expanded.
6363 42. If a MAIL command specified a SIZE value that was too large to fit into an
6364 int variable, the check against message_size_limit failed. Such values are
6365 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6366 day this will have to be increased, but I don't think I want to be around
6367 when emails are that large.
6374 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6375 iconv() is not standard in FreeBSD.
6377 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6378 IPv6 enabled. The observed symptom was a segmentation fault on return from
6379 the function os_common_find_running_interfaces() in src/os.c.
6381 3. In the check_special_case() function in daemon.c I had used "errno" as an
6382 argument name, which causes warnings on some systems. This was basically a
6383 typo, since it was named "eno" in the comments!
6385 4. The code that waits for the clock to tick (at a resolution of some fraction
6386 of a second) so as to ensure message-id uniqueness was always waiting for
6387 at least one whole tick, when it could have waited for less. [This is
6388 almost certainly not relevant at current processor speeds, where it is
6389 unlikely to ever wait at all. But we try to future-proof.]
6391 5. The function that sleeps for a time interval that includes fractions of a
6392 second contained a race. It did not block SIGALRM between setting the
6393 timer, and suspending (a couple of lines later). If the interval was short
6394 and the sigsuspend() was delayed until after it had expired, the suspension
6395 never ended. On busy systems this could lead to processes getting stuck for
6398 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6399 process, before it forks any delivery processes. The open lookup caching
6400 mechanism meant that the open file or database connection was passed into
6401 the delivery process. The problem was that delivery processes always tidy
6402 up cached lookup data. This could cause a problem for the next delivery
6403 process started by the queue runner, because the external queue runner
6404 process does not know about the closure. So the next delivery process
6405 still has data in the lookup cache. In the case of a file lookup, there was
6406 no problem because closing a file descriptor in a subprocess doesn't affect
6407 the parent. However, if the lookup was caching a connection to a database,
6408 the connection was closed, and the second delivery process was likely to
6409 see errors such as "PGSQL: query failed: server closed the connection
6410 unexpectedly". The problem has been fixed by closing all cached lookups
6411 in a queue runner before running a delivery process.
6413 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6414 seem to have the "const" qualifier which it has on other OS. I've
6417 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6418 *of the same type* (client or server) with the same public name that an
6419 error should be diagnosed.
6421 9. When Exim looked up a host name for an IP address, but failed to find the
6422 original IP address when looking up the host name (a safety check), it
6423 output the message "<ip address> does not match any IP for NULL", which was
6424 confusing, to say the least. The bug was that the host name should have
6425 appeared instead of "NULL".
6427 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6428 user that is built into the binary, and the -C or -D options is used, root
6429 privilege is dropped before the configuration file is read. In addition,
6430 logging is switched to stderr instead of the normal log files. If the
6431 configuration then re-defines the Exim user, the unprivileged environment
6432 is probably not what is expected, so Exim logs a panic warning message (but
6435 However, if deliver_drop_privilege is set, the unprivileged state may well
6436 be exactly what is intended, so the warning has been cut out in that case,
6437 and Exim is allowed to try to write to its normal log files.
6443 1. smtp_return_error_details was not giving details for temporary sender
6444 or receiver verification errors.
6446 2. Diagnose a configuration error if two authenticators have the same public
6449 3. Exim used not to create the message log file for a message until the first
6450 delivery attempt. This could be confusing when incoming messages were held
6451 for policy or load reasons. The message log file is now created at the time
6452 the message is received, and an initial "Received" line is written to it.
6454 4. The automatically generated man page for command line options had a minor
6455 bug that caused no ill effects; however, a more serious problem was that
6456 the procedure for building the man page automatically didn't always
6457 operate. Consequently, release 4.20 contains an out-of-date version. This
6458 shouldn't happen again.
6460 5. When building Exim with embedded Perl support, the script that builds the
6461 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6462 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6464 6. The freeze_tell option was not being used for messages that were frozen on
6465 arrival, either by an ACL or by local_scan().
6467 7. Added the smtp_incomplete_transaction log selector.
6469 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6470 was accepting AUTH without a new EHLO.
6472 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6473 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6476 10. Logging of TCP/IP connections (when configured) now happens in the main
6477 daemon process instead of the child process, so that the TCP/IP connection
6478 count is more accurate (but it can never be perfect).
6480 11. The use of "drop" in a nested ACL was not being handled correctly in the
6481 outer ACL. Now, if condition failure induced by the nested "drop" causes
6482 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6483 or "require"), the connection is dropped.
6485 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6486 that yield "discard" can now be used with an "accept" or a "discard" verb,
6487 but an error is generated for any others (because I can't see a useful way
6488 to define what should happen).
6490 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6491 are now processed in the same way as lines in the Exim configuration file.
6492 In particular, continuation lines are supported.
6494 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6496 15. Added -ti meaning -t -i.
6498 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6499 domains, and warn by logging if others are found.
6501 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6502 returned from GIFCONF when Exim is trying to find the list of interfaces on
6503 a host. The code in os.c has been modified to copy each ifreq to an aligned
6504 structure in all cases.
6506 Also, in some cases, the returned ifreq's were being copied to a 'struct
6507 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6508 means the last couple of bytes of an IPv6 address could be chopped if the
6509 ifreq contained only a normal sockaddr (14 bytes storage).
6511 18. Named domain lists were not supported in the hosts_treat_as_local option.
6512 An entry such as +xxxx was not recognized, and was treated as a literal
6515 19. Ensure that header lines added by a DATA ACL are included in the reject log
6516 if the ACL subsequently rejects the message.
6518 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6519 MD5 (which is deprecated).
6521 21. When testing a filter file using -bf, Exim was writing a message when it
6522 took the sender from a "From " line in the message, but it was not doing so
6523 when it took $return_path from a Return-Path: header line. It now does.
6525 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6526 with a valid header line field name (a series of printing characters
6527 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6529 23. Changed "disc" in the source to "disk" to conform to the documentation and
6530 the book and for uniformity.
6532 24. Ignore Sendmail's -Ooption=value command line item.
6534 25. When execve() failed while trying to run a command in a pipe transport,
6535 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6536 could be confused with a return value of 69 from the command itself. This
6537 has been changed to 127, the value the shell returns if it is asked to run
6538 a non-existent command. The wording for the related log line suggests a
6539 non-existent command as the problem.
6541 26. If received_header_text expands to an empty string, do not add a Received:
6542 header line to the message. (Well, it adds a token one on the spool, but
6543 marks it "old" so that it doesn't get used or transmitted.)
6545 27. Installed eximstats 1.28 (addition of -nt option).
6547 28. There was no check for failure on the call to getsockname() in the daemon
6548 code. This can fail if there is a shortage of resources on the system, with
6549 ENOMEM, for example. A temporary error is now given on failure.
6551 29. Contrary to the C standard, it seems that in some environments, the
6552 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6553 program. Exim now does this explicitly; it affects the formatting of
6554 timestamps using strftime().
6556 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6557 complaints. I've now initialized all the variables, to avoid this.
6559 32. Header lines added by a system filter were not being "seen" during
6560 transport-time rewrites.
6562 33. The info_callback() function passed to OpenSSL is set up with type void
6563 (*)(SSL *, int, int), as described somewhere. However, when calling the
6564 function (actually a macro) that sets it up, the type void(*)() is
6565 expected. I've put in a cast to prevent warnings from picky compilers.
6567 34. If a DNS black list lookup found a CNAME record, but there were no A
6568 records associated with the domain it pointed at, Exim crashed.
6570 35. If a DNS black list lookup returned more than one A record, Exim ignored
6571 all but the first. It now scans all returned addresses if a particular IP
6572 value is being sought. In this situation, the contents of the
6573 $dnslist_value variable are a list of all the addresses, separated by a
6576 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6577 to accept a host name and all its aliases if the forward lookup for any of
6578 them yielded the IP address of the incoming connection. Now it accepts only
6579 those names whose forward lookup yields the correct IP address. Any other
6580 names are discarded. This closes a loophole whereby a rogue DNS
6581 administrator could create reverse DNS records to break through a
6582 wildcarded host restriction in an ACL.
6584 37. If a user filter or a system filter that ran in a subprocess used any of
6585 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6586 the wrong values were passed to the pipe command ($thisaddress had the
6587 value of $0, $0 had the value of $1, etc). This bug was introduced by
6588 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6590 38. Improved the line breaking for long SMTP error messages from ACLs.
6591 Previously, if there was no break point between 40 and 75 characters, Exim
6592 left the rest of the message alone. Two changes have been made: (a) I've
6593 reduced the minimum length to 35 characters; (b) if it can't find a break
6594 point between 35 and 75 characters, it looks ahead and uses the first one
6595 that it finds. This may give the occasional overlong line, but at least the
6596 remaining text gets split now.
6598 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6599 file descriptors might be low, and that setting 1000 would always raise it.
6600 It turns out that in some environments, the limit is already over 1000 and
6601 that lowering it causes trouble. So now Exim takes care not to decrease it.
6603 40. When delivering a message, the value of $return_path is set to $sender_
6604 address at the start of routing (routers may change the value). By an
6605 oversight, this default was not being set up when an address was tested by
6606 -bt or -bv, which affected the outcome if any router or filter referred to
6609 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6610 log only when "message" or "log_message" are set. However, if one of the
6611 conditions was an address verification, or a call to a nested ACL, the
6612 messages generated by the underlying test were being passed through. This
6613 no longer happens. The underlying message is available in $acl_verify_
6614 message for both "message" and "log_message" expansions, so it can be
6615 passed through if needed.
6617 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6618 new expansion $bh_ to give the encoded byte string without charset
6619 translation. Translation happens only if iconv() is available; HAVE_ICONV
6620 indicates this at build time. HEADERS_CHARSET gives the charset to
6621 translate to; headers_charset can change it in the configuration, and
6622 "headers charset" can change it in an individual filter file.
6624 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6625 that creates RFC 2047 encoded "words" labels them as that charset instead
6626 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6627 expansion operator; (ii) when Exim creates a From: line for a local
6628 message; (iii) when a header line is rewritten to include a "phrase" part.
6630 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6631 buggy, causing it to skip the first lines of messages whose message ID
6632 ended in 'D'. This would not have bitten before Exim release 4.14, because
6633 message IDs were unlikely to end in 'D' before then. The effect was to have
6634 incorrect size information for certain domains.
6636 45. #include "config.h" was missing at the start of the crypt16.c module. This
6637 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6640 46. If there was a timeout during a "random" callout check, Exim treated it as
6641 a failure of the random address, and carried on sending RSET and the real
6642 address. If the delay was just some slowness somewhere, the response to the
6643 original RCPT would be taken as a response to RSET and so on, causing
6644 mayhem of various kinds.
6646 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6647 when I implemented it. It didn't allow for the fact that some option values
6648 may legitimately be negative (e.g. size_addition), and it didn't even do
6649 the right test for positive values.
6651 48. Domain names in DNS records are case-independent. Exim always looks them up
6652 in lower case. Some resolvers return domain names in exactly the case they
6653 appear in the zone file, that is, they may contain uppercase letters. Not
6654 all resolvers do this - some return always lower case. Exim was treating a
6655 change of case by a resolver as a change of domain, similar to a widening
6656 of a domain abbreviation. This triggered its re-routing code and so it was
6657 trying to route what was effectively the same domain again. This normally
6658 caused routing to fail (because the router wouldn't handle the domain
6659 twice). Now Exim checks for this case specially, and just changes the
6660 casing of the domain that it ultimately uses when it transmits the message
6663 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6666 50. If a filter generated a file delivery with a non-absolute name (possible if
6667 no home directory exists for the router), the forbid_file option was not
6670 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6671 the existing equality matching.
6673 52. Exim was using ints instead of ino_t variables in some places where it was
6674 dealing with inode numbers.
6676 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6677 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6678 called TMPDIR, and if it finds it is different, it changes its value.
6680 54. The smtp_printf() function is now made available to local_scan() so
6681 additional output lines can be written before returning. There is also an
6682 smtp_fflush() function to enable the detection of a dropped connection.
6683 The variables smtp_input and smtp_batched_input are exported to
6686 55. Changed the default runtime configuration: the message "Unknown user"
6687 has been removed from the ACL, and instead placed on the localuser router,
6688 using the cannot_route_message feature. This means that any verification
6689 failures that generate their own messages won't get overridden. Similarly,
6690 the "Unrouteable address" message that was in the ACL for unverifiable
6691 relay addresses has also been removed.
6693 56. Added hosts_avoid_esmtp to the smtp transport.
6695 57. The exicyclog script was not checking for the esoteric option
6696 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6697 will work only if exicyclog is run under the appropriate euid.
6699 58. Following a discussion on the list, the rules by which Exim recognises line
6700 endings on incoming messages have been changed. The -dropcr and drop_cr
6701 options are now no-ops, retained only for backwards compatibility. The
6702 following line terminators are recognized: LF CRLF CR. However, special
6703 processing applies to CR:
6705 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6706 nor a local message in the state where . is a terminator.
6708 (ii) If a bare CR is encountered in a header line, an extra space is added
6709 after the line terminator so as not to end the header. The reasoning
6710 behind this is that bare CRs in header lines are most likely either
6711 to be mistakes, or people trying to play silly games.
6713 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6714 was being incorrectly given as 18 bytes larger than it should have been.
6715 This is a VOB (very old bug).
6717 60. This may never have affected anything current, but just in case it has:
6718 When the local host is found other than at the start of a list of hosts,
6719 the local host, those with the same MX, and any that follow, are discarded.
6720 When the list in question was part of a longer list of hosts, the following
6721 hosts (not currently being processed) were also being discarded. This no
6722 longer happens. I'm not sure if this situation could ever has previously
6725 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6728 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6729 header. This is something that is recommended in a new Internet Draft, and
6730 is something that is documented as being done by Sendmail. There are two
6731 possible values. For messages generated by the autoreply transport, Exim
6734 Auto-Submitted: auto-replied
6736 whereas for all other generated messages (e.g. bounces) it adds
6738 Auto-Submitted: auto-generated
6740 63. The "personal" condition in filters now includes a test for the
6741 Auto-Submitted: header. If it contains the string "auto-" the message it
6742 not considered personal.
6744 64. Added rcpt_include_affixes as a generic transport option.
6746 65. Added queue_only_override (default true).
6748 66. Added the syslog_duplication option.
6750 67. If what should have been the first header line of a message consisted of
6751 a space followed by a colon, Exim was mis-interpreting it as a header line.
6752 It isn't of course - it is syntactically invalid and should therefore be
6753 treated as the start of the message body. The misbehaviour could have
6754 caused a number of strange effects, including loss of data in subsequent
6755 header lines, and spool format errors.
6757 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6758 client host had authenticated. This control can now be exercised by an ACL
6759 for more flexibility.
6761 69. By default, callouts do not happen when testing with -bh. There is now a
6762 variant, -bhc, which does actually run the callout code, including
6763 consulting and updating the callout cache.
6765 70. Added support for saslauthd authentication, courtesy of Alexander
6768 71. If statvfs() failed on the spool or log directories while checking their
6769 size for availability, Exim confusingly gave the error "space shortage".
6770 Furthermore, in debugging mode it crashed with a floating point exception.
6771 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6772 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6773 serious problem, Exim now writes to the main and panic logs when this
6774 happens, with details of the failure. It then refuses to accept the
6775 incoming message, giving the message "spool directory problem" or "log
6776 directory problem" with a 421 code for SMTP messages.
6778 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6779 0, 1, and 2 exist, because some OS complain for execs without them (see
6780 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6781 descriptors. However, the code omitted to check that the open succeeded,
6782 causing mysterious errors if for some reason the permissions on /dev/null
6783 got screwed. Now Exim writes a message to the main and panic logs, and
6784 bombs out if it can't open /dev/null.
6786 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6787 interact so that it is all more flexible. It is supposed to remain
6788 backwards compatible. Also added extra_local_interfaces.
6790 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6791 to bomb out with an assertion failure - to the client this appears as a
6792 connection drop. This problem occurs in the part of the code that was taken
6793 from the Samba project. Fortunately, the assertion is in a very simple
6794 function, so I have fixed this by reproducing the function inline in the
6795 one place where it is called, and arranging for authentication to fail
6796 instead of killing the process with assert().
6798 75. The SPA client code was not working when the server requested OEM rather
6799 than Unicode encoding.
6801 76. Added code to make require_files with a specific uid setting more usable in
6802 the case where statting the file as root fails - usually a non-root-mounted
6803 NFS file system. When this happens and the failure is EACCES, Exim now
6804 forks a subprocess and does the per-uid checking as the relevant uid.
6806 77. Added process_log_path.
6808 78. If log_file_path was not explicitly set, a setting of check_log_space or
6809 check_log_inodes was ignored.
6811 79. If a space check for the spool or log partitions fails, the incident is now
6812 logged. Of course, in the latter case the data may get lost...
6814 80. Added the %p formatting code to string_format() so that it can be used to
6815 print addresses in debug_print(). Adjusted all the address printing in the
6816 debugging in store.c to use %p rather than %d.
6818 81. There was a concern that a line of code in smtp_in.c could overflow a
6819 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6820 initially expressed, the concern was not well-founded, because trailing
6821 spaces are removed early. However, if the trailing spaces were followed by
6822 a NULL, they did not get removed, so the overflow was possible. Two fixes
6825 (a) I re-wrote the offending code in a cleaner fashion.
6826 (b) If an incoming SMTP command contains a NULL character, it is rejected
6829 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6830 runs initgroups(), so that if the Exim user is in any additional groups,
6831 they will be used during message reception.
6837 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob