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. Add support for FreeBSD also.
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 buffering.
81 PP/04 Add commandline_checks_require_admin option.
82 Exim drops privileges sanely, various checks such as -be aren't a
83 security problem, as long as you trust local users with access to their
84 own account. When invoked by services which pass untrusted data to
85 Exim, this might be an issue. Set this option in main configuration
86 AND make fixes to the calling application, such as using `--` to stop
89 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
90 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
91 responses to those, into a single TLS record each way (this usually means
92 a single packet). As a side issue, smtp_enforce_sync now works on TLS
99 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
100 than -2003 did; needs libidn2 in addition to libidn.
102 JH/02 The path option on a pipe transport is now expanded before use.
104 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
105 Patch provided by "Björn", documentation fix added too.
107 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
108 missing a wire-to-host endian conversion.
110 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
111 close after a BDAT command line could be taken as a following command,
112 giving a synch failure. Fix by only checking for synch immediately
113 before acknowledging the chunk.
115 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
116 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
117 macro. Patches provided by Josh Soref.
119 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
120 Previously we did not; the RFC seems ambiguous and VRFY is not listed
121 by IANA as a service extension. However, John Klensin suggests that we
124 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
125 the dkim code may be unix-mode line endings rather than smtp wire-format
126 CRLF, so prepend a CR to any bare LF.
128 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
129 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
131 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
132 the first were themselves being wrongly included in the feed into dkim
133 processing; with most chunk sizes in use this resulted in an incorrect
134 body hash calculated value.
136 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
137 DKIM signature block, for verification. Although advised against by
138 standards it is specifically not ruled illegal.
140 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
142 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
143 missing a body hash (the bh= tag).
145 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
146 It seems that HAProxy sends the Proxy Protocol information in clear and
147 only then does a TLS startup, so do the same.
149 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
150 TCP connections (such as for Spamd) unless the daemon successfully set
151 Fast Open mode on its listening sockets. This fixes breakage seen on
152 too-old kernels or those not configured for Fast Open, at the cost of
153 requiring both directions being enabled for TFO, and TFO never being used
154 by non-daemon-related Exim processes.
156 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
157 endings, at least on the first header line. Try to canonify any that get
158 past that check, despite the cost.
160 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
161 now limited to an arbitrary five deep, while parsing addresses with the
162 strip_excess_angle_brackets option enabled.
164 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
165 instead leave the unprompted TLS handshake in socket buffer for the
166 TLS library to consume.
168 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
170 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
172 JH/16 Drop variables when they go out of scope. Memory management drops a whole
173 region in one operation, for speed, and this leaves assigned pointers
174 dangling. Add checks run only under the testsuite which checks all
175 variables at a store-reset and panics on a dangling pointer; add code
176 explicitly nulling out all the variables discovered. Fixes one known
177 bug: a transport crash, where a dangling pointer for $sending_ip_address
178 originally assigned in a verify callout, is re-used.
180 PP/06 Drop '.' from @INC in various Perl scripts.
182 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
184 PP/08 Reduce a number of compilation warnings under clang; building with
185 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
186 should be warning-free.
188 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
190 HS/01 Fix portability problems introduced by PP/08 for platforms where
191 realloc(NULL) is not equivalent to malloc() [SunOS et al].
193 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
194 chunk. This allows us to accept broken chunked messages. We need a more
195 general solution here.
197 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
198 already-broken messages in the queue.
200 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
202 JH/19 Fix reference counting bug in routing-generated-address tracking.
208 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
209 supports it and a size is available (ie. the sending peer gave us one).
211 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
212 years of being deprecated). The replacements are the ACLs
213 acl_smtp_mime and acl_not_smtp_mime.
215 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
216 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
217 or even an in-clear connection were permitted. Now, if the host lookup
218 was dnssec and dane was requested then the host is only used if the
219 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
220 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
221 if one fails this test.
222 This means that a poorly-configured remote DNS will make it incommunicado;
223 but it protects against a DNS-interception attack on it.
225 JH/04 Bug 1810: make continued-use of an open smtp transport connection
226 non-noisy when a race steals the message being considered.
228 JH/05 If main configuration option tls_certificate is unset, generate a
229 self-signed certificate for inbound TLS connections.
231 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
232 in rewrites and routers.
234 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
235 and logged a warning sing 4.83; now they are a configuration file error.
237 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
238 (lacking @domain). Apply the same qualification processing as RCPT.
240 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
242 JH/10 Support ${sha256:} applied to a string (as well as the previous
245 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
246 a cutthrough deliver is pending, as we always want to make a connection.
247 This also avoids re-routing the message when later placing the cutthrough
248 connection after a verify cache hit.
249 Do not update it with the verify result either.
251 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
252 when routing results in more than one destination address.
254 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
255 signing (which inhibits the cutthrough capability). Previously only
256 the presence of an option was tested; now an expansion evaluating as
257 empty is permissible (obviously it should depend only on data available
258 when the cutthrough connection is made).
260 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
261 the relevant preceding SMTP command did not note the pipelining mode.
263 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
264 Previously they were not counted.
266 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
267 as one having no matching records. Previously we deferred the message
268 that needed the lookup.
270 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
271 distinguished as "(=".
273 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
274 for missing MX records. Previously it only worked for missing A records.
276 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
278 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
279 after the data-go-ahead and data-ack. Patch from Jason Betts.
281 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
282 even for a "none" policy. Patch from Tony Meyer.
284 JH/22 Fix continued use of a connection for further deliveries. If a port was
285 specified by a router, it must also match for the delivery to be
288 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
289 When one of the recipients of a message was unsuitable for the connection
290 (has no matching addresses), we lost track of needing to mark it
291 deferred. As a result mail would be lost.
293 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
295 JH/25 Decoding ACL controls is now done using a binary search; the source code
296 takes up less space and should be simpler to maintain. Merge the ACL
297 condition decode tables also, with similar effect.
299 JH/26 Fix problem with one_time used on a redirect router which returned the
300 parent address unchanged. A retry would see the parent address marked as
301 delivered, so not attempt the (identical) child. As a result mail would
304 JH/27 Fix a possible security hole, wherein a process operating with the Exim
305 UID can gain a root shell. Credit to http://www.halfdog.net/ for
306 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
309 JH/28 Enable {spool,log} filesystem space and inode checks as default.
310 Main config options check_{log,spool}_{inodes,space} are now
311 100 inodes, 10MB unless set otherwise in the configuration.
313 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
314 Previously it only applied to the main-section connection policy
317 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
319 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
320 by me. Added RFC7919 DH primes as an alternative.
322 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
323 are not in the system include path.
325 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
326 GnuTLS, when a session startup failed (eg because the client disconnected)
327 Exim did stdio operations after fclose. This was exposed by a recent
328 change which nulled out the file handle after the fclose.
330 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
331 signed directly by the cert-signing cert, rather than an intermediate
332 OCSP-signing cert. This is the model used by LetsEncrypt.
334 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
336 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
337 an incoming connection.
339 HS/02 Bug 1802: Do not half-close the connection after sending a request
342 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
343 fallback to "prime256v1".
345 JH/34 SECURITY: Use proper copy of DATA command in error message.
346 Could leak key material. Remotely exploitable. CVE-2016-9963.
352 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
353 and 3.4.4 - once the server is enabled to respond to an OCSP request
354 it does even when not requested, resulting in a stapling non-aware
355 client dropping the TLS connection.
357 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
358 support variable-length bit vectors. No functional change.
360 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
361 The I= interface field on outgoing lines is now after the H= remote
362 host field, same as incoming lines. There is a separate
363 outgoing_interface log selector which allows you to disable the
366 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
367 If not running log_selector +smtp_connection the mainlog would be held
368 open indefinitely after a "too many connections" event, including to a
369 deleted file after a log rotate. Leave the per net connection logging
370 leaving it open for efficiency as that will be quickly detected by the
371 check on the next write.
373 HS/01 Bug 1671: Fix post transport crash.
374 Processing the wait-<transport> messages could crash the delivery
375 process if the message IDs didn't exist for some reason. When
376 using 'split_spool_directory=yes' the construction of the spool
377 file name failed already, exposing the same netto behaviour.
379 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
380 mime_regex ACL conditions.
382 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
383 to DSN fail messages (bounces): remote IP, remote greeting, remote response
384 to HELO, local diagnostic string.
386 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
387 log line to debug. Even when configured with a tls authenticator many
388 client connections are expected to not authenticate in this way, so
389 an authenticate fail is not an error.
391 HS/02 Add the Exim version string to the process info. This way exiwhat
392 gives some more detail about the running daemon.
394 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
395 matter for fast-change records such as DNSBLs.
397 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
398 retry record, even if constant. There may be multiple transports with
399 different interface settings and the retry behaviour needs to be kept
402 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
404 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
406 JH/10 Bug 840: fix log_defer_output option of pipe transport
408 JH/11 Bug 830: use same host for all RCPTS of a message, even under
409 hosts_randomize. This matters a lot when combined with mua_wrapper.
411 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
412 ${quote_pgsql:<string>} operator.
414 JH/13 Bug 1708: avoid misaligned access in cached lookup.
416 JH/14 Change header file name for freeradius-client. Relevant if compiling
417 with Radius support; from the Gentoo tree and checked under Fedora.
419 JH/15 Bug 1712: Introduce $prdr_requested flag variable
421 JH/16 Bug 1714: Permit an empty string as expansion result for transport
422 option transport_filter, meaning no filtering.
424 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
426 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
427 defaults to "*" (all hosts). The variable is now available when not built
428 with TLS, default unset, mainly to enable keeping the testsuite sane.
429 If a server certificate is not supplied (via tls_certificate) an error is
430 logged, and clients will find TLS connections fail on startup. Presumably
431 they will retry in-clear.
432 Packagers of Exim are strongly encouraged to create a server certificate
433 at installation time.
435 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
436 with the $config_file variable.
438 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
439 in transport context, after the attempt, and per-recipient. The latter type
440 is per host attempted. The event data is the error message, and the errno
441 information encodes the lookup type (A vs. MX) used for the (first) host,
442 and the trailing two digits of the smtp 4xx response.
444 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
445 to write to mainlog (or rejectlog, paniclog) in the window between file
446 creation and permissions/ownership being changed. Particularly affects
447 installations where exicyclog is run as root, rather than exim user;
448 result is that the running daemon panics and dies.
450 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
452 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
453 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
454 "pri" and "weight". Note that the previous implicit priority given by the
455 list order is no longer honoured.
457 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
460 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
461 by defining SUPPORT_SOCKS.
463 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
464 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
465 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
466 variables are renamed to proxy_{local,external}_{address,port}.
468 JH/27 Move Internationalisation support from Experimental to mainline, enabled
469 for a build by defining SUPPORT_I18N
471 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
472 of the query string, and make ${quote_redis:} do that quoting.
474 JH/29 Move Events support from Experimental to mainline, enabled by default
475 and removable for a build by defining DISABLE_EVENT.
477 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
479 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
480 cached by the daemon.
482 JH/32 Move Redis support from Experimental to mainline, enabled for a build
483 by defining LOOKUP_REDIS. The libhiredis library is required.
485 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
486 keys are given for lookup.
488 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
489 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
490 only supported when built with TLS support. The PolarSSL SHA routines
491 are still used when the TLS library is too old for convenient support.
493 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
494 openssl_options), for security. OpenSSL forces this from version 1.1.0
495 server-side so match that on older versions.
497 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
498 allocation for $value could be released as the expansion processing
499 concluded, but leaving the global pointer active for it.
501 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
502 and to use the domains and local_parts ACL conditions.
504 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
505 incorrectly not doubled on cutthrough transmission, hence seen as a
506 body-termination at the receiving system - resulting in truncated mails.
507 Commonly the sender saw a TCP-level error, and retransmitted the message
508 via the normal store-and-forward channel. This could result in duplicates
509 received - but deduplicating mailstores were liable to retain only the
510 initial truncated version.
512 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
514 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
516 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
517 we're in there, support oversigning also; bug 1309.
519 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
521 HS/04 Add support for keep_environment and add_environment options.
523 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
524 either intentional arithmetic overflow during PRNG, or testing config-
527 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
528 delivery resulted in actual delivery. Cancel cutthrough before DATA
531 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
532 rejects a recipient: pass the reject to the originator.
534 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
535 Many were false-positives and ignorable, but it's worth fixing the
538 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
539 for the new environment-manipulation done at startup. Move the routines
540 from being local to tls.c to being global via the os.c file.
542 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
543 an extract embedded as result-arg for a map, the first arg for extract
544 is unavailable so we cannot tell if this is a numbered or keyed
545 extraction. Accept either.
551 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
554 JH/02 The smtp transport option "multi_domain" is now expanded.
556 JH/03 The smtp transport now requests PRDR by default, if the server offers
559 JH/04 Certificate name checking on server certificates, when exim is a client,
560 is now done by default. The transport option tls_verify_cert_hostnames
561 can be used to disable this per-host. The build option
562 EXPERIMENTAL_CERTNAMES is withdrawn.
564 JH/05 The value of the tls_verify_certificates smtp transport and main options
565 default to the word "system" to access the system default CA bundle.
566 For GnuTLS, only version 3.0.20 or later.
568 JH/06 Verification of the server certificate for a TLS connection is now tried
569 (but not required) by default. The verification status is now logged by
570 default, for both outbound TLS and client-certificate supplying inbound
573 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
576 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
577 Status Notification (bounce) messages are now MIME format per RFC 3464.
578 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
579 under the control of the dsn_advertise_hosts option, and routers may
580 have a dsn_lasthop option.
582 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
583 default, modifiable by a malware= option. The list separator for
584 the options can now be changed in the usual way. Bug 68.
586 JH/10 The smtp_receive_timeout main option is now expanded before use.
588 JH/11 The incoming_interface log option now also enables logging of the
589 local interface on delivery outgoing connections.
591 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
592 if the interface and destination host and port all match.
594 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
597 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
598 Patch from Andrew Lewis.
600 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
601 now supports optional time-restrictions, weighting, and priority
602 modifiers per server. Patch originally by <rommer@active.by>.
604 JH/16 The spamd_address main option now supports a mixed list of local
605 and remote servers. Remote servers can be IPv6 addresses, and
606 specify a port-range.
608 JH/17 Bug 68: The spamd_address main option now supports an optional
609 timeout value per server.
611 JH/18 Bug 1581: Router and transport options headers_add/remove can
612 now have the list separator specified.
614 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
617 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
620 JH/21 Support for the A6 type of dns record is withdrawn.
622 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
623 rather than the verbs used.
625 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
626 from 255 to 1024 chars.
628 JH/24 Verification callouts now attempt to use TLS by default.
630 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
631 are generic router options now. The defaults didn't change.
633 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
634 Original patch from Alexander Shikoff, worked over by JH.
636 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
637 files if ZCAT_COMMAND is not executable.
639 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
641 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
643 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
644 Normally benign, it bites when the pair was led to by a CNAME;
645 modern usage is to not canonicalize the domain to a CNAME target
646 (and we were inconsistent anyway for A-only vs AAAA+A).
648 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
650 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
651 when evaluating $sender_host_dnssec.
653 JH/31 Check the HELO verification lookup for DNSSEC, adding new
654 $sender_helo_dnssec variable.
656 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
658 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
660 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
662 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
663 documented as working, but never had. Support all but $spam_report.
665 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
666 added for tls authenticator.
668 HS/03 Add perl_taintmode main config option
674 TL/01 When running the test suite, the README says that variables such as
675 no_msglog_check are global and can be placed anywhere in a specific
676 test's script, however it was observed that placement needed to be near
677 the beginning for it to behave that way. Changed the runtest perl
678 script to read through the entire script once to detect and set these
679 variables, reset to the beginning of the script, and then run through
680 the script parsing/test process like normal.
682 TL/02 The BSD's have an arc4random API. One of the functions to induce
683 adding randomness was arc4random_stir(), but it has been removed in
684 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
685 function when detected.
687 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
688 cause callback expansion.
690 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
691 syntax errors in an expansion can be treated as a string instead of
692 logging or causing an error, due to the internal use of bool_lax
693 instead of bool when processing it.
695 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
696 server certificates when making smtp deliveries.
698 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
700 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
702 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
704 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
705 Merged patch from Sebastian Wiedenroth.
707 JH/05 Fix results-pipe from transport process. Several recipients, combined
708 with certificate use, exposed issues where response data items split
709 over buffer boundaries were not parsed properly. This eventually
710 resulted in duplicates being sent. This issue only became common enough
711 to notice due to the introduction of connection certificate information,
712 the item size being so much larger. Found and fixed by Wolfgang Breyha.
714 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
715 size buffer was used, resulting in syntax errors when an expansion
718 JH/07 Add support for directories of certificates when compiled with a GnuTLS
719 version 3.3.6 or later.
721 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
722 is EXPERIMENTAL_EVENT, the main-configuration and transport options
723 both become "event_action", the variables become $event_name, $event_data
724 and $event_defer_errno. There is a new variable $verify_mode, usable in
725 routers, transports and related events. The tls:cert event is now also
726 raised for inbound connections, if the main configuration event_action
729 TL/06 In test suite, disable OCSP for old versions of openssl which contained
730 early OCSP support, but no stapling (appears to be less than 1.0.0).
732 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
733 server certificate names available under the smtp transport option
734 "tls_verify_cert_hostname" now do not permit multi-component wildcard
737 JH/10 Time-related extraction expansions from certificates now use the main
738 option "timezone" setting for output formatting, and are consistent
739 between OpenSSL and GnuTLS compilations. Bug 1541.
741 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
742 encoded parameter in the incoming message. Bug 1558.
744 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
745 include certificate info, eximon was claiming there were spoolfile
748 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
750 JH/14 Log delivery-related information more consistently, using the sequence
751 "H=<name> [<ip>]" wherever possible.
753 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
754 are problematic for Debian distribution, omit them from the release
757 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
759 JH/16 Fix string representation of time values on 64bit time_t architectures.
762 JH/17 Fix a null-indirection in certextract expansions when a nondefault
763 output list separator was used.
768 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
769 checkers that were complaining about end of non-void function with no
772 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
773 This was a regression introduced in 4.83 by another bugfix.
775 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
777 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
778 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
784 TF/01 Correctly close the server side of TLS when forking for delivery.
786 When a message was received over SMTP with TLS, Exim failed to clear up
787 the incoming connection properly after forking off the child process to
788 deliver the message. In some situations the subsequent outgoing
789 delivery connection happened to have the same fd number as the incoming
790 connection previously had. Exim would try to use TLS and fail, logging
791 a "Bad file descriptor" error.
793 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
794 utilities have not been installed.
796 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
797 temporary space as the ACL may create new global variables.
799 TL/01 LDAP support uses per connection or global context settings, depending
800 upon the detected version of the libraries at build time.
802 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
803 to extract and use the src ip:port in logging and expansions as if it
804 were a direct connection from the outside internet. PPv2 support was
805 updated based on HAProxy spec change in May 2014.
807 JH/02 Add ${listextract {number}{list}{success}{fail}}.
809 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
810 Properly escape header and check for NULL return.
812 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
815 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
817 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
818 characters in header names, implemented as a verify condition.
819 Contributed by Michael Fischer v. Mollard.
821 TL/05 Rename SPF condition results err_perm and err_temp to standardized
822 results permerror and temperror. Previous values are deprecated but
823 still accepted. In a future release, err_perm and err_temp will be
824 completely removed, which will be a backward incompatibility if the
825 ACL tests for either of these two old results. Patch contributed by
826 user bes-internal on the mailing list.
828 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
830 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
831 selectors, in both main and reject logs.
833 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
836 JH/07 Add malware type "sock" for talking to simple daemon.
838 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
840 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
841 routers/transports under cutthrough routing.
843 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
844 numbers. Touch up "bool" conditional to keep the same definition.
846 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
848 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
850 JH/12 Expand items in router/transport headers_add or headers_remove lists
851 individually rather than the list as a whole. Bug 1452.
853 Required for reasonable handling of multiple headers_ options when
854 they may be empty; requires that headers_remove items with embedded
855 colons must have them doubled (or the list-separator changed).
857 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
858 view the policy declared in the DMARC record. Currently, $dmarc_status
859 is a combined value of both the record presence and the result of the
862 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
864 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
865 dnslookup router and the smtp transport (applying to the forward
868 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
869 of ldap servers used for a specific lookup. Patch provided by Heiko
872 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
873 New variable $lookup_dnssec_authenticated for observability.
875 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
876 Patch submitted by Lars Timman.
878 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
880 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
881 Requires trusted mode and valid format message id, aborts otherwise.
882 Patch contributed by Heiko Schlichting.
884 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
885 certextract with support for various fields. Bug 1358.
887 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
888 is requested by default, modifiable by smtp transport option
891 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
892 operate on certificate variables to give certificate fingerprints
893 Also new ${sha256:cert_variable}.
895 JH/23 The PRDR feature is moved from being Experimental into the mainline.
897 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
898 Christian Aistleitner.
900 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
902 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
903 file. Patch from Wolfgang Breyha.
905 JH/25 Expand the coverage of the delivery $host and $host_address to
906 client authenticators run in verify callout. Bug 1476.
908 JH/26 Port service names are now accepted for tls_on_connect_ports, to
909 align with daemon_smtp_ports. Bug 72.
911 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
912 support and error reporting did not work properly.
914 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
915 and is readable. Patch from Andrew Colin Kissa.
917 TL/14 Enhance documentation of ${run expansion and how it parses the
918 commandline after expansion, particularly in the case when an
919 unquoted variable expansion results in an empty value.
921 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
923 PP/02 Fix internal collision of T_APL on systems which support RFC3123
924 by renaming away from it. Addresses GH issue 15, reported by
927 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
929 TL/15 SECURITY: prevent double expansion in math comparison functions
930 (can expand unsanitized data). Not remotely exploitable.
937 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
939 PP/02 Make -n do something, by making it not do something.
940 When combined with -bP, the name of an option is not output.
942 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
945 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
946 $sender_host_name and config options to manage this, and basic check
949 PP/05 DSCP support for outbound connections and control modifier for inbound.
951 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
952 (Only plugin which currently uses this is kerberos4, which nobody should
953 be using, but we should make it available and other future plugins might
954 conceivably use it, even though it would break NAT; stuff *should* be
955 using channel bindings instead).
957 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
958 name; added for Sendmail compatibility; requires admin caller.
959 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
960 just ignore it); requires trusted caller.
961 Also parse but ignore: -Ac -Am -X<logfile>
964 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
966 TL/02 Add +smtp_confirmation as a default logging option.
968 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
969 Patch by Magnus Holmgren from 2007-02-20.
971 TL/04 Bugzilla 1281 - Spec typo.
972 Bugzilla 1283 - Spec typo.
973 Bugzilla 1290 - Spec grammar fixes.
975 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
977 TL/06 Add Experimental DMARC support using libopendmarc libraries.
979 TL/07 Fix an out of order global option causing a segfault. Reported to dev
980 mailing list by by Dmitry Isaikin.
982 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
984 JH/02 Support "G" suffix to numbers in ${if comparisons.
986 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
988 NM/01 Bugzilla 1197 - Spec typo
989 Bugzilla 1196 - Spec examples corrections
991 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
993 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
994 gnutls_enable_pkcs11, but renamed to more accurately indicate its
997 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
998 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1000 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1001 "acl {{name}{arg}...}", and optional args on acl condition
1004 JH/05 Permit multiple router/transport headers_add/remove lines.
1006 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1008 JH/07 Avoid using a waiting database for a single-message-only transport.
1009 Performance patch from Paul Fisher. Bugzilla 1262.
1011 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1014 JH/09 Add $headers_added variable, with content from use of ACL modifier
1015 add_header (but not yet added to the message). Bugzilla 199.
1017 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1018 Pulled from Bugzilla 817 by Wolfgang Breyha.
1020 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1022 (nb: this is the same fix as in Exim 4.80.1)
1024 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1027 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1030 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1032 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1033 advertise SMTP AUTH mechanism to us, instead of a generic
1034 protocol violation error. Also, make Exim more robust to bad
1035 data from the Dovecot auth socket.
1037 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1039 When a queue runner is handling a message, Exim first routes the
1040 recipient addresses, during which it prunes them based on the retry
1041 hints database. After that it attempts to deliver the message to
1042 any remaining recipients. It then updates the hints database using
1045 So if a recipient address works intermittently, it can get repeatedly
1046 deferred at routing time. The retry hints record remains fresh so the
1047 address never reaches the final cutoff time.
1049 This is a fairly common occurrence when a user is bumping up against
1050 their storage quota. Exim had some logic in its local delivery code
1051 to deal with this. However it did not apply to per-recipient defers
1052 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1054 This change adds a proper retry rule check during routing so that the
1055 final cutoff time is checked against the message's age. We only do
1056 this check if there is an address retry record and there is not a
1057 domain retry record; this implies that previous attempts to handle
1058 the address had the retry_use_local_parts option turned on. We use
1059 this as an approximation for the destination being like a local
1060 delivery, as in LMTP.
1062 I suspect this new check makes the old local delivery cutoff check
1063 redundant, but I have not verified this so I left the code in place.
1065 TF/02 Correct gecos expansion when From: is a prefix of the username.
1067 Test 0254 submits a message to Exim with the header
1071 When I ran the test suite under the user fanf2, Exim expanded
1072 the header to contain my full name, whereas it should have added
1073 a Resent-Sender: header. It erroneously treats any prefix of the
1074 username as equal to the username.
1076 This change corrects that bug.
1078 GF/01 DCC debug and logging tidyup
1079 Error conditions log to paniclog rather than rejectlog.
1080 Debug lines prefixed by "DCC: " to remove any ambiguity.
1082 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1084 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1085 Bug spotted by Jeremy Harris; was flawed since initial commit.
1086 Would have resulted in OCSP responses post-SNI triggering an Exim
1087 NULL dereference and crash.
1089 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1091 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1092 Bug detection, analysis and fix by Samuel Thibault.
1093 Bugzilla 1331, Debian bug #698092.
1095 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1097 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1098 Server implementation by Todd Lyons, client by JH.
1099 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1100 config variable "prdr_enable" controls whether the server
1101 advertises the facility. If the client requests PRDR a new
1102 acl_data_smtp_prdr ACL is called once for each recipient, after
1103 the body content is received and before the acl_smtp_data ACL.
1104 The client is controlled by both of: a hosts_try_prdr option
1105 on the smtp transport, and the server advertisement.
1106 Default client logging of deliveries and rejections involving
1107 PRDR are flagged with the string "PRDR".
1109 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1110 fclose(). Diagnosis by Todd Lyons.
1112 PP/17 Update configure.default to handle IPv6 localhost better.
1113 Patch by Alain Williams (plus minor tweaks).
1116 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1117 This is now consistent with GnuTLS, and is now documented: the
1118 previous undocumented portable approach to treating the option as
1119 unset was to force an expansion failure. That still works, and
1120 an empty string is now equivalent.
1122 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1123 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1124 not performing validation itself.
1126 PP/20 Added force_command boolean option to pipe transport.
1127 Patch from Nick Koston, of cPanel Inc.
1129 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1132 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1134 PP/21 Fix eximon continuous updating with timestamped log-files.
1135 Broken in a format-string cleanup in 4.80, missed when I repaired the
1136 other false fix of the same issue.
1137 Report and fix from Heiko Schlichting.
1140 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1141 Report from Prashanth Katuri.
1143 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1144 It's SecureTransport, so affects any MacOS clients which use the
1145 system-integrated TLS libraries, including email clients.
1147 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1148 using a MIME ACL for non-SMTP local injection.
1149 Report and assistance in diagnosis by Warren Baker.
1151 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1153 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1155 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1156 last id that failed so it may be referenced in subsequent ACL's.
1158 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1161 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1162 ldap library initialization, allowing self-signed CA's to be
1163 used. Also properly sets require_cert option later in code by
1164 using NULL (global ldap config) instead of ldap handle (per
1165 session). Bug diagnosis and testing by alxgomz.
1167 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1168 the src/util/ subdirectory.
1170 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1171 renamed to Transport Post Delivery Action by Jeremy Harris, as
1174 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1175 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1176 redis_servers = needs to be configured which will be used by the redis
1177 lookup. Patch from Warren Baker, of The Packet Hub.
1179 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1181 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1182 hostname or reverse DNS when processing a host list. Used suggestions
1183 from multiple comments on this bug.
1185 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1187 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1188 Missed a few lines, added it to make the runtest require no keyboard
1191 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1192 contains upper case chars. Make router use caseful_local_part.
1194 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1195 support when GnuTLS has been built with p11-kit.
1201 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1203 This, or similar/improved, will also be change PP/11 of 4.82.
1209 PP/01 Handle short writes when writing local log-files.
1210 In practice, only affects FreeBSD (8 onwards).
1211 Bugzilla 1053, with thanks to Dmitry Isaikin.
1213 NM/01 Bugzilla 949 - Documentation tweak
1215 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1218 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1220 PP/02 Implemented gsasl authenticator.
1222 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1224 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1225 `pkg-config foo` for cflags/libs.
1227 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1228 with rest of GSASL and with heimdal_gssapi.
1230 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1231 `pkg-config foo` for cflags/libs for the TLS implementation.
1233 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1234 properties get this fed in as external SSF. A number of robustness
1235 and debugging improvements to the cyrus_sasl authenticator.
1237 PP/08 cyrus_sasl server now expands the server_realm option.
1239 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1240 Patch by Jeremy Harris.
1242 PP/10 Added dbmjz lookup type.
1244 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1246 PP/12 MAIL args handles TAB as well as SP, for better interop with
1247 non-compliant senders.
1248 Analysis and variant patch by Todd Lyons.
1250 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1251 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1252 Patch from Dirk Mueller <dmueller@suse.com>
1254 PP/13 tls_peerdn now print-escaped for spool files.
1255 Observed some $tls_peerdn in wild which contained \n, which resulted
1256 in spool file corruption.
1258 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1259 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1260 or write after TLS renegotiation, which otherwise led to messages
1263 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1264 as a tracking header (ie: a signed header comes before the signature).
1265 Patch from Wolfgang Breyha.
1267 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1268 comma-sep list; embedded commas doubled.
1270 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1272 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1274 Report and patch from Dmitry Banschikov.
1276 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1277 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1278 failures. We appear to now support TLS1.1+ with Exim.
1280 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1281 lets Exim select keys and certificates based upon TLS SNI from client.
1282 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1283 before an outbound SMTP session. New log_selector, +tls_sni.
1285 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1286 NULL dereference. Report and patch from Alun Jones.
1288 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1289 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1290 Not seeing resolver debug output on NetBSD, but suspect this is a
1291 resolver implementation change.
1293 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1294 Left warnings. Added "eximon gdb" invocation mode.
1296 PP/21 Defaulting "accept_8bitmime" to true, not false.
1298 PP/22 Added -bw for inetd wait mode support.
1300 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1301 locate the relevant includes and libraries. Made this the default.
1303 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1304 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1306 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1307 This may cause build issues on older platforms.
1309 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1310 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1311 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1312 Added SNI support via GnuTLS too.
1313 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1315 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1317 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1318 Applied second patch from Janne, implementing suggestion to default
1319 multiple-strings-in-record handling to match SPF spec.
1321 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1323 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1324 read-only, out of scope).
1325 Patch from Wolfgang Breyha, report from Stuart Northfield.
1327 PP/29 Fix three issues highlighted by clang analyser static analysis.
1328 Only crash-plausible issue would require the Cambridge-specific
1329 iplookup router and a misconfiguration.
1330 Report from Marcin Mirosław.
1332 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1334 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1335 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1336 As part of this, removing so much warning spew let me fix some minor
1337 real issues in debug logging.
1339 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1340 assignment on my part. Fixed.
1342 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1343 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1344 Janne Snabb (who went above and beyond: thank you).
1346 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1347 string otherwise requires a connection and a bunch more work and it's
1348 relatively easy to get wrong. Should also expose TLS library linkage
1351 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1352 64-bit ${eval} (JH/03).
1354 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1355 GNU libc to support some of the 64-bit stuff, should not lead to
1356 conflicts. Defined before os.h is pulled in, so if a given platform
1357 needs to override this, it can.
1359 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1360 protection layer was required, which is not implemented.
1361 Bugzilla 1254, patch from Wolfgang Breyha.
1363 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1364 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1365 tls_dhparam take prime identifiers. Also unbreak combination of
1366 OpenSSL+DH_params+TLSSNI.
1368 PP/39 Disable SSLv2 by default in OpenSSL support.
1374 PP/01 Solaris build fix for Oracle's LDAP libraries.
1375 Bugzilla 1109, patch from Stephen Usher.
1377 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1379 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1382 TF/02 Fix a couple more cases where we did not log the error message
1383 when unlink() failed. See also change 4.74-TF/03.
1385 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1386 lock up or crash if it happened to be inside a call to libc when it
1387 got a SIGUSR1 from exiwhat.
1389 The SIGUSR1 handler appends the current process status to the process
1390 log which is later printed by exiwhat. It used to use the general
1391 purpose logging code to do this, but several functions it calls are
1392 not safe for signals.
1394 The new output code in the SIGUSR1 handler is specific to the process
1395 log, and simple enough that it's easy to inspect for signal safety.
1396 Removing some special cases also simplifies the general logging code.
1397 Removing the spurious timestamps from the process log simplifies
1400 TF/04 Improved ratelimit ACL condition.
1402 The /noupdate option has been deprecated in favour of /readonly which
1403 has clearer semantics. The /leaky, /strict, and /readonly update modes
1404 are mutually exclusive. The update mode is no longer included in the
1405 database key; it just determines when the database is updated. (This
1406 means that when you upgrade Exim will forget old rate measurements.)
1408 Exim now checks that the per_* options are used with an update mode that
1409 makes sense for the current ACL. For example, when Exim is processing a
1410 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1411 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1412 must specify per_mail/readonly. If you omit the update mode it defaults to
1413 /leaky where that makes sense (as before) or /readonly where required.
1415 The /noupdate option is now undocumented but still supported for
1416 backwards compatibility. It is equivalent to /readonly except that in
1417 ACLs where /readonly is required you may specify /leaky/noupdate or
1418 /strict/noupdate which are treated the same as /readonly.
1420 A useful new feature is the /count= option. This is a generalization
1421 of the per_byte option, so that you can measure the throughput of other
1422 aggregate values. For example, the per_byte option is now equivalent
1423 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1425 The per_rcpt option has been generalized using the /count= mechanism
1426 (though it's more complicated than the per_byte equivalence). When it is
1427 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1428 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1429 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1430 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1431 Note that using per_rcpt with a non-readonly update mode in more than
1432 one ACL will cause the recipients to be double-counted. (The per_mail
1433 and per_byte options don't have this problem.)
1435 The handling of very low rates has changed slightly. If the computed rate
1436 is less than the event's count (usually one) then this event is the first
1437 after a long gap. In this case the rate is set to the same as this event's
1438 count, so that the first message of a spam run is counted properly.
1440 The major new feature is a mechanism for counting the rate of unique
1441 events. The new per_addr option counts the number of different
1442 recipients that someone has sent messages to in the last time period. It
1443 behaves like per_rcpt if all the recipient addresses are different, but
1444 duplicate recipient addresses do not increase the measured rate. Like
1445 the /count= option this is a general mechanism, so the per_addr option
1446 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1447 example, measure the rate that a client uses different sender addresses
1448 with the options per_mail/unique=$sender_address. There are further
1449 details in the main documentation.
1451 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1453 TF/06 Removed a few PCRE remnants.
1455 TF/07 Automatically extract Exim's version number from tags in the git
1456 repository when doing development or release builds.
1458 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1459 Bugzilla 879. Patch from Paul Fisher.
1461 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1462 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1465 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1467 PP/05 Variable $av_failed, true if the AV scanner deferred.
1468 Bugzilla 1078. Patch from John Horne.
1470 PP/06 Stop make process more reliably on build failure.
1471 Bugzilla 1087. Patch from Heiko Schlittermann.
1473 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1474 Bugzilla 1089. Patch from Heiko Schlittermann.
1476 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1477 Bugzilla 1131. Patch from Holger Weiß.
1479 PP/09 Handle IPv6 addresses with SPF.
1480 Bugzilla 860. Patch from Wolfgang Breyha.
1482 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1484 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1487 PP/11 match_* no longer expand right-hand-side by default.
1488 New compile-time build option, EXPAND_LISTMATCH_RHS.
1489 New expansion conditions, "inlist", "inlisti".
1491 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1493 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1495 PP/14 fix log_write() format string regression from TF/03.
1496 Bugzilla 1152. Patch from Dmitry Isaikin.
1502 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1504 PP/02 Harmonised TLS library version reporting; only show if debugging.
1505 Layout now matches that introduced for other libraries in 4.74 PP/03.
1507 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1509 PP/04 New "dns_use_edns0" global option.
1511 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1514 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1515 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1517 TK/01 Updated PolarSSL code to 0.14.2.
1518 Bugzilla 1097. Patch from Andreas Metzler.
1520 PP/07 Catch divide-by-zero in ${eval:...}.
1521 Fixes bugzilla 1102.
1523 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1526 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1527 format-string attack -- SECURITY: remote arbitrary code execution.
1529 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1530 time unintentionally subject to list matching rules, letting the header
1531 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1532 arbitrary string expansion). This allowed for information disclosure.
1534 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1535 INT_MIN/-1 -- value coerced to INT_MAX.
1541 NM/01 Workaround for PCRE version dependency in version reporting
1544 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1545 This fixes portability to compilers other than gcc, notably
1546 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1548 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1549 makefiles for portability to HP-UX and POSIX correctness.
1551 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1552 Also via indented variable definition in the Makefile.
1553 (Debugging by Oliver Heesakkers).
1555 PP/02 Restore caching of spamd results with expanded spamd_address.
1556 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1558 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1559 Improves build reliability. Fix from: Frank Elsner
1561 NM/02 Fix wide character breakage in the rfc2047 coding
1562 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1564 NM/03 Allow underscore in dnslist lookups
1565 Fixes bug 1026. Patch from Graeme Fowler
1567 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1568 Code patches from Adam Ciarcinski of NetBSD.
1570 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1573 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1574 is logged, to avoid truncation. Patch from John Horne.
1576 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1577 Patch from Jakob Hirsch.
1579 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1580 SQL string expansion failure details.
1581 Patch from Andrey Oktyabrski.
1583 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1584 Patch from Simon Arlott.
1586 PP/09 New lookups functionality failed to compile on old gcc which rejects
1587 extern declarations in function scope.
1588 Patch from Oliver Fleischmann
1590 PP/10 Use sig_atomic_t for flags set from signal handlers.
1591 Check getgroups() return and improve debugging.
1592 Fixed developed for diagnosis in bug 927 (which turned out to be
1595 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1596 Patch from Mark Zealey.
1598 PP/12 Bugzilla 1056: Improved spamd server selection.
1599 Patch from Mark Zealey.
1601 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1602 Based on patch from Heiko Schlittermann.
1604 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1605 Patch from Uwe Doering, sign-off by Michael Haardt.
1607 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1608 variable declaration deep within a block. Bug and patch from
1611 PP/15 lookups-Makefile IRIX compatibility coercion.
1613 PP/16 Make DISABLE_DKIM build knob functional.
1615 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1616 Patch by Simon Arlott
1618 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1619 variable argument macros. Our copy now differs from upstream.
1625 TF/01 Failure to get a lock on a hints database can have serious
1626 consequences so log it to the panic log.
1628 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1629 controlled using the smtp_confirmation log selector.
1631 TF/03 Include the error message when we fail to unlink a spool file.
1633 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1634 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1635 for maintaining out-of-tree patches for some time.
1637 PP/01 Bugzilla 139: Documentation and portability issues.
1638 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1639 Handle per-OS dynamic-module compilation flags.
1641 PP/02 Let /dev/null have normal permissions.
1642 The 4.73 fixes were a little too stringent and complained about the
1643 permissions on /dev/null. Exempt it from some checks.
1644 Reported by Andreas M. Kirchwitz.
1646 PP/03 Report version information for many libraries, including
1647 Exim version information for dynamically loaded libraries. Created
1648 version.h, now support a version extension string for distributors
1649 who patch heavily. Dynamic module ABI change.
1651 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1652 privilege escalation vulnerability whereby the Exim run-time user
1653 can cause root to append content of the attacker's choosing to
1656 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1659 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1660 If dropping privileges for untrusted macros, we disabled normal logging
1661 on the basis that it would fail; for the Exim run-time user, this is not
1662 the case, and it resulted in successful deliveries going unlogged.
1663 Fixed. Reported by Andreas Metzler.
1669 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1670 only prepend for the Resent-* case. Fixes regression introduced in
1671 Exim 4.70 by NM/22 for Bugzilla 607.
1673 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1674 increasing numbers of administrators be bitten by this.
1676 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1678 PP/03 Bugzilla 994: added openssl_options main configuration option.
1680 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1682 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1684 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1686 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1687 without a peer certificate, leading to a segfault because of an
1688 assumption that peers always have certificates. Be a little more
1689 paranoid. Problem reported by Martin Tscholak.
1691 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1692 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1693 NB: ClamAV planning to remove STREAM in "middle of 2010".
1694 CL also introduces -bmalware, various -d+acl logging additions and
1695 more caution in buffer sizes.
1697 PP/09 Implemented reverse_ip expansion operator.
1699 PP/10 Bugzilla 937: provide a "debug" ACL control.
1701 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1703 PP/12 Bugzilla 973: Implement --version.
1705 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1707 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1709 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1711 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1712 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1713 ignore trailing whitespace.
1715 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1717 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1720 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1721 Notification from Dr Andrew Aitchison.
1723 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1724 ExtendedDetectionInfo response format.
1725 Notification from John Horne.
1727 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1730 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1731 XSL and documented dependency on system catalogs, with examples of how
1734 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1737 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1738 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1739 configuration file which is writeable by the Exim user or group.
1741 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1742 of configuration files to cover files specified with the -C option if
1743 they are going to be used with root privileges, not just the default
1746 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1747 option (effectively making it always true).
1749 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1750 files to be used while preserving root privileges.
1752 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1753 that rogue child processes cannot use them.
1755 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1756 run-time user, instead of root.
1758 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1759 Exim run-time user without dropping privileges.
1761 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1762 result string, instead of calling string_vformat() twice with the same
1765 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1766 for other users. Others should always drop root privileges if they use
1767 -C on the command line, even for a whitelisted configure file.
1769 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1771 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1777 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1778 $data_path, and $header_path variables; fixed documentation bugs and
1781 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1782 exipick to access non-standard spools, including the "frozen" queue
1785 NM/01 Bugzilla 965: Support mysql stored procedures.
1786 Patch from Alain Williams
1788 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1790 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1791 Patch from Andreas Metzler
1793 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1794 Patch from Kirill Miazine
1796 NM/05 Bugzilla 671: Added umask to procmail example.
1798 JJ/03 installed exipick 20100323.0, fixing doc bug
1800 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1801 directory. Notification and patch from Dan Rosenberg.
1803 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1805 TK/02 Improve log output when DKIM signing operation fails.
1807 MH/01 Treat the transport option dkim_domain as a colon separated
1808 list, not as a single string, and sign the message with each element,
1809 omitting multiple occurences of the same signer.
1811 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1812 Bugzilla 985, 986. Patch by Simon Arlott
1814 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1815 Patch by Simon Arlott
1817 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1818 MBX locking. Notification from Dan Rosenberg.
1824 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1826 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1828 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1830 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1832 NM/04 Bugzilla 915: Flag broken perl installation during build.
1838 TK/01 Added patch by Johannes Berg that expands the main option
1839 "spamd_address" if it starts with a dollar sign.
1841 TK/02 Write list of recipients to X-Envelope-Sender header when building
1842 the mbox-format spool file for content scanning (suggested by Jakob
1845 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1846 (http://www.dcc-servers.net/) support via dccifd. Activated by
1847 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1849 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1850 by Mark Daniel Reidel <mr@df.eu>.
1852 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1853 When building exim an external PCRE library is now needed -
1854 PCRE is a system library on the majority of modern systems.
1855 See entry on PCRE_LIBS in EDITME file.
1857 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1858 conversation. Added nologin parameter to request.
1859 Patch contributed by Kirill Miazine.
1861 TF/01 Do not log submission mode rewrites if they do not change the address.
1863 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1865 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1866 log files in place. Contributed by Roberto Lima.
1868 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1870 TF/03 Bugzilla 615: When checking the local_parts router precondition
1871 after a local_part_suffix or local_part_prefix option, Exim now
1872 does not use the address's named list lookup cache, since this
1873 contains cached lookups for the whole local part.
1875 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1876 Robert Millan. Documentation is in experimental-spec.txt.
1878 TF/04 Bugzilla 668: Fix parallel build (make -j).
1880 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1882 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1883 Patch provided by Jan Srzednicki.
1885 TF/05 Leading white space used to be stripped from $spam_report which
1886 wrecked the formatting. Now it is preserved.
1888 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1889 that they are available at delivery time.
1891 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1893 TF/08 TLS error reporting now respects the incoming_interface and
1894 incoming_port log selectors.
1896 TF/09 Produce a more useful error message if an SMTP transport's hosts
1897 setting expands to an empty string.
1899 NM/06 Bugzilla 744: EXPN did not work under TLS.
1900 Patch provided by Phil Pennock.
1902 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1903 Patch provided by Richard Godbee.
1905 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1906 acl_smtp_notquit, added index entry.
1908 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1909 Patch provided by Eugene Bujak.
1911 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1912 accept(). Patch provided by Maxim Dounin.
1914 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1915 Patch provided by Phil Pennock.
1917 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1919 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1920 Patch provided by Brad "anomie" Jorsch.
1922 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1923 Patch provided by Dean Brooks.
1925 TK/05 Add native DKIM support (does not depend on external libraries).
1927 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1928 Patch provided by Graeme Fowler.
1930 NM/16 Bugzilla 851: Documentation example syntax fix.
1932 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1934 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1937 NM/19 Bugzilla 745: TLS version reporting.
1938 Patch provided by Phil Pennock.
1940 NM/20 Bugzilla 167: bool: condition support.
1941 Patch provided by Phil Pennock.
1943 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1944 clients. Patch provided by Phil Pennock.
1946 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1947 Patch provided by Brad "anomie" Jorsch.
1949 NM/23 Bugzilla 687: Fix misparses in eximstats.
1950 Patch provided by Heiko Schlittermann.
1952 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1953 Patch provided by Heiko Schlittermann.
1955 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1956 plus update to original patch.
1958 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1960 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1961 Patch provided by David Brownlee.
1963 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1965 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1967 NM/30 Bugzilla 888: TLS documentation bugfixes.
1969 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1971 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1972 Unlike the original bugzilla I have changed all shell scripts in src tree.
1974 NM/33 Bugzilla 898: Transport filter timeout fix.
1975 Patch by Todd Rinaldo.
1977 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1978 Patch by Serge Demonchaux.
1980 NM/35 Bugzilla 39: Base64 decode bug fixes.
1981 Patch by Jakob Hirsch.
1983 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1985 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1987 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1989 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1995 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1996 ALT-N's libdkim that I have put here:
1997 http://duncanthrax.net/exim-experimental/
1999 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2000 were called 'true' and it seems that C99 defines that as a reserved
2001 keyword to be used with 'bool' variable types. That means you could
2002 not include C99-style headers which use bools without triggering
2003 build errors in sieve.c.
2005 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2006 as mailq or other aliases. Changed the --help handling significantly
2007 to do whats expected. exim_usage() emits usage/help information.
2009 SC/01 Added the -bylocaldomain option to eximstats.
2011 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2013 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2015 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2021 PH/01 Another patch from the Sieve maintainer.
2023 PH/02 When an IPv6 address is converted to a string for single-key lookup
2024 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2025 dots are used instead of colons so that keys in lsearch files need not
2026 contain colons. This was done some time before quoting was made available
2027 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2028 (notated using the quote facility) so as to distinguish them from IPv4
2029 keys. This meant that lookups for IP addresses in host lists did not work
2030 for iplsearch lookups.
2032 This has been fixed by arranging for IPv6 addresses to be expressed with
2033 colons if the lookup type is iplsearch. This is not incompatible, because
2034 previously such lookups could never work.
2036 The situation is now rather anomalous, since one *can* have colons in
2037 ordinary lsearch keys. However, making the change in all cases is
2038 incompatible and would probably break a number of configurations.
2040 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2043 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2044 conversion specification without a maximum field width, thereby enabling
2045 a rogue spamd server to cause a buffer overflow. While nobody in their
2046 right mind would setup Exim to query an untrusted spamd server, an
2047 attacker that gains access to a server running spamd could potentially
2048 exploit this vulnerability to run arbitrary code as the Exim user.
2050 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2051 $primary_hostname instead of what libspf2 thinks the hosts name is.
2053 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2054 a directory entry by the name of the lookup key. Previously, if a
2055 symlink pointed to a non-existing file or a file in a directory that
2056 Exim lacked permissions to read, a lookup for a key matching that
2057 symlink would fail. Now it is enough that a matching directory entry
2058 exists, symlink or not. (Bugzilla 503.)
2060 PH/03 The body_linecount and body_zerocount variables are now exported in the
2063 PH/04 Added the $dnslist_matched variable.
2065 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2066 This means they are set thereafter only if the connection becomes
2069 PH/06 Added the client_condition to authenticators so that some can be skipped
2070 by clients under certain conditions.
2072 PH/07 The error message for a badly-placed control=no_multiline_responses left
2073 "_responses" off the end of the name.
2075 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2077 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2078 (without spaces) instead of just copying the configuration text.
2080 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2082 PH/11 Added $max_received_linelength.
2084 PH/12 Added +ignore_defer and +include_defer to host lists.
2086 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2087 way in which PCRE > 7.0 is built.
2089 PH/14 Implemented queue_only_load_latch.
2091 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2092 MAIL command. The effect was to mangle the value on 64-bit systems.
2094 PH/16 Another patch from the Sieve maintainer.
2096 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2098 PH/18 If a system quota error occurred while trying to create the file for
2099 a maildir delivery, the message "Mailbox is full" was not appended to the
2100 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2101 only to a quota excession during the actual writing of the file.
2103 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2104 characters?) which causes problems in log lines. The DN values are now
2105 passed through string_printing() before being added to log lines.
2107 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2108 and InterBase are left for another time.)
2110 PH/21 Added message_body_newlines option.
2112 PH/22 Guard against possible overflow in moan_check_errorcopy().
2114 PH/23 POSIX allows open() to be a macro; guard against that.
2116 PH/24 If the recipient of an error message contained an @ in the local part
2117 (suitably quoted, of course), incorrect values were put in $domain and
2118 $local_part during the evaluation of errors_copy.
2124 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2125 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2128 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2129 issue a MAIL command.
2131 PH/02 In an ACL statement such as
2133 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2135 if a client was not listed at all, or was listed with a value other than
2136 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2137 the condition was not true (as it should be), so access was not denied.
2138 The bug was that the ! inversion was incorrectly passed on to the second
2139 item. This has been fixed.
2141 PH/03 Added additional dnslists conditions == and =& which are different from
2142 = and & when the dns lookup returns more than one IP address.
2144 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2145 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2147 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2148 FSYNC, which compiles an option called disable_fsync that allows for
2149 bypassing fsync(). The documentation is heavily laced with warnings.
2151 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2153 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2154 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2155 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2156 including adding "make clean"; (3) Added -fPIC when compiling the test
2157 dynamically loaded module, to get rid of a warning.
2159 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2160 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2161 = 0s. The bug is otherwise harmless.
2163 PH/07 There was a bug in the dovecot authenticator such that the value of
2164 $auth1 could be overwritten, and so not correctly preserved, after a
2165 successful authentication. This usually meant that the value preserved by
2166 the server_setid option was incorrect.
2168 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2170 PH/09 Installed PCRE release 7.0.
2172 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2173 run for batched SMTP input. It is now run at the start of every message
2174 in the batch. While fixing this I discovered that the process information
2175 (output by running exiwhat) was not always getting set for -bs and -bS
2176 input. This is fixed, and it now also says "batched" for BSMTP.
2178 PH/11 Added control=no_pipelining.
2180 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2181 patch, slightly modified), and move the expansion of helo_data till after
2182 the connection is made in the smtp transport (so it can use these
2185 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2187 PH/14 Added log_selector = +pid.
2189 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2191 PH/16 Add ${if forany and ${if forall.
2193 PH/17 Added dsn_from option to vary the From: line in DSNs.
2195 PH/18 Flush SMTP output before performing a callout, unless control =
2196 no_callout_flush is set.
2198 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2199 was true (the default) a successful delivery failed to delete the retry
2200 item, thus causing premature timeout of the address. The bug is now
2203 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2205 PH/21 Long custom messages for fakedefer and fakereject are now split up
2206 into multiline responses in the same way that messages for "deny" and
2207 other ACL rejections are.
2209 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2210 with slight modification.
2212 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2213 draft, changing the syntax and factoring some duplicate code".
2215 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2216 for deliveries of the second and subsequent messages over the same SMTP
2219 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2220 ${reduce, with only minor "tidies".
2222 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2224 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2225 expansion side effects.
2227 PH/27 When a message times out after an over-quota error from an Exim-imposed
2228 quota, the bounce message says "mailbox is full". This message was not
2229 being given when it was a system quota that was exceeded. It now should
2232 MH/03 Made $recipients available in local_scan(). local_scan() already has
2233 better access to the recipient list through recipients_list[], but
2234 $recipients can be useful in postmaster-provided expansion strings.
2236 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2237 in the case of a MAIL command with additional options following the
2238 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2239 were accidentally chopped off.
2241 PH/29 SMTP synchronization checks are implemented when a command is read -
2242 there is a check that no more input is waiting when there shouldn't be
2243 any. However, for some commands, a delay in an ACL can mean that it is
2244 some time before the response is written. In this time, more input might
2245 arrive, invalidly. So now there are extra checks after an ACL has run for
2246 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2247 pipelining has not been advertised.
2249 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2251 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2252 returned in the SMTP response when smtp_return_error_details was false.
2253 This has been fixed.
2255 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2256 instead of the C I/O that was originally supplied, because problems were
2257 reported on Solaris.
2259 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2260 Exim which did not show up earlier: it was assuming that a call to
2261 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2262 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2263 was a macro that became an assignment, so it seemed to work. This has
2264 changed to a proper function call with a void return, hence the compile
2265 error. Exim's code has been fixed.
2267 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2270 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2272 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2274 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2275 because it turns out to be too big or there is a timeout) there is no
2276 "Completed" line in the log. When some messages of this type were
2277 selected by exigrep, they were listed as "not completed". Others were
2278 picked up by some special patterns. I have improved the selection
2279 criteria to be more general.
2281 PH/38 The host_find_failed option in the manualroute router can now be set
2282 to "ignore", to completely ignore a host whose IP address cannot be
2283 found. If all hosts are ignored, the behaviour is controlled by the new
2284 host_all_ignored option.
2286 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2287 homing or because of multiple MX records with /mx) generated more than
2288 one IP address, and the following item turned out to be the local host,
2289 all the secondary addresses of the first item were incorrectly removed
2290 from the list, along with the local host and any following hosts (which
2291 is what is supposed to happen).
2293 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2294 whoever called Exim into the -H file. In the case of the daemon it was
2295 behaving confusingly. When first started, it used values for whoever
2296 started the daemon, but after a SIGHUP it used the Exim user (because it
2297 calls itself on a restart). I have changed the code so that it now always
2300 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2301 message are rejected with the same error (e.g. no authentication or bad
2302 sender address), and a DATA command is nevertheless sent (as can happen
2303 with PIPELINING or a stupid MUA), the error message that was given to the
2304 RCPT commands is included in the rejection of the DATA command. This is
2305 intended to be helpful for MUAs that show only the final error to their
2308 PH/42 Another patch from the Sieve maintainer.
2310 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2311 Eximstats - Fixed some broken HTML links and added missing column headers
2313 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2314 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2316 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2318 PH/43 Yet another patch from the Sieve maintainer.
2320 PH/44 I found a way to check for a TCP/IP connection going away before sending
2321 the response to the final '.' that terminates a message, but only in the
2322 case where the client has not sent further data following the '.'
2323 (unfortunately, this is allowed). However, in many cases there won't be
2324 any further data because there won't be any more messages to send. A call
2325 to select() can be used: if it shows that the input is "ready", there is
2326 either input waiting, or the socket has been closed. An attempt to read
2327 the next input character can distinguish the two cases. Previously, Exim
2328 would have sent an OK response which the client would never have see.
2329 This could lead to message repetition. This fix should cure that, at
2330 least in a lot of common cases.
2332 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2333 advertised in response to EHLO.
2339 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2340 fixed by 4.65/MH/01 (is this a record?) are fixed:
2342 (i) An empty string was always treated as zero by the numeric comparison
2343 operators. This behaviour has been restored.
2345 (ii) It is documented that the numeric comparison operators always treat
2346 their arguments as decimal numbers. This was broken in that numbers
2347 starting with 0 were being interpreted as octal.
2349 While fixing these problems I realized that there was another issue that
2350 hadn't been noticed. Values of message_size_limit (both the global option
2351 and the transport option) were treated as octal if they started with 0.
2352 The documentation was vague. These values are now always treated as
2353 decimal, and I will make that clear in the documentation.
2359 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2360 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2363 MH/01 Don't check that the operands of numeric comparison operators are
2364 integers when their expansion is in "skipping" mode (fixes bug
2365 introduced by 4.64-PH/07).
2367 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2368 child addresses, Exim now panics and dies. Previously, because the count
2369 is held in a short int, deliveries were likely to be lost. As such a
2370 large number of recipients for a single message is ridiculous
2371 (performance will be very, very poor), I have chosen to impose a limit
2372 rather than extend the field.
2378 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2379 leftover -K file (the existence of which was triggered by #402).
2380 While we were at it, introduced process PID as part of the -K
2381 filename. This should rule out race conditions when creating
2384 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2385 processing considerably. Previous code took too long for large mails,
2386 triggering a timeout which in turn triggers #401.
2388 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2389 in the DK code in transports.c. sendfile() is not really portable,
2390 hence the _LINUX specificness.
2392 TF/01 In the add_headers option to the mail command in an Exim filter,
2393 there was a bug that Exim would claim a syntax error in any
2394 header after the first one which had an odd number of characters
2397 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2398 callout verification, Exim cached a "reject" for the entire domain. This
2399 is correct for most verifications, but it is not correct for a recipient
2400 verification with use_sender or use_postmaster set, because in that case
2401 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2402 case of MAIL FROM:<> rejection from other early rejections (e.g.
2403 rejection of HELO). When verifying a recipient using a non-null MAIL
2404 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2405 Whatever the result of the callout, the value of the domain cache is
2406 left unchanged (for any other kind of callout, getting as far as trying
2407 RCPT means that the domain itself is ok).
2409 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2412 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2413 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2414 session, the daemon ignores ECONNECT errors and logs others; it now
2415 ignores EPIPE as well.
2417 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2418 (quoted-printable decoding).
2420 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2421 later the small subsequent patch to fix an introduced bug.
2423 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2425 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2427 PH/08 An error is now given if message_size_limit is specified negative.
2429 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2430 to be given (somewhat) arbitrary names.
2432 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2435 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2436 miscellaneous code fixes
2438 PH/10 Added the log_reject_target ACL modifier to specify where to log
2441 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2442 hostname. This is wrong, because it relates to the incoming message (and
2443 probably the interface on which it is arriving) and not to the outgoing
2444 callout (which could be using a different interface). This has been
2445 changed to use the value of the helo_data option from the smtp transport
2446 instead - this is what is used when a message is actually being sent. If
2447 there is no remote transport (possible with a router that sets up host
2448 addresses), $smtp_active_hostname is used.
2450 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2451 tweaks were necessary in order to get it to work (see also 21 below):
2452 (a) The code assumed that strncpy() returns a negative number on buffer
2453 overflow, which isn't the case. Replaced with Exim's string_format()
2455 (b) There were several signed/unsigned issues. I just did the minimum
2456 hacking in of casts. There is scope for a larger refactoring.
2457 (c) The code used strcasecmp() which is not a standard C function.
2458 Replaced with Exim's strcmpic() function.
2459 (d) The code set only $1; it now sets $auth1 as well.
2460 (e) A simple test gave the error "authentication client didn't specify
2461 service in request". It would seem that Dovecot has changed its
2462 interface. Fortunately there's a specification; I followed it and
2463 changed what the client sends and it appears to be working now.
2465 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2468 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2469 address A is aliased to B and C, where B exists and C does not. Without
2470 -v the output is "A verified" because verification stops after a
2471 successful redirection if more than one address is generated. However,
2472 with -v the child addresses are also verified. Exim was outputting "A
2473 failed to verify" and then showing the successful verification for C,
2474 with its parentage. It now outputs "B failed to verify", showing B's
2475 parentage before showing the successful verification of C.
2477 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2478 look up a TXT record in a specific list after matching in a combined
2481 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2482 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2483 they consult the DNS. I had assumed they would set it the way they
2484 wanted; and indeed my experiments on Linux seem to show that in some
2485 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2486 To be on the safe side, however, I have now made the interface to
2487 host_find_byname() similar to host_find_bydns(), with an argument
2488 containing the DNS resolver options. The host_find_byname() function now
2489 sets these options at its start, just as host_find_bydns() does. The smtp
2490 transport options dns_qualify_single and dns_search_parents are passed to
2491 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2492 of host_find_byname() use the default settings of RES_DEFNAMES
2493 (qualify_single) but not RES_DNSRCH (search_parents).
2495 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2496 spool_read_header() do less string testing, by means of a preliminary
2497 switch on the second character of optional "-foo" lines. (This is
2498 overdue, caused by the large number of possibilities that now exist.
2499 Originally there were few.) While I was there, I also converted the
2500 str(n)cmp tests so they don't re-test the leading "-" and the first
2501 character, in the hope this might squeeze out yet more improvement.
2503 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2504 flag allowing group syntax was set by the header_syntax check but not
2505 turned off, possible causing trouble later; (2) The flag was not being
2506 set at all for the header_verify test, causing "group"-style headers to
2507 be rejected. I have now set it in this case, and also caused header_
2508 verify to ignore an empty address taken from a group. While doing this, I
2509 came across some other cases where the code for allowing group syntax
2510 while scanning a header line wasn't quite right (mostly, not resetting
2511 the flag correctly in the right place). These bugs could have caused
2512 trouble for malformed header lines. I hope it is now all correct.
2514 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2515 with the "reply" argument non-NULL. The code, however (which originally
2516 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2517 but it didn't always do it. This confused somebody who was copying the
2518 code for some other use. I have removed all the tests.
2520 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2521 feature that was used to support insecure browsers during the U.S. crypto
2522 embargo. It requires special client support, and Exim is probably the
2523 only MTA that supported it -- and would never use it because real RSA is
2524 always available. This code has been removed, because it had the bad
2525 effect of slowing Exim down by computing (never used) parameters for the
2526 RSA_EXPORT functionality.
2528 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2529 authenticator to fail if there's a tab character in the incoming data
2530 (there should never be unless someone is messing about, as it's supposed
2531 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2532 if the connection is using TLS or if the remote IP is the same as the
2533 local IP, and the "valid-client-cert option" if a client certificate has
2536 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2537 authenticators. This can be used for authorization after authentication
2538 succeeds. (In the case of plaintext, it servers for both authentication
2541 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2542 if any retry times were supplied.
2544 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2545 connection, where there is no client IP address to check. In this
2546 situation, the verify now always succeeds.
2548 PH/25 Applied John Jetmore's -Mset patch.
2550 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2552 PH/27 In a string expansion for a processed (not raw) header when multiple
2553 headers of the same name were present, leading whitespace was being
2554 removed from all of them, but trailing whitespace was being removed only
2555 from the last one. Now trailing whitespace is removed from each header
2556 before concatenation. Completely empty headers in a concatenation (as
2557 before) are ignored.
2559 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2560 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2562 PH/29 [Removed. This was a change that I later backed out, and forgot to
2563 correct the ChangeLog entry (that I had efficiently created) before
2564 committing the later change.]
2566 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2567 address errors (4xx response to RCPT) over the same connection as other
2568 messages routed to the same hosts. Such deliveries are always "forced",
2569 so retry times are not inspected. This resulted in far too many retries
2570 for the affected addresses. The effect occurred only when there were more
2571 hosts than the hosts_max_try setting in the smtp transport when it had
2572 the 4xx errors. Those hosts that it had tried were not added to the list
2573 of hosts for which the message was waiting, so if all were tried, there
2574 was no problem. Two fixes have been applied:
2576 (i) If there are any address or message errors in an SMTP delivery, none
2577 of the hosts (tried or untried) are now added to the list of hosts
2578 for which the message is waiting, so the message should not be a
2579 candidate for sending over the same connection that was used for a
2580 successful delivery of some other message. This seems entirely
2581 reasonable: after all the message is NOT "waiting for some host".
2582 This is so "obvious" that I'm not sure why it wasn't done
2583 previously. Hope I haven't missed anything, but it can't do any
2584 harm, as the worst effect is to miss an optimization.
2586 (ii) If, despite (i), such a delivery is accidentally attempted, the
2587 routing retry time is respected, so at least it doesn't keep
2588 hammering the server.
2590 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2591 in ${readsocket because some servers need this prod.
2593 PH/32 Added some extra debug output when updating a wait-xxx database.
2595 PH/33 The hint "could be header name not terminated by colon", which has been
2596 given for certain expansion errors for a long time, was not being given
2597 for the ${if def:h_colon_omitted{... case.
2599 PH/34 The spec says: "With one important exception, whenever a domain list is
2600 being scanned, $domain contains the subject domain." There was at least
2601 one case where this was not true.
2603 PH/35 The error "getsockname() failed: connection reset by peer" was being
2604 written to the panic log as well as the main log, but it isn't really
2605 panic-worthy as it just means the connection died rather early on. I have
2606 removed the panic log writing for the ECONNRESET error when getsockname()
2609 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2610 runs only) independently of the message's sender address. This meant
2611 that, if the 4xx error was in fact related to the sender, a different
2612 message to the same recipient with a different sender could confuse
2613 things. In particular, this can happen when sending to a greylisting
2614 server, but other circumstances could also provoke similar problems.
2615 I have changed the default so that the retry time for these errors is now
2616 based a combination of the sender and recipient addresses. This change
2617 can be overridden by setting address_retry_include_sender=false in the
2620 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2621 remote server are returned as part of bounce messages. This was not
2622 happening for LMTP over a pipe (the lmtp transport), but now it is the
2623 same for both kinds of LMTP.
2625 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2626 in header lines that were in fact CNAMEs. This is no longer the case.
2628 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2629 and queue runs started by the daemon processed all messages. This has
2630 been fixed so that -R and -S can now usefully be given with -q<time>.
2632 PH/40 Import PCRE release 6.7 (fixes some bugs).
2634 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2636 PH/42 Give an error if -q is specified more than once.
2638 PH/43 Renamed the variables $interface_address and $interface_port as
2639 $received_ip_address and $received_port, to make it clear that these
2640 values apply to message reception, and not to the outgoing interface when
2641 a message is delivered. (The old names remain recognized, of course.)
2643 PH/44 There was no timeout on the connect() call when using a Unix domain
2644 socket in the ${readsocket expansion. There now is.
2646 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2647 be meaningful with "accept".
2649 SC/01 Eximstats V1.43
2650 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2652 SC/02 Eximstats V1.44
2653 Use a glob alias rather than an array ref in the generated
2654 parser. This improves both readability and performance.
2656 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2657 Collect SpamAssassin and rejection statistics.
2658 Don't display local sender or destination tables unless
2659 there is data to show.
2660 Added average volumes into the top table text output.
2662 SC/04 Eximstats V1.46
2663 Collect data on the number of addresses (recipients)
2664 as well as the number of messages.
2666 SC/05 Eximstats V1.47
2667 Added 'Message too big' to the list of mail rejection
2668 reasons (thanks to Marco Gaiarin).
2670 SC/06 Eximstats V1.48
2671 Mainlog lines which have GMT offsets and are too short to
2672 have a flag are now skipped.
2674 SC/07 Eximstats V1.49 (Alain Williams)
2675 Added the -emptyok flag.
2677 SC/08 Eximstats V1.50
2678 Fixes for obtaining the IP address from reject messages.
2680 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2681 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2682 whitespace changes from 4.64-PH/27
2684 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2687 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2688 are found, allow negative numbers in numeric criteria)
2690 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2692 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2693 to match changes made in 4.64-PH/43
2695 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2697 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2698 "domain" is actually an IP address, and force a failure. This locks out
2699 those revolvers/nameservers that support "A-for-A" lookups, in
2700 contravention of the specifications.
2702 PH/48 When a host name was looked up from an IP address, and the subsequent
2703 forward lookup of the name timed out, the host name was left in
2704 $sender_host_name, contrary to the specification.
2706 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2707 restricted to single-key lookups, Exim was not diagnosing an error if
2708 * or *@ was used with a query-style lookup.
2710 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2712 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2713 long ago, but noone interested enough thought of it. Let's just say that
2714 the "1.1" means that there are some new functions that weren't there at
2715 some point in the past.
2717 PH/51 Error processing for expansion failure of helo_data from an smtp
2718 transport during callout processing was broken.
2720 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2721 tested/used via the -bh/-bhc/-bs options.
2723 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2724 bug, fixed in subsequent PCRE releases).
2726 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2727 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2729 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2735 SC/01 Use a glob alias rather than an array ref in eximstats generated
2736 parser. This improves both readability and performance.
2738 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2739 Don't display local sender or destination tables in eximstats unless
2740 there is data to show.
2741 Added average volumes into the eximstats top table text output.
2743 SC/03 Collect data on the number of addresses (recipients) as well
2744 as the number of messages in eximstats.
2746 TF/01 Correct an error in the documentation for the redirect router. Exim
2747 does (usually) call initgroups() when daemonizing.
2749 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2750 with consistent privilege compared to when running as a daemon.
2752 TF/03 Note in the spec that $authenticated_id is not set for local
2753 submissions from trusted users.
2755 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2756 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2758 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2759 by adding some example configuration directives to the default
2760 configuration file. A little bit of work is required to uncomment the
2761 directives and define how usernames and passwords are checked, but
2762 there is now a framework to start from.
2764 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2765 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2766 without this. I don't know how relevant this is to other LDAP libraries.
2768 PH/02 Add the verb name to the "unknown ACL verb" error.
2770 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2772 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2774 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2775 directory not expanded when it should be if an expanded home directory
2776 was set for the address (which is overridden by the transport).
2778 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2781 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2782 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2783 because it is too late at that time, and has no effect.
2785 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2786 security issue with \' (bugzilla #107). I could not use the
2787 PQescapeStringConn() function, because it needs a PGconn value as one of
2790 PH/08 When testing addresses using -bt, indicate those final addresses that
2791 are duplicates that would not cause an additional delivery. At least one
2792 person was confused, thinking that -bt output corresponded to deliveries.
2793 (Suppressing duplicates isn't a good idea as you lose the information
2794 about possibly different redirections that led to the duplicates.)
2796 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2797 systems where poll() doesn't work, in particular OS X.
2799 PH/10 Added more information to debugging output for retry time not reached.
2801 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2802 operations in malware.c.
2804 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2807 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2808 syslog_duplication set false, log lines that would normally be written
2809 both the the main log and to the reject log were not written to syslog at
2812 PH/14 In the default configuration, change the use of "message" in ACL warn
2813 statements to "add_header".
2815 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2816 not followed by a command (e.g. "seen endif").
2818 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2819 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2822 PH/17 Added extra conditions to the default value of delay_warning_condition
2826 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2827 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2828 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2831 The Auto-Submitted: and various List- headers are standardised, whereas I
2832 don't think Precedence: ever was.
2834 PH/18 Refactored debugging code in route_finduser() to show more information,
2835 in particular, the error code if getpwnam() issues one.
2837 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2838 This is apparently needed in addition to the PH/07 change above to avoid
2839 any possible encoding problems.
2841 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2842 but not after initializing Perl.
2844 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2845 output them only if debugging. By default they are written stderr,
2846 apparently, which is not desirable.
2848 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2851 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2854 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2856 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2857 authenticated or an ident call has been made. Suppress the default
2858 values for $authenticated_id and $authenticated_sender (but permit -oMai
2859 and -oMas) when testing with -bh.
2861 PH/24 Re-jigged the order of the tests in the default configuration so that the
2862 tests for valid domains and recipients precede the DNS black list and CSA
2863 tests, on the grounds that those ones are more expensive.
2865 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2866 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2867 This bug exists in every version of Exim that I still have, right back to
2870 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2871 However, an attempt to turn on case-sensitivity in a regex key by
2872 including (?-i) didn't work because the subject string was already
2873 lowercased, and the effects were non-intuitive. It turns out that a
2874 one-line patch can be used to allow (?-i) to work as expected.
2880 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2881 other effects) broke the use of negated acl sub-conditions.
2883 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2886 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2887 "Deny" causes Exim to reject the incoming connection with a 554 error.
2888 Unfortunately, if there is a major crisis, such as a disk failure,
2889 tcp-wrappers gives "deny", whereas what one would like would be some
2890 kind of temporary error. A kludge has been added to help with this.
2891 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2892 554 error is used if errno is still zero or contains ENOENT (which occurs
2893 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2896 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2898 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2899 errors. Otherwise a message that provokes a temporary error (when other
2900 messages do not) can cause a whole host to time out.
2902 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2903 addresses were routed directly to files or pipes from a redirect router.
2904 File deliveries just didn't batch; pipe deliveries might have suffered
2907 PH/06 A failure to get a lock for a hints database would erroneously always say
2908 "Failed to get write lock", even when it was really a read lock.
2910 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2911 of 0600. This has been changed to use the value of the lockfile_mode
2912 option (which defaults to 0600).
2914 PH/08 Applied small patch from the Sieve maintainer.
2916 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2917 folder from quota calculations, a direct delivery into this folder messed
2918 up the contents of the maildirsize file. This was because the regex was
2919 used only to exclude .Trash (or whatever) when the size of the mailbox
2920 was calculated. There was no check that a delivery was happening into an
2921 excluded directory. This bug has been fixed by ignoring all quota
2922 processing for deliveries into excluded directories.
2924 PH/10 Added the maildirfolder_create_regex option to appendfile.
2930 PH/01 The code for finding all the local interface addresses on a FreeBSD
2931 system running IPv6 was broken. This may well have applied to all BSD
2932 systems, as well as to others that have similar system calls. The broken
2933 code found IPv4 interfaces correctly, but gave incorrect values for the
2934 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2935 that it would not match correctly against @[] and not recognize the IPv6
2938 PH/02 The ipliteral router was not recognizing addresses of the form user@
2939 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2941 PH/03 Added disable_ipv6.
2943 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2944 decoded form, because it is most often used to construct To: headers
2945 lines in autoreplies, and the decoded form may well be syntactically
2946 invalid. However, $reply_address has leading white space removed, and all
2947 newlines turned into spaces so that the autoreply transport does not
2950 PH/05 If group was specified without a user on a router, and no group or user
2951 was specified on a transport, the group from the router was ignored.
2953 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2954 for visible compile-time settings that can be used to change these
2955 numbers, for those that want even more. Backwards compatibility with old
2956 spool files has been maintained. However, going back to a previous Exim
2957 release will lost any variables that are in spool files.
2959 PH/07 Two small changes when running in the test harness: increase delay when
2960 passing a TCP/IP connection to a new process, in case the original
2961 process has to generate a bounce, and remove special handling of
2962 127.0.0.2 (sic), which is no longer necessary.
2964 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2965 be the same on different OS.
2967 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2970 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2971 whether --show-vars was specified or not
2973 JJ/02 exipick: Added support for new ACL variable spool format introduced
2976 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2977 syntactically invalid From: or Reply-to: line, and a filter used this to
2978 generate an autoreply, and therefore failed to obtain an address for the
2979 autoreply, Exim could try to deliver to a non-existent relative file
2980 name, causing unrelated and misleading errors. What now happens is that
2981 it logs this as a hard delivery error, but does not attempt to create a
2984 PH/11 The exinext utility has a -C option for testing purposes, but although
2985 the given file was scanned by exinext itself; it wasn't being passed on
2986 when Exim was called.
2988 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2989 an end-of-file indication when reading a command response.
2991 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2992 compiled. In many other places in Exim, IPv6 addresses are always
2993 recognized, so I have changed this. It also means that IPv4 domain
2994 literals of the form [IPV4:n.n.n.n] are now always recognized.
2996 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2997 used if the router is not running as root, for example, when verifying at
2998 ACL time, or when using -bh. The debugging output from this situation was
2999 non-existent - all you got was a failure to exec. I have made two
3002 (a) Failures to set uid/gid, the current directory, or a process leader
3003 in a subprocess such as that created by queryprogram now generate
3004 suitable debugging output when -d is set.
3006 (b) The queryprogram router detects when it is not running as root,
3007 outputs suitable debugging information if -d is set, and then runs
3008 the subprocess without attempting to change uid/gid.
3010 PH/15 Minor change to Makefile for building test_host (undocumented testing
3013 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3014 additional section of a DNS packet that returns MX or SRV records.
3015 Instead, it always explicitly searches for A/AAAA records. This avoids
3016 major problems that occur when a DNS server includes only records of one
3017 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3018 fixed another bug: if SRV records were looked up and the corresponding
3019 address records were *not* found in the additional section, the port
3020 values from the SRV records were lost.
3022 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3023 using the correct key (the original address) when searching the retry
3024 rules in order to find which one to use for generating the retry hint.
3026 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3027 adding the default one. Similarly, if it contains a Reply-To: header, the
3028 errors_reply_to option, if set, is not used.
3030 PH/19 When calculating a retry time, Exim used to measure the "time since
3031 failure" by looking at the "first failed" field in the retry record. Now
3032 it does not use this if it is later than than the arrival time of the
3033 message. Instead it uses the arrival time. This makes for better
3034 behaviour in cases where some deliveries succeed, thus re-setting the
3035 "first failed" field. An example is a quota failure for a huge message
3036 when small messages continue to be delivered. Without this change, the
3037 "time since failure" will always be short, possible causing more frequent
3038 delivery attempts for the huge message than are intended.
3039 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3041 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3042 $1, $2, $3) because the numerical variables can be reset during some
3043 expansion items (e.g. "match"), thereby losing the authentication data.
3045 PH/21 Make -bV show the size of off_t variables so that the test suite can
3046 decide whether to run tests for quotas > 2G.
3048 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3049 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3050 filecount value is greater than 2G or if a quota value is greater than 2G
3051 on a system where the size of off_t is not greater than 4, a panic error
3054 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3055 never match. The debug and -bh output now contains an explicit error
3056 message indicating a malformed IPv4 address or mask.
3058 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3059 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3060 PH/23 above applies.
3062 PH/25 Do not write to syslog when running in the test harness. The only
3063 occasion when this arises is a failure to open the main or panic logs
3064 (for which there is an explicit test).
3066 PH/26 Added the /no_tell option to "control=freeze".
3068 PH/27 If a host name lookup failed very early in a connection, for example, if
3069 the IP address matched host_lookup and the reverse lookup yielded a name
3070 that did not have a forward lookup, an error message of the form "no IP
3071 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3072 could be logged. Now it outputs the IP address instead of "NULL".
3074 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3075 allows the sender and the authenticated sender to be set when
3076 submitting a message from within Exim. Since child_open_exim() is
3077 documented for local_scan(), the new function should be too.
3079 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3080 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3081 results in an empty string is now treated as unset.
3083 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3085 PH/31 Added sender_verify_fail logging option.
3087 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3088 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3089 tidied the source and removed it altogether.
3091 PH/33 When a queue run was abandoned because the load average was too high, a
3092 log line was always written; now it is written only if the queue_run log
3093 selector is set. In addition, the log line for abandonment now contains
3094 information about the queue run such as the pid. This is always present
3095 in "start" and "stop" lines but was omitted from the "abandon" line.
3097 PH/34 Omit spaces between a header name and the colon in the error message that
3098 is given when verify = headers_syntax fails (if there are lots of them,
3099 the message gets confusing).
3101 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3102 names, as there are now some PTR records that contain slashes. This check
3103 is only to protect against broken name servers that fall over on strange
3104 characters, so the fact that it applies to all lookups doesn't matter.
3106 PH/36 Now that the new test suite is complete, we can remove some of the
3107 special code in Exim that was needed for the old test suite. For example,
3108 sorting DNS records because real resolvers return them in an arbitrary
3109 order. The new test suite's fake resolver always returns records in the
3112 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3113 bounces) except when queue_only is set, to avoid logging races between
3114 the different processes.
3116 PH/38 Panic-die if .include specifies a non-absolute path.
3118 PH/39 A tweak to the "H" retry rule from its user.
3120 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3121 a label. They prevented compilation on older perls.
3123 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3124 a warning to be raised on newish perls.
3126 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3127 on queue. Changes to match documented behaviour of showing count of
3128 messages matching specified criteria.
3130 PH/40 Changed the default ident timeout from 30s to 5s.
3132 PH/41 Added support for the use of login_cap features, on those BSD systems
3133 that have them, for controlling the resources used by pipe deliveries.
3135 PH/42 The content-scanning code uses fopen() to create files in which to put
3136 message data. Previously it was not paying any attention to the mode of
3137 the files. Exim runs with umask(0) because the rest of the code creates
3138 files with open(), and sets the required mode explicitly. Thus, these
3139 files were ending up world-writeable. This was not a big issue, because,
3140 being within the spool directory, they were not world-accessible. I have
3141 created a function called modefopen, which takes an additional mode
3142 argument. It sets umask(777), creates the file, chmods it to the required
3143 mode, then resets the umask. All the relevant calls to fopen() in the
3144 content scanning code have been changed to use this function.
3146 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3147 to 24 hours. This avoids potential overflow problems when processing G
3148 and H retry rules. I suspect nobody ever tinkers with this value.
3150 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3152 PH/45 When the plaintext authenticator is running as a client, the server's
3153 challenges are checked to ensure they are valid base64 strings. By
3154 default, the authentication attempt is cancelled if an invalid string is
3155 received. Setting client_ignore_invalid_base64 true ignores these errors.
3156 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3157 they are received. Thus, the responses can be made to depend on the
3158 challenges. If an invalid string is ignored, an empty string is placed in
3161 PH/46 Messages that are created by the autoreply transport now contains a
3162 References: header, in accordance with RFCs 2822 and 3834.
3164 PH/47 Added authenticated_sender_force to the smtp transport.
3166 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3168 PH/49 Installed latest patch from the Sieve maintainer.
3170 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3171 was also set, the appendfile transport was unnecessarily scanning a
3172 directory of message files (e.g. for maildir delivery) to find the count
3173 of files (along with the size), even though it did not need this
3174 information. It now does the scan only if it needs to find either the
3175 size of the count of files.
3177 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3179 PH/52 Two bugs concerned with error handling when the smtp transport is
3182 (i) Exim was not creating retry information for temporary errors given
3183 for individual recipients after the DATA command when the smtp transport
3184 was used in LMTP mode. This meant that they could be retried too
3185 frequently, and not timed out correctly.
3187 (ii) Exim was setting the flag that allows error details to be returned
3188 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3189 recipients that were returned after the DATA command.
3191 PH/53 This is related to PH/52, but is more general: for any failing address,
3192 when detailed error information was permitted to be returned to the
3193 sender, but the error was temporary, then after the final timeout, only
3194 "retry timeout exceeded" was returned. Now it returns the full error as
3195 well as "retry timeout exceeded".
3197 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3198 do this, and (what is worse) MTAs that accept it.
3200 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3201 will now be deprecated.
3203 PH/56 New os.c-cygwin from the Cygwin maintainer.
3205 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3206 formats (previously only available in exim formats via -bpr, -bpru,
3207 and -bpra. Now also available in native and exiqgrep formats)
3209 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3210 with very large, slow to parse queues
3212 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3214 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3216 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3217 responses to authentication challenges, though it was showing the
3218 challenges; (ii) I've removed the CR characters from the debug output for
3221 PH/58 Allow for the insertion of a newline as well as a space when a string
3222 is turned into more than one encoded-word during RFC 2047 encoding. The
3223 Sieve code now uses this.
3225 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3226 data_4xx, lost_connection, tls_required.
3228 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3229 message was being sent as an SMTP response.
3231 PH/61 Add -l and -k options to exicyclog.
3233 PH/62 When verifying, if an address was redirected to one new address, so that
3234 verification continued, and the new address failed or deferred after
3235 having set something in $address_data, the value of $address_data was not
3236 passed back to the ACL. This was different to the case when no
3237 redirection occurred. The value is now passed back in both cases.
3239 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3240 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3241 use login_cap.h, so on its own it isn't the distinguishing feature. The
3242 new name refers directly to the setclassresources() function.
3244 PH/65 Added configuration files for NetBSD3.
3246 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3248 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3249 is preferred over IPv4.
3251 PH/68 The bounce_return_message and bounce_return_body options were not being
3252 honoured for bounces generated during the reception of non-SMTP messages.
3253 In particular, this applied to messages rejected by the ACL. This bug has
3254 been fixed. However, if bounce_return_message is true and bounce_return_
3255 body is false, the headers that are returned for a non-SMTP message
3256 include only those that have been read before the error was detected.
3257 (In the case of an ACL rejection, they have all been read.)
3259 PH/69 The HTML version of the specification is now built in a directory called
3260 spec_html instead of spec.html, because the latter looks like a path with
3261 a MIME-type, and this confuses some software.
3263 PH/70 Catch two compiler warnings in sieve.c.
3265 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3266 function verify_get_ident() calls ip_connect() to connect a socket, but
3267 if the "connect()" function timed out, ip_connect() used to close the
3268 socket. However, verify_get_ident() also closes the socket later, and in
3269 between Exim writes to the log, which may get opened at this point. When
3270 the socket was closed in ip_connect(), the log could get the same file
3271 descriptor number as the socket. This naturally causes chaos. The fix is
3272 not to close the socket in ip_connect(); the socket should be closed by
3273 the function that creates it. There was only one place in the code where
3274 this was missing, in the iplookup router, which I don't think anybody now
3275 uses, but I've fixed it anyway.
3277 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3278 well as to direct DNS lookups. Otherwise the handling of names in host
3279 lists is inconsistent and therefore confusing.
3285 PH/01 Two changes to the default runtime configuration:
3287 (1) Move the checks for relay_from_hosts and authenticated clients from
3288 after to before the (commented out) DNS black list checks.
3290 (2) Add control=submission to the relay_from_hosts and authenticated
3291 clients checks, on the grounds that messages accepted by these
3292 statements are most likely to be submissions.
3294 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3296 (1) Generate an error if the third argument for the ${prvs expansion is
3299 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3302 (3) Reset the variables that are obtained from the first argument of
3303 ${prvscheck and used in the second argument before leaving the code,
3304 because their memory is reclaimed, so using them afterwards may do
3307 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3308 one (it's much easier than Tom thought :-).
3310 (5) Because of (4), we can now allow for the use of $prvscheck_result
3311 inside the third argument.
3313 PH/03 For some reason, the default setting of PATH when running a command from
3314 a pipe transport was just "/usr/bin". I have changed it to
3317 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3318 anything to be listed in the output from -bV.
3320 PH/05 When a filter generated an autoreply, the entire To: header line was
3321 quoted in the delivery log line, like this:
3323 => >A.N.Other <ano@some.domain> <original@ddress> ...
3325 This has been changed so that it extracts the operative address. There
3326 may be more than one such address. If so, they are comma-separated, like
3329 => >ano@some.domain,ona@other.domain <original@ddress> ...
3331 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3332 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3333 not being looked up in the rDNS to get a host name, Exim was showing the
3334 IP address twice in Received: lines, even though the IP addresses were
3335 identical. For example:
3337 Received: from [1.2.3.4] (helo=[1.2.3.4])
3339 However, if the real host name was known, it was omitting the HELO data
3340 if it matched the actual IP address. This has been tidied up so that it
3341 doesn't show the same IP address twice.
3343 PH/07 When both +timestamp and +memory debugging was on, the value given by
3344 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3345 called by the debug printing, thereby overwriting the timestamp buffer.
3346 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3348 PH/08 When the original message was included in an autoreply transport, it
3349 always said "this is a copy of the message, including all the headers",
3350 even if body_only or headers_only was set. It now gives an appropriate
3353 PH/09 Applied a patch from the Sieve maintainer which:
3355 o fixes some comments
3356 o adds the (disabled) notify extension core
3357 o adds some debug output for the result of if/elsif tests
3358 o points to the current vacation draft in the documentation
3359 and documents the missing references header update
3363 o fixes a bug in processing the envelope test (when testing
3364 multiple envelope elements, the last element determined the
3367 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3368 Electronic Mail") by including:
3370 Auto-submitted: auto-generated
3372 in the messages that it generates (bounce messages and others, such as
3373 warnings). In the case of bounce messages for non-SMTP messages, there was
3374 also a typo: it was using "Auto_submitted" (underscore instead of
3375 hyphen). Since every message generated by Exim is necessarily in response
3376 to another message, thes have all been changed to:
3378 Auto-Submitted: auto-replied
3380 in accordance with these statements in the RFC:
3382 The auto-replied keyword:
3384 - SHOULD be used on messages sent in direct response to another
3385 message by an automatic process,
3387 - MUST NOT be used on manually-generated messages,
3389 - MAY be used on Delivery Status Notifications (DSNs) and Message
3390 Disposition Notifications (MDNs),
3392 - MUST NOT be used on messages generated by automatic or periodic
3393 processes, except for messages which are automatic responses to
3396 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3397 to the default Received: header definition.
3399 PH/12 Added log selector acl_warn_skipped (default on).
3401 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3402 variables because (a) they are in the wrong storage pool and (b) even if
3403 they were copied, it wouldn't work properly because of the caching.
3405 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3406 checking when decoding. Apparently there are clients that generate
3407 overlong encoded strings. Why am I not surprised?
3409 PH/15 If the first argument of "${if match_address" was not empty, but did not
3410 contain an "@" character, Exim crashed. Now it writes a panic log message
3411 and treats the condition as false.
3413 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3415 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3416 extension "envelope-auth". The code is finished and in agreement with
3417 other implementations, but there is no documentation so far and in fact,
3418 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3419 not changing the active code.
3421 Print executed "if" and "elsif" statements when debugging is used. This
3422 helps a great deal to understand what a filter does.
3424 Document more things not specified clearly in RFC3028. I had all this
3425 sorted out, when out of a sudden new issues came to my mind. Oops."
3427 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3430 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3431 canonical form (as documented). However, after a host name lookup from
3432 the IP address, check_host() was doing a simple string comparison with
3433 addresses acquired from the DNS when checking that the found name did
3434 have the original IP as one of its addresses. Since any found IPv6
3435 addresses are likely to be in abbreviated form, the comparison could
3436 fail. Luckily, there already exists a function for doing the comparison
3437 by converting both addresses to binary, so now that is used instead of
3438 the text comparison.
3440 PH/20 There was another similar case to PH/19, when a complete host name was
3441 given in a host list; looking up its IP address could give an abbreviated
3442 form, whereas the current host's name might or might not be abbreviated.
3443 The same fix has been applied.
3449 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3450 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3453 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3454 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3456 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3458 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3459 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3460 Also, for "Auto-Submitted", treat anything other than "no" as
3461 non-personal, in accordance with RFC 3834. (Previously it treated
3462 anything starting "auto-" as non-personal.)
3464 TF/01 The control=submission/name=... option had a problem with syntax
3465 errors if the name included a slash character. The /name= option
3466 now slurps the rest of the string, so it can include any characters
3467 but it must come last in the list of options (after /sender_retain
3470 PH/05 Some modifications to the interface to the fake nameserver for the new
3478 TK/01 Added the "success_on_redirect" address verification option. See
3479 NewStuff for rationale and an example.
3481 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3483 PH/02 Patch to exigrep to allow it to work on syslog lines.
3485 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3486 fread() to skip over the body file's header line, because in Cygwin the
3487 header line is locked and is inaccessible.
3489 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3490 co-exist for some time) to make it clear that it is the Exim ID that is
3491 referenced, not the Message-ID: header line.
3493 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3494 string_format() function, because snprintf() does not exist on all
3497 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3498 ${stat: expansion item.
3500 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3501 protocol synchronization error", to keep the pedants happy.
3503 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3504 well as for IRIX systems, when gcc is being used. See the host.c source
3507 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3509 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3512 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3513 transport and to the smtp transport in LMTP mode.
3515 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3517 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3518 run a filter in a subprocess. This could lead to confusion in subsequent
3519 lookups in the parent process. There should also be a search_tidyup() at
3520 the end of the subprocess.
3522 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3523 only if the host matched helo_try_verify_hosts, which caused the
3524 verification to occur when the EHLO/HELO command was issued. The ACL just
3525 tested the remembered result. Now, if a previous verification attempt has
3526 not happened, "verify = helo" does it there and then.
3528 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3530 TK/03 Fix log output including CR from clamd.
3532 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3533 exist provoked a memory error which could cause a segfault.
3535 PH/15 Installed PCRE 6.2
3537 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3539 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3540 of the problem. Specifically, suggested +O2 rather than +O1 for the
3543 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3545 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3546 record, the port was not correctly logged when the outgoing_port log
3547 selector was set (it logged the transort's default port).
3549 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3550 fallback_hosts, and "hosts" in the smtp transport.
3552 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3553 host errors such as "Connection refused".
3555 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3556 authentication with radiusclient 0.4.9:
3558 - Error returned from rc_read_config was caught wrongly
3559 - Username/password not passed on to radius server due to wrong length.
3561 The presumption is that some radiusclient API changes for 4.51/PH/17
3562 were not taken care of correctly. The code is still untested by me (my
3563 Linux distribution still has 0.3.2 of radiusclient), but it was
3564 contributed by a Radius user.
3566 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3567 expanding the "port" option of the smtp transport.
3569 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3570 while reading a MIME header. Thanks to Tom Hughes for a patch.
3572 PH/24 Include config.h inside local_scan.h so that configuration settings are
3575 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3576 that in an ACL for RCPT (for example), you can examine exactly what was
3579 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3580 commands, but it was not correctly comparing the address with the actual
3581 client host address. Thus, it would show the EHLO address in Received:
3582 header lines when this was not necessary.
3584 PH/27 Added the % operator to ${eval:}.
3586 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3587 it should be ignoring failures (because with -C, for example, it has lost
3588 privilege). It wasn't ignoring creation failures other than "already
3591 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3594 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3595 because an input line was too long, either on its own, or by virtue of
3596 too many continuations, the temporary file was not being removed, and the
3597 return code was incorrect.
3599 PH/31 Missing "BOOL" in function definition in filtertest.c.
3601 PH/32 Applied Sieve patches from the maintainer.
3603 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3605 PH/33 Added "verify = not_blind".
3607 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3608 Local/Makefile (with some defaults set). These are used in built scripts
3609 such as exicyclog, but they have never been used in the exim_install
3610 script (though there are many overriding facilities there). I have
3611 arranged that the exim_install script now takes note of these two
3614 PH/35 Installed configuration files for Dragonfly.
3616 PH/36 When a locally submitted message by a trusted user did not contain a
3617 From: header, and the sender address was obtained from -f or from an SMTP
3618 MAIL command, and the trusted user did not use -F to supply a sender
3619 name, $originator_name was incorrectly used when constructing a From:
3620 header. Furthermore, $originator_name was used for submission mode
3621 messages from external hosts without From: headers in a similar way,
3622 which is clearly wrong.
3624 PH/37 Added control=suppress_local_fixups.
3626 PH/38 When log_selector = +received_sender was set, and the addition of the
3627 sender made the log line's construction buffer exactly full, or one byte
3628 less than full, an overflow happened when the terminating "\n" was
3631 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3632 when the result of a list match is failure because a DNS lookup failed.
3634 PH/40 RM_COMMAND is now used in the building process.
3636 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3637 the "build-* directories that it finds.
3639 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3640 address in a domain literal was a prefix of an interface address.
3642 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3643 when verifying a sender address, unless rewrite_headers is false.
3645 PH/44 Wrote a long comment about why errors_to addresses are verified as
3646 recipients, not senders.
3648 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3649 the ratelimit ACL was added.
3651 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3653 PH/46 Added extra information about PostgreSQL errors to the error string.
3655 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3656 avoiding the need to install special zones in a real server. This is
3657 backwards compatible; if it can't find the fake resolver, it drops back.
3658 Thus, both old and new test suites can be run.
3660 TF/02 Added util/ratelimit.pl
3662 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3663 clock is set back in time.
3665 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3666 Candler <B.Candler@pobox.com>.
3668 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3669 OK for addresses that are the result of redirections.
3671 PH/48 A number of further additions for the benefit of the new test suite,
3672 including a fake gethostbyname() that interfaces to the fake DNS resolver
3675 TF/06 The fix for widen_domains has also been applied to qualify_single and
3676 search_parents which are the other dnslookup options that can cause
3679 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3682 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3684 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3685 service" if it encountered an MX record with an empty target hostname.
3686 The message is now "an MX or SRV record indicated no SMTP service".
3688 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3689 if the DNS of the sending site is misconfigured. This is quite a
3690 common situation. This change restores the behaviour of treating a
3691 helo verification defer as a failure.
3693 PH/51 If self=fail was set on a router, the bounce message did not include the
3694 actual error message.
3700 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3702 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3703 command itself ended in error, the underlying message about the transport
3704 filter timeout was being overwritten with the pipe command error. Now the
3705 underlying error message should be appended to the second error message.
3707 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3709 PH/02 Reduce the amount of output that "make" produces by default. Full output
3710 can still be requested.
3712 PH/03 The warning log line about a condition test deferring for a "warn" verb
3713 was being output only once per connection, rather than after each
3714 occurrence (because it was using the same function as for successful
3715 "warn" verbs). This seems wrong, so I have changed it.
3717 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3718 it should not have, which might have caused a crash in the right
3719 circumstances, but probably never did.
3721 PH/04 Installed a modified version of Tony Finch's patch to make submission
3722 mode fix the return path as well as the Sender: header line, and to
3723 add a /name= option so that you can make the user's friendly name appear
3726 TF/03 Added the control = fakedefer ACL modifier.
3728 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3729 Mark Lowes for thorough testing.
3731 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3733 TK/03 Merged latest SRS patch from Miles Wilton.
3735 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3736 with the definition in sysexits.h (which is #included earlier).
3737 Fortunately, Exim does not actually use EX_OK. The code used to try to
3738 preserve the sysexits.h value, by assuming that macro definitions were
3739 scanned for macro replacements. I have been disabused of this notion,
3740 so now the code just undefines EX_OK before #including unistd.h.
3742 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3743 in the smtp transport. When a block could not be written in a single
3744 write() function, the timeout was being re-applied to each part-write.
3745 This seems wrong - if the receiver was accepting one byte at a time it
3746 would take for ever. The timeout is now adjusted when this happens. It
3747 doesn't have to be particularly precise.
3749 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3750 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3752 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3753 without a domain if the check to <postmaster@domain> fails.
3755 SC/01 Eximstats: added -xls and the ability to specify output files
3756 (patch written by Frank Heydlauf).
3758 SC/02 Eximstats: use FileHandles for outputting results.
3760 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3762 SC/04 Eximstats: fixed display of large numbers with -nvr option
3764 SC/05 Eximstats: fixed merging of reports with empty tables.
3766 SC/06 Eximstats: added the -include_original_destination flag
3768 SC/07 Eximstats: removed tabs and trailing whitespace.
3770 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3772 TK/06 MBOX spool code: Add real "From " MBOX separator line
3773 so the .eml file is really in mbox format (even though
3774 most programs do not really care). Patch from Alex Miller.
3776 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3777 The latter is generated from $received_to and is only set if the
3778 message has one envelope recipient. SA can use these headers,
3779 obviously out-of-the-box. Patch from Alex Miller.
3781 PH/08 The ${def test on a variable was returning false if the variable's
3782 value was "0", contrary to what the specification has always said!
3783 The result should be true unless the variable is empty.
3785 PH/09 The syntax error of a character other than { following "${if
3786 def:variable_name" (after optional whitespace) was not being diagnosed.
3787 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3788 accidental colon was present, for example, could give incorrect results.
3790 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3791 result is used (not including appendfile, where other changes are about
3794 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3795 This involved changing a lot of size variables from int to off_t. It
3796 should work with maildirs and everything.
3798 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3799 spamd dying while we are connected to it.
3801 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3804 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3805 function for BDB 4.3.
3807 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3809 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3810 the sources for PCRE have been re-arranged and more files are now
3813 PH/15 The code I had for printing potentially long long variables in PH/11
3814 above was not the best (it lost precision). The length of off_t variables
3815 is now inspected at build time, and an appropriate printing format (%ld
3816 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3817 to be "long long int" or "long int". This is needed for the internal
3818 formatting function string_vformat().
3820 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3821 the configuration file to be ":syslog", then the script "guesses" where
3822 the logs files are, rather than using the compiled in default. In our
3823 case the guess is not the same as the compiled default, so the script
3824 suddenly stopped working when I started to use syslog. The patch checks
3825 to see if log_file_path is "". If so, it attempts to read it from exim
3826 with no configuration file to get the compiled in version, before it
3827 falls back to the previous guessing code."
3829 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3830 implementing BATV in an Exim configuration. See NewStuff for the gory
3833 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3834 Makefile that are specific to HP-UX.
3836 PH/18 If the "use_postmaster" option was set for a recipient callout together
3837 with the "random" option, the postmaster address was used as the MAIL
3838 FROM address for the random test, but not for the subsequent recipient
3839 test. It is now used for both.
3841 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3842 patch removes a few documentation additions to RFC 3028, because the
3843 latest draft now contains them. It adds the new en;ascii-case comparator
3844 and a new error check for 8bit text in MIME parts. Comparator and
3845 require names are now matched exactly. I enabled the subaddress
3846 extension, but it is not well tested yet (read: it works for me)."
3848 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3849 rework some of the code of TK/09 above to avoid the hardwired use of
3850 "%lld" and "long long". Replaced the call to snprintf() with a call to
3853 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3854 records point to non-existent hosts", "retry timeout exceeded", and
3855 "retry time not reached for any host after a long failure period".
3857 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3858 experimental DomainKeys support:
3860 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3861 (2) On an error such as an illegally used "control", the wrong name for
3862 the control was given.
3864 These problems did NOT occur unless DomainKeys support was compiled.
3866 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3868 PH/24 Added ${if match_ip condition.
3870 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3871 databases so that it will be absolutely obvious if a crash occurs in the
3872 DB library. This is a regular occurrence (often caused by mis-matched
3875 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3876 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3877 picked up on a user's system that detects such things. There doesn't seem
3878 to be a gcc warning option for this - only an attribute that has to be
3879 put on the function's prototype. It seems that in Fedora Core 4 they have
3880 set this on a number of new functions. No doubt there will be more in due
3883 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3884 specify a transport. However, if an address that was verified by such a
3885 router was the subject of a callout, Exim crashed because it tried to
3886 read the rcpt_include_affixes from the non-existent transport. Now it
3887 just assumes that the setting of that option is false. This bug was
3888 introduced by 4.51/PH/31.
3890 PH/28 Changed -d+all to exclude +memory, because that information is very
3891 rarely of interest, but it makes the output a lot bigger. People tend to
3892 do -d+all out of habit.
3894 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3895 code in os-type was giving problems when libc.so lives in lib64, like on
3898 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3899 aren't the modern standard, and it seems that some systems' include files
3900 don't always have them. Exim was already checking for some of the newer
3901 ones like T_AAAA, and defining it itself. I've added checks for all the
3902 record types that Exim uses.
3904 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3905 not automatically generating a new one, as it is supposed to. This
3906 prevented TLS from working. If the file did exist, but contained invalid
3907 data, a new version was generated, as expected. It was only the case of a
3908 non-existent file that was broken.
3910 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3911 with a change in libdomainkeys > 0.64.
3913 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3914 from DNS. If the selector record carries the flag, it now has
3915 precedence over the domain-wide flag.
3917 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3919 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3920 the use of a port name that isn't defined in /etc/services) occurred, the
3921 message was deferred as in a normal delivery, and thus remained on the
3922 spool, instead of being failed because of the mua_wrapper setting. This
3923 is now fixed, and I tidied up some of the mua_wrapper messages at the
3926 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3927 the messages in a hash of arrays rather than using individual hashes.
3928 This is a bit cleaner and results in dramatic memory savings, albeit
3929 at a slight CPU cost.
3931 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3932 as requested by Marc Sherman.
3934 SC/10 Eximstats: added histograms for user specified patterns as requested
3937 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3939 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3940 fopen() in the content-scanning modules that did not already have it.
3946 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3947 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3949 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3951 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3953 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3954 to test Sieve filters that use "vacation".
3956 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3957 that changes the way the GnuTLS parameters are stored in the cache file.
3958 The new format can be generated externally. For backward compatibility,
3959 if the data in the cache doesn't make sense, Exim assumes it has read an
3960 old-format file, and it generates new data and writes a new file. This
3961 means that you can't go back to an older release without removing the
3964 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3965 work if there are any delivery delays because "one_time" forces the
3966 parent to be marked "delivered", so its unseen clone is never tried
3967 again. For this reason, Exim now forbids the simultaneous setting of
3970 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3971 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3972 entry if you want to know the details. The fix, however, neglected to
3973 consider the case when local delivery batching is involved. The test for
3974 "previously delivered" was not happening when checking to see if an
3975 address could be batched with a previous (undelivered) one; under
3976 certain circumstances this could lead to multiple deliveries to the same
3979 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3980 in its include files, and this causes problems building Exim.
3982 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3983 header_syntax) but Exim was just ignoring anything given after a slash.
3984 In particular, this caused confusion with an attempt to use "verify =
3985 reverse_host_lookup/defer_ok". An error is now given when options are
3986 supplied for verify items that do not have them. (Maybe reverse_host_
3987 lookup should have a defer_ok option, but that's a different point.)
3989 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3990 defined by RFC 821) to 2048, because there were problems with some AUTH
3991 commands, and RFC 1869 says the size should be increased for extended
3992 SMTP commands that take arguments.
3994 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3997 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3998 "unknown" error; now it says that the functionality isn't in the binary.
4000 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4001 an address' error message when a string expansion fails (syntax or
4002 whatever). Otherwise the password may appear in the log. Following change
4003 PH/42 below, there is no longer a chance of it appearing in a bounce
4006 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4008 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4009 of its addresses was ever tried. (Bugzilla bug #2.)
4011 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4012 the result incorrectly in the debug output. (It correctly added a newline
4013 to what was transported.)
4015 TF/01 Added $received_time.
4017 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4018 commented out examples of how to interface to a virus scanner and to
4019 SpamAssassin. Also added commented examples of av_scanner and
4020 spamd_address settings.
4022 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4023 and controls are allowed in which ACLs. There were a couple of minor
4024 errors. Some of the entries in the conditions table (which is a table of
4025 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4026 negation of where the condition IS allowed.
4028 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4030 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4031 header file does not have a version number, so I've had to invent a new
4032 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4033 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4034 radiusclient), but it was contributed by a Radius user.
4036 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4037 files or directories, for OpenSSL.
4039 PH/19 When an Exim process that is running as root has to create an Exim log
4040 file, it does so in a subprocess that runs as exim:exim so as to get the
4041 ownership right at creation (otherwise, other Exim processes might see
4042 the file with the wrong ownership). There was no test for failure of this
4043 fork() call, which would lead to the process getting stuck as it waited
4044 for a non-existent subprocess. Forks do occasionally fail when resources
4045 run out. I reviewed all the other calls to fork(); they all seem to check
4048 PH/20 When checking for unexpected SMTP input at connect time (before writing
4049 the banner), Exim was not dealing correctly with a non-positive return
4050 from the read() function. If the client had disconnected by this time,
4051 the result was a log entry for a synchronization error with an empty
4052 string after "input=" when read() returned zero. If read() returned -1
4053 (an event I could not check), uninitialized data bytes were printed.
4054 There were reports of junk text (parts of files, etc) appearing after
4057 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4059 PH/22 Added support for macro redefinition, and (re)definition in between
4060 driver and ACL definitions.
4062 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4063 forgetting to use the resulting value; it was using the unexpanded value.
4065 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4066 hadn't been configured. The fix is from Juergen Kreileder, who
4067 understands it better than I do:
4069 "Here's what I see happening with three configured cyrus_sasl
4070 authenticators configured (plain, login, cram-md5):
4072 On startup auth_cyrus_sasl_init() gets called for each of these.
4073 This means three calls to sasl_listmech() without a specified mech_list.
4074 => SASL tests which mechs of all available mechs actually work
4075 => three warnings about OTP not working
4076 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4078 With the patch, sasl_listmech() also gets called three times. But now
4079 SASL's mech_list option is set to the server_mech specified in the the
4080 authenticator. Or in other words, the answer from sasl_listmech()
4081 gets limited to just the mech you're testing for (which is different
4083 => the return list contains just 'plain' or 'login', 'cram-md5' or
4084 nothing depending on the value of ob->server_mech.
4086 I've just tested the patch: Authentication still works fine,
4087 unavailable mechs specified in the exim configuration are still
4088 caught, and the auth.log warnings about OTP are gone."
4090 PH/25 When debugging is enabled, the contents of the command line are added
4091 to the debugging output, even when log_selector=+arguments is not
4094 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4095 answer is "GNU", and only if the return is "GNU/something" is the answer
4098 PH/27 $acl_verify_message is now set immediately after the failure of a
4099 verification in an ACL, and so is available in subsequent modifiers. In
4100 particular, the message can be preserved by coding like this:
4102 warn !verify = sender
4103 set acl_m0 = $acl_verify_message
4105 Previously, $acl_verify_message was set only while expanding "message"
4106 and "log_message" when a very denied access.
4108 PH/28 Modified OS/os.c-Linux with
4110 -#ifndef OS_LOAD_AVERAGE
4111 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4113 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4114 nomenclature these days.)
4116 PH/29 Installed patch from the Sieve maintainer that adds the options
4117 sieve_useraddress and sieve_subaddress to the redirect router.
4119 PH/30 In these circumstances:
4120 . Two addresses routed to the same list of hosts;
4121 . First host does not offer TLS;
4122 . First host accepts first address;
4123 . First host gives temporary error to second address;
4124 . Second host offers TLS and a TLS session is established;
4125 . Second host accepts second address.
4126 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4127 and peerdn, if requested) that were in fact used only for the second
4130 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4131 attention to any local part prefix or suffix that was matched by the
4132 router that accepted the address. It now behaves in the same way as it
4133 does for delivery: the affixes are removed from the local part unless
4134 rcpt_include_affixes is set on the transport.
4136 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4137 timeout during the DATA phase of an incoming message.
4139 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4140 applied a patch sanctioned by the Sieve maintainer.
4142 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4143 the uid or gid is negative. A case of a negative gid caused this to be
4144 noticed. The fix allows for either to be negative.
4146 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4147 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4150 PH/36 The change PH/12 above was broken. Fixed it.
4152 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4153 the grounds that routing the same address twice would always produce the
4154 same answer. This might have been true once, but it is certainly no
4155 longer true now. Routing a child address may depend on the previous
4156 routing that produced that child. Some complicated redirection strategies
4157 went wrong when messages had multiple recipients, and made Exim's
4158 behaviour dependent on the order in which the addresses were given.
4160 I have moved the duplicate checking until after the routing is complete.
4161 Exim scans the addresses that are assigned to local and remote
4162 transports, and removes any duplicates. This means that more work will be
4163 done, as duplicates will always all be routed, but duplicates are
4164 presumably rare, so I don't expect this is of any significance.
4166 For deliveries to pipes, files, and autoreplies, the duplicate checking
4167 still happens during the routing process, since they are not going to be
4170 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4171 It corrects a timeout issue with spamd. This is Ian's comment: "The
4172 background is that sometimes spamd either never reads data from a
4173 connection it has accepted, or it never writes response data. The exiscan
4174 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4175 blindly assumes that writes won't block so it may never time out."
4177 PH/39 Allow G after quota size as well as K and M.
4179 PH/40 The value set for $authenticated_id in an authenticator may not contain
4180 binary zeroes or newlines because the value is written to log lines and
4181 to spool files. There was no check on this. Now the value is run through
4182 the string_printing() function so that such characters are converted to
4183 printable escape sequences.
4185 PH/41 $message_linecount is a new variable that contains the total number of
4186 lines in the message. Compare $body_linecount, which is the count for the
4189 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4190 bounce and delay warning messages, except in certain special cases, which
4193 (a) An SMTP error message from a remote host;
4194 (b) A message specified in a :fail: redirection;
4195 (c) A message specified in a "fail" command in a system filter;
4196 (d) A message specified in a FAIL return from the queryprogram router;
4197 (e) A message specified by the cannot_route_message router option.
4199 In these cases only, Exim does include the error details in bounce and
4200 warning messages. There are also a few cases where bland messages such
4201 as "unrouteable address" or "local delivery error" are given.
4203 PH/43 $value is now also set for the "else" part of a ${run expansion.
4205 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4206 being worked on, but at least Exim now implements the latest version to
4209 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4210 process to complete was treated as a delivery failure, a timeout while
4211 writing the message to the pipe was logged, but erroneously treated as a
4212 successful delivery. Such timeouts include transport filter timeouts. For
4213 consistency with the overall process timeout, these timeouts are now
4214 treated as errors, giving rise to delivery failures by default. However,
4215 there is now a new Boolean option for the pipe transport called
4216 timeout_defer, which, if set TRUE, converts the failures into defers for
4217 both kinds of timeout. A transport filter timeout is now identified in
4220 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4221 systems where "make" and "gmake" are different, calling "gmake" at top
4222 level broke things. I've arranged for the value of $(MAKE) to be passed
4223 from the Makefile to this script so that it can call the same version of
4227 A note about Exim versions 4.44 and 4.50
4228 ----------------------------------------
4230 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4231 changes of various kinds. As a consequence, a big documentation update was
4232 needed. This delayed the release for rather longer than seemed good, especially
4233 in the light of a couple of (minor) security issues. Therefore, the changes
4234 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4235 release. So 4.44 and 4.50 are in effect two different branches that both start
4238 I have left the 4.50 change log unchanged; it contains all the changes since
4239 4.43. The change log for 4.44 is below; many of its items are identical to
4240 those for 4.50. This seems to be the most sensible way to preserve the
4241 historical information.
4247 1. Minor wording change to the doc/README.SIEVE file.
4249 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4250 computation of the current number of files was incorrect.
4252 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4253 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4256 4. Give more explanation in the error message when the command for a transport
4257 filter fails to execute.
4259 5. There are several places where Exim runs a non-Exim command in a
4260 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4261 was being done only for the command run by the queryprogram router. It is
4262 now done for all such subprocesses. The other cases are: ${run, transport
4263 filters, and the commands run by the lmtp and pipe transports.
4265 6. Added CONFIGURE_GROUP build-time option.
4267 7. Some older OS have a limit of 256 on the maximum number of file
4268 descriptors. Exim was using setrlimit() to set 1000 as a large value
4269 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4270 systems. I've change it so that if it can't get 1000, it tries for 256.
4272 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4273 was an oversight, and furthermore, ever since the addition of extra
4274 controls (e.g. 4.43/32), the checks on when to allow different forms of
4275 "control" were broken. There should now be diagnostics for all cases when a
4276 control that does not make sense is encountered.
4278 9. Added the /retain_sender option to "control=submission".
4280 10. $recipients is now available in the predata ACL (oversight).
4282 11. Tidy the search cache before the fork to do a delivery from a message
4283 received from the command line. Otherwise the child will trigger a lookup
4284 failure and thereby defer the delivery if it tries to use (for example) a
4285 cached ldap connection that the parent has called unbind on.
4287 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4288 of $address_data from the recipient verification was clobbered by the
4289 sender verification.
4291 13. The value of address_data from a sender verification is now available in
4292 $sender_address_data in subsequent conditions in the ACL statement.
4294 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4296 15. Added a new option "connect=<time>" to callout options, to set a different
4299 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4300 was its contents. (It was OK if the option was not defined at all.)
4302 17. A "Completed" log line is now written for messages that are removed from
4303 the spool by the -Mrm option.
4305 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4306 information about exactly what failed.
4308 19. Added -dd to debug only the daemon process.
4310 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4311 handles timeouts, both on the server side and network timeouts. Renamed the
4312 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4314 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4315 It is now set to "smtps".
4317 22. $host_address is now set to the target address during the checking of
4318 ignore_target_hosts.
4320 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4321 being passed; this would have caused $sender_host_name to have been used if
4322 matching the list had actually called for a host name (not very likely,
4323 since this list is usually IP addresses). A host name is now passed as
4326 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4327 code that specifies a non-restarting handler (typically sigaction() in
4328 modern systems) in an attempt to fix a rare and obscure crash bug.
4330 25. Narrowed the window for a race in the daemon that could cause it to ignore
4331 SIGCHLD signals. This is not a major problem, because they are used only to
4332 wake it up if nothing else does.
4334 26. A malformed maildirsize file could cause Exim to calculate negative values
4335 for the mailbox size or file count. Odd effects could occur as a result.
4336 The maildirsize information is now recalculated if the size or filecount
4339 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4340 support for a long time. Removed HAVE_SYS_VFS_H.
4342 28. Installed the latest version of exipick from John Jetmore.
4344 29. In an address list, if the pattern was not a regular expression, an empty
4345 subject address (from a bounce message) matched only if the pattern was an
4346 empty string. Non-empty patterns were not even tested. This was the wrong
4347 because it is perfectly reasonable to use an empty address as part of a
4348 database query. An empty address is now tested by patterns that are
4349 lookups. However, all the other forms of pattern expect the subject to
4350 contain a local part and a domain, and therefore, for them, an empty
4351 address still always fails if the pattern is not itself empty.
4353 30. Exim went into a mad DNS loop when attempting to do a callout where the
4354 host was specified on an smtp transport, and looking it up yielded more
4355 than one IP address.
4357 31. Re-factored the code for checking spool and log partition space into a
4358 function that finds that data and another that does the check. The former
4359 is then used to implement four new variables: $spool_space, $log_space,
4360 $spool_inodes, and $log_inodes.
4362 32. The RFC2047 encoding function was originally intended for short strings
4363 such as real names; it was not keeping to the 75-character limit for
4364 encoded words that the RFC imposes. It now respects the limit, and
4365 generates multiple encoded words if necessary. To be on the safe side, I
4366 have increased the buffer size for the ${rfc2047: expansion operator from
4369 33. It is now permitted to omit both strings after an "if" condition; if the
4370 condition is true, the result is "true". As before, when the second string
4371 is omitted, a false condition yields an empty string. This makes it less
4372 cumbersome to write custom ACL and router conditions.
4374 34. Failure to deliver a bounce message always caused it to be frozen, even if
4375 there was an errors_to setting on the router. The errors_to setting is now
4378 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4379 canonical form (fully expanded) before being placed in
4380 $sender_host_address.
4382 36. The table in the code that translates DNS record types into text (T_A to
4383 "A" for instance) was missing entries for NS and CNAME. It is just possible
4384 that this could have caused confusion if both these types were looked up
4385 for the same domain, because the text type is used as part of Exim's
4386 per-process caching. But the chance of anyone hitting this buglet seems
4389 37. The dnsdb lookup has been extended in a number of ways.
4391 (1) There is a new type, "zns", which walks up the domain tree until it
4392 finds some nameserver records. It should be used with care.
4394 (2) There is a new type, "mxh", which is like "mx" except that it returns
4395 just the host names, not the priorities.
4397 (3) It is now possible to give a list of domains (or IP addresses) to be
4398 looked up. The behaviour when one of the lookups defers can be
4399 controlled by a keyword.
4401 (4) It is now possible to specify the separator character for use when
4402 multiple records are returned.
4404 38. The dnslists ACL condition has been extended: it is now possible to supply
4405 a list of IP addresses and/or domains to be looked up in a particular DNS
4408 39. Added log_selector=+queue_time_overall.
4410 40. When running the queue in the test harness, wait just a tad after forking a
4411 delivery process, to get repeatability of debugging output.
4413 41. Include certificate and key file names in error message when GnuTLS fails
4414 to set them up, because the GnuTLS error message doesn't include the name
4415 of the failing file when there is a problem reading it.
4417 42. Allow both -bf and -bF in the same test run.
4419 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4421 44. The "Exiscan patch" is now merged into the mainline Exim source.
4423 45. Sometimes the final signoff response after QUIT could fail to get
4424 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4425 < 0 before doing a fflush(). This bug looks as though it goes back to the
4426 introduction of TLS in release 3.20, but "sometimes" must have been rare
4427 because the tests only now provoked it.
4429 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4430 (this can affect the format of dates).
4432 47. exim_tidydb, when checking for the continued existence of a message for
4433 which it has found a message-specific retry record, was not finding
4434 messages that were in split spool directories. Consequently, it was
4435 deleting retry records that should have stayed in existence.
4437 48. Steve fixed some bugs in eximstats.
4439 49. The SPA authentication driver was not abandoning authentication and moving
4440 on to the next authenticator when an expansion was forced to fail,
4441 contradicting the general specification for all authenticators. Instead it
4442 was generating a temporary error. It now behaves as specified.
4444 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4445 (the order specifies the preference for clients). The order is now AES256,
4446 AES128, 3DES, ARCFOUR128.
4448 51. Small patch to Sieve code - explicitly set From: when generating an
4451 52. Exim crashed if a remote delivery caused a very long error message to be
4452 recorded - for instance if somebody sent an entire SpamAssassin report back
4453 as a large number of 550 error lines. This bug was coincidentally fixed by
4454 increasing the size of one of Exim's internal buffers (big_buffer) that
4455 happened as part of the Exiscan merge. However, to be on the safe side, I
4456 have made the code more robust (and fixed the comments that describe what
4459 53. Now that there can be additional text after "Completed" in log lines (if
4460 the queue_time_overall log selector is set), a one-byte patch to exigrep
4461 was needed to allow it to recognize "Completed" as not the last thing in
4464 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4465 patch that reportedly fixes this has been added. I am not expert enough to
4466 create a test for it. This is what the patch creator wrote:
4468 "I found a little strange behaviour of ldap code when working with
4469 Windows 2003 AD Domain, where users was placed in more than one
4470 Organization Units. When I tried to give exim partial DN, the exit code
4471 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4472 But simultaneously result of request was absolutely normal ldap result,
4473 so I produce this patch..."
4475 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4476 REFERENCE, so I have modified the code to exclude the patch when that macro
4479 55. Some experimental protocols are using DNS PTR records for new purposes. The
4480 keys for these records are domain names, not reversed IP addresses. The
4481 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4482 leaves it alone. Component reversal etc. now happens only for IP addresses.
4485 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4487 57. Double the size of the debug message buffer (to 2048) so that more of very
4488 long debug lines gets shown.
4490 58. The exicyclog utility now does better if the number of log files to keep
4491 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4493 59. Two changes related to the smtp_active_hostname option:
4495 (1) $smtp_active_hostname is now available as a variable.
4496 (2) The default for smtp_banner uses $smtp_active_hostname instead
4497 of $primary_hostname.
4499 60. The host_aton() function is supposed to be passed a string that is known
4500 to be a valid IP address. However, in the case of IPv6 addresses, it was
4501 not checking this. This is a hostage to fortune. Exim now panics and dies
4502 if the condition is not met. A case was found where this could be provoked
4503 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4504 components; fortuitously, this particular loophole had already been fixed
4505 by change 4.50/55 above.
4507 If there are any other similar loopholes, the new check in host_aton()
4508 itself should stop them being exploited. The report I received stated that
4509 data on the command line could provoke the exploit when Exim was running as
4510 exim, but did not say which command line option was involved. All I could
4511 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4512 running as the user.
4515 61. There was a buffer overflow vulnerability in the SPA authentication code
4516 (which came originally from the Samba project). I have added a test to the
4517 spa_base64_to_bits() function which I hope fixes it.
4520 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4521 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4523 63. The daemon start-up calls getloadavg() while still root for those OS that
4524 need the first call to be done as root, but it missed one case: when
4525 deliver_queue_load_max is set with deliver_drop_privilege. This is
4526 necessary for the benefit of the queue runner, because there is no re-exec
4527 when deliver_drop_privilege is set.
4529 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4530 This has been fixed.
4532 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4533 was in use, was not putting the data itself into the right store pool;
4534 consequently, it could be overwritten for a subsequent message in the same
4535 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4538 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4540 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4541 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4542 the function were treating the return as a boolean value, which happened to
4543 work because 0=false and not-0=true, but is not correct code.
4545 68. The host_aton() function was not handling scoped IPv6 addresses (those
4546 with, for example, "%eth0" on the end) correctly.
4548 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4549 negated items (that is, ~something) in unsigned ints. Some compilers
4550 apparently mutter when there is no cast.
4552 70. If an address verification called from an ACL failed, and did not produce a
4553 user-specific message (i.e. there was only a "system" message), nothing was
4554 put in $acl_verify_message. In this situation, it now puts the system
4557 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4558 session; change 4.31/43 added the unwanted input to the log line - except
4559 that it did not do this in the start of session case. It now does.
4561 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4562 This is wrong and can cause the other end to generate a synchronization
4563 error if it is another Exim or anything else that does the synchronization
4564 check. A QUIT command is no longer sent after a timeout.
4566 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4567 during host lookups.
4569 74. The defer_ok option of callout verification was not working if it was used
4570 when verifying addresses in header lines, that is, for this case:
4572 verify = header_sender/callout=defer_ok
4574 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4575 those file descriptors could be used for SMTP connections. If anything
4576 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4577 could be sent to the SMTP client, causing chaos. The daemon now opens
4578 stdin, stdout, and stderr to /dev/null when it puts itself into the
4581 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4582 log by default. The user can override this with suitable Perl magic.
4584 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4585 the log message when discard triggers, was not working for the DATA ACL or
4586 for the non-SMTP ACL.
4588 78. Error message wording change in sieve.c.
4590 79. If smtp_accept_max_per_host was set, the number of connections could be
4591 restricted to fewer than expected, because the daemon was trying to set up
4592 a new connection before checking whether the processes handling previous
4593 connections had finished. The check for completed processes is now done
4594 earlier. On busy systems, this bug wouldn't be noticed because something
4595 else would have woken the daemon, and it would have reaped the completed
4598 80. If a message was submitted locally by a user whose login name contained one
4599 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4600 It caused a spool format error. I have fixed the spool reading code. A
4601 related problem was that the "from" clause in the Received: line became
4602 illegal because of the space(s). It is now covered by ${quote_local_part.
4604 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4607 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4609 83. Patch from Sieve maintainer to fix unterminated string problem in
4610 "vacation" handling.
4612 84. Some minor changes to the Linux configuration files to help with other
4613 OS variants using glibc.
4615 85. One more patch for Sieve to update vacation handling to latest spec.
4618 ----------------------------------------------------
4619 See the note above about the 4.44 and 4.50 releases.
4620 ----------------------------------------------------
4626 1. Change 4.43/35 introduced a bug that caused file counts to be
4627 incorrectly computed when quota_filecount was set in an appendfile
4630 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4631 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4634 3. Give more explanation in the error message when the command for a transport
4635 filter fails to execute.
4637 4. There are several places where Exim runs a non-Exim command in a
4638 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4639 was being done only for the command run by the queryprogram router. It is
4640 now done for all such subprocesses. The other cases are: ${run, transport
4641 filters, and the commands run by the lmtp and pipe transports.
4643 5. Some older OS have a limit of 256 on the maximum number of file
4644 descriptors. Exim was using setrlimit() to set 1000 as a large value
4645 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4646 systems. I've change it so that if it can't get 1000, it tries for 256.
4648 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4649 was an oversight, and furthermore, ever since the addition of extra
4650 controls (e.g. 4.43/32), the checks on when to allow different forms of
4651 "control" were broken. There should now be diagnostics for all cases when a
4652 control that does not make sense is encountered.
4654 7. $recipients is now available in the predata ACL (oversight).
4656 8. Tidy the search cache before the fork to do a delivery from a message
4657 received from the command line. Otherwise the child will trigger a lookup
4658 failure and thereby defer the delivery if it tries to use (for example) a
4659 cached ldap connection that the parent has called unbind on.
4661 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4662 of $address_data from the recipient verification was clobbered by the
4663 sender verification.
4665 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4666 was its contents. (It was OK if the option was not defined at all.)
4668 11. A "Completed" log line is now written for messages that are removed from
4669 the spool by the -Mrm option.
4671 12. $host_address is now set to the target address during the checking of
4672 ignore_target_hosts.
4674 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4675 being passed; this would have caused $sender_host_name to have been used if
4676 matching the list had actually called for a host name (not very likely,
4677 since this list is usually IP addresses). A host name is now passed as
4680 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4681 code that specifies a non-restarting handler (typically sigaction() in
4682 modern systems) in an attempt to fix a rare and obscure crash bug.
4684 15. Narrowed the window for a race in the daemon that could cause it to ignore
4685 SIGCHLD signals. This is not a major problem, because they are used only to
4686 wake it up if nothing else does.
4688 16. A malformed maildirsize file could cause Exim to calculate negative values
4689 for the mailbox size or file count. Odd effects could occur as a result.
4690 The maildirsize information is now recalculated if the size or filecount
4693 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4694 support for a long time. Removed HAVE_SYS_VFS_H.
4696 18. Updated exipick to current release from John Jetmore.
4698 19. Allow an empty sender to be matched against a lookup in an address list.
4699 Previously the only cases considered were a regular expression, or an
4702 20. Exim went into a mad DNS lookup loop when doing a callout where the
4703 host was specified on the transport, if the DNS lookup yielded more than
4706 21. The RFC2047 encoding function was originally intended for short strings
4707 such as real names; it was not keeping to the 75-character limit for
4708 encoded words that the RFC imposes. It now respects the limit, and
4709 generates multiple encoded words if necessary. To be on the safe side, I
4710 have increased the buffer size for the ${rfc2047: expansion operator from
4713 22. Failure to deliver a bounce message always caused it to be frozen, even if
4714 there was an errors_to setting on the router. The errors_to setting is now
4717 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4718 canonical form (fully expanded) before being placed in
4719 $sender_host_address.
4721 24. Updated eximstats to version 1.33
4723 25. Include certificate and key file names in error message when GnuTLS fails
4724 to set them up, because the GnuTLS error message doesn't include the name
4725 of the failing file when there is a problem reading it.
4727 26. Expand error message when OpenSSL has problems setting up cert/key files.
4730 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4731 (this can affect the format of dates).
4733 28. exim_tidydb, when checking for the continued existence of a message for
4734 which it has found a message-specific retry record, was not finding
4735 messages that were in split spool directories. Consequently, it was
4736 deleting retry records that should have stayed in existence.
4738 29. eximstats updated to version 1.35
4739 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4740 1.35 - bugfix such that pie charts by volume are generated correctly
4742 30. The SPA authentication driver was not abandoning authentication and moving
4743 on to the next authenticator when an expansion was forced to fail,
4744 contradicting the general specification for all authenticators. Instead it
4745 was generating a temporary error. It now behaves as specified.
4747 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4748 (the order specifies the preference for clients). The order is now AES256,
4749 AES128, 3DES, ARCFOUR128.
4751 31. Small patch to Sieve code - explicitly set From: when generating an
4754 32. Exim crashed if a remote delivery caused a very long error message to be
4755 recorded - for instance if somebody sent an entire SpamAssassin report back
4756 as a large number of 550 error lines. This bug was coincidentally fixed by
4757 increasing the size of one of Exim's internal buffers (big_buffer) that
4758 happened as part of the Exiscan merge. However, to be on the safe side, I
4759 have made the code more robust (and fixed the comments that describe what
4762 33. Some experimental protocols are using DNS PTR records for new purposes. The
4763 keys for these records are domain names, not reversed IP addresses. The
4764 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4765 leaves it alone. Component reversal etc. now happens only for IP addresses.
4768 34. The host_aton() function is supposed to be passed a string that is known
4769 to be a valid IP address. However, in the case of IPv6 addresses, it was
4770 not checking this. This is a hostage to fortune. Exim now panics and dies
4771 if the condition is not met. A case was found where this could be provoked
4772 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4773 components; fortuitously, this particular loophole had already been fixed
4774 by change 4.50/55 or 4.44/33 above.
4776 If there are any other similar loopholes, the new check in host_aton()
4777 itself should stop them being exploited. The report I received stated that
4778 data on the command line could provoke the exploit when Exim was running as
4779 exim, but did not say which command line option was involved. All I could
4780 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4781 running as the user.
4784 35. There was a buffer overflow vulnerability in the SPA authentication code
4785 (which came originally from the Samba project). I have added a test to the
4786 spa_base64_to_bits() function which I hope fixes it.
4789 36. The daemon start-up calls getloadavg() while still root for those OS that
4790 need the first call to be done as root, but it missed one case: when
4791 deliver_queue_load_max is set with deliver_drop_privilege. This is
4792 necessary for the benefit of the queue runner, because there is no re-exec
4793 when deliver_drop_privilege is set.
4795 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4796 was in use, was not putting the data itself into the right store pool;
4797 consequently, it could be overwritten for a subsequent message in the same
4798 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4801 38. Sometimes the final signoff response after QUIT could fail to get
4802 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4803 < 0 before doing a fflush(). This bug looks as though it goes back to the
4804 introduction of TLS in release 3.20, but "sometimes" must have been rare
4805 because the tests only now provoked it.
4811 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4812 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4813 Now it returns an integer. A number of calls were still expecting a T/F
4814 return. Fortuitously, in all cases, the tests worked in OK situations,
4815 which is the norm. However, things would have gone wrong on any write
4816 failures on the smtp file descriptor. This function is used when sending
4817 messages over SMTP and also when doing verify callouts.
4819 2. When Exim is called to do synchronous delivery of a locally submitted
4820 message (the -odf or -odi options), it no longer closes stderr before doing
4823 3. Implemented the mua_wrapper option.
4825 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4827 5. Implemented the functions header_remove(), header_testname(),
4828 header_add_at_position(), and receive_remove_recipient(), and exported them
4831 6. If an ACL "warn" statement specified the addition of headers, Exim already
4832 inserted X-ACL-Warn: at the start if there was no header name. However, it
4833 was not making this test for the second and subsequent header lines if
4834 there were newlines in the string. This meant that an invalid header could
4835 be inserted if Exim was badly configured.
4837 7. Allow an ACL "warn" statement to add header lines at the start or after all
4838 the Received: headers, as well as at the end.
4840 8. Added the rcpt_4xx retry error code.
4842 9. Added postmaster_mailfrom=xxx to callout verification option.
4844 10. Added mailfrom=xxxx to the callout verification option, for verify=
4847 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4848 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4850 12. Inserted some casts to stop certain compilers warning when using pointer
4851 differences as field lengths or precisions in printf-type calls (mostly
4852 affecting debugging statements).
4854 13. Added optional readline() support for -be (dynamically loaded).
4856 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4857 same clock tick as a message's arrival, so that its received time was the
4858 same as the "first fail" time on the retry record, and that message
4859 remained on the queue past the ultimate address timeout, every queue runner
4860 would try a delivery (because it was past the ultimate address timeout) but
4861 after another failure, the ultimate address timeout, which should have then
4862 bounced the address, did not kick in. This was a "< instead of <=" error;
4863 in most cases the first failure would have been in the next clock tick
4864 after the received time, and all would be well.
4866 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4867 being recognized when the domain list was tested by the match_domain
4868 condition in an expansion string.
4870 16. Added the ${str2b64: operator.
4872 17. Exim was always calling setrlimit() to set a large limit for the number of
4873 processes, without checking whether the existing limit was already
4874 adequate. (It did check for the limit on file descriptors.) Furthermore,
4875 errors from getrlimit() and setrlimit() were being ignored. Now they are
4876 logged to the main and panic logs, but Exim does carry on, to try to do its
4877 job under whatever limits there are.
4879 18. Imported PCRE 5.0.
4881 19. Trivial typo in log message " temporarily refused connection" (the leading
4884 20. If the log selector return_path_on_delivery was set and an address was
4885 redirected to /dev/null, the delivery process crashed because it assumed
4886 that a return path would always be set for a "successful" delivery. In this
4887 case, the whole delivery is bypassed as an optimization, and therefore no
4890 21. Internal re-arrangement: the function for sending a challenge and reading
4891 a response while authentication was assuming a zero-terminated challenge
4892 string. It's now changed to take a pointer and a length, to allow for
4893 binary data in such strings.
4895 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4897 23. Exim was not respecting finduser_retries when seeking the login of the
4898 uid under which it was called; it was always trying 10 times. (The default
4899 setting of finduser_retries is zero.) Also, it was sleeping after the final
4900 failure, which is pointless.
4902 24. Implemented tls_on_connect_ports.
4904 25. Implemented acl_smtp_predata.
4906 26. If the domain in control=submission is set empty, Exim assumes that the
4907 authenticated id is a complete email address when it generates From: or
4908 Sender: header lines.
4910 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4911 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4912 chown and chgrp in /bin and hostname in /usr/bin.
4914 28. Exim was keeping the "process log" file open after each use, just as it
4915 does for the main log. This opens the possibility of it remaining open for
4916 long periods when the USR1 signal hits a daemon. Occasional processlog
4917 errors were reported, that could have been caused by this. Anyway, it seems
4918 much more sensible not to leave this file open at all, so that is what now
4921 29. The long-running daemon process does not normally write to the log once it
4922 has entered its main loop, and it closes the log before doing so. This is
4923 so that log files can straightforwardly be renamed and moved. However,
4924 there are a couple of unusual error situations where the daemon does write
4925 log entries, and I had neglected to close the log afterwards.
4927 30. The text of an SMTP error response that was received during a remote
4928 delivery was being truncated at 512 bytes. This is too short for some of
4929 the long messages that one sometimes sees. I've increased the limit to
4932 31. It is now possible to make retry rules that apply only when a message has a
4933 specific sender, in particular, an empty sender.
4935 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4936 it possible to be selective about when SMTP synchronization is enforced.
4938 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4940 32. Implemented hosts_connection_nolog.
4942 33. Added an ACL for QUIT.
4944 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4947 35. Added mailbox_size and mailbox_filecount to appendfile.
4949 36. Added control = no_multiline_responses to ACLs.
4951 37. There was a bug in the logic of the code that waits for the clock to tick
4952 in the case where the clock went backwards by a substantial amount such
4953 that the microsecond fraction of "now" was more than the microsecond
4954 fraction of "then" (but the whole seconds number was less).
4956 38. Added support for the libradius Radius client library this is found on
4957 FreeBSD (previously only the radiusclient library was supported).
4963 1. When certain lookups returned multiple values in the form name=value, the
4964 quoting of the values was not always being done properly. Specifically:
4965 (a) If the value started with a double quote, but contained no whitespace,
4967 (b) If the value contained whitespace other than a space character (i.e.
4968 tabs or newlines or carriage returns) it was not quoted.
4969 This fix has been applied to the mysql and pgsql lookups by writing a
4970 separate quoting function and calling it from the lookup code. The fix
4971 should probably also be applied to nisplus, ibase and oracle lookups, but
4972 since I cannot test any of those, I have not disturbed their existing code.
4974 2. A hit in the callout cache for a specific address caused a log line with no
4975 reason for rejecting RCPT. Now it says "Previous (cached) callout
4976 verification failure".
4978 3. There was an off-by-one bug in the queryprogram router. An over-long
4979 return line was truncated at 256 instead of 255 characters, thereby
4980 overflowing its buffer with the terminating zero. As well as fixing this, I
4981 have increased the buffer size to 1024 (and made a note to document this).
4983 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4984 when Exim is waiting for an SMTP response from a remote server, Exim
4985 restarts its select() call on the socket, thereby resetting its timeout.
4986 This is not a problem when such interrupts are rare. Somebody set up a cron
4987 job to run exiwhat every 2 minutes, which is less than the normal select()
4988 timeout (5 or 10 minutes). This meant that the select() timeout never
4989 kicked in because it was always reset. I have fixed this by comparing the
4990 time when an interrupt arrives with the time at the start of the first call
4991 to select(). If more time than the timeout has elapsed, the interrupt is
4992 treated as a timeout.
4994 5. Some internal re-factoring in preparation for the addition of Sieve
4995 extensions (by MH). In particular, the "personal" test is moved to a
4996 separate function, and given an option for scanning Cc: and Bcc: (which is
4997 not set for Exim filters).
4999 6. When Exim created an email address using the login of the caller as the
5000 local part (e.g. when creating a From: or Sender: header line), it was not
5001 quoting the local part when it contained special characters such as @.
5003 7. Installed new OpenBSD configuration files.
5005 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5006 try to make them clearer.
5008 9. Callout options, other than the timeout value, were being ignored when
5009 verifying sender addresses in header lines. For example, when using
5011 verify = header_sender/callout=no_cache
5013 the cache was (incorrectly) being used.
5015 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5016 only the Cygwin environment.
5018 11. When return_path_on_delivery was set as a log selector, if different remote
5019 addresses in the same message used different return paths and parallel
5020 remote delivery occurred, the wrong values would sometimes be logged.
5021 (Whenever a remote delivery process finished, the return path value from
5022 the most recently started remote delivery process was logged.)
5024 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5025 lines when AUTH and/or TLS are in use. This is the "received protocol"
5026 field. Exim used to use "asmtp" for authenticated SMTP, without any
5027 indication (in the protocol name) for TLS use. Now it follows the RFC and
5028 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5029 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5030 names appear in log lines as well as in Received: header lines.
5032 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5033 extensions, and comparison tests, and to fix some bugs.
5035 14. Changes to the "personal" filter test:
5037 (1) The test was buggy in that it was just doing the equivalent of
5038 "contains" tests on header lines. For example, if a user's address was
5039 anne@some.where, the "personal" test would incorrectly be true for
5041 To: susanne@some.where
5043 This test is now done by extracting each address from the header in turn,
5044 and checking the entire address. Other tests that are part of "personal"
5045 are now done using regular expressions (for example, to check local parts
5046 of addresses in From: header lines).
5048 (2) The list of non-personal local parts in From: addresses has been
5049 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5050 taken from the Sieve specification recommendations.
5052 (3) If the message contains any header line starting with "List-" it is
5053 treated as non-personal.
5055 (4) The test for "circular" in the Subject: header line has been removed
5056 because it now seems ill-conceived.
5058 15. Minor typos in src/EDITME comments corrected.
5060 16. Installed latest exipick from John Jetmore.
5062 17. If headers_add on a router specified a text string that was too long for
5063 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5064 of string_sprintf() is now avoided.
5066 18. $message_body_size was not set (it was always zero) when running the DATA
5067 ACL and the local_scan() function.
5069 19. For the "mail" command in an Exim filter, no default was being set for
5070 the once_repeat time, causing a random time value to be used if "once" was
5071 specified. (If the value happened to be <= 0, no repeat happened.) The
5072 default is now 0s, meaning "never repeat". The "vacation" command was OK
5073 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5074 (I found it when inspecting the code).
5076 20. There is now an overall timeout for performing a callout verification. It
5077 defaults to 4 times the callout timeout, which applies to individual SMTP
5078 commands during the callout. The overall timeout applies when there is more
5079 than one host that can be tried. The timeout is checked before trying the
5080 next host. This prevents very long delays if there are a large number of
5081 hosts and all are timing out (e.g. when the network connections are timing
5082 out). The value of the overall timeout can be changed by specifying an
5083 additional sub-option for "callout", called "maxwait". For example:
5085 verify = sender/callout=5s,maxwait=20s
5087 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5088 to the end before writing, but this should make it even safer).
5090 22. Exim was forgetting that it had advertised PIPELINING for the second and
5091 subsequent messages on an SMTP connection. It was also not resetting its
5092 memory on STARTTLS and an internal HELO.
5094 23. When Exim logs an SMTP synchronization error within a session, it now
5095 records whether PIPELINING has been advertised or not.
5097 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5098 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5099 rather than long int.
5101 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5103 26. Added the never_mail option to autoreply.
5109 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5110 crash if the getsockname() call failed; this can happen if a connection is
5111 closed very soon after it is established. The problem was simply in the
5112 order in which certain operations were done, causing Exim to try to write
5113 to the SMTP stream before it had set up the file descriptor. The bug has
5114 been fixed by making things happen in the correct order.
5120 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5121 before the connection was closed, thus losing the rejection response.
5123 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5124 some early Solaris releases, but causes trouble in current releases where
5125 socklen_t is defined.
5127 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5130 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5133 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5134 that releasing the top of it at the end releases what was used for sub-
5135 expansions (unless the block got too big). However, discard this block if
5136 the first thing is a variable or header, so that we can use its block when
5137 it is dynamic (useful for very large $message_headers, for example).
5139 6. Lookups now cache *every* query, not just the most recent. A new, separate
5140 store pool is used for this. It can be recovered when all lookup caches are
5141 flushed. Lookups now release memory at the end of their result strings.
5142 This has involved some general refactoring of the lookup sources.
5144 7. Some code has been added to the store_xxx() functions to reduce the amount
5145 of flapping under certain conditions.
5147 8. log_incoming_interface used to affect only the <= reception log lines. Now
5148 it causes the local interface and port to be added to several more SMTP log
5149 lines, for example "SMTP connection from", and rejection lines.
5151 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5153 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5155 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5156 could be overwritten at the end of the current message (or the start of a
5157 new message if it was set in a HELO ACL). The value is now preserved for
5158 the duration of the SMTP connection.
5160 12. If a transport had a headers_rewrite setting, and a matching header line
5161 contained an unqualified address, that address was qualified, even if it
5162 did not match any rewriting rules. The underlying bug was that the values
5163 of the flags that permit the existence of unqualified sender and recipient
5164 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5165 non-local messages, and by -bnq for local messages) were not being
5166 preserved with the message after it was received.
5168 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5169 "next input=" as part of the text comprising the host identity instead of
5170 the correct text. The code was using the same buffer for two different
5171 strings. However, depending on which order the printing function evaluated
5172 its arguments, the bug did not always show up. Under Linux, for example, my
5173 test suite worked just fine.
5175 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5176 doesn't work with some older versions of Perl. It has been changed to "my",
5177 which in any case is probably the better facility to use.
5179 15. A really picky compiler found some instances of statements for creating
5180 error messages that either had too many or two few arguments for the format
5183 16. The size of the buffer for calls to the DNS resolver has been increased
5184 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5185 for addresses that have a lot of PTR records. This alleviates a problem; it
5186 does not fully solve it.
5188 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5189 buffer now truncates the list and logs the incident, which is the same
5190 action as happens when Exim is looking up a host name and its aliases.
5191 Previously in this situation something unpredictable would happen;
5192 sometimes it was "internal error: store_reset failed".
5194 18. If a server dropped the connection unexpectedly when an Exim client was
5195 using GnuTLS and trying to read a response, the client delivery process
5196 crashed while trying to generate an error log message.
5198 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5199 string, for example:
5201 warn message = H1: something\nH2: something
5203 the text was added as a single header line from Exim's point of view
5204 though it ended up OK in the delivered message. However, searching for the
5205 second and subsequent header lines using $h_h2: did not work. This has been
5206 fixed. Similarly, if a system filter added multiple headers in this way,
5207 the routers could not see them.
5209 20. Expanded the error message when iplsearch is called with an invalid key to
5210 suggest using net-iplsearch in a host list.
5212 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5213 ACLs are no longer actually imposed (and a message to that effect is
5216 22. If a "gecos" field in a passwd entry contained escaped characters, in
5217 particular, if it contained a \" sequence, Exim got it wrong when building
5218 a From: or a Sender: header from that name. A second bug also caused
5219 incorrect handling when an unquoted " was present following a character
5220 that needed quoting.
5222 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5223 was not being matched caselessly.
5225 24. Arranged for all hyphens in the exim.8 source to be escaped with
5228 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5229 properly. Recipient callout cache records were still being keyed to include
5230 the sender, even when use_sender was set false. This led to far more
5231 callouts that were necessary. The sender is no longer included in the key
5232 when use_sender is false.
5234 26. Added "control = submission" modifier to ACLs.
5236 27. Added the ${base62d: operator to decode base 62 numbers.
5238 28. dnsdb lookups can now access SRV records.
5240 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5241 the configuration file.
5243 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5244 (-v) mode. This makes the output for a verbose queue run more intelligible.
5246 31. Added a use_postmaster feature to recipient callouts.
5248 32. Added the $body_zerocount variable, containing the number of binary zero
5249 bytes in the message body.
5251 33. The time of last modification of the "new" subdirectory is now used as the
5252 "mailbox time last read" when there is a quota error for a maildir
5255 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5257 35. Added +ignore_unknown as a special item in host lists.
5259 36. Code for decoding IPv6 addresses in host lists is now included, even if
5260 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5261 address was recognized as an IP address, but was then not correctly decoded
5262 into binary, causing unexpected and incorrect effects when compared with
5269 1. Very minor rewording of debugging text in manualroute to say "list of
5270 hosts" instead of "hostlist".
5272 2. If verify=header_syntax was set, and a header line with an unqualified
5273 address (no domain) and a large number of spaces between the end of the
5274 name and the colon was received, the reception process suffered a buffer
5275 overflow, and (when I tested it) crashed. This was caused by some obsolete
5276 code that should have been removed. The fix is to remove it!
5278 3. When running in the test harness, delay a bit after writing a bounce
5279 message to get a bit more predictability in the log output.
5281 4. Added a call to search_tidyup() just before forking a reception process. In
5282 theory, someone could use a lookup in the expansion of smtp_accept_max_
5283 per_host which, without the tidyup, could leave open a database connection.
5285 5. Added the variables $recipient_data and $sender_data which get set from a
5286 lookup success in an ACL "recipients" or "senders" condition, or a router
5287 "senders" option, similar to $domain_data and $local_part_data.
5289 6. Moved the writing of debug_print from before to after the "senders" test
5292 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5293 problems for message scanning, either using a data ACL, or using
5294 local_scan() because the Received: header was not generated till after they
5295 were called (in order to set the time as the time of reception completion).
5296 I have revised the way this works. The header is now generated after the
5297 body is received, but before the ACL or local_scan() are called. After they
5298 are run, the timestamp in the header is updated.
5304 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5305 before starting a queue runner without re-exec. This happened only when
5306 deliver_drop_privilege was set or when the Exim user was set to root. The
5307 effect of the bug was that timeouts during subsequent deliveries caused
5308 crashes instead of being properly handled. The handler is now left at its
5309 default (and expected) setting.
5311 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5312 message, again when deliver_drop_privilege is set or Exim is run as root.
5313 The bug described in (1) was not present in this case, but the tidying up
5314 of the other signals was missing. I have made the two cases consistent.
5316 3. The ignore_target_hosts setting on a manualroute router was being ignored
5317 for hosts that were looked up using the /MX notation.
5319 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5322 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5323 operated on the sender address. After changing the $sender_address to <>
5324 for the sender address verify, Exim was re-instated it as the original
5325 (before rewriting) address, but remembering that it had rewritten it, so it
5326 wasn't rewriting it again. This bug also had the effect of breaking the
5327 sender address verification caching when the sender address was rewritten.
5329 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5330 This has been changed so that if the ip literal address matches
5331 ignore_target_hosts, the router declines.
5333 7. Added expansion conditions match_domain, match_address, and match_local_
5334 part (NOT match_host).
5336 8. The placeholder for the Received: header didn't have a length field set.
5338 9. Added code to Exim itself and to exim_lock to test for a specific race
5339 condition that could lead to file corruption when using MBX delivery. The
5340 issue is with the lockfile that is created in /tmp. If this file is removed
5341 after a process has opened it but before that process has acquired a lock,
5342 there is the potential for a second process to recreate the file and also
5343 acquire a lock. This could lead to two Exim processes writing to the file
5344 at the same time. The added code performs the same test as UW imapd; it
5345 checks after acquiring the lock that its file descriptor still refers to
5346 the same named file.
5348 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5349 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5352 11. Added the smtp_active_hostname option. If used, this will typically be made
5353 to depend on the incoming interface address. Because $interface_address is
5354 not set up until the daemon has forked a reception process, error responses
5355 that can happen earlier (such as "too many connections") no longer contain
5358 12. If an expansion in a condition on a "warn" statement fails because a lookup
5359 defers, the "warn" statement is abandoned, and the next ACL statement is
5360 processed. Previously this caused the whole ACL to be aborted.
5362 13. Added the iplsearch lookup type.
5364 14. Added ident_timeout as a log selector.
5366 15. Added tls_certificate_verified as a log selector.
5368 16. Added a global option tls_require_ciphers (compare the smtp transport
5369 option of the same name). This controls incoming TLS connections.
5371 17. I finally figured out how to make tls_require_ciphers do a similar thing
5372 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5373 before starting the TLS session.
5375 18. Tabs are now shown as \t in -bP output.
5377 19. If the log selector return_path_on_delivery was set, Exim crashed when
5378 bouncing a message because it had too many Received: header lines.
5380 20. If two routers both had headers_remove settings, and the first one included
5381 a superfluous trailing colon, the final name in the first list and the
5382 first name in the second list were incorrectly joined into one item (with a
5383 colon in the middle).
5389 1. Added -C and -D options to the exinext utility, mainly to make it easier
5390 to include in the automated testing, but these could be helpful when
5391 multiple configurations are in use.
5393 2. The exinext utility was not formatting the output nicely when there was
5394 an alternate port involved in the retry record key, nor when there was a
5395 message id as well (for retries that were specific to a specific message
5396 and a specific host). It was also confused by IPv6 addresses, because of
5397 the additional colons they contain. I have fixed the IPv4 problem, and
5398 patched it up to do a reasonable job for IPv6.
5400 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5401 delivery, the log line now contains "pipelined" if PIPELINING was used.
5403 4. An SMTP transport process used to panic and die if the bind() call to set
5404 an explicit outgoing interface failed. This has been changed; it is now
5405 treated in the same way as a connect() failure.
5407 5. A reference to $sender_host_name in the part of a conditional expansion
5408 that was being skipped was still causing a DNS lookup. This no longer
5411 6. The def: expansion condition was not recognizing references to header lines
5412 that used bh_ and bheader_.
5414 7. Added the _cache feature to named lists.
5416 8. The code for checking quota_filecount in the appendfile transport was
5417 allowing one more file than it should have been.
5419 9. For compatibility with Sendmail, the command line option
5427 and sets the incoming protocol and host name (for trusted callers). The
5428 host name and its colon can be omitted when only the protocol is to be set.
5429 Note the Exim already has two private options, -pd and -ps, that refer to
5430 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5431 "s", but I don't think that's a major issue.
5433 10. A number of refactoring changes to the code, none of which should affect
5436 (a) The number of logging options was getting close to filling up the
5437 32-bit word that was used as a bit map. I have split them into two classes:
5438 those that are passed in the argument to log_write(), and those that are
5439 only ever tested independently outside of that function. These are now in
5440 separate 32-bit words, so there is plenty of room for expansion again.
5441 There is no change in the user interface or the logging behaviour.
5443 (b) When building, for example, log lines, the code previously used a
5444 macro that called string_cat() twice, in order to add two strings. This is
5445 not really sufficiently general. Furthermore, there was one instance where
5446 it was actually wrong because one of the argument was used twice, and in
5447 one call a function was used. (As it happened, calling the function twice
5448 did not affect the overall behaviour.) The macro has been replaced by a
5449 function that can join an arbitrary number of extra strings onto a growing
5452 (c) The code for expansion conditions now uses a table and a binary chop
5453 instead of a serial search (which was left over from when there were very
5454 few conditions). Also, it now recognizes conditions like "pam" even when
5455 the relevant support is not compiled in: a suitably worded error message is
5456 given if an attempt is made to use such a condition.
5458 11. Added ${time_interval:xxxxx}.
5460 12. A bug was causing one of the ddress fields not to be passed back correctly
5461 from remote delivery subprocesses. The field in question was not being
5462 subsequently used, so this caused to problems in practice.
5464 13. Added new log selectors queue_time and deliver_time.
5466 14. Might have fixed a bug in maildirsizefile handling that threw up
5467 "unexpected character" debug warnings, and recalculated the data
5468 unnecessarily. In any case, I expanded the warning message to give more
5471 15. Added the message "Restricted characters in address" to the statements in
5472 the default ACL that block characters like @ and % in local parts.
5474 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5475 Three changes have been made:
5477 (a) There was a serious bug; a negative response to MAIL caused the whole
5478 recipient domain to be cached as invalid, thereby blocking all messages
5479 to all local parts at the same domain, from all senders. This bug has
5480 been fixed. The domain is no longer cached after a negative response to
5481 MAIL if the sender used is not empty.
5483 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5486 (c) A new callout option, "use_sender" has been added for people who want
5487 the modified behaviour.
5493 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5496 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5497 indeed breaks things for older releases.
5499 3. Added additional logging to the case where there is a problem reading data
5500 from a filter that is running in a subprocess using a pipe, in order to
5501 try to track down a specific problem.
5503 4. Testing facility fudge: when running in the test harness and attempting
5504 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5505 getting "No route to host". Convert this to a timeout.
5507 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5510 6. Some OS don't have socklen_t but use size_t instead. This affects the
5511 fifth argument of getsockopt() amongst other things. This is now
5512 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5513 can be set for individual OS. I have set it for SunOS5, OSF1, and
5514 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5515 some earlier ones do not.
5517 7. Change 4.30/15 was not doing the test caselessly.
5519 8. The standard form for an IPv6 address literal was being rejected by address
5520 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5521 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5522 this, as well as the form without the "IPv6" on the front (but only when
5523 address literals are enabled, of course).
5525 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5527 10. Exim crashed if a message with an empty sender address specified by -f
5528 encountered a router with an errors_to setting. This could be provoked only
5529 by a command such as
5533 where an empty string was supplied; "<>" did not hit this bug.
5535 11. Installed PCRE release 4.5.
5537 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5538 remained set. It is now erased.
5540 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5541 times from message ids (which are base 36 rather than the normal 62).
5543 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5544 were being counted as actual protocol errors, and logged if the log
5545 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5546 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5547 and DATA following a set of rejected RCPTs do not count as protocol errors.
5548 In other words, Exim assumes they were pipelined, though this may not
5549 actually be the case. Of course, in all cases the client gets an
5550 appropriate error code.
5552 15. If a lookup fails in an ACL condition, a message about the failure may
5553 be available; it is used if testing the ACL cannot continue, because most
5554 such messages specify what the cause of the deferral is. However, some
5555 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5556 that caused an old message to be retained and used if a later statement
5557 caused a defer, replacing the real cause of the deferral.
5559 16. If an IP address had so many PTR records that the DNS lookup buffer
5560 was not large enough to hold them, Exim could crash while trying to process
5561 the truncated data. It now detects and logs this case.
5563 17. Further to 4.21/58, another change has been made: if (and only if) the
5564 first line of a message (the first header line) ends with CRLF, a bare LF
5565 in a subsequent header line has a space inserted after it, so as not to
5566 terminate the header.
5568 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5569 unnecessarily, used atoi() instead of strtol(), and didn't check the
5570 termination when getting file sizes from file names by regex.
5572 19. Completely re-implemented the support for maildirsize files, in the light
5573 of a number of problems with the previous contributed implementation
5574 (4.30/29). In particular:
5576 . If the quota is zero, the maildirsize file is maintained, but no quota is
5579 . If the maildir directory does not exist, it is created before any attempt
5580 to write a maildirsize file.
5582 . The quota value in the file is just a cache; if the quota is changed in
5583 the transport, the new value overrides.
5585 . A regular expression is available for excluding directories from the
5588 20. The autoreply transport checks the characters in options that define the
5589 message's headers; it allows continued headers, but it was checking with
5590 isspace() after an embedded newline instead of explicitly looking for a
5593 21. If all the "regular" hosts to which an address was routed had passed their
5594 expiry times, and had not reached their retry times, the address was
5595 bounced, even if fallback hosts were defined. Now Exim should go on to try
5598 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5599 equivalent code in the SMTP transport. Some hosts send humungous responses
5600 to HELO/EHLO, more than 1024 it seems.
5602 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5603 gives compiler warnings in some environments. I've now done it "properly",
5606 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5607 (because of problems with the built-in one) was declared to return uschar *
5608 instead of char *, causing compiler failure.
5610 25. Fixed a file descriptor leak when processing alias/forward files.
5612 26. Fixed a minor format string issue in dbfn.c.
5614 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5616 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5617 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5620 29. When a sender address is verified, it is cached, to save repeating the test
5621 when there is more than one recipient in a message. However, when the
5622 verification involves a callout, it is possible for different callout
5623 options to be set for different recipients. It is too complicated to keep
5624 track of this in the cache, so now Exim always runs a verification when a
5625 callout is required, relying on the callout cache for the optimization.
5626 The overhead is duplication of the address routing, but this should not be
5629 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5630 to be verified with callout=postmaster, and the main callout worked but the
5631 postmaster check failed, the verification correctly failed. However, if a
5632 subsequent RCPT command asked for sender verification *without* the
5633 postmaster check, incorrect caching caused this verification also to fail,
5636 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5637 it was not caching the DNS options (qualify_single, search_parents) that
5638 were used when the lookup failed. A subsequent lookup with different
5639 options therefore always gave the same answer, though there were cases
5640 where it should not have. (Example: a "domains = !$mx_any" option on a
5641 dnslookup router: the "domains" option is always processed without any
5642 widening, but the router might have qualify_single set.) Now Exim uses the
5643 cached value only when the same options are set.
5645 32. Added John Jetmore's "exipick" utility to the distribution.
5647 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5648 than a timeout (e.g. a certificate is required, and is not provided), an
5649 Exim server now closes the connection immediately. Previously it waited for
5650 the client to close - but if the client is SSL, it seems that they each
5651 wait for each other, leading to a delay before one of them times out.
5653 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5654 maintained 0.8.x compatibility because I don't think many are using it, and
5655 it is clearly obsolete.
5657 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5660 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5661 client certificate was expired. A simple patch fixes this, though I don't
5662 understand the full logic of why the verify callback is called multiple
5665 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5666 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5667 which causes problems with some clients (such as the Certicom SSL Plus
5668 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5669 disables the coutermeasure allowing Eudora to connect."
5671 38. Exim was not checking that a write() to a log file succeeded. This could
5672 lead to Bad Things if a log got too big, in particular if it hit a file
5673 size limit. Exim now panics and dies if it cannot write to a log file, just
5674 as it does if it cannot open a log file.
5676 39. Modified OS/Makefile-Linux so that it now contains
5678 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5680 The two -D definitions ensure that Exim is compiled with large file
5681 support, which makes it possible to handle log files that are bigger than
5684 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5685 instance) a domain was checked against a named list that involved a lookup,
5686 causing $domain_data to be set, then another domain was checked against the
5687 same list, then the first domain was re-checked, the value of $domain_data
5688 after the final check could be wrong. In particular, if the second check
5689 failed, it could be set empty. This bug probably also applied to
5692 41. The strip_trailing_dot option was not being applied to the address given
5693 with the -f command-line option.
5695 42. The code for reading a message's header from the spool was incrementing
5696 $received_count, but never initializing it. This meant that the value was
5697 incorrect (doubled) while delivering a message in the same process in which
5698 it was received. In the most common configuration of Exim, this never
5699 happens - a fresh exec is done - but it can happen when
5700 deliver_drop_privilege is set.
5702 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5703 it now includes up to 150 characters of the unexpected data in the log
5706 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5707 and building data strings. The size of both of these buffers was 10 000
5708 bytes - far larger than anybody would *ever* want, thought I. Needless to
5709 say, somebody hit the limit. I have increased the maximum line length to
5710 20 000 and the maximum data length of concatenated lines to 100 000. I have
5711 also fixed two bugs, because there was no checking on these buffers. Tsk,
5712 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5713 buffer is too small.
5715 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5716 lsearch lookups. Now it does.
5718 46. When parsing a route_list item in a manualroute router, a fixed-length
5719 buffer was used for the list of hosts. I made this 1024 bytes long,
5720 thinking that nobody would ever have a list of hosts that long. Wrong.
5721 Somebody had a whole pile of complicated expansion conditions, and the
5722 string was silently truncated, leading to an expansion error. It turns out
5723 that it is easier to change to an unlimited length (owing to other changes
5724 that have happened since this code was originally written) than to build
5725 structure for giving a limitation error. The length of the item that
5726 expands into the list of hosts is now unlimited.
5728 47. The lsearch lookup could not handle data where the length of text line was
5729 more than 4095 characters. Such lines were truncated, leading to shortened
5730 data being returned. It should now handle lines of any length.
5732 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5733 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5736 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5737 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5738 no longer happens in comment lines. A list of replacements is now placed
5739 at the head of all of the source files, except those whose only change is
5740 to replace PERL_COMMAND in the very first #! line.
5742 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5743 messages on the queue, with a bottom-up merge sort, using code contributed
5744 by Michael Haardt. This should make operations like -bp somewhat faster on
5745 large queues. It won't affect queue runners, except when queue_run_in_order
5748 51. Installed eximstats 1.31 in the distribution.
5750 52. Added support for SRV lookups to the dnslookup router.
5752 53. If an ACL referred to $message_body or $message_body_end, the value was not
5753 reset for any messages that followed in the same SMTP session.
5755 54. The store-handling optimization for building very long strings was not
5756 differentiating between the different store pools. I don't think this
5757 actually made any difference in practice, but I've tidied it.
5759 55. While running the routers to verify a sender address, $sender_address
5760 was still set to the sender address. This is wrong, because when routing to
5761 send a bounce to the sender, it would be empty. Therefore, I have changed
5762 it so that, while verifying a sender address, $sender_address is set to <>.
5763 (There is no change to what happens when verifying a recipient address.)
5765 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5766 target A or AAAA records (if not already returned) without resetting the
5767 qualify_single or search_parents options of the DNS resolver. These are
5768 inappropriate in this case because the targets of MX and SRV records must
5769 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5770 target that, when qualified, matched something in the local domain. These
5771 two options are now turned off when doing these lookups.
5773 57. It seems that at least some releases of Reiserfs (which does not have the
5774 concept of a fixed number of inodes) returns zero and not -1 for the
5775 number of available inodes. This interacted badly with check_spool_inodes,
5776 which assumed that -1 was the "no such thing" setting. What I have done is
5777 to check that the total number of inodes is greater than zero before doing
5778 the test of how many are available.
5780 58. When a "warn" ACL statement has a log_message modifier, the message is
5781 remembered, and not repeated. This is to avoid a lot of repetition when a
5782 message has many recipients that cause the same warning to be written.
5783 However, Exim was preserving the list of already written lines for an
5784 entire SMTP session, which doesn't seem right. The memory is now reset if a
5785 new message is started.
5787 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5788 debugging output unless log_rewrite was also set.
5790 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5791 of (int)(handle) when we know that handle contains (void *)(-1).
5793 61. The Exim daemon panic-logs an error return when it closes the incoming
5794 connection. However "connection reset by peer" seems to be common, and
5795 isn't really an error worthy of noting specially, so that particular error
5798 62. When Exim is trying to find all the local interfaces, it used to panic and
5799 die if the ioctl to get the interface flags failed. However, it seems that
5800 on at least one OS (Solaris 9) it is possible to have an interface that is
5801 included in the list of interfaces, but for which you get a failure error
5802 for this call. This happens when the interface is not "plumbed" into a
5803 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5804 failure of the "get flags" call assumes that the interface is down.
5806 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5807 makes life easier for people who are doing arithmetic on fields extracted
5808 from dates, where you often get leading zeros that should not be
5809 interpreted as octal.
5811 64. Added qualify_domain to the redirect router, to override the global
5814 65. If a pathologically long header line contained very many addresses (the
5815 report of this problem mentioned 10 000) and each of them was rewritten,
5816 Exim could use up a very large amount of memory. (It kept on making new
5817 copies of the header line as it rewrote, and never released the old ones.)
5818 At the expense of a bit more processing, the header rewriting function has
5819 been changed so that it no longer eats memory in this way.
5821 66. The generation of the Received: header has been moved from the time that a
5822 message starts to be received, to the time that it finishes. The timestamp
5823 in the Received: header should now be very close to that of the <= log
5824 line. There are two side-effects of this change:
5826 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5827 logged header lines no longer include the local Received: line, because
5828 it has not yet been created. The same applies to a copy of the message
5829 that is returned to a non-SMTP sender when a message is rejected.
5831 (b) When a filter file is tested using -bf, no additional Received: header
5832 is added to the test message. After some thought, I decided that this
5835 This change does not affect the value of $received_for. It is still set
5836 after address rewriting, but before local_scan() is called.
5838 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5840 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5841 gave an unhelpful panic error message, and a defer error. I have managed to
5842 change this behaviour so that it now rejects any supplied certificate,
5843 which seems right, as the list of acceptable certificates is empty.
5845 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5846 gave an unhelpful defer error. I have not managed to make this reject any
5847 supplied certificates, but the error message it gives is "no certificate
5848 supplied", which is not helpful.
5850 70. exigrep's output now also includes lines that are not associated with any
5851 message, but which match the given pattern. Implemented by a patch from
5852 Martin Sluka, which also tidied up the Perl a bit.
5854 71. Recipient callout verification, like sender verification, was using <> in
5855 the MAIL FROM command. This isn't really the right thing, since the actual
5856 sender may affect whether the remote host accepts the recipient or not. I
5857 have changed it to use the actual sender in the callout; this means that
5858 the cache record is now keyed on a recipient/sender pair, not just the
5859 recipient address. There doesn't seem to be a real danger of callout loops,
5860 since a callout by the remote host to check the sender would use <>.
5861 [SEE ABOVE: changed after hitting problems.]
5863 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5864 temporary errors. However, in the case of such a code being given after
5865 the end of a data transmission (i.e. after ".") Exim was failing to write
5866 a retry record for the message. (Yes, there was some broken host that was
5867 actually sending 8xx at this point.)
5869 73. An unknown lookup type in a host list could cause Exim to panic-die when
5870 the list was checked. (An example that provoked this was putting <; in the
5871 middle of a list instead of at the start.) If this happened during a DATA
5872 ACL check, a -D file could be left lying around. This kind of configuration
5873 error no longer causes Exim to die; instead it causes a defer error. The
5874 incident is still logged to the main and panic logs.
5876 74. Buglet left over from Exim 3 conversion. The message "too many messages
5877 in one connection" was written to the rejectlog but not the mainlog, except
5878 when address rewriting (yes!) was being logged.
5880 75. Added write_rejectlog option.
5882 76. When a system filter was run not as root (that is, when system_filter_user
5883 was set), the values of the $n variables were not being returned to the
5884 main process; thus, they were not subsequently available in the $sn
5887 77. Added +return_path_on_delivery log selector.
5889 78. A connection timeout was being treated differently from recipients deferred
5890 when testing hosts_max_try with a message that was older than the host's
5891 retry timeout. (The host should not be counted, thus allowing all hosts to
5892 be tried at least once before bouncing.) This may have been the cause of an
5893 occasionally reported bug whereby a message would remain on the queue
5894 longer than the retry timeout, but would be bounced if a delivery was
5895 forced. I say "may" because I never totally pinned down the problem;
5896 setting up timeout/retry tests is difficult. See also the next item.
5898 79. The ultimate address timeout was not being applied to errors that involved
5899 a combination of host plus message (for example, a timeout on a MAIL
5900 command). When an address resolved to a number of possible hosts, and they
5901 were not all tried for each delivery (e.g. because of hosts_max_try), a
5902 message could remain on the queue longer than the retry timeout.
5904 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5907 81. Fixed an obscure SMTP outgoing bug which required at least the following
5908 conditions: (a) there was another message waiting for the same server;
5909 (b) the server returned 5xx to all RCPT commands in the first message so
5910 that the message was not completed; (c) the server dropped the connection
5911 or gave a negative response to the RSET that Exim sends to abort the
5912 transaction. The observed case was a dropped connection after DATA that had
5913 been sent in pipelining mode. That is, the server had advertised PIPELINING
5914 but was not implementing it correctly. The effect of the bug was incorrect
5915 behaviour, such as trying another host, and this could lead to a crash.
5921 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5922 and daemon.c were passed as pointers to ints; they should have been
5923 pointers to socklen_t variables (which are typically unsigned ints).
5925 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5928 3. Fixed a really odd bug that affected only the testing scheme; patching a
5929 certain fixed string in the binary changed the value of another string that
5930 happened to be identical to the end of the original first string.
5932 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5933 name", it returns that address as the IP address. On some operating
5934 systems (e.g. Solaris), it also passes back the IP address string as the
5935 "host name". However, on others (e.g. Linux), it passes back an empty
5936 string. Exim wasn't checking for this, and was changing the host name to an
5937 empty string, assuming it had been canonicalized.
5939 5. Although rare, it is permitted to have more than one PTR record for a given
5940 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5941 all the names associated with an address, because they do in Solaris.
5942 However, it seems that they do not in Linux for data that comes from the
5943 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5944 I found this out when I moved to a new Linux workstation and tried to run
5945 the Exim test suite.
5947 To get round this problem I have changed the code so that it now does its
5948 own call to the DNS to look up PTR records when searching for a host name.
5949 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5950 addresses that are only in /etc/hosts are still found.
5952 This behaviour is, however, controlled by an option called host_lookup_
5953 order, which defaults to "bydns:byaddr". If people want to use the other
5954 order, or indeed, just use one or the other means of lookup, they can
5955 specify it in this variable.
5957 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5958 some operating systems, this comes back from gethostbyaddr() as an empty
5959 string, and this is what Exim used to test for. However, it seems that in
5960 other systems, "." is yielded. Exim now tests for this case too.
5962 7. The values of check_spool_space and check_log_space are now held internally
5963 as a number of kilobytes instead of an absolute number of bytes. If a
5964 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5965 kilobyte. This means that much larger values can be stored.
5967 8. Exim monitor: an attempt to get the action menu when not actually pointing
5968 at a message produces an empty menu entitled "No message selected". This
5969 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5970 no entries in it ("Shell widget menu has zero width and/or height"). So I
5971 have added a single, blank menu entry in this case.
5973 9. Added ${quote_local_part.
5975 10. MIME decoding is now applied to the contents of Subject: header lines when
5978 11. Now that a reference to $sender_host_address automatically causes a reverse
5979 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5980 host lookup before query-style lookups in lists that might use this
5981 variable. This has therefore been abolished, and the "net-" prefix is no
5982 longer necessary for query-style lookups.
5984 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5985 have been a typo for LFLAGS, so it has been changed.
5987 13. The install script calls Exim with "-C /dev/null" in order to find the
5988 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5989 to be output. However, since Exim outputs its version number before the
5990 error, it didn't break the script. It just looked ugly. I fixed this by
5991 always allowing "-C /dev/null" if the caller is root.
5993 14. Ignore overlarge ACL variable number when reading spool file - insurance
5994 against a later release with more variables having written the file.
5996 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5997 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5998 this, as well as the form without the "IPv6" on the front.
6000 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6001 OS/Makefile-Darwin file.
6003 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6004 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6006 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6007 variable. However, some OS use ints for inodes. I've added cast to long int
6008 to get rid of the compiler warning.
6010 19. I had forgotten to lock out "/../" in configuration file names when
6011 ALT_CONFIG_PREFIX was set.
6013 20. Routers used for verification do not need to specify transports. However,
6014 if such a router generated a host list, and callout was configured, Exim
6015 crashed, because it could not find a port number from the (non-existent)
6016 transport. It now assumes port 25 in this circumstance.
6018 21. Added the -t option to exigrep.
6020 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6021 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6022 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6023 them is set, LOOKUP_LSEARCH is forced.
6025 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6027 24. Added sender and host information to the "rejected by local_scan()" log
6028 line; previously there was no indication of these.
6030 25. Added .include_if_exists.
6032 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6033 Linux. It turns out that not all file systems support this. Apparently some
6034 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6035 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6036 not supported on the file descriptor, is now ignored when Exim is trying to
6037 sync a directory. This applies only to Linux.
6039 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6041 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6042 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6045 29. Added support for maildirsize files from supplied patch (modified a bit).
6047 30. The use of :fail: followed by an empty string could lead Exim to respond to
6048 sender verification failures with (e.g.):
6050 550 Verification failed for <xxx>
6051 550 Sender verify failed
6053 where the first response line was missing the '-' that indicates it is not
6054 the final line of the response.
6056 31. The loop for finding the name of the user that called Exim had a hardwired
6057 limit of 10; it now uses the value of finduser_retries, which is used for
6058 all other user lookups.
6060 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6063 33. Exim was neglecting to zero errno before one call of strtol() when
6064 expanding a string and expecting an integer value. On some systems this
6065 resulted in spurious "integer overflow" errors. Also, it was casting the
6066 result into an int without checking.
6068 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6069 did not work. The code looks as if it has *never* worked, though it appears
6070 to have been documented since at least release 1.62. I have made it work.
6072 35. The "timeout_DNS" error in retry rules, also documented since at least
6073 1.62, also never worked. As it isn't clear exactly what this means, and
6074 clearly it isn't a major issue, I have abolished the feature by treating it
6075 as "timeout", and writing a warning to the main and panic logs.
6077 36. The display of retry rules for -brt wasn't always showing the error code
6080 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6081 timeout_connect_A, timeout_connect_MX.
6083 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6084 to the empty sender.
6086 39. The daemon was not analysing the content of -oX till after it had closed
6087 stderr and disconnected from the controlling terminal. This meant that any
6088 syntax errors were only noted on the panic log, and the return code from
6089 the command was 0. By re-arranging the code a little, I've made the
6090 decoding happen first, so such errors now appear on stderr, and the return
6091 code is 1. However, the actual setting up of the sockets still happens in
6092 the disconnected process, so errors there are still only recorded on the
6095 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6096 connections (as happens on some IP stacks) was logged at start up time as
6097 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6098 it from "IPv6 and IPv4", which means that two separate sockets are being
6101 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6102 says whether AF_INET or AF_INET6 was passed as an argument.
6104 42. Exiwhat output was messed up when time zones were included in log
6107 43. Exiwhat now gives more information about the daemon's listening ports,
6108 and whether -tls-on-connect was used.
6110 44. The "port" option of the smtp transport is now expanded.
6112 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6113 silently ignored. Now an error message is written to the main and panic
6116 46. There's a new ACL modifier called "logwrite" which writes to a log file
6117 as soon as it is encountered.
6119 47. Added $local_user_uid and $local_user_gid at routing time.
6121 48. Exim crashed when trying to verify a sender address that was being
6124 49. Exim was recognizing only a space character after ".include". It now also
6125 recognizes a tab character.
6127 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6128 extracting the relevant information from the specification. The man page no
6129 longer contains scrambled data for the -d option, and I've added a section
6130 at the front about calling Exim under different names.
6132 51. Added "extra_headers" argument to the "mail" command in filter files.
6134 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6137 53. Installed eximstats 1.29.
6139 54. Added transport_filter_timeout as a generic transport option.
6141 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6142 Cc: header lines. This was required by RFC 822, but it not required by RFC
6145 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6146 incoming messages that did not have them. Now it does so only if the
6147 message originates locally, that is, if there is no associated remote host
6148 address. When Resent- header lines are present, this applies to the Resent-
6149 lines rather than the non-Resent- lines.
6151 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6152 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6154 58. Messages for configuration errors now include the name of the main
6155 configuration file - useful now that there may be more than one file in a
6156 list (.included file names were always shown).
6158 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6159 for those rare installations that do not start the daemon as root or run it
6160 setuid root. I've cut out the call to initgroups() if the daemon is not
6163 60. The Exim user and group can now be bound into the binary as text strings
6164 that are looked up at the start of Exim's processing.
6166 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6168 62. Added $mailstore_basename variable.
6170 63. Installed patch to sieve.c from Michael Haardt.
6172 64. When Exim failed to open the panic log after failing to open the main log,
6173 the original message it was trying to log was written to stderr and debug
6174 output, but if they were not available (the usual case in production), it
6175 was lost. Now it is written to syslog before the two lines that record the
6176 failures to open the logs.
6178 65. Users' Exim filters run in subprocesses under the user's uid. It is
6179 possible for a "deliver" command or an alias in a "personal" command to
6180 provoke an address rewrite. If logging of address rewriting is configured,
6181 this fails because the process is not running as root or exim. There may be
6182 a better way of dealing with this, but for the moment (because 4.30 needs
6183 to be released), I have disabled address rewrite logging when running a
6184 filter in a non-root, non-exim process.
6190 1. The buildconfig auxiliary program wasn't quoting the value set for
6191 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6192 not defined. This bug was masked in 4.22 by the effect that was fixed in
6195 2. Some messages that were rejected after a message id was allocated were
6196 shown as "incomplete" by exigrep. It no longer does this for messages that
6197 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6199 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6200 have allow_domain_literals set, the ID did not get logged in the <= line.
6201 Domain literals are now always recognized in Message-ID: header lines.
6203 4. The first argument for a ${extract expansion item is the key name or field
6204 number. Leading and trailing spaces in this item were not being ignored,
6205 causing some misleading effects.
6207 5. When deliver_drop_privilege was set, single queue runner processes started
6208 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6209 same command in the process it spins off) were not dropping privilege.
6211 6. When the daemon running as "exim" started a queue runner, it always
6212 re-executed Exim in the spun-off process. This is a waste of effort when
6213 deliver_drop_privilege is set. The new process now just calls the
6214 queue-runner function directly.
6220 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6223 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6224 ignored. Though the use of -oP was forcing the writing of a pid file, it
6225 was always written to the default place.
6227 3. If the message "no IP address found for host xxxx" is generated during
6228 incoming verification, it is now followed by identification of the incoming
6229 connection (so you can more easily find what provoked it).
6231 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6233 5. Added some features to "harden" Exim a bit more against certain attacks:
6235 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6236 be put in Local/Makefile. This is like the never_users runtime option,
6237 but it cannot be overridden. The default setting is "root".
6239 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6240 prefix string with which any file named in a -C command line option
6243 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6244 is retained for -C and -D only if the caller of Exim is root. Without
6245 it, the exim user may also use -C and -D and retain privilege.
6247 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6248 command line option is disabled.
6250 6. Macro names set by the -D option must start with an upper case letter, just
6251 like macro names defined in the configuration file.
6253 7. Added "dereference=" facility to LDAP.
6255 8. Two instances of the typo "uknown" in the source files are fixed.
6257 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6258 the Configure-Makefile script screwed up while processing it.
6260 10. Incorporated PCRE 4.4.
6262 11. The SMTP synchronization check was not operating right at the start of an
6263 SMTP session. For example, it could not catch a HELO sent before the client
6264 waited for the greeting. There is now a check for outstanding input at the
6265 point when the greeting is written. Because of the duplex, asynchronous
6266 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6267 way, but not yet received, when the check is performed.
6269 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6270 on TCP/IP sockets, because this apparently causes some broken clients to
6273 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6274 unchanged) from the Cygwin maintainer.
6276 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6277 was supported instead of testing for mbx. Effectively a typo.
6279 15. The spa authenticator server code was not checking that the input it
6280 received was valid base64.
6282 16. The debug output line for the "set" modifier in ACLs was not showing the
6283 name of the variable that was being set.
6285 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6287 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6288 lookup on its own. Something else was needed to trigger the lookup. For
6289 example, a match in host_lookup or the need for a host name in a host list.
6290 Now, if $sender_host_name is referenced and the host name has not yet been
6291 looked up, a lookup is performed. If the lookup fails, the variable remains
6292 empty, and $host_lookup_failed is set to "1".
6294 19. Added "eqi" as a case-independent comparison operator.
6296 20. The saslauthd authentication condition could segfault if neither service
6297 nor realm was specified.
6299 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6300 error message that was logged was misleading, and incoming SMTP
6301 connections were dropped. The message is now more accurate, and temporary
6302 errors are given to SMTP connections.
6304 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6305 (except RSET) with a 421 error, until QUIT is received. However, it was
6306 failing to send a response to QUIT.
6308 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6309 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6310 ACL is now run after the helo_try_verify_hosts code.
6312 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6313 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6314 case-independent, so other case variants are also recognized). Apparently
6315 some systems use these upper case variants.
6317 25. If more than two messages were waiting for the same host, and a transport
6318 filter was specified for the transport, Exim sent two messages over the
6319 same TCP/IP connection, and then failed with "socket operation on non-
6320 socket" when it tried to send the third.
6322 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6324 27. The extern definition of crypt16() in expand.c was not being excluded when
6325 the OS had its own crypt16() function.
6327 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6328 as a preferred synonym for return_size_limit, both as an option and as an
6331 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6333 30. Changed the default configuration ACL to relax the local part checking rule
6334 for addresses that are not in any local domains. For these addresses,
6335 slashes and pipe symbols are allowed within local parts, but the sequence
6336 /../ is explicitly forbidden.
6338 31. SPA server authentication was not clearing the challenge buffer before
6341 32. log_message in a "warn" ACL statement was writing to the reject log as
6342 well as to the main log, which contradicts the documentation and doesn't
6343 seem right (because no rejection is happening). So I have stopped it.
6345 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6346 However, I am unable to do any testing of this.
6348 34. Fixed an infelicity in the appendfile transport. When checking directories
6349 for a mailbox, to see if any needed to be created, it was accidentally
6350 using path names with one or more superfluous leading slashes; tracing
6351 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6353 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6354 contents are added to the log line that is written for every discarded
6355 recipient. (Previously a log_message setting was ignored.)
6357 36. The ${quote: operator now quotes the string if it is empty.
6359 37. The install script runs exim in order to find its version number. If for
6360 some reason other than non-existence or emptiness, which it checks, it
6361 could not run './exim', it was installing it with an empty version number,
6362 i.e. as "exim-". This error state is now caught, and the installation is
6365 38. An argument was missing from the function that creates an error message
6366 when Exim fails to connect to the socket for saslauthd authentication.
6367 This could cause Exim to crash, or give a corrupted message.
6369 39. Added isip, isip4, and isip6 to ${if conditions.
6371 40. The ACL variables $acl_xx are now saved with the message, and can be
6372 accessed later in routers, transports, and filters.
6374 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6375 strings in the file are not string-expanded.
6377 42. If a MAIL command specified a SIZE value that was too large to fit into an
6378 int variable, the check against message_size_limit failed. Such values are
6379 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6380 day this will have to be increased, but I don't think I want to be around
6381 when emails are that large.
6388 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6389 iconv() is not standard in FreeBSD.
6391 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6392 IPv6 enabled. The observed symptom was a segmentation fault on return from
6393 the function os_common_find_running_interfaces() in src/os.c.
6395 3. In the check_special_case() function in daemon.c I had used "errno" as an
6396 argument name, which causes warnings on some systems. This was basically a
6397 typo, since it was named "eno" in the comments!
6399 4. The code that waits for the clock to tick (at a resolution of some fraction
6400 of a second) so as to ensure message-id uniqueness was always waiting for
6401 at least one whole tick, when it could have waited for less. [This is
6402 almost certainly not relevant at current processor speeds, where it is
6403 unlikely to ever wait at all. But we try to future-proof.]
6405 5. The function that sleeps for a time interval that includes fractions of a
6406 second contained a race. It did not block SIGALRM between setting the
6407 timer, and suspending (a couple of lines later). If the interval was short
6408 and the sigsuspend() was delayed until after it had expired, the suspension
6409 never ended. On busy systems this could lead to processes getting stuck for
6412 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6413 process, before it forks any delivery processes. The open lookup caching
6414 mechanism meant that the open file or database connection was passed into
6415 the delivery process. The problem was that delivery processes always tidy
6416 up cached lookup data. This could cause a problem for the next delivery
6417 process started by the queue runner, because the external queue runner
6418 process does not know about the closure. So the next delivery process
6419 still has data in the lookup cache. In the case of a file lookup, there was
6420 no problem because closing a file descriptor in a subprocess doesn't affect
6421 the parent. However, if the lookup was caching a connection to a database,
6422 the connection was closed, and the second delivery process was likely to
6423 see errors such as "PGSQL: query failed: server closed the connection
6424 unexpectedly". The problem has been fixed by closing all cached lookups
6425 in a queue runner before running a delivery process.
6427 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6428 seem to have the "const" qualifier which it has on other OS. I've
6431 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6432 *of the same type* (client or server) with the same public name that an
6433 error should be diagnosed.
6435 9. When Exim looked up a host name for an IP address, but failed to find the
6436 original IP address when looking up the host name (a safety check), it
6437 output the message "<ip address> does not match any IP for NULL", which was
6438 confusing, to say the least. The bug was that the host name should have
6439 appeared instead of "NULL".
6441 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6442 user that is built into the binary, and the -C or -D options is used, root
6443 privilege is dropped before the configuration file is read. In addition,
6444 logging is switched to stderr instead of the normal log files. If the
6445 configuration then re-defines the Exim user, the unprivileged environment
6446 is probably not what is expected, so Exim logs a panic warning message (but
6449 However, if deliver_drop_privilege is set, the unprivileged state may well
6450 be exactly what is intended, so the warning has been cut out in that case,
6451 and Exim is allowed to try to write to its normal log files.
6457 1. smtp_return_error_details was not giving details for temporary sender
6458 or receiver verification errors.
6460 2. Diagnose a configuration error if two authenticators have the same public
6463 3. Exim used not to create the message log file for a message until the first
6464 delivery attempt. This could be confusing when incoming messages were held
6465 for policy or load reasons. The message log file is now created at the time
6466 the message is received, and an initial "Received" line is written to it.
6468 4. The automatically generated man page for command line options had a minor
6469 bug that caused no ill effects; however, a more serious problem was that
6470 the procedure for building the man page automatically didn't always
6471 operate. Consequently, release 4.20 contains an out-of-date version. This
6472 shouldn't happen again.
6474 5. When building Exim with embedded Perl support, the script that builds the
6475 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6476 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6478 6. The freeze_tell option was not being used for messages that were frozen on
6479 arrival, either by an ACL or by local_scan().
6481 7. Added the smtp_incomplete_transaction log selector.
6483 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6484 was accepting AUTH without a new EHLO.
6486 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6487 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6490 10. Logging of TCP/IP connections (when configured) now happens in the main
6491 daemon process instead of the child process, so that the TCP/IP connection
6492 count is more accurate (but it can never be perfect).
6494 11. The use of "drop" in a nested ACL was not being handled correctly in the
6495 outer ACL. Now, if condition failure induced by the nested "drop" causes
6496 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6497 or "require"), the connection is dropped.
6499 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6500 that yield "discard" can now be used with an "accept" or a "discard" verb,
6501 but an error is generated for any others (because I can't see a useful way
6502 to define what should happen).
6504 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6505 are now processed in the same way as lines in the Exim configuration file.
6506 In particular, continuation lines are supported.
6508 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6510 15. Added -ti meaning -t -i.
6512 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6513 domains, and warn by logging if others are found.
6515 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6516 returned from GIFCONF when Exim is trying to find the list of interfaces on
6517 a host. The code in os.c has been modified to copy each ifreq to an aligned
6518 structure in all cases.
6520 Also, in some cases, the returned ifreq's were being copied to a 'struct
6521 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6522 means the last couple of bytes of an IPv6 address could be chopped if the
6523 ifreq contained only a normal sockaddr (14 bytes storage).
6525 18. Named domain lists were not supported in the hosts_treat_as_local option.
6526 An entry such as +xxxx was not recognized, and was treated as a literal
6529 19. Ensure that header lines added by a DATA ACL are included in the reject log
6530 if the ACL subsequently rejects the message.
6532 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6533 MD5 (which is deprecated).
6535 21. When testing a filter file using -bf, Exim was writing a message when it
6536 took the sender from a "From " line in the message, but it was not doing so
6537 when it took $return_path from a Return-Path: header line. It now does.
6539 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6540 with a valid header line field name (a series of printing characters
6541 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6543 23. Changed "disc" in the source to "disk" to conform to the documentation and
6544 the book and for uniformity.
6546 24. Ignore Sendmail's -Ooption=value command line item.
6548 25. When execve() failed while trying to run a command in a pipe transport,
6549 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6550 could be confused with a return value of 69 from the command itself. This
6551 has been changed to 127, the value the shell returns if it is asked to run
6552 a non-existent command. The wording for the related log line suggests a
6553 non-existent command as the problem.
6555 26. If received_header_text expands to an empty string, do not add a Received:
6556 header line to the message. (Well, it adds a token one on the spool, but
6557 marks it "old" so that it doesn't get used or transmitted.)
6559 27. Installed eximstats 1.28 (addition of -nt option).
6561 28. There was no check for failure on the call to getsockname() in the daemon
6562 code. This can fail if there is a shortage of resources on the system, with
6563 ENOMEM, for example. A temporary error is now given on failure.
6565 29. Contrary to the C standard, it seems that in some environments, the
6566 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6567 program. Exim now does this explicitly; it affects the formatting of
6568 timestamps using strftime().
6570 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6571 complaints. I've now initialized all the variables, to avoid this.
6573 32. Header lines added by a system filter were not being "seen" during
6574 transport-time rewrites.
6576 33. The info_callback() function passed to OpenSSL is set up with type void
6577 (*)(SSL *, int, int), as described somewhere. However, when calling the
6578 function (actually a macro) that sets it up, the type void(*)() is
6579 expected. I've put in a cast to prevent warnings from picky compilers.
6581 34. If a DNS black list lookup found a CNAME record, but there were no A
6582 records associated with the domain it pointed at, Exim crashed.
6584 35. If a DNS black list lookup returned more than one A record, Exim ignored
6585 all but the first. It now scans all returned addresses if a particular IP
6586 value is being sought. In this situation, the contents of the
6587 $dnslist_value variable are a list of all the addresses, separated by a
6590 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6591 to accept a host name and all its aliases if the forward lookup for any of
6592 them yielded the IP address of the incoming connection. Now it accepts only
6593 those names whose forward lookup yields the correct IP address. Any other
6594 names are discarded. This closes a loophole whereby a rogue DNS
6595 administrator could create reverse DNS records to break through a
6596 wildcarded host restriction in an ACL.
6598 37. If a user filter or a system filter that ran in a subprocess used any of
6599 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6600 the wrong values were passed to the pipe command ($thisaddress had the
6601 value of $0, $0 had the value of $1, etc). This bug was introduced by
6602 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6604 38. Improved the line breaking for long SMTP error messages from ACLs.
6605 Previously, if there was no break point between 40 and 75 characters, Exim
6606 left the rest of the message alone. Two changes have been made: (a) I've
6607 reduced the minimum length to 35 characters; (b) if it can't find a break
6608 point between 35 and 75 characters, it looks ahead and uses the first one
6609 that it finds. This may give the occasional overlong line, but at least the
6610 remaining text gets split now.
6612 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6613 file descriptors might be low, and that setting 1000 would always raise it.
6614 It turns out that in some environments, the limit is already over 1000 and
6615 that lowering it causes trouble. So now Exim takes care not to decrease it.
6617 40. When delivering a message, the value of $return_path is set to $sender_
6618 address at the start of routing (routers may change the value). By an
6619 oversight, this default was not being set up when an address was tested by
6620 -bt or -bv, which affected the outcome if any router or filter referred to
6623 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6624 log only when "message" or "log_message" are set. However, if one of the
6625 conditions was an address verification, or a call to a nested ACL, the
6626 messages generated by the underlying test were being passed through. This
6627 no longer happens. The underlying message is available in $acl_verify_
6628 message for both "message" and "log_message" expansions, so it can be
6629 passed through if needed.
6631 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6632 new expansion $bh_ to give the encoded byte string without charset
6633 translation. Translation happens only if iconv() is available; HAVE_ICONV
6634 indicates this at build time. HEADERS_CHARSET gives the charset to
6635 translate to; headers_charset can change it in the configuration, and
6636 "headers charset" can change it in an individual filter file.
6638 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6639 that creates RFC 2047 encoded "words" labels them as that charset instead
6640 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6641 expansion operator; (ii) when Exim creates a From: line for a local
6642 message; (iii) when a header line is rewritten to include a "phrase" part.
6644 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6645 buggy, causing it to skip the first lines of messages whose message ID
6646 ended in 'D'. This would not have bitten before Exim release 4.14, because
6647 message IDs were unlikely to end in 'D' before then. The effect was to have
6648 incorrect size information for certain domains.
6650 45. #include "config.h" was missing at the start of the crypt16.c module. This
6651 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6654 46. If there was a timeout during a "random" callout check, Exim treated it as
6655 a failure of the random address, and carried on sending RSET and the real
6656 address. If the delay was just some slowness somewhere, the response to the
6657 original RCPT would be taken as a response to RSET and so on, causing
6658 mayhem of various kinds.
6660 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6661 when I implemented it. It didn't allow for the fact that some option values
6662 may legitimately be negative (e.g. size_addition), and it didn't even do
6663 the right test for positive values.
6665 48. Domain names in DNS records are case-independent. Exim always looks them up
6666 in lower case. Some resolvers return domain names in exactly the case they
6667 appear in the zone file, that is, they may contain uppercase letters. Not
6668 all resolvers do this - some return always lower case. Exim was treating a
6669 change of case by a resolver as a change of domain, similar to a widening
6670 of a domain abbreviation. This triggered its re-routing code and so it was
6671 trying to route what was effectively the same domain again. This normally
6672 caused routing to fail (because the router wouldn't handle the domain
6673 twice). Now Exim checks for this case specially, and just changes the
6674 casing of the domain that it ultimately uses when it transmits the message
6677 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6680 50. If a filter generated a file delivery with a non-absolute name (possible if
6681 no home directory exists for the router), the forbid_file option was not
6684 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6685 the existing equality matching.
6687 52. Exim was using ints instead of ino_t variables in some places where it was
6688 dealing with inode numbers.
6690 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6691 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6692 called TMPDIR, and if it finds it is different, it changes its value.
6694 54. The smtp_printf() function is now made available to local_scan() so
6695 additional output lines can be written before returning. There is also an
6696 smtp_fflush() function to enable the detection of a dropped connection.
6697 The variables smtp_input and smtp_batched_input are exported to
6700 55. Changed the default runtime configuration: the message "Unknown user"
6701 has been removed from the ACL, and instead placed on the localuser router,
6702 using the cannot_route_message feature. This means that any verification
6703 failures that generate their own messages won't get overridden. Similarly,
6704 the "Unrouteable address" message that was in the ACL for unverifiable
6705 relay addresses has also been removed.
6707 56. Added hosts_avoid_esmtp to the smtp transport.
6709 57. The exicyclog script was not checking for the esoteric option
6710 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6711 will work only if exicyclog is run under the appropriate euid.
6713 58. Following a discussion on the list, the rules by which Exim recognises line
6714 endings on incoming messages have been changed. The -dropcr and drop_cr
6715 options are now no-ops, retained only for backwards compatibility. The
6716 following line terminators are recognized: LF CRLF CR. However, special
6717 processing applies to CR:
6719 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6720 nor a local message in the state where . is a terminator.
6722 (ii) If a bare CR is encountered in a header line, an extra space is added
6723 after the line terminator so as not to end the header. The reasoning
6724 behind this is that bare CRs in header lines are most likely either
6725 to be mistakes, or people trying to play silly games.
6727 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6728 was being incorrectly given as 18 bytes larger than it should have been.
6729 This is a VOB (very old bug).
6731 60. This may never have affected anything current, but just in case it has:
6732 When the local host is found other than at the start of a list of hosts,
6733 the local host, those with the same MX, and any that follow, are discarded.
6734 When the list in question was part of a longer list of hosts, the following
6735 hosts (not currently being processed) were also being discarded. This no
6736 longer happens. I'm not sure if this situation could ever has previously
6739 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6742 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6743 header. This is something that is recommended in a new Internet Draft, and
6744 is something that is documented as being done by Sendmail. There are two
6745 possible values. For messages generated by the autoreply transport, Exim
6748 Auto-Submitted: auto-replied
6750 whereas for all other generated messages (e.g. bounces) it adds
6752 Auto-Submitted: auto-generated
6754 63. The "personal" condition in filters now includes a test for the
6755 Auto-Submitted: header. If it contains the string "auto-" the message it
6756 not considered personal.
6758 64. Added rcpt_include_affixes as a generic transport option.
6760 65. Added queue_only_override (default true).
6762 66. Added the syslog_duplication option.
6764 67. If what should have been the first header line of a message consisted of
6765 a space followed by a colon, Exim was mis-interpreting it as a header line.
6766 It isn't of course - it is syntactically invalid and should therefore be
6767 treated as the start of the message body. The misbehaviour could have
6768 caused a number of strange effects, including loss of data in subsequent
6769 header lines, and spool format errors.
6771 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6772 client host had authenticated. This control can now be exercised by an ACL
6773 for more flexibility.
6775 69. By default, callouts do not happen when testing with -bh. There is now a
6776 variant, -bhc, which does actually run the callout code, including
6777 consulting and updating the callout cache.
6779 70. Added support for saslauthd authentication, courtesy of Alexander
6782 71. If statvfs() failed on the spool or log directories while checking their
6783 size for availability, Exim confusingly gave the error "space shortage".
6784 Furthermore, in debugging mode it crashed with a floating point exception.
6785 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6786 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6787 serious problem, Exim now writes to the main and panic logs when this
6788 happens, with details of the failure. It then refuses to accept the
6789 incoming message, giving the message "spool directory problem" or "log
6790 directory problem" with a 421 code for SMTP messages.
6792 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6793 0, 1, and 2 exist, because some OS complain for execs without them (see
6794 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6795 descriptors. However, the code omitted to check that the open succeeded,
6796 causing mysterious errors if for some reason the permissions on /dev/null
6797 got screwed. Now Exim writes a message to the main and panic logs, and
6798 bombs out if it can't open /dev/null.
6800 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6801 interact so that it is all more flexible. It is supposed to remain
6802 backwards compatible. Also added extra_local_interfaces.
6804 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6805 to bomb out with an assertion failure - to the client this appears as a
6806 connection drop. This problem occurs in the part of the code that was taken
6807 from the Samba project. Fortunately, the assertion is in a very simple
6808 function, so I have fixed this by reproducing the function inline in the
6809 one place where it is called, and arranging for authentication to fail
6810 instead of killing the process with assert().
6812 75. The SPA client code was not working when the server requested OEM rather
6813 than Unicode encoding.
6815 76. Added code to make require_files with a specific uid setting more usable in
6816 the case where statting the file as root fails - usually a non-root-mounted
6817 NFS file system. When this happens and the failure is EACCES, Exim now
6818 forks a subprocess and does the per-uid checking as the relevant uid.
6820 77. Added process_log_path.
6822 78. If log_file_path was not explicitly set, a setting of check_log_space or
6823 check_log_inodes was ignored.
6825 79. If a space check for the spool or log partitions fails, the incident is now
6826 logged. Of course, in the latter case the data may get lost...
6828 80. Added the %p formatting code to string_format() so that it can be used to
6829 print addresses in debug_print(). Adjusted all the address printing in the
6830 debugging in store.c to use %p rather than %d.
6832 81. There was a concern that a line of code in smtp_in.c could overflow a
6833 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6834 initially expressed, the concern was not well-founded, because trailing
6835 spaces are removed early. However, if the trailing spaces were followed by
6836 a NULL, they did not get removed, so the overflow was possible. Two fixes
6839 (a) I re-wrote the offending code in a cleaner fashion.
6840 (b) If an incoming SMTP command contains a NULL character, it is rejected
6843 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6844 runs initgroups(), so that if the Exim user is in any additional groups,
6845 they will be used during message reception.
6851 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob