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 buferring.
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
93 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
94 than -2003 did; needs libidn2 in addition to libidn.
96 JH/02 The path option on a pipe transport is now expanded before use.
98 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
99 Patch provided by "Björn", documentation fix added too.
101 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
102 missing a wire-to-host endian conversion.
104 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
105 close after a BDAT command line could be taken as a following command,
106 giving a synch failure. Fix by only checking for synch immediately
107 before acknowledging the chunk.
109 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
110 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
111 macro. Patches provided by Josh Soref.
113 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
114 Previously we did not; the RFC seems ambiguous and VRFY is not listed
115 by IANA as a service extension. However, John Klensin suggests that we
118 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
119 the dkim code may be unix-mode line endings rather than smtp wire-format
120 CRLF, so prepend a CR to any bare LF.
122 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
123 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
125 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
126 the first were themselves being wrongly included in the feed into dkim
127 processing; with most chunk sizes in use this resulted in an incorrect
128 body hash calculated value.
130 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
131 DKIM signature block, for verification. Although advised against by
132 standards it is specifically not ruled illegal.
134 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
136 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
137 missing a body hash (the bh= tag).
139 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
140 It seems that HAProxy sends the Proxy Protocol information in clear and
141 only then does a TLS startup, so do the same.
143 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
144 TCP connections (such as for Spamd) unless the daemon successfully set
145 Fast Open mode on its listening sockets. This fixes breakage seen on
146 too-old kernels or those not configured for Fast Open, at the cost of
147 requiring both directions being enabled for TFO, and TFO never being used
148 by non-daemon-related Exim processes.
150 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
151 endings, at least on the first header line. Try to canonify any that get
152 past that check, despite the cost.
154 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
155 now limited to an arbitrary five deep, while parsing addresses with the
156 strip_excess_angle_brackets option enabled.
158 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
159 instead leave the unprompted TLS handshake in socket buffer for the
160 TLS library to consume.
162 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
164 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
166 JH/16 Drop variables when they go out of scope. Memory management drops a whole
167 region in one operation, for speed, and this leaves assigned pointers
168 dangling. Add checks run only under the testsuite which checks all
169 variables at a store-reset and panics on a dangling pointer; add code
170 explicitly nulling out all the variables discovered. Fixes one known
171 bug: a transport crash, where a dangling pointer for $sending_ip_address
172 originally assigned in a verify callout, is re-used.
174 PP/06 Drop '.' from @INC in various Perl scripts.
176 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
178 PP/08 Reduce a number of compilation warnings under clang; building with
179 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
180 should be warning-free.
182 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
184 HS/01 Fix portability problems introduced by PP/08 for platforms where
185 realloc(NULL) is not equivalent to malloc() [SunOS et al].
187 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
188 chunk. This allows us to accept broken chunked messages. We need a more
189 general solution here.
191 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
192 already-broken messages in the queue.
194 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
196 JH/19 Fix reference counting bug in routing-generated-address tracking.
202 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
203 supports it and a size is available (ie. the sending peer gave us one).
205 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
206 years of being deprecated). The replacements are the ACLs
207 acl_smtp_mime and acl_not_smtp_mime.
209 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
210 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
211 or even an in-clear connection were permitted. Now, if the host lookup
212 was dnssec and dane was requested then the host is only used if the
213 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
214 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
215 if one fails this test.
216 This means that a poorly-configured remote DNS will make it incommunicado;
217 but it protects against a DNS-interception attack on it.
219 JH/04 Bug 1810: make continued-use of an open smtp transport connection
220 non-noisy when a race steals the message being considered.
222 JH/05 If main configuration option tls_certificate is unset, generate a
223 self-signed certificate for inbound TLS connections.
225 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
226 in rewrites and routers.
228 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
229 and logged a warning sing 4.83; now they are a configuration file error.
231 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
232 (lacking @domain). Apply the same qualification processing as RCPT.
234 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
236 JH/10 Support ${sha256:} applied to a string (as well as the previous
239 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
240 a cutthrough deliver is pending, as we always want to make a connection.
241 This also avoids re-routing the message when later placing the cutthrough
242 connection after a verify cache hit.
243 Do not update it with the verify result either.
245 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
246 when routing results in more than one destination address.
248 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
249 signing (which inhibits the cutthrough capability). Previously only
250 the presence of an option was tested; now an expansion evaluating as
251 empty is permissible (obviously it should depend only on data available
252 when the cutthrough connection is made).
254 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
255 the relevant preceding SMTP command did not note the pipelining mode.
257 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
258 Previously they were not counted.
260 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
261 as one having no matching records. Previously we deferred the message
262 that needed the lookup.
264 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
265 distinguished as "(=".
267 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
268 for missing MX records. Previously it only worked for missing A records.
270 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
272 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
273 after the data-go-ahead and data-ack. Patch from Jason Betts.
275 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
276 even for a "none" policy. Patch from Tony Meyer.
278 JH/22 Fix continued use of a connection for further deliveries. If a port was
279 specified by a router, it must also match for the delivery to be
282 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
283 When one of the recipients of a message was unsuitable for the connection
284 (has no matching addresses), we lost track of needing to mark it
285 deferred. As a result mail would be lost.
287 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
289 JH/25 Decoding ACL controls is now done using a binary search; the source code
290 takes up less space and should be simpler to maintain. Merge the ACL
291 condition decode tables also, with similar effect.
293 JH/26 Fix problem with one_time used on a redirect router which returned the
294 parent address unchanged. A retry would see the parent address marked as
295 delivered, so not attempt the (identical) child. As a result mail would
298 JH/27 Fix a possible security hole, wherein a process operating with the Exim
299 UID can gain a root shell. Credit to http://www.halfdog.net/ for
300 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
303 JH/28 Enable {spool,log} filesystem space and inode checks as default.
304 Main config options check_{log,spool}_{inodes,space} are now
305 100 inodes, 10MB unless set otherwise in the configuration.
307 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
308 Previously it only applied to the main-section connection policy
311 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
313 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
314 by me. Added RFC7919 DH primes as an alternative.
316 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
317 are not in the system include path.
319 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
320 GnuTLS, when a session startup failed (eg because the client disconnected)
321 Exim did stdio operations after fclose. This was exposed by a recent
322 change which nulled out the file handle after the fclose.
324 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
325 signed directly by the cert-signing cert, rather than an intermediate
326 OCSP-signing cert. This is the model used by LetsEncrypt.
328 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
330 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
331 an incoming connection.
333 HS/02 Bug 1802: Do not half-close the connection after sending a request
336 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
337 fallback to "prime256v1".
339 JH/34 SECURITY: Use proper copy of DATA command in error message.
340 Could leak key material. Remotely exploitable. CVE-2016-9963.
346 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
347 and 3.4.4 - once the server is enabled to respond to an OCSP request
348 it does even when not requested, resulting in a stapling non-aware
349 client dropping the TLS connection.
351 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
352 support variable-length bit vectors. No functional change.
354 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
355 The I= interface field on outgoing lines is now after the H= remote
356 host field, same as incoming lines. There is a separate
357 outgoing_interface log selector which allows you to disable the
360 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
361 If not running log_selector +smtp_connection the mainlog would be held
362 open indefinitely after a "too many connections" event, including to a
363 deleted file after a log rotate. Leave the per net connection logging
364 leaving it open for efficiency as that will be quickly detected by the
365 check on the next write.
367 HS/01 Bug 1671: Fix post transport crash.
368 Processing the wait-<transport> messages could crash the delivery
369 process if the message IDs didn't exist for some reason. When
370 using 'split_spool_directory=yes' the construction of the spool
371 file name failed already, exposing the same netto behaviour.
373 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
374 mime_regex ACL conditions.
376 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
377 to DSN fail messages (bounces): remote IP, remote greeting, remote response
378 to HELO, local diagnostic string.
380 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
381 log line to debug. Even when configured with a tls authenticator many
382 client connections are expected to not authenticate in this way, so
383 an authenticate fail is not an error.
385 HS/02 Add the Exim version string to the process info. This way exiwhat
386 gives some more detail about the running daemon.
388 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
389 matter for fast-change records such as DNSBLs.
391 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
392 retry record, even if constant. There may be multiple transports with
393 different interface settings and the retry behaviour needs to be kept
396 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
398 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
400 JH/10 Bug 840: fix log_defer_output option of pipe transport
402 JH/11 Bug 830: use same host for all RCPTS of a message, even under
403 hosts_randomize. This matters a lot when combined with mua_wrapper.
405 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
406 ${quote_pgsql:<string>} operator.
408 JH/13 Bug 1708: avoid misaligned access in cached lookup.
410 JH/14 Change header file name for freeradius-client. Relevant if compiling
411 with Radius support; from the Gentoo tree and checked under Fedora.
413 JH/15 Bug 1712: Introduce $prdr_requested flag variable
415 JH/16 Bug 1714: Permit an empty string as expansion result for transport
416 option transport_filter, meaning no filtering.
418 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
420 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
421 defaults to "*" (all hosts). The variable is now available when not built
422 with TLS, default unset, mainly to enable keeping the testsuite sane.
423 If a server certificate is not supplied (via tls_certificate) an error is
424 logged, and clients will find TLS connections fail on startup. Presumably
425 they will retry in-clear.
426 Packagers of Exim are strongly encouraged to create a server certificate
427 at installation time.
429 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
430 with the $config_file variable.
432 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
433 in transport context, after the attempt, and per-recipient. The latter type
434 is per host attempted. The event data is the error message, and the errno
435 information encodes the lookup type (A vs. MX) used for the (first) host,
436 and the trailing two digits of the smtp 4xx response.
438 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
439 to write to mainlog (or rejectlog, paniclog) in the window between file
440 creation and permissions/ownership being changed. Particularly affects
441 installations where exicyclog is run as root, rather than exim user;
442 result is that the running daemon panics and dies.
444 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
446 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
447 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
448 "pri" and "weight". Note that the previous implicit priority given by the
449 list order is no longer honoured.
451 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
454 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
455 by defining SUPPORT_SOCKS.
457 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
458 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
459 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
460 variables are renamed to proxy_{local,external}_{address,port}.
462 JH/27 Move Internationalisation support from Experimental to mainline, enabled
463 for a build by defining SUPPORT_I18N
465 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
466 of the query string, and make ${quote_redis:} do that quoting.
468 JH/29 Move Events support from Experimental to mainline, enabled by default
469 and removable for a build by defining DISABLE_EVENT.
471 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
473 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
474 cached by the daemon.
476 JH/32 Move Redis support from Experimental to mainline, enabled for a build
477 by defining LOOKUP_REDIS. The libhiredis library is required.
479 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
480 keys are given for lookup.
482 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
483 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
484 only supported when built with TLS support. The PolarSSL SHA routines
485 are still used when the TLS library is too old for convenient support.
487 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
488 openssl_options), for security. OpenSSL forces this from version 1.1.0
489 server-side so match that on older versions.
491 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
492 allocation for $value could be released as the expansion processing
493 concluded, but leaving the global pointer active for it.
495 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
496 and to use the domains and local_parts ACL conditions.
498 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
499 incorrectly not doubled on cutthrough transmission, hence seen as a
500 body-termination at the receiving system - resulting in truncated mails.
501 Commonly the sender saw a TCP-level error, and retransmitted the message
502 via the normal store-and-forward channel. This could result in duplicates
503 received - but deduplicating mailstores were liable to retain only the
504 initial truncated version.
506 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
508 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
510 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
511 we're in there, support oversigning also; bug 1309.
513 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
515 HS/04 Add support for keep_environment and add_environment options.
517 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
518 either intentional arithmetic overflow during PRNG, or testing config-
521 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
522 delivery resulted in actual delivery. Cancel cutthrough before DATA
525 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
526 rejects a recipient: pass the reject to the originator.
528 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
529 Many were false-positives and ignorable, but it's worth fixing the
532 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
533 for the new environment-manipulation done at startup. Move the routines
534 from being local to tls.c to being global via the os.c file.
536 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
537 an extract embedded as result-arg for a map, the first arg for extract
538 is unavailable so we cannot tell if this is a numbered or keyed
539 extraction. Accept either.
545 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
548 JH/02 The smtp transport option "multi_domain" is now expanded.
550 JH/03 The smtp transport now requests PRDR by default, if the server offers
553 JH/04 Certificate name checking on server certificates, when exim is a client,
554 is now done by default. The transport option tls_verify_cert_hostnames
555 can be used to disable this per-host. The build option
556 EXPERIMENTAL_CERTNAMES is withdrawn.
558 JH/05 The value of the tls_verify_certificates smtp transport and main options
559 default to the word "system" to access the system default CA bundle.
560 For GnuTLS, only version 3.0.20 or later.
562 JH/06 Verification of the server certificate for a TLS connection is now tried
563 (but not required) by default. The verification status is now logged by
564 default, for both outbound TLS and client-certificate supplying inbound
567 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
570 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
571 Status Notification (bounce) messages are now MIME format per RFC 3464.
572 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
573 under the control of the dsn_advertise_hosts option, and routers may
574 have a dsn_lasthop option.
576 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
577 default, modifiable by a malware= option. The list separator for
578 the options can now be changed in the usual way. Bug 68.
580 JH/10 The smtp_receive_timeout main option is now expanded before use.
582 JH/11 The incoming_interface log option now also enables logging of the
583 local interface on delivery outgoing connections.
585 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
586 if the interface and destination host and port all match.
588 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
591 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
592 Patch from Andrew Lewis.
594 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
595 now supports optional time-restrictions, weighting, and priority
596 modifiers per server. Patch originally by <rommer@active.by>.
598 JH/16 The spamd_address main option now supports a mixed list of local
599 and remote servers. Remote servers can be IPv6 addresses, and
600 specify a port-range.
602 JH/17 Bug 68: The spamd_address main option now supports an optional
603 timeout value per server.
605 JH/18 Bug 1581: Router and transport options headers_add/remove can
606 now have the list separator specified.
608 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
611 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
614 JH/21 Support for the A6 type of dns record is withdrawn.
616 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
617 rather than the verbs used.
619 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
620 from 255 to 1024 chars.
622 JH/24 Verification callouts now attempt to use TLS by default.
624 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
625 are generic router options now. The defaults didn't change.
627 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
628 Original patch from Alexander Shikoff, worked over by JH.
630 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
631 files if ZCAT_COMMAND is not executable.
633 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
635 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
637 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
638 Normally benign, it bites when the pair was led to by a CNAME;
639 modern usage is to not canonicalize the domain to a CNAME target
640 (and we were inconsistent anyway for A-only vs AAAA+A).
642 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
644 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
645 when evaluating $sender_host_dnssec.
647 JH/31 Check the HELO verification lookup for DNSSEC, adding new
648 $sender_helo_dnssec variable.
650 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
652 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
654 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
656 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
657 documented as working, but never had. Support all but $spam_report.
659 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
660 added for tls authenticator.
662 HS/03 Add perl_taintmode main config option
668 TL/01 When running the test suite, the README says that variables such as
669 no_msglog_check are global and can be placed anywhere in a specific
670 test's script, however it was observed that placement needed to be near
671 the beginning for it to behave that way. Changed the runtest perl
672 script to read through the entire script once to detect and set these
673 variables, reset to the beginning of the script, and then run through
674 the script parsing/test process like normal.
676 TL/02 The BSD's have an arc4random API. One of the functions to induce
677 adding randomness was arc4random_stir(), but it has been removed in
678 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
679 function when detected.
681 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
682 cause callback expansion.
684 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
685 syntax errors in an expansion can be treated as a string instead of
686 logging or causing an error, due to the internal use of bool_lax
687 instead of bool when processing it.
689 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
690 server certificates when making smtp deliveries.
692 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
694 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
696 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
698 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
699 Merged patch from Sebastian Wiedenroth.
701 JH/05 Fix results-pipe from transport process. Several recipients, combined
702 with certificate use, exposed issues where response data items split
703 over buffer boundaries were not parsed properly. This eventually
704 resulted in duplicates being sent. This issue only became common enough
705 to notice due to the introduction of connection certificate information,
706 the item size being so much larger. Found and fixed by Wolfgang Breyha.
708 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
709 size buffer was used, resulting in syntax errors when an expansion
712 JH/07 Add support for directories of certificates when compiled with a GnuTLS
713 version 3.3.6 or later.
715 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
716 is EXPERIMENTAL_EVENT, the main-configuration and transport options
717 both become "event_action", the variables become $event_name, $event_data
718 and $event_defer_errno. There is a new variable $verify_mode, usable in
719 routers, transports and related events. The tls:cert event is now also
720 raised for inbound connections, if the main configuration event_action
723 TL/06 In test suite, disable OCSP for old versions of openssl which contained
724 early OCSP support, but no stapling (appears to be less than 1.0.0).
726 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
727 server certificate names available under the smtp transport option
728 "tls_verify_cert_hostname" now do not permit multi-component wildcard
731 JH/10 Time-related extraction expansions from certificates now use the main
732 option "timezone" setting for output formatting, and are consistent
733 between OpenSSL and GnuTLS compilations. Bug 1541.
735 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
736 encoded parameter in the incoming message. Bug 1558.
738 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
739 include certificate info, eximon was claiming there were spoolfile
742 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
744 JH/14 Log delivery-related information more consistently, using the sequence
745 "H=<name> [<ip>]" wherever possible.
747 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
748 are problematic for Debian distribution, omit them from the release
751 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
753 JH/16 Fix string representation of time values on 64bit time_t architectures.
756 JH/17 Fix a null-indirection in certextract expansions when a nondefault
757 output list separator was used.
762 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
763 checkers that were complaining about end of non-void function with no
766 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
767 This was a regression introduced in 4.83 by another bugfix.
769 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
771 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
772 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
778 TF/01 Correctly close the server side of TLS when forking for delivery.
780 When a message was received over SMTP with TLS, Exim failed to clear up
781 the incoming connection properly after forking off the child process to
782 deliver the message. In some situations the subsequent outgoing
783 delivery connection happened to have the same fd number as the incoming
784 connection previously had. Exim would try to use TLS and fail, logging
785 a "Bad file descriptor" error.
787 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
788 utilities have not been installed.
790 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
791 temporary space as the ACL may create new global variables.
793 TL/01 LDAP support uses per connection or global context settings, depending
794 upon the detected version of the libraries at build time.
796 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
797 to extract and use the src ip:port in logging and expansions as if it
798 were a direct connection from the outside internet. PPv2 support was
799 updated based on HAProxy spec change in May 2014.
801 JH/02 Add ${listextract {number}{list}{success}{fail}}.
803 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
804 Properly escape header and check for NULL return.
806 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
809 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
811 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
812 characters in header names, implemented as a verify condition.
813 Contributed by Michael Fischer v. Mollard.
815 TL/05 Rename SPF condition results err_perm and err_temp to standardized
816 results permerror and temperror. Previous values are deprecated but
817 still accepted. In a future release, err_perm and err_temp will be
818 completely removed, which will be a backward incompatibility if the
819 ACL tests for either of these two old results. Patch contributed by
820 user bes-internal on the mailing list.
822 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
824 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
825 selectors, in both main and reject logs.
827 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
830 JH/07 Add malware type "sock" for talking to simple daemon.
832 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
834 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
835 routers/transports under cutthrough routing.
837 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
838 numbers. Touch up "bool" conditional to keep the same definition.
840 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
842 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
844 JH/12 Expand items in router/transport headers_add or headers_remove lists
845 individually rather than the list as a whole. Bug 1452.
847 Required for reasonable handling of multiple headers_ options when
848 they may be empty; requires that headers_remove items with embedded
849 colons must have them doubled (or the list-separator changed).
851 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
852 view the policy declared in the DMARC record. Currently, $dmarc_status
853 is a combined value of both the record presence and the result of the
856 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
858 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
859 dnslookup router and the smtp transport (applying to the forward
862 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
863 of ldap servers used for a specific lookup. Patch provided by Heiko
866 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
867 New variable $lookup_dnssec_authenticated for observability.
869 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
870 Patch submitted by Lars Timman.
872 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
874 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
875 Requires trusted mode and valid format message id, aborts otherwise.
876 Patch contributed by Heiko Schlichting.
878 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
879 certextract with support for various fields. Bug 1358.
881 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
882 is requested by default, modifiable by smtp transport option
885 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
886 operate on certificate variables to give certificate fingerprints
887 Also new ${sha256:cert_variable}.
889 JH/23 The PRDR feature is moved from being Experimental into the mainline.
891 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
892 Christian Aistleitner.
894 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
896 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
897 file. Patch from Wolfgang Breyha.
899 JH/25 Expand the coverage of the delivery $host and $host_address to
900 client authenticators run in verify callout. Bug 1476.
902 JH/26 Port service names are now accepted for tls_on_connect_ports, to
903 align with daemon_smtp_ports. Bug 72.
905 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
906 support and error reporting did not work properly.
908 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
909 and is readable. Patch from Andrew Colin Kissa.
911 TL/14 Enhance documentation of ${run expansion and how it parses the
912 commandline after expansion, particularly in the case when an
913 unquoted variable expansion results in an empty value.
915 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
917 PP/02 Fix internal collision of T_APL on systems which support RFC3123
918 by renaming away from it. Addresses GH issue 15, reported by
921 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
923 TL/15 SECURITY: prevent double expansion in math comparison functions
924 (can expand unsanitized data). Not remotely exploitable.
931 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
933 PP/02 Make -n do something, by making it not do something.
934 When combined with -bP, the name of an option is not output.
936 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
939 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
940 $sender_host_name and config options to manage this, and basic check
943 PP/05 DSCP support for outbound connections and control modifier for inbound.
945 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
946 (Only plugin which currently uses this is kerberos4, which nobody should
947 be using, but we should make it available and other future plugins might
948 conceivably use it, even though it would break NAT; stuff *should* be
949 using channel bindings instead).
951 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
952 name; added for Sendmail compatibility; requires admin caller.
953 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
954 just ignore it); requires trusted caller.
955 Also parse but ignore: -Ac -Am -X<logfile>
958 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
960 TL/02 Add +smtp_confirmation as a default logging option.
962 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
963 Patch by Magnus Holmgren from 2007-02-20.
965 TL/04 Bugzilla 1281 - Spec typo.
966 Bugzilla 1283 - Spec typo.
967 Bugzilla 1290 - Spec grammar fixes.
969 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
971 TL/06 Add Experimental DMARC support using libopendmarc libraries.
973 TL/07 Fix an out of order global option causing a segfault. Reported to dev
974 mailing list by by Dmitry Isaikin.
976 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
978 JH/02 Support "G" suffix to numbers in ${if comparisons.
980 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
982 NM/01 Bugzilla 1197 - Spec typo
983 Bugzilla 1196 - Spec examples corrections
985 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
987 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
988 gnutls_enable_pkcs11, but renamed to more accurately indicate its
991 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
992 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
994 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
995 "acl {{name}{arg}...}", and optional args on acl condition
998 JH/05 Permit multiple router/transport headers_add/remove lines.
1000 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1002 JH/07 Avoid using a waiting database for a single-message-only transport.
1003 Performance patch from Paul Fisher. Bugzilla 1262.
1005 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1008 JH/09 Add $headers_added variable, with content from use of ACL modifier
1009 add_header (but not yet added to the message). Bugzilla 199.
1011 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1012 Pulled from Bugzilla 817 by Wolfgang Breyha.
1014 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1016 (nb: this is the same fix as in Exim 4.80.1)
1018 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1021 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1024 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1026 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1027 advertise SMTP AUTH mechanism to us, instead of a generic
1028 protocol violation error. Also, make Exim more robust to bad
1029 data from the Dovecot auth socket.
1031 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1033 When a queue runner is handling a message, Exim first routes the
1034 recipient addresses, during which it prunes them based on the retry
1035 hints database. After that it attempts to deliver the message to
1036 any remaining recipients. It then updates the hints database using
1039 So if a recipient address works intermittently, it can get repeatedly
1040 deferred at routing time. The retry hints record remains fresh so the
1041 address never reaches the final cutoff time.
1043 This is a fairly common occurrence when a user is bumping up against
1044 their storage quota. Exim had some logic in its local delivery code
1045 to deal with this. However it did not apply to per-recipient defers
1046 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1048 This change adds a proper retry rule check during routing so that the
1049 final cutoff time is checked against the message's age. We only do
1050 this check if there is an address retry record and there is not a
1051 domain retry record; this implies that previous attempts to handle
1052 the address had the retry_use_local_parts option turned on. We use
1053 this as an approximation for the destination being like a local
1054 delivery, as in LMTP.
1056 I suspect this new check makes the old local delivery cutoff check
1057 redundant, but I have not verified this so I left the code in place.
1059 TF/02 Correct gecos expansion when From: is a prefix of the username.
1061 Test 0254 submits a message to Exim with the header
1065 When I ran the test suite under the user fanf2, Exim expanded
1066 the header to contain my full name, whereas it should have added
1067 a Resent-Sender: header. It erroneously treats any prefix of the
1068 username as equal to the username.
1070 This change corrects that bug.
1072 GF/01 DCC debug and logging tidyup
1073 Error conditions log to paniclog rather than rejectlog.
1074 Debug lines prefixed by "DCC: " to remove any ambiguity.
1076 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1078 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1079 Bug spotted by Jeremy Harris; was flawed since initial commit.
1080 Would have resulted in OCSP responses post-SNI triggering an Exim
1081 NULL dereference and crash.
1083 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1085 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1086 Bug detection, analysis and fix by Samuel Thibault.
1087 Bugzilla 1331, Debian bug #698092.
1089 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1091 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1092 Server implementation by Todd Lyons, client by JH.
1093 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1094 config variable "prdr_enable" controls whether the server
1095 advertises the facility. If the client requests PRDR a new
1096 acl_data_smtp_prdr ACL is called once for each recipient, after
1097 the body content is received and before the acl_smtp_data ACL.
1098 The client is controlled by both of: a hosts_try_prdr option
1099 on the smtp transport, and the server advertisement.
1100 Default client logging of deliveries and rejections involving
1101 PRDR are flagged with the string "PRDR".
1103 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1104 fclose(). Diagnosis by Todd Lyons.
1106 PP/17 Update configure.default to handle IPv6 localhost better.
1107 Patch by Alain Williams (plus minor tweaks).
1110 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1111 This is now consistent with GnuTLS, and is now documented: the
1112 previous undocumented portable approach to treating the option as
1113 unset was to force an expansion failure. That still works, and
1114 an empty string is now equivalent.
1116 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1117 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1118 not performing validation itself.
1120 PP/20 Added force_command boolean option to pipe transport.
1121 Patch from Nick Koston, of cPanel Inc.
1123 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1126 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1128 PP/21 Fix eximon continuous updating with timestamped log-files.
1129 Broken in a format-string cleanup in 4.80, missed when I repaired the
1130 other false fix of the same issue.
1131 Report and fix from Heiko Schlichting.
1134 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1135 Report from Prashanth Katuri.
1137 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1138 It's SecureTransport, so affects any MacOS clients which use the
1139 system-integrated TLS libraries, including email clients.
1141 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1142 using a MIME ACL for non-SMTP local injection.
1143 Report and assistance in diagnosis by Warren Baker.
1145 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1147 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1149 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1150 last id that failed so it may be referenced in subsequent ACL's.
1152 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1155 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1156 ldap library initialization, allowing self-signed CA's to be
1157 used. Also properly sets require_cert option later in code by
1158 using NULL (global ldap config) instead of ldap handle (per
1159 session). Bug diagnosis and testing by alxgomz.
1161 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1162 the src/util/ subdirectory.
1164 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1165 renamed to Transport Post Delivery Action by Jeremy Harris, as
1168 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1169 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1170 redis_servers = needs to be configured which will be used by the redis
1171 lookup. Patch from Warren Baker, of The Packet Hub.
1173 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1175 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1176 hostname or reverse DNS when processing a host list. Used suggestions
1177 from multiple comments on this bug.
1179 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1181 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1182 Missed a few lines, added it to make the runtest require no keyboard
1185 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1186 contains upper case chars. Make router use caseful_local_part.
1188 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1189 support when GnuTLS has been built with p11-kit.
1195 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1197 This, or similar/improved, will also be change PP/11 of 4.82.
1203 PP/01 Handle short writes when writing local log-files.
1204 In practice, only affects FreeBSD (8 onwards).
1205 Bugzilla 1053, with thanks to Dmitry Isaikin.
1207 NM/01 Bugzilla 949 - Documentation tweak
1209 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1212 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1214 PP/02 Implemented gsasl authenticator.
1216 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1218 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1219 `pkg-config foo` for cflags/libs.
1221 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1222 with rest of GSASL and with heimdal_gssapi.
1224 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1225 `pkg-config foo` for cflags/libs for the TLS implementation.
1227 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1228 properties get this fed in as external SSF. A number of robustness
1229 and debugging improvements to the cyrus_sasl authenticator.
1231 PP/08 cyrus_sasl server now expands the server_realm option.
1233 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1234 Patch by Jeremy Harris.
1236 PP/10 Added dbmjz lookup type.
1238 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1240 PP/12 MAIL args handles TAB as well as SP, for better interop with
1241 non-compliant senders.
1242 Analysis and variant patch by Todd Lyons.
1244 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1245 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1246 Patch from Dirk Mueller <dmueller@suse.com>
1248 PP/13 tls_peerdn now print-escaped for spool files.
1249 Observed some $tls_peerdn in wild which contained \n, which resulted
1250 in spool file corruption.
1252 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1253 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1254 or write after TLS renegotiation, which otherwise led to messages
1257 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1258 as a tracking header (ie: a signed header comes before the signature).
1259 Patch from Wolfgang Breyha.
1261 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1262 comma-sep list; embedded commas doubled.
1264 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1266 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1268 Report and patch from Dmitry Banschikov.
1270 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1271 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1272 failures. We appear to now support TLS1.1+ with Exim.
1274 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1275 lets Exim select keys and certificates based upon TLS SNI from client.
1276 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1277 before an outbound SMTP session. New log_selector, +tls_sni.
1279 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1280 NULL dereference. Report and patch from Alun Jones.
1282 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1283 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1284 Not seeing resolver debug output on NetBSD, but suspect this is a
1285 resolver implementation change.
1287 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1288 Left warnings. Added "eximon gdb" invocation mode.
1290 PP/21 Defaulting "accept_8bitmime" to true, not false.
1292 PP/22 Added -bw for inetd wait mode support.
1294 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1295 locate the relevant includes and libraries. Made this the default.
1297 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1298 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1300 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1301 This may cause build issues on older platforms.
1303 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1304 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1305 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1306 Added SNI support via GnuTLS too.
1307 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1309 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1311 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1312 Applied second patch from Janne, implementing suggestion to default
1313 multiple-strings-in-record handling to match SPF spec.
1315 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1317 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1318 read-only, out of scope).
1319 Patch from Wolfgang Breyha, report from Stuart Northfield.
1321 PP/29 Fix three issues highlighted by clang analyser static analysis.
1322 Only crash-plausible issue would require the Cambridge-specific
1323 iplookup router and a misconfiguration.
1324 Report from Marcin Mirosław.
1326 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1328 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1329 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1330 As part of this, removing so much warning spew let me fix some minor
1331 real issues in debug logging.
1333 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1334 assignment on my part. Fixed.
1336 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1337 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1338 Janne Snabb (who went above and beyond: thank you).
1340 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1341 string otherwise requires a connection and a bunch more work and it's
1342 relatively easy to get wrong. Should also expose TLS library linkage
1345 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1346 64-bit ${eval} (JH/03).
1348 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1349 GNU libc to support some of the 64-bit stuff, should not lead to
1350 conflicts. Defined before os.h is pulled in, so if a given platform
1351 needs to override this, it can.
1353 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1354 protection layer was required, which is not implemented.
1355 Bugzilla 1254, patch from Wolfgang Breyha.
1357 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1358 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1359 tls_dhparam take prime identifiers. Also unbreak combination of
1360 OpenSSL+DH_params+TLSSNI.
1362 PP/39 Disable SSLv2 by default in OpenSSL support.
1368 PP/01 Solaris build fix for Oracle's LDAP libraries.
1369 Bugzilla 1109, patch from Stephen Usher.
1371 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1373 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1376 TF/02 Fix a couple more cases where we did not log the error message
1377 when unlink() failed. See also change 4.74-TF/03.
1379 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1380 lock up or crash if it happened to be inside a call to libc when it
1381 got a SIGUSR1 from exiwhat.
1383 The SIGUSR1 handler appends the current process status to the process
1384 log which is later printed by exiwhat. It used to use the general
1385 purpose logging code to do this, but several functions it calls are
1386 not safe for signals.
1388 The new output code in the SIGUSR1 handler is specific to the process
1389 log, and simple enough that it's easy to inspect for signal safety.
1390 Removing some special cases also simplifies the general logging code.
1391 Removing the spurious timestamps from the process log simplifies
1394 TF/04 Improved ratelimit ACL condition.
1396 The /noupdate option has been deprecated in favour of /readonly which
1397 has clearer semantics. The /leaky, /strict, and /readonly update modes
1398 are mutually exclusive. The update mode is no longer included in the
1399 database key; it just determines when the database is updated. (This
1400 means that when you upgrade Exim will forget old rate measurements.)
1402 Exim now checks that the per_* options are used with an update mode that
1403 makes sense for the current ACL. For example, when Exim is processing a
1404 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1405 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1406 must specify per_mail/readonly. If you omit the update mode it defaults to
1407 /leaky where that makes sense (as before) or /readonly where required.
1409 The /noupdate option is now undocumented but still supported for
1410 backwards compatibility. It is equivalent to /readonly except that in
1411 ACLs where /readonly is required you may specify /leaky/noupdate or
1412 /strict/noupdate which are treated the same as /readonly.
1414 A useful new feature is the /count= option. This is a generalization
1415 of the per_byte option, so that you can measure the throughput of other
1416 aggregate values. For example, the per_byte option is now equivalent
1417 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1419 The per_rcpt option has been generalized using the /count= mechanism
1420 (though it's more complicated than the per_byte equivalence). When it is
1421 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1422 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1423 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1424 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1425 Note that using per_rcpt with a non-readonly update mode in more than
1426 one ACL will cause the recipients to be double-counted. (The per_mail
1427 and per_byte options don't have this problem.)
1429 The handling of very low rates has changed slightly. If the computed rate
1430 is less than the event's count (usually one) then this event is the first
1431 after a long gap. In this case the rate is set to the same as this event's
1432 count, so that the first message of a spam run is counted properly.
1434 The major new feature is a mechanism for counting the rate of unique
1435 events. The new per_addr option counts the number of different
1436 recipients that someone has sent messages to in the last time period. It
1437 behaves like per_rcpt if all the recipient addresses are different, but
1438 duplicate recipient addresses do not increase the measured rate. Like
1439 the /count= option this is a general mechanism, so the per_addr option
1440 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1441 example, measure the rate that a client uses different sender addresses
1442 with the options per_mail/unique=$sender_address. There are further
1443 details in the main documentation.
1445 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1447 TF/06 Removed a few PCRE remnants.
1449 TF/07 Automatically extract Exim's version number from tags in the git
1450 repository when doing development or release builds.
1452 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1453 Bugzilla 879. Patch from Paul Fisher.
1455 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1456 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1459 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1461 PP/05 Variable $av_failed, true if the AV scanner deferred.
1462 Bugzilla 1078. Patch from John Horne.
1464 PP/06 Stop make process more reliably on build failure.
1465 Bugzilla 1087. Patch from Heiko Schlittermann.
1467 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1468 Bugzilla 1089. Patch from Heiko Schlittermann.
1470 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1471 Bugzilla 1131. Patch from Holger Weiß.
1473 PP/09 Handle IPv6 addresses with SPF.
1474 Bugzilla 860. Patch from Wolfgang Breyha.
1476 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1478 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1481 PP/11 match_* no longer expand right-hand-side by default.
1482 New compile-time build option, EXPAND_LISTMATCH_RHS.
1483 New expansion conditions, "inlist", "inlisti".
1485 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1487 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1489 PP/14 fix log_write() format string regression from TF/03.
1490 Bugzilla 1152. Patch from Dmitry Isaikin.
1496 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1498 PP/02 Harmonised TLS library version reporting; only show if debugging.
1499 Layout now matches that introduced for other libraries in 4.74 PP/03.
1501 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1503 PP/04 New "dns_use_edns0" global option.
1505 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1508 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1509 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1511 TK/01 Updated PolarSSL code to 0.14.2.
1512 Bugzilla 1097. Patch from Andreas Metzler.
1514 PP/07 Catch divide-by-zero in ${eval:...}.
1515 Fixes bugzilla 1102.
1517 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1520 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1521 format-string attack -- SECURITY: remote arbitrary code execution.
1523 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1524 time unintentionally subject to list matching rules, letting the header
1525 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1526 arbitrary string expansion). This allowed for information disclosure.
1528 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1529 INT_MIN/-1 -- value coerced to INT_MAX.
1535 NM/01 Workaround for PCRE version dependency in version reporting
1538 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1539 This fixes portability to compilers other than gcc, notably
1540 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1542 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1543 makefiles for portability to HP-UX and POSIX correctness.
1545 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1546 Also via indented variable definition in the Makefile.
1547 (Debugging by Oliver Heesakkers).
1549 PP/02 Restore caching of spamd results with expanded spamd_address.
1550 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1552 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1553 Improves build reliability. Fix from: Frank Elsner
1555 NM/02 Fix wide character breakage in the rfc2047 coding
1556 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1558 NM/03 Allow underscore in dnslist lookups
1559 Fixes bug 1026. Patch from Graeme Fowler
1561 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1562 Code patches from Adam Ciarcinski of NetBSD.
1564 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1567 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1568 is logged, to avoid truncation. Patch from John Horne.
1570 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1571 Patch from Jakob Hirsch.
1573 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1574 SQL string expansion failure details.
1575 Patch from Andrey Oktyabrski.
1577 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1578 Patch from Simon Arlott.
1580 PP/09 New lookups functionality failed to compile on old gcc which rejects
1581 extern declarations in function scope.
1582 Patch from Oliver Fleischmann
1584 PP/10 Use sig_atomic_t for flags set from signal handlers.
1585 Check getgroups() return and improve debugging.
1586 Fixed developed for diagnosis in bug 927 (which turned out to be
1589 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1590 Patch from Mark Zealey.
1592 PP/12 Bugzilla 1056: Improved spamd server selection.
1593 Patch from Mark Zealey.
1595 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1596 Based on patch from Heiko Schlittermann.
1598 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1599 Patch from Uwe Doering, sign-off by Michael Haardt.
1601 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1602 variable declaration deep within a block. Bug and patch from
1605 PP/15 lookups-Makefile IRIX compatibility coercion.
1607 PP/16 Make DISABLE_DKIM build knob functional.
1609 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1610 Patch by Simon Arlott
1612 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1613 variable argument macros. Our copy now differs from upstream.
1619 TF/01 Failure to get a lock on a hints database can have serious
1620 consequences so log it to the panic log.
1622 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1623 controlled using the smtp_confirmation log selector.
1625 TF/03 Include the error message when we fail to unlink a spool file.
1627 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1628 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1629 for maintaining out-of-tree patches for some time.
1631 PP/01 Bugzilla 139: Documentation and portability issues.
1632 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1633 Handle per-OS dynamic-module compilation flags.
1635 PP/02 Let /dev/null have normal permissions.
1636 The 4.73 fixes were a little too stringent and complained about the
1637 permissions on /dev/null. Exempt it from some checks.
1638 Reported by Andreas M. Kirchwitz.
1640 PP/03 Report version information for many libraries, including
1641 Exim version information for dynamically loaded libraries. Created
1642 version.h, now support a version extension string for distributors
1643 who patch heavily. Dynamic module ABI change.
1645 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1646 privilege escalation vulnerability whereby the Exim run-time user
1647 can cause root to append content of the attacker's choosing to
1650 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1653 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1654 If dropping privileges for untrusted macros, we disabled normal logging
1655 on the basis that it would fail; for the Exim run-time user, this is not
1656 the case, and it resulted in successful deliveries going unlogged.
1657 Fixed. Reported by Andreas Metzler.
1663 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1664 only prepend for the Resent-* case. Fixes regression introduced in
1665 Exim 4.70 by NM/22 for Bugzilla 607.
1667 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1668 increasing numbers of administrators be bitten by this.
1670 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1672 PP/03 Bugzilla 994: added openssl_options main configuration option.
1674 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1676 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1678 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1680 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1681 without a peer certificate, leading to a segfault because of an
1682 assumption that peers always have certificates. Be a little more
1683 paranoid. Problem reported by Martin Tscholak.
1685 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1686 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1687 NB: ClamAV planning to remove STREAM in "middle of 2010".
1688 CL also introduces -bmalware, various -d+acl logging additions and
1689 more caution in buffer sizes.
1691 PP/09 Implemented reverse_ip expansion operator.
1693 PP/10 Bugzilla 937: provide a "debug" ACL control.
1695 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1697 PP/12 Bugzilla 973: Implement --version.
1699 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1701 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1703 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1705 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1706 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1707 ignore trailing whitespace.
1709 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1711 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1714 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1715 Notification from Dr Andrew Aitchison.
1717 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1718 ExtendedDetectionInfo response format.
1719 Notification from John Horne.
1721 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1724 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1725 XSL and documented dependency on system catalogs, with examples of how
1728 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1731 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1732 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1733 configuration file which is writeable by the Exim user or group.
1735 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1736 of configuration files to cover files specified with the -C option if
1737 they are going to be used with root privileges, not just the default
1740 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1741 option (effectively making it always true).
1743 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1744 files to be used while preserving root privileges.
1746 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1747 that rogue child processes cannot use them.
1749 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1750 run-time user, instead of root.
1752 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1753 Exim run-time user without dropping privileges.
1755 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1756 result string, instead of calling string_vformat() twice with the same
1759 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1760 for other users. Others should always drop root privileges if they use
1761 -C on the command line, even for a whitelisted configure file.
1763 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1765 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1771 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1772 $data_path, and $header_path variables; fixed documentation bugs and
1775 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1776 exipick to access non-standard spools, including the "frozen" queue
1779 NM/01 Bugzilla 965: Support mysql stored procedures.
1780 Patch from Alain Williams
1782 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1784 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1785 Patch from Andreas Metzler
1787 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1788 Patch from Kirill Miazine
1790 NM/05 Bugzilla 671: Added umask to procmail example.
1792 JJ/03 installed exipick 20100323.0, fixing doc bug
1794 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1795 directory. Notification and patch from Dan Rosenberg.
1797 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1799 TK/02 Improve log output when DKIM signing operation fails.
1801 MH/01 Treat the transport option dkim_domain as a colon separated
1802 list, not as a single string, and sign the message with each element,
1803 omitting multiple occurences of the same signer.
1805 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1806 Bugzilla 985, 986. Patch by Simon Arlott
1808 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1809 Patch by Simon Arlott
1811 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1812 MBX locking. Notification from Dan Rosenberg.
1818 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1820 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1822 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1824 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1826 NM/04 Bugzilla 915: Flag broken perl installation during build.
1832 TK/01 Added patch by Johannes Berg that expands the main option
1833 "spamd_address" if it starts with a dollar sign.
1835 TK/02 Write list of recipients to X-Envelope-Sender header when building
1836 the mbox-format spool file for content scanning (suggested by Jakob
1839 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1840 (http://www.dcc-servers.net/) support via dccifd. Activated by
1841 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1843 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1844 by Mark Daniel Reidel <mr@df.eu>.
1846 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1847 When building exim an external PCRE library is now needed -
1848 PCRE is a system library on the majority of modern systems.
1849 See entry on PCRE_LIBS in EDITME file.
1851 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1852 conversation. Added nologin parameter to request.
1853 Patch contributed by Kirill Miazine.
1855 TF/01 Do not log submission mode rewrites if they do not change the address.
1857 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1859 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1860 log files in place. Contributed by Roberto Lima.
1862 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1864 TF/03 Bugzilla 615: When checking the local_parts router precondition
1865 after a local_part_suffix or local_part_prefix option, Exim now
1866 does not use the address's named list lookup cache, since this
1867 contains cached lookups for the whole local part.
1869 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1870 Robert Millan. Documentation is in experimental-spec.txt.
1872 TF/04 Bugzilla 668: Fix parallel build (make -j).
1874 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1876 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1877 Patch provided by Jan Srzednicki.
1879 TF/05 Leading white space used to be stripped from $spam_report which
1880 wrecked the formatting. Now it is preserved.
1882 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1883 that they are available at delivery time.
1885 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1887 TF/08 TLS error reporting now respects the incoming_interface and
1888 incoming_port log selectors.
1890 TF/09 Produce a more useful error message if an SMTP transport's hosts
1891 setting expands to an empty string.
1893 NM/06 Bugzilla 744: EXPN did not work under TLS.
1894 Patch provided by Phil Pennock.
1896 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1897 Patch provided by Richard Godbee.
1899 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1900 acl_smtp_notquit, added index entry.
1902 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1903 Patch provided by Eugene Bujak.
1905 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1906 accept(). Patch provided by Maxim Dounin.
1908 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1909 Patch provided by Phil Pennock.
1911 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1913 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1914 Patch provided by Brad "anomie" Jorsch.
1916 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1917 Patch provided by Dean Brooks.
1919 TK/05 Add native DKIM support (does not depend on external libraries).
1921 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1922 Patch provided by Graeme Fowler.
1924 NM/16 Bugzilla 851: Documentation example syntax fix.
1926 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1928 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1931 NM/19 Bugzilla 745: TLS version reporting.
1932 Patch provided by Phil Pennock.
1934 NM/20 Bugzilla 167: bool: condition support.
1935 Patch provided by Phil Pennock.
1937 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1938 clients. Patch provided by Phil Pennock.
1940 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1941 Patch provided by Brad "anomie" Jorsch.
1943 NM/23 Bugzilla 687: Fix misparses in eximstats.
1944 Patch provided by Heiko Schlittermann.
1946 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1947 Patch provided by Heiko Schlittermann.
1949 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1950 plus update to original patch.
1952 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1954 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1955 Patch provided by David Brownlee.
1957 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1959 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1961 NM/30 Bugzilla 888: TLS documentation bugfixes.
1963 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1965 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1966 Unlike the original bugzilla I have changed all shell scripts in src tree.
1968 NM/33 Bugzilla 898: Transport filter timeout fix.
1969 Patch by Todd Rinaldo.
1971 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1972 Patch by Serge Demonchaux.
1974 NM/35 Bugzilla 39: Base64 decode bug fixes.
1975 Patch by Jakob Hirsch.
1977 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1979 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1981 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1983 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1989 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1990 ALT-N's libdkim that I have put here:
1991 http://duncanthrax.net/exim-experimental/
1993 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1994 were called 'true' and it seems that C99 defines that as a reserved
1995 keyword to be used with 'bool' variable types. That means you could
1996 not include C99-style headers which use bools without triggering
1997 build errors in sieve.c.
1999 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2000 as mailq or other aliases. Changed the --help handling significantly
2001 to do whats expected. exim_usage() emits usage/help information.
2003 SC/01 Added the -bylocaldomain option to eximstats.
2005 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2007 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2009 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2015 PH/01 Another patch from the Sieve maintainer.
2017 PH/02 When an IPv6 address is converted to a string for single-key lookup
2018 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2019 dots are used instead of colons so that keys in lsearch files need not
2020 contain colons. This was done some time before quoting was made available
2021 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2022 (notated using the quote facility) so as to distinguish them from IPv4
2023 keys. This meant that lookups for IP addresses in host lists did not work
2024 for iplsearch lookups.
2026 This has been fixed by arranging for IPv6 addresses to be expressed with
2027 colons if the lookup type is iplsearch. This is not incompatible, because
2028 previously such lookups could never work.
2030 The situation is now rather anomalous, since one *can* have colons in
2031 ordinary lsearch keys. However, making the change in all cases is
2032 incompatible and would probably break a number of configurations.
2034 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2037 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2038 conversion specification without a maximum field width, thereby enabling
2039 a rogue spamd server to cause a buffer overflow. While nobody in their
2040 right mind would setup Exim to query an untrusted spamd server, an
2041 attacker that gains access to a server running spamd could potentially
2042 exploit this vulnerability to run arbitrary code as the Exim user.
2044 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2045 $primary_hostname instead of what libspf2 thinks the hosts name is.
2047 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2048 a directory entry by the name of the lookup key. Previously, if a
2049 symlink pointed to a non-existing file or a file in a directory that
2050 Exim lacked permissions to read, a lookup for a key matching that
2051 symlink would fail. Now it is enough that a matching directory entry
2052 exists, symlink or not. (Bugzilla 503.)
2054 PH/03 The body_linecount and body_zerocount variables are now exported in the
2057 PH/04 Added the $dnslist_matched variable.
2059 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2060 This means they are set thereafter only if the connection becomes
2063 PH/06 Added the client_condition to authenticators so that some can be skipped
2064 by clients under certain conditions.
2066 PH/07 The error message for a badly-placed control=no_multiline_responses left
2067 "_responses" off the end of the name.
2069 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2071 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2072 (without spaces) instead of just copying the configuration text.
2074 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2076 PH/11 Added $max_received_linelength.
2078 PH/12 Added +ignore_defer and +include_defer to host lists.
2080 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2081 way in which PCRE > 7.0 is built.
2083 PH/14 Implemented queue_only_load_latch.
2085 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2086 MAIL command. The effect was to mangle the value on 64-bit systems.
2088 PH/16 Another patch from the Sieve maintainer.
2090 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2092 PH/18 If a system quota error occurred while trying to create the file for
2093 a maildir delivery, the message "Mailbox is full" was not appended to the
2094 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2095 only to a quota excession during the actual writing of the file.
2097 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2098 characters?) which causes problems in log lines. The DN values are now
2099 passed through string_printing() before being added to log lines.
2101 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2102 and InterBase are left for another time.)
2104 PH/21 Added message_body_newlines option.
2106 PH/22 Guard against possible overflow in moan_check_errorcopy().
2108 PH/23 POSIX allows open() to be a macro; guard against that.
2110 PH/24 If the recipient of an error message contained an @ in the local part
2111 (suitably quoted, of course), incorrect values were put in $domain and
2112 $local_part during the evaluation of errors_copy.
2118 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2119 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2122 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2123 issue a MAIL command.
2125 PH/02 In an ACL statement such as
2127 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2129 if a client was not listed at all, or was listed with a value other than
2130 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2131 the condition was not true (as it should be), so access was not denied.
2132 The bug was that the ! inversion was incorrectly passed on to the second
2133 item. This has been fixed.
2135 PH/03 Added additional dnslists conditions == and =& which are different from
2136 = and & when the dns lookup returns more than one IP address.
2138 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2139 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2141 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2142 FSYNC, which compiles an option called disable_fsync that allows for
2143 bypassing fsync(). The documentation is heavily laced with warnings.
2145 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2147 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2148 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2149 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2150 including adding "make clean"; (3) Added -fPIC when compiling the test
2151 dynamically loaded module, to get rid of a warning.
2153 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2154 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2155 = 0s. The bug is otherwise harmless.
2157 PH/07 There was a bug in the dovecot authenticator such that the value of
2158 $auth1 could be overwritten, and so not correctly preserved, after a
2159 successful authentication. This usually meant that the value preserved by
2160 the server_setid option was incorrect.
2162 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2164 PH/09 Installed PCRE release 7.0.
2166 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2167 run for batched SMTP input. It is now run at the start of every message
2168 in the batch. While fixing this I discovered that the process information
2169 (output by running exiwhat) was not always getting set for -bs and -bS
2170 input. This is fixed, and it now also says "batched" for BSMTP.
2172 PH/11 Added control=no_pipelining.
2174 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2175 patch, slightly modified), and move the expansion of helo_data till after
2176 the connection is made in the smtp transport (so it can use these
2179 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2181 PH/14 Added log_selector = +pid.
2183 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2185 PH/16 Add ${if forany and ${if forall.
2187 PH/17 Added dsn_from option to vary the From: line in DSNs.
2189 PH/18 Flush SMTP output before performing a callout, unless control =
2190 no_callout_flush is set.
2192 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2193 was true (the default) a successful delivery failed to delete the retry
2194 item, thus causing premature timeout of the address. The bug is now
2197 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2199 PH/21 Long custom messages for fakedefer and fakereject are now split up
2200 into multiline responses in the same way that messages for "deny" and
2201 other ACL rejections are.
2203 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2204 with slight modification.
2206 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2207 draft, changing the syntax and factoring some duplicate code".
2209 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2210 for deliveries of the second and subsequent messages over the same SMTP
2213 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2214 ${reduce, with only minor "tidies".
2216 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2218 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2219 expansion side effects.
2221 PH/27 When a message times out after an over-quota error from an Exim-imposed
2222 quota, the bounce message says "mailbox is full". This message was not
2223 being given when it was a system quota that was exceeded. It now should
2226 MH/03 Made $recipients available in local_scan(). local_scan() already has
2227 better access to the recipient list through recipients_list[], but
2228 $recipients can be useful in postmaster-provided expansion strings.
2230 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2231 in the case of a MAIL command with additional options following the
2232 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2233 were accidentally chopped off.
2235 PH/29 SMTP synchronization checks are implemented when a command is read -
2236 there is a check that no more input is waiting when there shouldn't be
2237 any. However, for some commands, a delay in an ACL can mean that it is
2238 some time before the response is written. In this time, more input might
2239 arrive, invalidly. So now there are extra checks after an ACL has run for
2240 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2241 pipelining has not been advertised.
2243 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2245 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2246 returned in the SMTP response when smtp_return_error_details was false.
2247 This has been fixed.
2249 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2250 instead of the C I/O that was originally supplied, because problems were
2251 reported on Solaris.
2253 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2254 Exim which did not show up earlier: it was assuming that a call to
2255 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2256 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2257 was a macro that became an assignment, so it seemed to work. This has
2258 changed to a proper function call with a void return, hence the compile
2259 error. Exim's code has been fixed.
2261 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2264 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2266 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2268 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2269 because it turns out to be too big or there is a timeout) there is no
2270 "Completed" line in the log. When some messages of this type were
2271 selected by exigrep, they were listed as "not completed". Others were
2272 picked up by some special patterns. I have improved the selection
2273 criteria to be more general.
2275 PH/38 The host_find_failed option in the manualroute router can now be set
2276 to "ignore", to completely ignore a host whose IP address cannot be
2277 found. If all hosts are ignored, the behaviour is controlled by the new
2278 host_all_ignored option.
2280 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2281 homing or because of multiple MX records with /mx) generated more than
2282 one IP address, and the following item turned out to be the local host,
2283 all the secondary addresses of the first item were incorrectly removed
2284 from the list, along with the local host and any following hosts (which
2285 is what is supposed to happen).
2287 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2288 whoever called Exim into the -H file. In the case of the daemon it was
2289 behaving confusingly. When first started, it used values for whoever
2290 started the daemon, but after a SIGHUP it used the Exim user (because it
2291 calls itself on a restart). I have changed the code so that it now always
2294 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2295 message are rejected with the same error (e.g. no authentication or bad
2296 sender address), and a DATA command is nevertheless sent (as can happen
2297 with PIPELINING or a stupid MUA), the error message that was given to the
2298 RCPT commands is included in the rejection of the DATA command. This is
2299 intended to be helpful for MUAs that show only the final error to their
2302 PH/42 Another patch from the Sieve maintainer.
2304 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2305 Eximstats - Fixed some broken HTML links and added missing column headers
2307 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2308 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2310 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2312 PH/43 Yet another patch from the Sieve maintainer.
2314 PH/44 I found a way to check for a TCP/IP connection going away before sending
2315 the response to the final '.' that terminates a message, but only in the
2316 case where the client has not sent further data following the '.'
2317 (unfortunately, this is allowed). However, in many cases there won't be
2318 any further data because there won't be any more messages to send. A call
2319 to select() can be used: if it shows that the input is "ready", there is
2320 either input waiting, or the socket has been closed. An attempt to read
2321 the next input character can distinguish the two cases. Previously, Exim
2322 would have sent an OK response which the client would never have see.
2323 This could lead to message repetition. This fix should cure that, at
2324 least in a lot of common cases.
2326 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2327 advertised in response to EHLO.
2333 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2334 fixed by 4.65/MH/01 (is this a record?) are fixed:
2336 (i) An empty string was always treated as zero by the numeric comparison
2337 operators. This behaviour has been restored.
2339 (ii) It is documented that the numeric comparison operators always treat
2340 their arguments as decimal numbers. This was broken in that numbers
2341 starting with 0 were being interpreted as octal.
2343 While fixing these problems I realized that there was another issue that
2344 hadn't been noticed. Values of message_size_limit (both the global option
2345 and the transport option) were treated as octal if they started with 0.
2346 The documentation was vague. These values are now always treated as
2347 decimal, and I will make that clear in the documentation.
2353 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2354 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2357 MH/01 Don't check that the operands of numeric comparison operators are
2358 integers when their expansion is in "skipping" mode (fixes bug
2359 introduced by 4.64-PH/07).
2361 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2362 child addresses, Exim now panics and dies. Previously, because the count
2363 is held in a short int, deliveries were likely to be lost. As such a
2364 large number of recipients for a single message is ridiculous
2365 (performance will be very, very poor), I have chosen to impose a limit
2366 rather than extend the field.
2372 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2373 leftover -K file (the existence of which was triggered by #402).
2374 While we were at it, introduced process PID as part of the -K
2375 filename. This should rule out race conditions when creating
2378 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2379 processing considerably. Previous code took too long for large mails,
2380 triggering a timeout which in turn triggers #401.
2382 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2383 in the DK code in transports.c. sendfile() is not really portable,
2384 hence the _LINUX specificness.
2386 TF/01 In the add_headers option to the mail command in an Exim filter,
2387 there was a bug that Exim would claim a syntax error in any
2388 header after the first one which had an odd number of characters
2391 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2392 callout verification, Exim cached a "reject" for the entire domain. This
2393 is correct for most verifications, but it is not correct for a recipient
2394 verification with use_sender or use_postmaster set, because in that case
2395 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2396 case of MAIL FROM:<> rejection from other early rejections (e.g.
2397 rejection of HELO). When verifying a recipient using a non-null MAIL
2398 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2399 Whatever the result of the callout, the value of the domain cache is
2400 left unchanged (for any other kind of callout, getting as far as trying
2401 RCPT means that the domain itself is ok).
2403 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2406 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2407 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2408 session, the daemon ignores ECONNECT errors and logs others; it now
2409 ignores EPIPE as well.
2411 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2412 (quoted-printable decoding).
2414 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2415 later the small subsequent patch to fix an introduced bug.
2417 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2419 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2421 PH/08 An error is now given if message_size_limit is specified negative.
2423 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2424 to be given (somewhat) arbitrary names.
2426 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2429 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2430 miscellaneous code fixes
2432 PH/10 Added the log_reject_target ACL modifier to specify where to log
2435 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2436 hostname. This is wrong, because it relates to the incoming message (and
2437 probably the interface on which it is arriving) and not to the outgoing
2438 callout (which could be using a different interface). This has been
2439 changed to use the value of the helo_data option from the smtp transport
2440 instead - this is what is used when a message is actually being sent. If
2441 there is no remote transport (possible with a router that sets up host
2442 addresses), $smtp_active_hostname is used.
2444 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2445 tweaks were necessary in order to get it to work (see also 21 below):
2446 (a) The code assumed that strncpy() returns a negative number on buffer
2447 overflow, which isn't the case. Replaced with Exim's string_format()
2449 (b) There were several signed/unsigned issues. I just did the minimum
2450 hacking in of casts. There is scope for a larger refactoring.
2451 (c) The code used strcasecmp() which is not a standard C function.
2452 Replaced with Exim's strcmpic() function.
2453 (d) The code set only $1; it now sets $auth1 as well.
2454 (e) A simple test gave the error "authentication client didn't specify
2455 service in request". It would seem that Dovecot has changed its
2456 interface. Fortunately there's a specification; I followed it and
2457 changed what the client sends and it appears to be working now.
2459 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2462 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2463 address A is aliased to B and C, where B exists and C does not. Without
2464 -v the output is "A verified" because verification stops after a
2465 successful redirection if more than one address is generated. However,
2466 with -v the child addresses are also verified. Exim was outputting "A
2467 failed to verify" and then showing the successful verification for C,
2468 with its parentage. It now outputs "B failed to verify", showing B's
2469 parentage before showing the successful verification of C.
2471 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2472 look up a TXT record in a specific list after matching in a combined
2475 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2476 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2477 they consult the DNS. I had assumed they would set it the way they
2478 wanted; and indeed my experiments on Linux seem to show that in some
2479 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2480 To be on the safe side, however, I have now made the interface to
2481 host_find_byname() similar to host_find_bydns(), with an argument
2482 containing the DNS resolver options. The host_find_byname() function now
2483 sets these options at its start, just as host_find_bydns() does. The smtp
2484 transport options dns_qualify_single and dns_search_parents are passed to
2485 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2486 of host_find_byname() use the default settings of RES_DEFNAMES
2487 (qualify_single) but not RES_DNSRCH (search_parents).
2489 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2490 spool_read_header() do less string testing, by means of a preliminary
2491 switch on the second character of optional "-foo" lines. (This is
2492 overdue, caused by the large number of possibilities that now exist.
2493 Originally there were few.) While I was there, I also converted the
2494 str(n)cmp tests so they don't re-test the leading "-" and the first
2495 character, in the hope this might squeeze out yet more improvement.
2497 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2498 flag allowing group syntax was set by the header_syntax check but not
2499 turned off, possible causing trouble later; (2) The flag was not being
2500 set at all for the header_verify test, causing "group"-style headers to
2501 be rejected. I have now set it in this case, and also caused header_
2502 verify to ignore an empty address taken from a group. While doing this, I
2503 came across some other cases where the code for allowing group syntax
2504 while scanning a header line wasn't quite right (mostly, not resetting
2505 the flag correctly in the right place). These bugs could have caused
2506 trouble for malformed header lines. I hope it is now all correct.
2508 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2509 with the "reply" argument non-NULL. The code, however (which originally
2510 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2511 but it didn't always do it. This confused somebody who was copying the
2512 code for some other use. I have removed all the tests.
2514 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2515 feature that was used to support insecure browsers during the U.S. crypto
2516 embargo. It requires special client support, and Exim is probably the
2517 only MTA that supported it -- and would never use it because real RSA is
2518 always available. This code has been removed, because it had the bad
2519 effect of slowing Exim down by computing (never used) parameters for the
2520 RSA_EXPORT functionality.
2522 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2523 authenticator to fail if there's a tab character in the incoming data
2524 (there should never be unless someone is messing about, as it's supposed
2525 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2526 if the connection is using TLS or if the remote IP is the same as the
2527 local IP, and the "valid-client-cert option" if a client certificate has
2530 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2531 authenticators. This can be used for authorization after authentication
2532 succeeds. (In the case of plaintext, it servers for both authentication
2535 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2536 if any retry times were supplied.
2538 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2539 connection, where there is no client IP address to check. In this
2540 situation, the verify now always succeeds.
2542 PH/25 Applied John Jetmore's -Mset patch.
2544 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2546 PH/27 In a string expansion for a processed (not raw) header when multiple
2547 headers of the same name were present, leading whitespace was being
2548 removed from all of them, but trailing whitespace was being removed only
2549 from the last one. Now trailing whitespace is removed from each header
2550 before concatenation. Completely empty headers in a concatenation (as
2551 before) are ignored.
2553 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2554 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2556 PH/29 [Removed. This was a change that I later backed out, and forgot to
2557 correct the ChangeLog entry (that I had efficiently created) before
2558 committing the later change.]
2560 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2561 address errors (4xx response to RCPT) over the same connection as other
2562 messages routed to the same hosts. Such deliveries are always "forced",
2563 so retry times are not inspected. This resulted in far too many retries
2564 for the affected addresses. The effect occurred only when there were more
2565 hosts than the hosts_max_try setting in the smtp transport when it had
2566 the 4xx errors. Those hosts that it had tried were not added to the list
2567 of hosts for which the message was waiting, so if all were tried, there
2568 was no problem. Two fixes have been applied:
2570 (i) If there are any address or message errors in an SMTP delivery, none
2571 of the hosts (tried or untried) are now added to the list of hosts
2572 for which the message is waiting, so the message should not be a
2573 candidate for sending over the same connection that was used for a
2574 successful delivery of some other message. This seems entirely
2575 reasonable: after all the message is NOT "waiting for some host".
2576 This is so "obvious" that I'm not sure why it wasn't done
2577 previously. Hope I haven't missed anything, but it can't do any
2578 harm, as the worst effect is to miss an optimization.
2580 (ii) If, despite (i), such a delivery is accidentally attempted, the
2581 routing retry time is respected, so at least it doesn't keep
2582 hammering the server.
2584 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2585 in ${readsocket because some servers need this prod.
2587 PH/32 Added some extra debug output when updating a wait-xxx database.
2589 PH/33 The hint "could be header name not terminated by colon", which has been
2590 given for certain expansion errors for a long time, was not being given
2591 for the ${if def:h_colon_omitted{... case.
2593 PH/34 The spec says: "With one important exception, whenever a domain list is
2594 being scanned, $domain contains the subject domain." There was at least
2595 one case where this was not true.
2597 PH/35 The error "getsockname() failed: connection reset by peer" was being
2598 written to the panic log as well as the main log, but it isn't really
2599 panic-worthy as it just means the connection died rather early on. I have
2600 removed the panic log writing for the ECONNRESET error when getsockname()
2603 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2604 runs only) independently of the message's sender address. This meant
2605 that, if the 4xx error was in fact related to the sender, a different
2606 message to the same recipient with a different sender could confuse
2607 things. In particular, this can happen when sending to a greylisting
2608 server, but other circumstances could also provoke similar problems.
2609 I have changed the default so that the retry time for these errors is now
2610 based a combination of the sender and recipient addresses. This change
2611 can be overridden by setting address_retry_include_sender=false in the
2614 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2615 remote server are returned as part of bounce messages. This was not
2616 happening for LMTP over a pipe (the lmtp transport), but now it is the
2617 same for both kinds of LMTP.
2619 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2620 in header lines that were in fact CNAMEs. This is no longer the case.
2622 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2623 and queue runs started by the daemon processed all messages. This has
2624 been fixed so that -R and -S can now usefully be given with -q<time>.
2626 PH/40 Import PCRE release 6.7 (fixes some bugs).
2628 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2630 PH/42 Give an error if -q is specified more than once.
2632 PH/43 Renamed the variables $interface_address and $interface_port as
2633 $received_ip_address and $received_port, to make it clear that these
2634 values apply to message reception, and not to the outgoing interface when
2635 a message is delivered. (The old names remain recognized, of course.)
2637 PH/44 There was no timeout on the connect() call when using a Unix domain
2638 socket in the ${readsocket expansion. There now is.
2640 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2641 be meaningful with "accept".
2643 SC/01 Eximstats V1.43
2644 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2646 SC/02 Eximstats V1.44
2647 Use a glob alias rather than an array ref in the generated
2648 parser. This improves both readability and performance.
2650 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2651 Collect SpamAssassin and rejection statistics.
2652 Don't display local sender or destination tables unless
2653 there is data to show.
2654 Added average volumes into the top table text output.
2656 SC/04 Eximstats V1.46
2657 Collect data on the number of addresses (recipients)
2658 as well as the number of messages.
2660 SC/05 Eximstats V1.47
2661 Added 'Message too big' to the list of mail rejection
2662 reasons (thanks to Marco Gaiarin).
2664 SC/06 Eximstats V1.48
2665 Mainlog lines which have GMT offsets and are too short to
2666 have a flag are now skipped.
2668 SC/07 Eximstats V1.49 (Alain Williams)
2669 Added the -emptyok flag.
2671 SC/08 Eximstats V1.50
2672 Fixes for obtaining the IP address from reject messages.
2674 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2675 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2676 whitespace changes from 4.64-PH/27
2678 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2681 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2682 are found, allow negative numbers in numeric criteria)
2684 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2686 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2687 to match changes made in 4.64-PH/43
2689 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2691 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2692 "domain" is actually an IP address, and force a failure. This locks out
2693 those revolvers/nameservers that support "A-for-A" lookups, in
2694 contravention of the specifications.
2696 PH/48 When a host name was looked up from an IP address, and the subsequent
2697 forward lookup of the name timed out, the host name was left in
2698 $sender_host_name, contrary to the specification.
2700 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2701 restricted to single-key lookups, Exim was not diagnosing an error if
2702 * or *@ was used with a query-style lookup.
2704 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2706 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2707 long ago, but noone interested enough thought of it. Let's just say that
2708 the "1.1" means that there are some new functions that weren't there at
2709 some point in the past.
2711 PH/51 Error processing for expansion failure of helo_data from an smtp
2712 transport during callout processing was broken.
2714 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2715 tested/used via the -bh/-bhc/-bs options.
2717 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2718 bug, fixed in subsequent PCRE releases).
2720 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2721 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2723 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2729 SC/01 Use a glob alias rather than an array ref in eximstats generated
2730 parser. This improves both readability and performance.
2732 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2733 Don't display local sender or destination tables in eximstats unless
2734 there is data to show.
2735 Added average volumes into the eximstats top table text output.
2737 SC/03 Collect data on the number of addresses (recipients) as well
2738 as the number of messages in eximstats.
2740 TF/01 Correct an error in the documentation for the redirect router. Exim
2741 does (usually) call initgroups() when daemonizing.
2743 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2744 with consistent privilege compared to when running as a daemon.
2746 TF/03 Note in the spec that $authenticated_id is not set for local
2747 submissions from trusted users.
2749 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2750 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2752 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2753 by adding some example configuration directives to the default
2754 configuration file. A little bit of work is required to uncomment the
2755 directives and define how usernames and passwords are checked, but
2756 there is now a framework to start from.
2758 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2759 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2760 without this. I don't know how relevant this is to other LDAP libraries.
2762 PH/02 Add the verb name to the "unknown ACL verb" error.
2764 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2766 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2768 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2769 directory not expanded when it should be if an expanded home directory
2770 was set for the address (which is overridden by the transport).
2772 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2775 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2776 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2777 because it is too late at that time, and has no effect.
2779 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2780 security issue with \' (bugzilla #107). I could not use the
2781 PQescapeStringConn() function, because it needs a PGconn value as one of
2784 PH/08 When testing addresses using -bt, indicate those final addresses that
2785 are duplicates that would not cause an additional delivery. At least one
2786 person was confused, thinking that -bt output corresponded to deliveries.
2787 (Suppressing duplicates isn't a good idea as you lose the information
2788 about possibly different redirections that led to the duplicates.)
2790 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2791 systems where poll() doesn't work, in particular OS X.
2793 PH/10 Added more information to debugging output for retry time not reached.
2795 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2796 operations in malware.c.
2798 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2801 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2802 syslog_duplication set false, log lines that would normally be written
2803 both the the main log and to the reject log were not written to syslog at
2806 PH/14 In the default configuration, change the use of "message" in ACL warn
2807 statements to "add_header".
2809 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2810 not followed by a command (e.g. "seen endif").
2812 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2813 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2816 PH/17 Added extra conditions to the default value of delay_warning_condition
2820 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2821 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2822 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2825 The Auto-Submitted: and various List- headers are standardised, whereas I
2826 don't think Precedence: ever was.
2828 PH/18 Refactored debugging code in route_finduser() to show more information,
2829 in particular, the error code if getpwnam() issues one.
2831 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2832 This is apparently needed in addition to the PH/07 change above to avoid
2833 any possible encoding problems.
2835 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2836 but not after initializing Perl.
2838 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2839 output them only if debugging. By default they are written stderr,
2840 apparently, which is not desirable.
2842 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2845 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2848 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2850 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2851 authenticated or an ident call has been made. Suppress the default
2852 values for $authenticated_id and $authenticated_sender (but permit -oMai
2853 and -oMas) when testing with -bh.
2855 PH/24 Re-jigged the order of the tests in the default configuration so that the
2856 tests for valid domains and recipients precede the DNS black list and CSA
2857 tests, on the grounds that those ones are more expensive.
2859 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2860 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2861 This bug exists in every version of Exim that I still have, right back to
2864 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2865 However, an attempt to turn on case-sensitivity in a regex key by
2866 including (?-i) didn't work because the subject string was already
2867 lowercased, and the effects were non-intuitive. It turns out that a
2868 one-line patch can be used to allow (?-i) to work as expected.
2874 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2875 other effects) broke the use of negated acl sub-conditions.
2877 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2880 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2881 "Deny" causes Exim to reject the incoming connection with a 554 error.
2882 Unfortunately, if there is a major crisis, such as a disk failure,
2883 tcp-wrappers gives "deny", whereas what one would like would be some
2884 kind of temporary error. A kludge has been added to help with this.
2885 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2886 554 error is used if errno is still zero or contains ENOENT (which occurs
2887 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2890 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2892 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2893 errors. Otherwise a message that provokes a temporary error (when other
2894 messages do not) can cause a whole host to time out.
2896 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2897 addresses were routed directly to files or pipes from a redirect router.
2898 File deliveries just didn't batch; pipe deliveries might have suffered
2901 PH/06 A failure to get a lock for a hints database would erroneously always say
2902 "Failed to get write lock", even when it was really a read lock.
2904 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2905 of 0600. This has been changed to use the value of the lockfile_mode
2906 option (which defaults to 0600).
2908 PH/08 Applied small patch from the Sieve maintainer.
2910 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2911 folder from quota calculations, a direct delivery into this folder messed
2912 up the contents of the maildirsize file. This was because the regex was
2913 used only to exclude .Trash (or whatever) when the size of the mailbox
2914 was calculated. There was no check that a delivery was happening into an
2915 excluded directory. This bug has been fixed by ignoring all quota
2916 processing for deliveries into excluded directories.
2918 PH/10 Added the maildirfolder_create_regex option to appendfile.
2924 PH/01 The code for finding all the local interface addresses on a FreeBSD
2925 system running IPv6 was broken. This may well have applied to all BSD
2926 systems, as well as to others that have similar system calls. The broken
2927 code found IPv4 interfaces correctly, but gave incorrect values for the
2928 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2929 that it would not match correctly against @[] and not recognize the IPv6
2932 PH/02 The ipliteral router was not recognizing addresses of the form user@
2933 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2935 PH/03 Added disable_ipv6.
2937 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2938 decoded form, because it is most often used to construct To: headers
2939 lines in autoreplies, and the decoded form may well be syntactically
2940 invalid. However, $reply_address has leading white space removed, and all
2941 newlines turned into spaces so that the autoreply transport does not
2944 PH/05 If group was specified without a user on a router, and no group or user
2945 was specified on a transport, the group from the router was ignored.
2947 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2948 for visible compile-time settings that can be used to change these
2949 numbers, for those that want even more. Backwards compatibility with old
2950 spool files has been maintained. However, going back to a previous Exim
2951 release will lost any variables that are in spool files.
2953 PH/07 Two small changes when running in the test harness: increase delay when
2954 passing a TCP/IP connection to a new process, in case the original
2955 process has to generate a bounce, and remove special handling of
2956 127.0.0.2 (sic), which is no longer necessary.
2958 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2959 be the same on different OS.
2961 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2964 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2965 whether --show-vars was specified or not
2967 JJ/02 exipick: Added support for new ACL variable spool format introduced
2970 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2971 syntactically invalid From: or Reply-to: line, and a filter used this to
2972 generate an autoreply, and therefore failed to obtain an address for the
2973 autoreply, Exim could try to deliver to a non-existent relative file
2974 name, causing unrelated and misleading errors. What now happens is that
2975 it logs this as a hard delivery error, but does not attempt to create a
2978 PH/11 The exinext utility has a -C option for testing purposes, but although
2979 the given file was scanned by exinext itself; it wasn't being passed on
2980 when Exim was called.
2982 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2983 an end-of-file indication when reading a command response.
2985 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2986 compiled. In many other places in Exim, IPv6 addresses are always
2987 recognized, so I have changed this. It also means that IPv4 domain
2988 literals of the form [IPV4:n.n.n.n] are now always recognized.
2990 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2991 used if the router is not running as root, for example, when verifying at
2992 ACL time, or when using -bh. The debugging output from this situation was
2993 non-existent - all you got was a failure to exec. I have made two
2996 (a) Failures to set uid/gid, the current directory, or a process leader
2997 in a subprocess such as that created by queryprogram now generate
2998 suitable debugging output when -d is set.
3000 (b) The queryprogram router detects when it is not running as root,
3001 outputs suitable debugging information if -d is set, and then runs
3002 the subprocess without attempting to change uid/gid.
3004 PH/15 Minor change to Makefile for building test_host (undocumented testing
3007 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3008 additional section of a DNS packet that returns MX or SRV records.
3009 Instead, it always explicitly searches for A/AAAA records. This avoids
3010 major problems that occur when a DNS server includes only records of one
3011 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3012 fixed another bug: if SRV records were looked up and the corresponding
3013 address records were *not* found in the additional section, the port
3014 values from the SRV records were lost.
3016 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3017 using the correct key (the original address) when searching the retry
3018 rules in order to find which one to use for generating the retry hint.
3020 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3021 adding the default one. Similarly, if it contains a Reply-To: header, the
3022 errors_reply_to option, if set, is not used.
3024 PH/19 When calculating a retry time, Exim used to measure the "time since
3025 failure" by looking at the "first failed" field in the retry record. Now
3026 it does not use this if it is later than than the arrival time of the
3027 message. Instead it uses the arrival time. This makes for better
3028 behaviour in cases where some deliveries succeed, thus re-setting the
3029 "first failed" field. An example is a quota failure for a huge message
3030 when small messages continue to be delivered. Without this change, the
3031 "time since failure" will always be short, possible causing more frequent
3032 delivery attempts for the huge message than are intended.
3033 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3035 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3036 $1, $2, $3) because the numerical variables can be reset during some
3037 expansion items (e.g. "match"), thereby losing the authentication data.
3039 PH/21 Make -bV show the size of off_t variables so that the test suite can
3040 decide whether to run tests for quotas > 2G.
3042 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3043 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3044 filecount value is greater than 2G or if a quota value is greater than 2G
3045 on a system where the size of off_t is not greater than 4, a panic error
3048 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3049 never match. The debug and -bh output now contains an explicit error
3050 message indicating a malformed IPv4 address or mask.
3052 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3053 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3054 PH/23 above applies.
3056 PH/25 Do not write to syslog when running in the test harness. The only
3057 occasion when this arises is a failure to open the main or panic logs
3058 (for which there is an explicit test).
3060 PH/26 Added the /no_tell option to "control=freeze".
3062 PH/27 If a host name lookup failed very early in a connection, for example, if
3063 the IP address matched host_lookup and the reverse lookup yielded a name
3064 that did not have a forward lookup, an error message of the form "no IP
3065 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3066 could be logged. Now it outputs the IP address instead of "NULL".
3068 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3069 allows the sender and the authenticated sender to be set when
3070 submitting a message from within Exim. Since child_open_exim() is
3071 documented for local_scan(), the new function should be too.
3073 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3074 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3075 results in an empty string is now treated as unset.
3077 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3079 PH/31 Added sender_verify_fail logging option.
3081 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3082 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3083 tidied the source and removed it altogether.
3085 PH/33 When a queue run was abandoned because the load average was too high, a
3086 log line was always written; now it is written only if the queue_run log
3087 selector is set. In addition, the log line for abandonment now contains
3088 information about the queue run such as the pid. This is always present
3089 in "start" and "stop" lines but was omitted from the "abandon" line.
3091 PH/34 Omit spaces between a header name and the colon in the error message that
3092 is given when verify = headers_syntax fails (if there are lots of them,
3093 the message gets confusing).
3095 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3096 names, as there are now some PTR records that contain slashes. This check
3097 is only to protect against broken name servers that fall over on strange
3098 characters, so the fact that it applies to all lookups doesn't matter.
3100 PH/36 Now that the new test suite is complete, we can remove some of the
3101 special code in Exim that was needed for the old test suite. For example,
3102 sorting DNS records because real resolvers return them in an arbitrary
3103 order. The new test suite's fake resolver always returns records in the
3106 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3107 bounces) except when queue_only is set, to avoid logging races between
3108 the different processes.
3110 PH/38 Panic-die if .include specifies a non-absolute path.
3112 PH/39 A tweak to the "H" retry rule from its user.
3114 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3115 a label. They prevented compilation on older perls.
3117 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3118 a warning to be raised on newish perls.
3120 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3121 on queue. Changes to match documented behaviour of showing count of
3122 messages matching specified criteria.
3124 PH/40 Changed the default ident timeout from 30s to 5s.
3126 PH/41 Added support for the use of login_cap features, on those BSD systems
3127 that have them, for controlling the resources used by pipe deliveries.
3129 PH/42 The content-scanning code uses fopen() to create files in which to put
3130 message data. Previously it was not paying any attention to the mode of
3131 the files. Exim runs with umask(0) because the rest of the code creates
3132 files with open(), and sets the required mode explicitly. Thus, these
3133 files were ending up world-writeable. This was not a big issue, because,
3134 being within the spool directory, they were not world-accessible. I have
3135 created a function called modefopen, which takes an additional mode
3136 argument. It sets umask(777), creates the file, chmods it to the required
3137 mode, then resets the umask. All the relevant calls to fopen() in the
3138 content scanning code have been changed to use this function.
3140 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3141 to 24 hours. This avoids potential overflow problems when processing G
3142 and H retry rules. I suspect nobody ever tinkers with this value.
3144 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3146 PH/45 When the plaintext authenticator is running as a client, the server's
3147 challenges are checked to ensure they are valid base64 strings. By
3148 default, the authentication attempt is cancelled if an invalid string is
3149 received. Setting client_ignore_invalid_base64 true ignores these errors.
3150 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3151 they are received. Thus, the responses can be made to depend on the
3152 challenges. If an invalid string is ignored, an empty string is placed in
3155 PH/46 Messages that are created by the autoreply transport now contains a
3156 References: header, in accordance with RFCs 2822 and 3834.
3158 PH/47 Added authenticated_sender_force to the smtp transport.
3160 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3162 PH/49 Installed latest patch from the Sieve maintainer.
3164 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3165 was also set, the appendfile transport was unnecessarily scanning a
3166 directory of message files (e.g. for maildir delivery) to find the count
3167 of files (along with the size), even though it did not need this
3168 information. It now does the scan only if it needs to find either the
3169 size of the count of files.
3171 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3173 PH/52 Two bugs concerned with error handling when the smtp transport is
3176 (i) Exim was not creating retry information for temporary errors given
3177 for individual recipients after the DATA command when the smtp transport
3178 was used in LMTP mode. This meant that they could be retried too
3179 frequently, and not timed out correctly.
3181 (ii) Exim was setting the flag that allows error details to be returned
3182 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3183 recipients that were returned after the DATA command.
3185 PH/53 This is related to PH/52, but is more general: for any failing address,
3186 when detailed error information was permitted to be returned to the
3187 sender, but the error was temporary, then after the final timeout, only
3188 "retry timeout exceeded" was returned. Now it returns the full error as
3189 well as "retry timeout exceeded".
3191 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3192 do this, and (what is worse) MTAs that accept it.
3194 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3195 will now be deprecated.
3197 PH/56 New os.c-cygwin from the Cygwin maintainer.
3199 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3200 formats (previously only available in exim formats via -bpr, -bpru,
3201 and -bpra. Now also available in native and exiqgrep formats)
3203 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3204 with very large, slow to parse queues
3206 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3208 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3210 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3211 responses to authentication challenges, though it was showing the
3212 challenges; (ii) I've removed the CR characters from the debug output for
3215 PH/58 Allow for the insertion of a newline as well as a space when a string
3216 is turned into more than one encoded-word during RFC 2047 encoding. The
3217 Sieve code now uses this.
3219 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3220 data_4xx, lost_connection, tls_required.
3222 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3223 message was being sent as an SMTP response.
3225 PH/61 Add -l and -k options to exicyclog.
3227 PH/62 When verifying, if an address was redirected to one new address, so that
3228 verification continued, and the new address failed or deferred after
3229 having set something in $address_data, the value of $address_data was not
3230 passed back to the ACL. This was different to the case when no
3231 redirection occurred. The value is now passed back in both cases.
3233 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3234 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3235 use login_cap.h, so on its own it isn't the distinguishing feature. The
3236 new name refers directly to the setclassresources() function.
3238 PH/65 Added configuration files for NetBSD3.
3240 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3242 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3243 is preferred over IPv4.
3245 PH/68 The bounce_return_message and bounce_return_body options were not being
3246 honoured for bounces generated during the reception of non-SMTP messages.
3247 In particular, this applied to messages rejected by the ACL. This bug has
3248 been fixed. However, if bounce_return_message is true and bounce_return_
3249 body is false, the headers that are returned for a non-SMTP message
3250 include only those that have been read before the error was detected.
3251 (In the case of an ACL rejection, they have all been read.)
3253 PH/69 The HTML version of the specification is now built in a directory called
3254 spec_html instead of spec.html, because the latter looks like a path with
3255 a MIME-type, and this confuses some software.
3257 PH/70 Catch two compiler warnings in sieve.c.
3259 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3260 function verify_get_ident() calls ip_connect() to connect a socket, but
3261 if the "connect()" function timed out, ip_connect() used to close the
3262 socket. However, verify_get_ident() also closes the socket later, and in
3263 between Exim writes to the log, which may get opened at this point. When
3264 the socket was closed in ip_connect(), the log could get the same file
3265 descriptor number as the socket. This naturally causes chaos. The fix is
3266 not to close the socket in ip_connect(); the socket should be closed by
3267 the function that creates it. There was only one place in the code where
3268 this was missing, in the iplookup router, which I don't think anybody now
3269 uses, but I've fixed it anyway.
3271 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3272 well as to direct DNS lookups. Otherwise the handling of names in host
3273 lists is inconsistent and therefore confusing.
3279 PH/01 Two changes to the default runtime configuration:
3281 (1) Move the checks for relay_from_hosts and authenticated clients from
3282 after to before the (commented out) DNS black list checks.
3284 (2) Add control=submission to the relay_from_hosts and authenticated
3285 clients checks, on the grounds that messages accepted by these
3286 statements are most likely to be submissions.
3288 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3290 (1) Generate an error if the third argument for the ${prvs expansion is
3293 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3296 (3) Reset the variables that are obtained from the first argument of
3297 ${prvscheck and used in the second argument before leaving the code,
3298 because their memory is reclaimed, so using them afterwards may do
3301 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3302 one (it's much easier than Tom thought :-).
3304 (5) Because of (4), we can now allow for the use of $prvscheck_result
3305 inside the third argument.
3307 PH/03 For some reason, the default setting of PATH when running a command from
3308 a pipe transport was just "/usr/bin". I have changed it to
3311 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3312 anything to be listed in the output from -bV.
3314 PH/05 When a filter generated an autoreply, the entire To: header line was
3315 quoted in the delivery log line, like this:
3317 => >A.N.Other <ano@some.domain> <original@ddress> ...
3319 This has been changed so that it extracts the operative address. There
3320 may be more than one such address. If so, they are comma-separated, like
3323 => >ano@some.domain,ona@other.domain <original@ddress> ...
3325 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3326 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3327 not being looked up in the rDNS to get a host name, Exim was showing the
3328 IP address twice in Received: lines, even though the IP addresses were
3329 identical. For example:
3331 Received: from [1.2.3.4] (helo=[1.2.3.4])
3333 However, if the real host name was known, it was omitting the HELO data
3334 if it matched the actual IP address. This has been tidied up so that it
3335 doesn't show the same IP address twice.
3337 PH/07 When both +timestamp and +memory debugging was on, the value given by
3338 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3339 called by the debug printing, thereby overwriting the timestamp buffer.
3340 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3342 PH/08 When the original message was included in an autoreply transport, it
3343 always said "this is a copy of the message, including all the headers",
3344 even if body_only or headers_only was set. It now gives an appropriate
3347 PH/09 Applied a patch from the Sieve maintainer which:
3349 o fixes some comments
3350 o adds the (disabled) notify extension core
3351 o adds some debug output for the result of if/elsif tests
3352 o points to the current vacation draft in the documentation
3353 and documents the missing references header update
3357 o fixes a bug in processing the envelope test (when testing
3358 multiple envelope elements, the last element determined the
3361 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3362 Electronic Mail") by including:
3364 Auto-submitted: auto-generated
3366 in the messages that it generates (bounce messages and others, such as
3367 warnings). In the case of bounce messages for non-SMTP messages, there was
3368 also a typo: it was using "Auto_submitted" (underscore instead of
3369 hyphen). Since every message generated by Exim is necessarily in response
3370 to another message, thes have all been changed to:
3372 Auto-Submitted: auto-replied
3374 in accordance with these statements in the RFC:
3376 The auto-replied keyword:
3378 - SHOULD be used on messages sent in direct response to another
3379 message by an automatic process,
3381 - MUST NOT be used on manually-generated messages,
3383 - MAY be used on Delivery Status Notifications (DSNs) and Message
3384 Disposition Notifications (MDNs),
3386 - MUST NOT be used on messages generated by automatic or periodic
3387 processes, except for messages which are automatic responses to
3390 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3391 to the default Received: header definition.
3393 PH/12 Added log selector acl_warn_skipped (default on).
3395 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3396 variables because (a) they are in the wrong storage pool and (b) even if
3397 they were copied, it wouldn't work properly because of the caching.
3399 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3400 checking when decoding. Apparently there are clients that generate
3401 overlong encoded strings. Why am I not surprised?
3403 PH/15 If the first argument of "${if match_address" was not empty, but did not
3404 contain an "@" character, Exim crashed. Now it writes a panic log message
3405 and treats the condition as false.
3407 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3409 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3410 extension "envelope-auth". The code is finished and in agreement with
3411 other implementations, but there is no documentation so far and in fact,
3412 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3413 not changing the active code.
3415 Print executed "if" and "elsif" statements when debugging is used. This
3416 helps a great deal to understand what a filter does.
3418 Document more things not specified clearly in RFC3028. I had all this
3419 sorted out, when out of a sudden new issues came to my mind. Oops."
3421 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3424 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3425 canonical form (as documented). However, after a host name lookup from
3426 the IP address, check_host() was doing a simple string comparison with
3427 addresses acquired from the DNS when checking that the found name did
3428 have the original IP as one of its addresses. Since any found IPv6
3429 addresses are likely to be in abbreviated form, the comparison could
3430 fail. Luckily, there already exists a function for doing the comparison
3431 by converting both addresses to binary, so now that is used instead of
3432 the text comparison.
3434 PH/20 There was another similar case to PH/19, when a complete host name was
3435 given in a host list; looking up its IP address could give an abbreviated
3436 form, whereas the current host's name might or might not be abbreviated.
3437 The same fix has been applied.
3443 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3444 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3447 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3448 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3450 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3452 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3453 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3454 Also, for "Auto-Submitted", treat anything other than "no" as
3455 non-personal, in accordance with RFC 3834. (Previously it treated
3456 anything starting "auto-" as non-personal.)
3458 TF/01 The control=submission/name=... option had a problem with syntax
3459 errors if the name included a slash character. The /name= option
3460 now slurps the rest of the string, so it can include any characters
3461 but it must come last in the list of options (after /sender_retain
3464 PH/05 Some modifications to the interface to the fake nameserver for the new
3472 TK/01 Added the "success_on_redirect" address verification option. See
3473 NewStuff for rationale and an example.
3475 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3477 PH/02 Patch to exigrep to allow it to work on syslog lines.
3479 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3480 fread() to skip over the body file's header line, because in Cygwin the
3481 header line is locked and is inaccessible.
3483 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3484 co-exist for some time) to make it clear that it is the Exim ID that is
3485 referenced, not the Message-ID: header line.
3487 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3488 string_format() function, because snprintf() does not exist on all
3491 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3492 ${stat: expansion item.
3494 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3495 protocol synchronization error", to keep the pedants happy.
3497 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3498 well as for IRIX systems, when gcc is being used. See the host.c source
3501 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3503 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3506 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3507 transport and to the smtp transport in LMTP mode.
3509 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3511 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3512 run a filter in a subprocess. This could lead to confusion in subsequent
3513 lookups in the parent process. There should also be a search_tidyup() at
3514 the end of the subprocess.
3516 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3517 only if the host matched helo_try_verify_hosts, which caused the
3518 verification to occur when the EHLO/HELO command was issued. The ACL just
3519 tested the remembered result. Now, if a previous verification attempt has
3520 not happened, "verify = helo" does it there and then.
3522 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3524 TK/03 Fix log output including CR from clamd.
3526 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3527 exist provoked a memory error which could cause a segfault.
3529 PH/15 Installed PCRE 6.2
3531 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3533 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3534 of the problem. Specifically, suggested +O2 rather than +O1 for the
3537 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3539 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3540 record, the port was not correctly logged when the outgoing_port log
3541 selector was set (it logged the transort's default port).
3543 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3544 fallback_hosts, and "hosts" in the smtp transport.
3546 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3547 host errors such as "Connection refused".
3549 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3550 authentication with radiusclient 0.4.9:
3552 - Error returned from rc_read_config was caught wrongly
3553 - Username/password not passed on to radius server due to wrong length.
3555 The presumption is that some radiusclient API changes for 4.51/PH/17
3556 were not taken care of correctly. The code is still untested by me (my
3557 Linux distribution still has 0.3.2 of radiusclient), but it was
3558 contributed by a Radius user.
3560 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3561 expanding the "port" option of the smtp transport.
3563 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3564 while reading a MIME header. Thanks to Tom Hughes for a patch.
3566 PH/24 Include config.h inside local_scan.h so that configuration settings are
3569 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3570 that in an ACL for RCPT (for example), you can examine exactly what was
3573 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3574 commands, but it was not correctly comparing the address with the actual
3575 client host address. Thus, it would show the EHLO address in Received:
3576 header lines when this was not necessary.
3578 PH/27 Added the % operator to ${eval:}.
3580 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3581 it should be ignoring failures (because with -C, for example, it has lost
3582 privilege). It wasn't ignoring creation failures other than "already
3585 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3588 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3589 because an input line was too long, either on its own, or by virtue of
3590 too many continuations, the temporary file was not being removed, and the
3591 return code was incorrect.
3593 PH/31 Missing "BOOL" in function definition in filtertest.c.
3595 PH/32 Applied Sieve patches from the maintainer.
3597 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3599 PH/33 Added "verify = not_blind".
3601 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3602 Local/Makefile (with some defaults set). These are used in built scripts
3603 such as exicyclog, but they have never been used in the exim_install
3604 script (though there are many overriding facilities there). I have
3605 arranged that the exim_install script now takes note of these two
3608 PH/35 Installed configuration files for Dragonfly.
3610 PH/36 When a locally submitted message by a trusted user did not contain a
3611 From: header, and the sender address was obtained from -f or from an SMTP
3612 MAIL command, and the trusted user did not use -F to supply a sender
3613 name, $originator_name was incorrectly used when constructing a From:
3614 header. Furthermore, $originator_name was used for submission mode
3615 messages from external hosts without From: headers in a similar way,
3616 which is clearly wrong.
3618 PH/37 Added control=suppress_local_fixups.
3620 PH/38 When log_selector = +received_sender was set, and the addition of the
3621 sender made the log line's construction buffer exactly full, or one byte
3622 less than full, an overflow happened when the terminating "\n" was
3625 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3626 when the result of a list match is failure because a DNS lookup failed.
3628 PH/40 RM_COMMAND is now used in the building process.
3630 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3631 the "build-* directories that it finds.
3633 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3634 address in a domain literal was a prefix of an interface address.
3636 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3637 when verifying a sender address, unless rewrite_headers is false.
3639 PH/44 Wrote a long comment about why errors_to addresses are verified as
3640 recipients, not senders.
3642 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3643 the ratelimit ACL was added.
3645 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3647 PH/46 Added extra information about PostgreSQL errors to the error string.
3649 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3650 avoiding the need to install special zones in a real server. This is
3651 backwards compatible; if it can't find the fake resolver, it drops back.
3652 Thus, both old and new test suites can be run.
3654 TF/02 Added util/ratelimit.pl
3656 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3657 clock is set back in time.
3659 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3660 Candler <B.Candler@pobox.com>.
3662 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3663 OK for addresses that are the result of redirections.
3665 PH/48 A number of further additions for the benefit of the new test suite,
3666 including a fake gethostbyname() that interfaces to the fake DNS resolver
3669 TF/06 The fix for widen_domains has also been applied to qualify_single and
3670 search_parents which are the other dnslookup options that can cause
3673 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3676 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3678 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3679 service" if it encountered an MX record with an empty target hostname.
3680 The message is now "an MX or SRV record indicated no SMTP service".
3682 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3683 if the DNS of the sending site is misconfigured. This is quite a
3684 common situation. This change restores the behaviour of treating a
3685 helo verification defer as a failure.
3687 PH/51 If self=fail was set on a router, the bounce message did not include the
3688 actual error message.
3694 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3696 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3697 command itself ended in error, the underlying message about the transport
3698 filter timeout was being overwritten with the pipe command error. Now the
3699 underlying error message should be appended to the second error message.
3701 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3703 PH/02 Reduce the amount of output that "make" produces by default. Full output
3704 can still be requested.
3706 PH/03 The warning log line about a condition test deferring for a "warn" verb
3707 was being output only once per connection, rather than after each
3708 occurrence (because it was using the same function as for successful
3709 "warn" verbs). This seems wrong, so I have changed it.
3711 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3712 it should not have, which might have caused a crash in the right
3713 circumstances, but probably never did.
3715 PH/04 Installed a modified version of Tony Finch's patch to make submission
3716 mode fix the return path as well as the Sender: header line, and to
3717 add a /name= option so that you can make the user's friendly name appear
3720 TF/03 Added the control = fakedefer ACL modifier.
3722 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3723 Mark Lowes for thorough testing.
3725 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3727 TK/03 Merged latest SRS patch from Miles Wilton.
3729 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3730 with the definition in sysexits.h (which is #included earlier).
3731 Fortunately, Exim does not actually use EX_OK. The code used to try to
3732 preserve the sysexits.h value, by assuming that macro definitions were
3733 scanned for macro replacements. I have been disabused of this notion,
3734 so now the code just undefines EX_OK before #including unistd.h.
3736 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3737 in the smtp transport. When a block could not be written in a single
3738 write() function, the timeout was being re-applied to each part-write.
3739 This seems wrong - if the receiver was accepting one byte at a time it
3740 would take for ever. The timeout is now adjusted when this happens. It
3741 doesn't have to be particularly precise.
3743 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3744 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3746 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3747 without a domain if the check to <postmaster@domain> fails.
3749 SC/01 Eximstats: added -xls and the ability to specify output files
3750 (patch written by Frank Heydlauf).
3752 SC/02 Eximstats: use FileHandles for outputting results.
3754 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3756 SC/04 Eximstats: fixed display of large numbers with -nvr option
3758 SC/05 Eximstats: fixed merging of reports with empty tables.
3760 SC/06 Eximstats: added the -include_original_destination flag
3762 SC/07 Eximstats: removed tabs and trailing whitespace.
3764 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3766 TK/06 MBOX spool code: Add real "From " MBOX separator line
3767 so the .eml file is really in mbox format (even though
3768 most programs do not really care). Patch from Alex Miller.
3770 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3771 The latter is generated from $received_to and is only set if the
3772 message has one envelope recipient. SA can use these headers,
3773 obviously out-of-the-box. Patch from Alex Miller.
3775 PH/08 The ${def test on a variable was returning false if the variable's
3776 value was "0", contrary to what the specification has always said!
3777 The result should be true unless the variable is empty.
3779 PH/09 The syntax error of a character other than { following "${if
3780 def:variable_name" (after optional whitespace) was not being diagnosed.
3781 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3782 accidental colon was present, for example, could give incorrect results.
3784 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3785 result is used (not including appendfile, where other changes are about
3788 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3789 This involved changing a lot of size variables from int to off_t. It
3790 should work with maildirs and everything.
3792 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3793 spamd dying while we are connected to it.
3795 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3798 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3799 function for BDB 4.3.
3801 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3803 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3804 the sources for PCRE have been re-arranged and more files are now
3807 PH/15 The code I had for printing potentially long long variables in PH/11
3808 above was not the best (it lost precision). The length of off_t variables
3809 is now inspected at build time, and an appropriate printing format (%ld
3810 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3811 to be "long long int" or "long int". This is needed for the internal
3812 formatting function string_vformat().
3814 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3815 the configuration file to be ":syslog", then the script "guesses" where
3816 the logs files are, rather than using the compiled in default. In our
3817 case the guess is not the same as the compiled default, so the script
3818 suddenly stopped working when I started to use syslog. The patch checks
3819 to see if log_file_path is "". If so, it attempts to read it from exim
3820 with no configuration file to get the compiled in version, before it
3821 falls back to the previous guessing code."
3823 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3824 implementing BATV in an Exim configuration. See NewStuff for the gory
3827 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3828 Makefile that are specific to HP-UX.
3830 PH/18 If the "use_postmaster" option was set for a recipient callout together
3831 with the "random" option, the postmaster address was used as the MAIL
3832 FROM address for the random test, but not for the subsequent recipient
3833 test. It is now used for both.
3835 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3836 patch removes a few documentation additions to RFC 3028, because the
3837 latest draft now contains them. It adds the new en;ascii-case comparator
3838 and a new error check for 8bit text in MIME parts. Comparator and
3839 require names are now matched exactly. I enabled the subaddress
3840 extension, but it is not well tested yet (read: it works for me)."
3842 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3843 rework some of the code of TK/09 above to avoid the hardwired use of
3844 "%lld" and "long long". Replaced the call to snprintf() with a call to
3847 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3848 records point to non-existent hosts", "retry timeout exceeded", and
3849 "retry time not reached for any host after a long failure period".
3851 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3852 experimental DomainKeys support:
3854 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3855 (2) On an error such as an illegally used "control", the wrong name for
3856 the control was given.
3858 These problems did NOT occur unless DomainKeys support was compiled.
3860 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3862 PH/24 Added ${if match_ip condition.
3864 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3865 databases so that it will be absolutely obvious if a crash occurs in the
3866 DB library. This is a regular occurrence (often caused by mis-matched
3869 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3870 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3871 picked up on a user's system that detects such things. There doesn't seem
3872 to be a gcc warning option for this - only an attribute that has to be
3873 put on the function's prototype. It seems that in Fedora Core 4 they have
3874 set this on a number of new functions. No doubt there will be more in due
3877 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3878 specify a transport. However, if an address that was verified by such a
3879 router was the subject of a callout, Exim crashed because it tried to
3880 read the rcpt_include_affixes from the non-existent transport. Now it
3881 just assumes that the setting of that option is false. This bug was
3882 introduced by 4.51/PH/31.
3884 PH/28 Changed -d+all to exclude +memory, because that information is very
3885 rarely of interest, but it makes the output a lot bigger. People tend to
3886 do -d+all out of habit.
3888 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3889 code in os-type was giving problems when libc.so lives in lib64, like on
3892 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3893 aren't the modern standard, and it seems that some systems' include files
3894 don't always have them. Exim was already checking for some of the newer
3895 ones like T_AAAA, and defining it itself. I've added checks for all the
3896 record types that Exim uses.
3898 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3899 not automatically generating a new one, as it is supposed to. This
3900 prevented TLS from working. If the file did exist, but contained invalid
3901 data, a new version was generated, as expected. It was only the case of a
3902 non-existent file that was broken.
3904 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3905 with a change in libdomainkeys > 0.64.
3907 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3908 from DNS. If the selector record carries the flag, it now has
3909 precedence over the domain-wide flag.
3911 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3913 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3914 the use of a port name that isn't defined in /etc/services) occurred, the
3915 message was deferred as in a normal delivery, and thus remained on the
3916 spool, instead of being failed because of the mua_wrapper setting. This
3917 is now fixed, and I tidied up some of the mua_wrapper messages at the
3920 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3921 the messages in a hash of arrays rather than using individual hashes.
3922 This is a bit cleaner and results in dramatic memory savings, albeit
3923 at a slight CPU cost.
3925 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3926 as requested by Marc Sherman.
3928 SC/10 Eximstats: added histograms for user specified patterns as requested
3931 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3933 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3934 fopen() in the content-scanning modules that did not already have it.
3940 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3941 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3943 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3945 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3947 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3948 to test Sieve filters that use "vacation".
3950 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3951 that changes the way the GnuTLS parameters are stored in the cache file.
3952 The new format can be generated externally. For backward compatibility,
3953 if the data in the cache doesn't make sense, Exim assumes it has read an
3954 old-format file, and it generates new data and writes a new file. This
3955 means that you can't go back to an older release without removing the
3958 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3959 work if there are any delivery delays because "one_time" forces the
3960 parent to be marked "delivered", so its unseen clone is never tried
3961 again. For this reason, Exim now forbids the simultaneous setting of
3964 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3965 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3966 entry if you want to know the details. The fix, however, neglected to
3967 consider the case when local delivery batching is involved. The test for
3968 "previously delivered" was not happening when checking to see if an
3969 address could be batched with a previous (undelivered) one; under
3970 certain circumstances this could lead to multiple deliveries to the same
3973 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3974 in its include files, and this causes problems building Exim.
3976 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3977 header_syntax) but Exim was just ignoring anything given after a slash.
3978 In particular, this caused confusion with an attempt to use "verify =
3979 reverse_host_lookup/defer_ok". An error is now given when options are
3980 supplied for verify items that do not have them. (Maybe reverse_host_
3981 lookup should have a defer_ok option, but that's a different point.)
3983 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3984 defined by RFC 821) to 2048, because there were problems with some AUTH
3985 commands, and RFC 1869 says the size should be increased for extended
3986 SMTP commands that take arguments.
3988 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3991 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3992 "unknown" error; now it says that the functionality isn't in the binary.
3994 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3995 an address' error message when a string expansion fails (syntax or
3996 whatever). Otherwise the password may appear in the log. Following change
3997 PH/42 below, there is no longer a chance of it appearing in a bounce
4000 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4002 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4003 of its addresses was ever tried. (Bugzilla bug #2.)
4005 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4006 the result incorrectly in the debug output. (It correctly added a newline
4007 to what was transported.)
4009 TF/01 Added $received_time.
4011 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4012 commented out examples of how to interface to a virus scanner and to
4013 SpamAssassin. Also added commented examples of av_scanner and
4014 spamd_address settings.
4016 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4017 and controls are allowed in which ACLs. There were a couple of minor
4018 errors. Some of the entries in the conditions table (which is a table of
4019 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4020 negation of where the condition IS allowed.
4022 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4024 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4025 header file does not have a version number, so I've had to invent a new
4026 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4027 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4028 radiusclient), but it was contributed by a Radius user.
4030 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4031 files or directories, for OpenSSL.
4033 PH/19 When an Exim process that is running as root has to create an Exim log
4034 file, it does so in a subprocess that runs as exim:exim so as to get the
4035 ownership right at creation (otherwise, other Exim processes might see
4036 the file with the wrong ownership). There was no test for failure of this
4037 fork() call, which would lead to the process getting stuck as it waited
4038 for a non-existent subprocess. Forks do occasionally fail when resources
4039 run out. I reviewed all the other calls to fork(); they all seem to check
4042 PH/20 When checking for unexpected SMTP input at connect time (before writing
4043 the banner), Exim was not dealing correctly with a non-positive return
4044 from the read() function. If the client had disconnected by this time,
4045 the result was a log entry for a synchronization error with an empty
4046 string after "input=" when read() returned zero. If read() returned -1
4047 (an event I could not check), uninitialized data bytes were printed.
4048 There were reports of junk text (parts of files, etc) appearing after
4051 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4053 PH/22 Added support for macro redefinition, and (re)definition in between
4054 driver and ACL definitions.
4056 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4057 forgetting to use the resulting value; it was using the unexpanded value.
4059 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4060 hadn't been configured. The fix is from Juergen Kreileder, who
4061 understands it better than I do:
4063 "Here's what I see happening with three configured cyrus_sasl
4064 authenticators configured (plain, login, cram-md5):
4066 On startup auth_cyrus_sasl_init() gets called for each of these.
4067 This means three calls to sasl_listmech() without a specified mech_list.
4068 => SASL tests which mechs of all available mechs actually work
4069 => three warnings about OTP not working
4070 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4072 With the patch, sasl_listmech() also gets called three times. But now
4073 SASL's mech_list option is set to the server_mech specified in the the
4074 authenticator. Or in other words, the answer from sasl_listmech()
4075 gets limited to just the mech you're testing for (which is different
4077 => the return list contains just 'plain' or 'login', 'cram-md5' or
4078 nothing depending on the value of ob->server_mech.
4080 I've just tested the patch: Authentication still works fine,
4081 unavailable mechs specified in the exim configuration are still
4082 caught, and the auth.log warnings about OTP are gone."
4084 PH/25 When debugging is enabled, the contents of the command line are added
4085 to the debugging output, even when log_selector=+arguments is not
4088 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4089 answer is "GNU", and only if the return is "GNU/something" is the answer
4092 PH/27 $acl_verify_message is now set immediately after the failure of a
4093 verification in an ACL, and so is available in subsequent modifiers. In
4094 particular, the message can be preserved by coding like this:
4096 warn !verify = sender
4097 set acl_m0 = $acl_verify_message
4099 Previously, $acl_verify_message was set only while expanding "message"
4100 and "log_message" when a very denied access.
4102 PH/28 Modified OS/os.c-Linux with
4104 -#ifndef OS_LOAD_AVERAGE
4105 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4107 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4108 nomenclature these days.)
4110 PH/29 Installed patch from the Sieve maintainer that adds the options
4111 sieve_useraddress and sieve_subaddress to the redirect router.
4113 PH/30 In these circumstances:
4114 . Two addresses routed to the same list of hosts;
4115 . First host does not offer TLS;
4116 . First host accepts first address;
4117 . First host gives temporary error to second address;
4118 . Second host offers TLS and a TLS session is established;
4119 . Second host accepts second address.
4120 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4121 and peerdn, if requested) that were in fact used only for the second
4124 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4125 attention to any local part prefix or suffix that was matched by the
4126 router that accepted the address. It now behaves in the same way as it
4127 does for delivery: the affixes are removed from the local part unless
4128 rcpt_include_affixes is set on the transport.
4130 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4131 timeout during the DATA phase of an incoming message.
4133 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4134 applied a patch sanctioned by the Sieve maintainer.
4136 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4137 the uid or gid is negative. A case of a negative gid caused this to be
4138 noticed. The fix allows for either to be negative.
4140 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4141 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4144 PH/36 The change PH/12 above was broken. Fixed it.
4146 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4147 the grounds that routing the same address twice would always produce the
4148 same answer. This might have been true once, but it is certainly no
4149 longer true now. Routing a child address may depend on the previous
4150 routing that produced that child. Some complicated redirection strategies
4151 went wrong when messages had multiple recipients, and made Exim's
4152 behaviour dependent on the order in which the addresses were given.
4154 I have moved the duplicate checking until after the routing is complete.
4155 Exim scans the addresses that are assigned to local and remote
4156 transports, and removes any duplicates. This means that more work will be
4157 done, as duplicates will always all be routed, but duplicates are
4158 presumably rare, so I don't expect this is of any significance.
4160 For deliveries to pipes, files, and autoreplies, the duplicate checking
4161 still happens during the routing process, since they are not going to be
4164 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4165 It corrects a timeout issue with spamd. This is Ian's comment: "The
4166 background is that sometimes spamd either never reads data from a
4167 connection it has accepted, or it never writes response data. The exiscan
4168 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4169 blindly assumes that writes won't block so it may never time out."
4171 PH/39 Allow G after quota size as well as K and M.
4173 PH/40 The value set for $authenticated_id in an authenticator may not contain
4174 binary zeroes or newlines because the value is written to log lines and
4175 to spool files. There was no check on this. Now the value is run through
4176 the string_printing() function so that such characters are converted to
4177 printable escape sequences.
4179 PH/41 $message_linecount is a new variable that contains the total number of
4180 lines in the message. Compare $body_linecount, which is the count for the
4183 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4184 bounce and delay warning messages, except in certain special cases, which
4187 (a) An SMTP error message from a remote host;
4188 (b) A message specified in a :fail: redirection;
4189 (c) A message specified in a "fail" command in a system filter;
4190 (d) A message specified in a FAIL return from the queryprogram router;
4191 (e) A message specified by the cannot_route_message router option.
4193 In these cases only, Exim does include the error details in bounce and
4194 warning messages. There are also a few cases where bland messages such
4195 as "unrouteable address" or "local delivery error" are given.
4197 PH/43 $value is now also set for the "else" part of a ${run expansion.
4199 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4200 being worked on, but at least Exim now implements the latest version to
4203 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4204 process to complete was treated as a delivery failure, a timeout while
4205 writing the message to the pipe was logged, but erroneously treated as a
4206 successful delivery. Such timeouts include transport filter timeouts. For
4207 consistency with the overall process timeout, these timeouts are now
4208 treated as errors, giving rise to delivery failures by default. However,
4209 there is now a new Boolean option for the pipe transport called
4210 timeout_defer, which, if set TRUE, converts the failures into defers for
4211 both kinds of timeout. A transport filter timeout is now identified in
4214 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4215 systems where "make" and "gmake" are different, calling "gmake" at top
4216 level broke things. I've arranged for the value of $(MAKE) to be passed
4217 from the Makefile to this script so that it can call the same version of
4221 A note about Exim versions 4.44 and 4.50
4222 ----------------------------------------
4224 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4225 changes of various kinds. As a consequence, a big documentation update was
4226 needed. This delayed the release for rather longer than seemed good, especially
4227 in the light of a couple of (minor) security issues. Therefore, the changes
4228 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4229 release. So 4.44 and 4.50 are in effect two different branches that both start
4232 I have left the 4.50 change log unchanged; it contains all the changes since
4233 4.43. The change log for 4.44 is below; many of its items are identical to
4234 those for 4.50. This seems to be the most sensible way to preserve the
4235 historical information.
4241 1. Minor wording change to the doc/README.SIEVE file.
4243 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4244 computation of the current number of files was incorrect.
4246 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4247 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4250 4. Give more explanation in the error message when the command for a transport
4251 filter fails to execute.
4253 5. There are several places where Exim runs a non-Exim command in a
4254 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4255 was being done only for the command run by the queryprogram router. It is
4256 now done for all such subprocesses. The other cases are: ${run, transport
4257 filters, and the commands run by the lmtp and pipe transports.
4259 6. Added CONFIGURE_GROUP build-time option.
4261 7. Some older OS have a limit of 256 on the maximum number of file
4262 descriptors. Exim was using setrlimit() to set 1000 as a large value
4263 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4264 systems. I've change it so that if it can't get 1000, it tries for 256.
4266 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4267 was an oversight, and furthermore, ever since the addition of extra
4268 controls (e.g. 4.43/32), the checks on when to allow different forms of
4269 "control" were broken. There should now be diagnostics for all cases when a
4270 control that does not make sense is encountered.
4272 9. Added the /retain_sender option to "control=submission".
4274 10. $recipients is now available in the predata ACL (oversight).
4276 11. Tidy the search cache before the fork to do a delivery from a message
4277 received from the command line. Otherwise the child will trigger a lookup
4278 failure and thereby defer the delivery if it tries to use (for example) a
4279 cached ldap connection that the parent has called unbind on.
4281 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4282 of $address_data from the recipient verification was clobbered by the
4283 sender verification.
4285 13. The value of address_data from a sender verification is now available in
4286 $sender_address_data in subsequent conditions in the ACL statement.
4288 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4290 15. Added a new option "connect=<time>" to callout options, to set a different
4293 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4294 was its contents. (It was OK if the option was not defined at all.)
4296 17. A "Completed" log line is now written for messages that are removed from
4297 the spool by the -Mrm option.
4299 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4300 information about exactly what failed.
4302 19. Added -dd to debug only the daemon process.
4304 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4305 handles timeouts, both on the server side and network timeouts. Renamed the
4306 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4308 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4309 It is now set to "smtps".
4311 22. $host_address is now set to the target address during the checking of
4312 ignore_target_hosts.
4314 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4315 being passed; this would have caused $sender_host_name to have been used if
4316 matching the list had actually called for a host name (not very likely,
4317 since this list is usually IP addresses). A host name is now passed as
4320 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4321 code that specifies a non-restarting handler (typically sigaction() in
4322 modern systems) in an attempt to fix a rare and obscure crash bug.
4324 25. Narrowed the window for a race in the daemon that could cause it to ignore
4325 SIGCHLD signals. This is not a major problem, because they are used only to
4326 wake it up if nothing else does.
4328 26. A malformed maildirsize file could cause Exim to calculate negative values
4329 for the mailbox size or file count. Odd effects could occur as a result.
4330 The maildirsize information is now recalculated if the size or filecount
4333 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4334 support for a long time. Removed HAVE_SYS_VFS_H.
4336 28. Installed the latest version of exipick from John Jetmore.
4338 29. In an address list, if the pattern was not a regular expression, an empty
4339 subject address (from a bounce message) matched only if the pattern was an
4340 empty string. Non-empty patterns were not even tested. This was the wrong
4341 because it is perfectly reasonable to use an empty address as part of a
4342 database query. An empty address is now tested by patterns that are
4343 lookups. However, all the other forms of pattern expect the subject to
4344 contain a local part and a domain, and therefore, for them, an empty
4345 address still always fails if the pattern is not itself empty.
4347 30. Exim went into a mad DNS loop when attempting to do a callout where the
4348 host was specified on an smtp transport, and looking it up yielded more
4349 than one IP address.
4351 31. Re-factored the code for checking spool and log partition space into a
4352 function that finds that data and another that does the check. The former
4353 is then used to implement four new variables: $spool_space, $log_space,
4354 $spool_inodes, and $log_inodes.
4356 32. The RFC2047 encoding function was originally intended for short strings
4357 such as real names; it was not keeping to the 75-character limit for
4358 encoded words that the RFC imposes. It now respects the limit, and
4359 generates multiple encoded words if necessary. To be on the safe side, I
4360 have increased the buffer size for the ${rfc2047: expansion operator from
4363 33. It is now permitted to omit both strings after an "if" condition; if the
4364 condition is true, the result is "true". As before, when the second string
4365 is omitted, a false condition yields an empty string. This makes it less
4366 cumbersome to write custom ACL and router conditions.
4368 34. Failure to deliver a bounce message always caused it to be frozen, even if
4369 there was an errors_to setting on the router. The errors_to setting is now
4372 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4373 canonical form (fully expanded) before being placed in
4374 $sender_host_address.
4376 36. The table in the code that translates DNS record types into text (T_A to
4377 "A" for instance) was missing entries for NS and CNAME. It is just possible
4378 that this could have caused confusion if both these types were looked up
4379 for the same domain, because the text type is used as part of Exim's
4380 per-process caching. But the chance of anyone hitting this buglet seems
4383 37. The dnsdb lookup has been extended in a number of ways.
4385 (1) There is a new type, "zns", which walks up the domain tree until it
4386 finds some nameserver records. It should be used with care.
4388 (2) There is a new type, "mxh", which is like "mx" except that it returns
4389 just the host names, not the priorities.
4391 (3) It is now possible to give a list of domains (or IP addresses) to be
4392 looked up. The behaviour when one of the lookups defers can be
4393 controlled by a keyword.
4395 (4) It is now possible to specify the separator character for use when
4396 multiple records are returned.
4398 38. The dnslists ACL condition has been extended: it is now possible to supply
4399 a list of IP addresses and/or domains to be looked up in a particular DNS
4402 39. Added log_selector=+queue_time_overall.
4404 40. When running the queue in the test harness, wait just a tad after forking a
4405 delivery process, to get repeatability of debugging output.
4407 41. Include certificate and key file names in error message when GnuTLS fails
4408 to set them up, because the GnuTLS error message doesn't include the name
4409 of the failing file when there is a problem reading it.
4411 42. Allow both -bf and -bF in the same test run.
4413 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4415 44. The "Exiscan patch" is now merged into the mainline Exim source.
4417 45. Sometimes the final signoff response after QUIT could fail to get
4418 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4419 < 0 before doing a fflush(). This bug looks as though it goes back to the
4420 introduction of TLS in release 3.20, but "sometimes" must have been rare
4421 because the tests only now provoked it.
4423 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4424 (this can affect the format of dates).
4426 47. exim_tidydb, when checking for the continued existence of a message for
4427 which it has found a message-specific retry record, was not finding
4428 messages that were in split spool directories. Consequently, it was
4429 deleting retry records that should have stayed in existence.
4431 48. Steve fixed some bugs in eximstats.
4433 49. The SPA authentication driver was not abandoning authentication and moving
4434 on to the next authenticator when an expansion was forced to fail,
4435 contradicting the general specification for all authenticators. Instead it
4436 was generating a temporary error. It now behaves as specified.
4438 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4439 (the order specifies the preference for clients). The order is now AES256,
4440 AES128, 3DES, ARCFOUR128.
4442 51. Small patch to Sieve code - explicitly set From: when generating an
4445 52. Exim crashed if a remote delivery caused a very long error message to be
4446 recorded - for instance if somebody sent an entire SpamAssassin report back
4447 as a large number of 550 error lines. This bug was coincidentally fixed by
4448 increasing the size of one of Exim's internal buffers (big_buffer) that
4449 happened as part of the Exiscan merge. However, to be on the safe side, I
4450 have made the code more robust (and fixed the comments that describe what
4453 53. Now that there can be additional text after "Completed" in log lines (if
4454 the queue_time_overall log selector is set), a one-byte patch to exigrep
4455 was needed to allow it to recognize "Completed" as not the last thing in
4458 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4459 patch that reportedly fixes this has been added. I am not expert enough to
4460 create a test for it. This is what the patch creator wrote:
4462 "I found a little strange behaviour of ldap code when working with
4463 Windows 2003 AD Domain, where users was placed in more than one
4464 Organization Units. When I tried to give exim partial DN, the exit code
4465 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4466 But simultaneously result of request was absolutely normal ldap result,
4467 so I produce this patch..."
4469 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4470 REFERENCE, so I have modified the code to exclude the patch when that macro
4473 55. Some experimental protocols are using DNS PTR records for new purposes. The
4474 keys for these records are domain names, not reversed IP addresses. The
4475 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4476 leaves it alone. Component reversal etc. now happens only for IP addresses.
4479 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4481 57. Double the size of the debug message buffer (to 2048) so that more of very
4482 long debug lines gets shown.
4484 58. The exicyclog utility now does better if the number of log files to keep
4485 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4487 59. Two changes related to the smtp_active_hostname option:
4489 (1) $smtp_active_hostname is now available as a variable.
4490 (2) The default for smtp_banner uses $smtp_active_hostname instead
4491 of $primary_hostname.
4493 60. The host_aton() function is supposed to be passed a string that is known
4494 to be a valid IP address. However, in the case of IPv6 addresses, it was
4495 not checking this. This is a hostage to fortune. Exim now panics and dies
4496 if the condition is not met. A case was found where this could be provoked
4497 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4498 components; fortuitously, this particular loophole had already been fixed
4499 by change 4.50/55 above.
4501 If there are any other similar loopholes, the new check in host_aton()
4502 itself should stop them being exploited. The report I received stated that
4503 data on the command line could provoke the exploit when Exim was running as
4504 exim, but did not say which command line option was involved. All I could
4505 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4506 running as the user.
4509 61. There was a buffer overflow vulnerability in the SPA authentication code
4510 (which came originally from the Samba project). I have added a test to the
4511 spa_base64_to_bits() function which I hope fixes it.
4514 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4515 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4517 63. The daemon start-up calls getloadavg() while still root for those OS that
4518 need the first call to be done as root, but it missed one case: when
4519 deliver_queue_load_max is set with deliver_drop_privilege. This is
4520 necessary for the benefit of the queue runner, because there is no re-exec
4521 when deliver_drop_privilege is set.
4523 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4524 This has been fixed.
4526 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4527 was in use, was not putting the data itself into the right store pool;
4528 consequently, it could be overwritten for a subsequent message in the same
4529 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4532 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4534 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4535 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4536 the function were treating the return as a boolean value, which happened to
4537 work because 0=false and not-0=true, but is not correct code.
4539 68. The host_aton() function was not handling scoped IPv6 addresses (those
4540 with, for example, "%eth0" on the end) correctly.
4542 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4543 negated items (that is, ~something) in unsigned ints. Some compilers
4544 apparently mutter when there is no cast.
4546 70. If an address verification called from an ACL failed, and did not produce a
4547 user-specific message (i.e. there was only a "system" message), nothing was
4548 put in $acl_verify_message. In this situation, it now puts the system
4551 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4552 session; change 4.31/43 added the unwanted input to the log line - except
4553 that it did not do this in the start of session case. It now does.
4555 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4556 This is wrong and can cause the other end to generate a synchronization
4557 error if it is another Exim or anything else that does the synchronization
4558 check. A QUIT command is no longer sent after a timeout.
4560 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4561 during host lookups.
4563 74. The defer_ok option of callout verification was not working if it was used
4564 when verifying addresses in header lines, that is, for this case:
4566 verify = header_sender/callout=defer_ok
4568 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4569 those file descriptors could be used for SMTP connections. If anything
4570 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4571 could be sent to the SMTP client, causing chaos. The daemon now opens
4572 stdin, stdout, and stderr to /dev/null when it puts itself into the
4575 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4576 log by default. The user can override this with suitable Perl magic.
4578 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4579 the log message when discard triggers, was not working for the DATA ACL or
4580 for the non-SMTP ACL.
4582 78. Error message wording change in sieve.c.
4584 79. If smtp_accept_max_per_host was set, the number of connections could be
4585 restricted to fewer than expected, because the daemon was trying to set up
4586 a new connection before checking whether the processes handling previous
4587 connections had finished. The check for completed processes is now done
4588 earlier. On busy systems, this bug wouldn't be noticed because something
4589 else would have woken the daemon, and it would have reaped the completed
4592 80. If a message was submitted locally by a user whose login name contained one
4593 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4594 It caused a spool format error. I have fixed the spool reading code. A
4595 related problem was that the "from" clause in the Received: line became
4596 illegal because of the space(s). It is now covered by ${quote_local_part.
4598 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4601 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4603 83. Patch from Sieve maintainer to fix unterminated string problem in
4604 "vacation" handling.
4606 84. Some minor changes to the Linux configuration files to help with other
4607 OS variants using glibc.
4609 85. One more patch for Sieve to update vacation handling to latest spec.
4612 ----------------------------------------------------
4613 See the note above about the 4.44 and 4.50 releases.
4614 ----------------------------------------------------
4620 1. Change 4.43/35 introduced a bug that caused file counts to be
4621 incorrectly computed when quota_filecount was set in an appendfile
4624 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4625 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4628 3. Give more explanation in the error message when the command for a transport
4629 filter fails to execute.
4631 4. There are several places where Exim runs a non-Exim command in a
4632 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4633 was being done only for the command run by the queryprogram router. It is
4634 now done for all such subprocesses. The other cases are: ${run, transport
4635 filters, and the commands run by the lmtp and pipe transports.
4637 5. Some older OS have a limit of 256 on the maximum number of file
4638 descriptors. Exim was using setrlimit() to set 1000 as a large value
4639 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4640 systems. I've change it so that if it can't get 1000, it tries for 256.
4642 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4643 was an oversight, and furthermore, ever since the addition of extra
4644 controls (e.g. 4.43/32), the checks on when to allow different forms of
4645 "control" were broken. There should now be diagnostics for all cases when a
4646 control that does not make sense is encountered.
4648 7. $recipients is now available in the predata ACL (oversight).
4650 8. Tidy the search cache before the fork to do a delivery from a message
4651 received from the command line. Otherwise the child will trigger a lookup
4652 failure and thereby defer the delivery if it tries to use (for example) a
4653 cached ldap connection that the parent has called unbind on.
4655 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4656 of $address_data from the recipient verification was clobbered by the
4657 sender verification.
4659 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4660 was its contents. (It was OK if the option was not defined at all.)
4662 11. A "Completed" log line is now written for messages that are removed from
4663 the spool by the -Mrm option.
4665 12. $host_address is now set to the target address during the checking of
4666 ignore_target_hosts.
4668 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4669 being passed; this would have caused $sender_host_name to have been used if
4670 matching the list had actually called for a host name (not very likely,
4671 since this list is usually IP addresses). A host name is now passed as
4674 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4675 code that specifies a non-restarting handler (typically sigaction() in
4676 modern systems) in an attempt to fix a rare and obscure crash bug.
4678 15. Narrowed the window for a race in the daemon that could cause it to ignore
4679 SIGCHLD signals. This is not a major problem, because they are used only to
4680 wake it up if nothing else does.
4682 16. A malformed maildirsize file could cause Exim to calculate negative values
4683 for the mailbox size or file count. Odd effects could occur as a result.
4684 The maildirsize information is now recalculated if the size or filecount
4687 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4688 support for a long time. Removed HAVE_SYS_VFS_H.
4690 18. Updated exipick to current release from John Jetmore.
4692 19. Allow an empty sender to be matched against a lookup in an address list.
4693 Previously the only cases considered were a regular expression, or an
4696 20. Exim went into a mad DNS lookup loop when doing a callout where the
4697 host was specified on the transport, if the DNS lookup yielded more than
4700 21. The RFC2047 encoding function was originally intended for short strings
4701 such as real names; it was not keeping to the 75-character limit for
4702 encoded words that the RFC imposes. It now respects the limit, and
4703 generates multiple encoded words if necessary. To be on the safe side, I
4704 have increased the buffer size for the ${rfc2047: expansion operator from
4707 22. Failure to deliver a bounce message always caused it to be frozen, even if
4708 there was an errors_to setting on the router. The errors_to setting is now
4711 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4712 canonical form (fully expanded) before being placed in
4713 $sender_host_address.
4715 24. Updated eximstats to version 1.33
4717 25. Include certificate and key file names in error message when GnuTLS fails
4718 to set them up, because the GnuTLS error message doesn't include the name
4719 of the failing file when there is a problem reading it.
4721 26. Expand error message when OpenSSL has problems setting up cert/key files.
4724 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4725 (this can affect the format of dates).
4727 28. exim_tidydb, when checking for the continued existence of a message for
4728 which it has found a message-specific retry record, was not finding
4729 messages that were in split spool directories. Consequently, it was
4730 deleting retry records that should have stayed in existence.
4732 29. eximstats updated to version 1.35
4733 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4734 1.35 - bugfix such that pie charts by volume are generated correctly
4736 30. The SPA authentication driver was not abandoning authentication and moving
4737 on to the next authenticator when an expansion was forced to fail,
4738 contradicting the general specification for all authenticators. Instead it
4739 was generating a temporary error. It now behaves as specified.
4741 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4742 (the order specifies the preference for clients). The order is now AES256,
4743 AES128, 3DES, ARCFOUR128.
4745 31. Small patch to Sieve code - explicitly set From: when generating an
4748 32. Exim crashed if a remote delivery caused a very long error message to be
4749 recorded - for instance if somebody sent an entire SpamAssassin report back
4750 as a large number of 550 error lines. This bug was coincidentally fixed by
4751 increasing the size of one of Exim's internal buffers (big_buffer) that
4752 happened as part of the Exiscan merge. However, to be on the safe side, I
4753 have made the code more robust (and fixed the comments that describe what
4756 33. Some experimental protocols are using DNS PTR records for new purposes. The
4757 keys for these records are domain names, not reversed IP addresses. The
4758 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4759 leaves it alone. Component reversal etc. now happens only for IP addresses.
4762 34. The host_aton() function is supposed to be passed a string that is known
4763 to be a valid IP address. However, in the case of IPv6 addresses, it was
4764 not checking this. This is a hostage to fortune. Exim now panics and dies
4765 if the condition is not met. A case was found where this could be provoked
4766 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4767 components; fortuitously, this particular loophole had already been fixed
4768 by change 4.50/55 or 4.44/33 above.
4770 If there are any other similar loopholes, the new check in host_aton()
4771 itself should stop them being exploited. The report I received stated that
4772 data on the command line could provoke the exploit when Exim was running as
4773 exim, but did not say which command line option was involved. All I could
4774 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4775 running as the user.
4778 35. There was a buffer overflow vulnerability in the SPA authentication code
4779 (which came originally from the Samba project). I have added a test to the
4780 spa_base64_to_bits() function which I hope fixes it.
4783 36. The daemon start-up calls getloadavg() while still root for those OS that
4784 need the first call to be done as root, but it missed one case: when
4785 deliver_queue_load_max is set with deliver_drop_privilege. This is
4786 necessary for the benefit of the queue runner, because there is no re-exec
4787 when deliver_drop_privilege is set.
4789 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4790 was in use, was not putting the data itself into the right store pool;
4791 consequently, it could be overwritten for a subsequent message in the same
4792 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4795 38. Sometimes the final signoff response after QUIT could fail to get
4796 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4797 < 0 before doing a fflush(). This bug looks as though it goes back to the
4798 introduction of TLS in release 3.20, but "sometimes" must have been rare
4799 because the tests only now provoked it.
4805 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4806 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4807 Now it returns an integer. A number of calls were still expecting a T/F
4808 return. Fortuitously, in all cases, the tests worked in OK situations,
4809 which is the norm. However, things would have gone wrong on any write
4810 failures on the smtp file descriptor. This function is used when sending
4811 messages over SMTP and also when doing verify callouts.
4813 2. When Exim is called to do synchronous delivery of a locally submitted
4814 message (the -odf or -odi options), it no longer closes stderr before doing
4817 3. Implemented the mua_wrapper option.
4819 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4821 5. Implemented the functions header_remove(), header_testname(),
4822 header_add_at_position(), and receive_remove_recipient(), and exported them
4825 6. If an ACL "warn" statement specified the addition of headers, Exim already
4826 inserted X-ACL-Warn: at the start if there was no header name. However, it
4827 was not making this test for the second and subsequent header lines if
4828 there were newlines in the string. This meant that an invalid header could
4829 be inserted if Exim was badly configured.
4831 7. Allow an ACL "warn" statement to add header lines at the start or after all
4832 the Received: headers, as well as at the end.
4834 8. Added the rcpt_4xx retry error code.
4836 9. Added postmaster_mailfrom=xxx to callout verification option.
4838 10. Added mailfrom=xxxx to the callout verification option, for verify=
4841 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4842 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4844 12. Inserted some casts to stop certain compilers warning when using pointer
4845 differences as field lengths or precisions in printf-type calls (mostly
4846 affecting debugging statements).
4848 13. Added optional readline() support for -be (dynamically loaded).
4850 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4851 same clock tick as a message's arrival, so that its received time was the
4852 same as the "first fail" time on the retry record, and that message
4853 remained on the queue past the ultimate address timeout, every queue runner
4854 would try a delivery (because it was past the ultimate address timeout) but
4855 after another failure, the ultimate address timeout, which should have then
4856 bounced the address, did not kick in. This was a "< instead of <=" error;
4857 in most cases the first failure would have been in the next clock tick
4858 after the received time, and all would be well.
4860 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4861 being recognized when the domain list was tested by the match_domain
4862 condition in an expansion string.
4864 16. Added the ${str2b64: operator.
4866 17. Exim was always calling setrlimit() to set a large limit for the number of
4867 processes, without checking whether the existing limit was already
4868 adequate. (It did check for the limit on file descriptors.) Furthermore,
4869 errors from getrlimit() and setrlimit() were being ignored. Now they are
4870 logged to the main and panic logs, but Exim does carry on, to try to do its
4871 job under whatever limits there are.
4873 18. Imported PCRE 5.0.
4875 19. Trivial typo in log message " temporarily refused connection" (the leading
4878 20. If the log selector return_path_on_delivery was set and an address was
4879 redirected to /dev/null, the delivery process crashed because it assumed
4880 that a return path would always be set for a "successful" delivery. In this
4881 case, the whole delivery is bypassed as an optimization, and therefore no
4884 21. Internal re-arrangement: the function for sending a challenge and reading
4885 a response while authentication was assuming a zero-terminated challenge
4886 string. It's now changed to take a pointer and a length, to allow for
4887 binary data in such strings.
4889 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4891 23. Exim was not respecting finduser_retries when seeking the login of the
4892 uid under which it was called; it was always trying 10 times. (The default
4893 setting of finduser_retries is zero.) Also, it was sleeping after the final
4894 failure, which is pointless.
4896 24. Implemented tls_on_connect_ports.
4898 25. Implemented acl_smtp_predata.
4900 26. If the domain in control=submission is set empty, Exim assumes that the
4901 authenticated id is a complete email address when it generates From: or
4902 Sender: header lines.
4904 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4905 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4906 chown and chgrp in /bin and hostname in /usr/bin.
4908 28. Exim was keeping the "process log" file open after each use, just as it
4909 does for the main log. This opens the possibility of it remaining open for
4910 long periods when the USR1 signal hits a daemon. Occasional processlog
4911 errors were reported, that could have been caused by this. Anyway, it seems
4912 much more sensible not to leave this file open at all, so that is what now
4915 29. The long-running daemon process does not normally write to the log once it
4916 has entered its main loop, and it closes the log before doing so. This is
4917 so that log files can straightforwardly be renamed and moved. However,
4918 there are a couple of unusual error situations where the daemon does write
4919 log entries, and I had neglected to close the log afterwards.
4921 30. The text of an SMTP error response that was received during a remote
4922 delivery was being truncated at 512 bytes. This is too short for some of
4923 the long messages that one sometimes sees. I've increased the limit to
4926 31. It is now possible to make retry rules that apply only when a message has a
4927 specific sender, in particular, an empty sender.
4929 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4930 it possible to be selective about when SMTP synchronization is enforced.
4932 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4934 32. Implemented hosts_connection_nolog.
4936 33. Added an ACL for QUIT.
4938 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4941 35. Added mailbox_size and mailbox_filecount to appendfile.
4943 36. Added control = no_multiline_responses to ACLs.
4945 37. There was a bug in the logic of the code that waits for the clock to tick
4946 in the case where the clock went backwards by a substantial amount such
4947 that the microsecond fraction of "now" was more than the microsecond
4948 fraction of "then" (but the whole seconds number was less).
4950 38. Added support for the libradius Radius client library this is found on
4951 FreeBSD (previously only the radiusclient library was supported).
4957 1. When certain lookups returned multiple values in the form name=value, the
4958 quoting of the values was not always being done properly. Specifically:
4959 (a) If the value started with a double quote, but contained no whitespace,
4961 (b) If the value contained whitespace other than a space character (i.e.
4962 tabs or newlines or carriage returns) it was not quoted.
4963 This fix has been applied to the mysql and pgsql lookups by writing a
4964 separate quoting function and calling it from the lookup code. The fix
4965 should probably also be applied to nisplus, ibase and oracle lookups, but
4966 since I cannot test any of those, I have not disturbed their existing code.
4968 2. A hit in the callout cache for a specific address caused a log line with no
4969 reason for rejecting RCPT. Now it says "Previous (cached) callout
4970 verification failure".
4972 3. There was an off-by-one bug in the queryprogram router. An over-long
4973 return line was truncated at 256 instead of 255 characters, thereby
4974 overflowing its buffer with the terminating zero. As well as fixing this, I
4975 have increased the buffer size to 1024 (and made a note to document this).
4977 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4978 when Exim is waiting for an SMTP response from a remote server, Exim
4979 restarts its select() call on the socket, thereby resetting its timeout.
4980 This is not a problem when such interrupts are rare. Somebody set up a cron
4981 job to run exiwhat every 2 minutes, which is less than the normal select()
4982 timeout (5 or 10 minutes). This meant that the select() timeout never
4983 kicked in because it was always reset. I have fixed this by comparing the
4984 time when an interrupt arrives with the time at the start of the first call
4985 to select(). If more time than the timeout has elapsed, the interrupt is
4986 treated as a timeout.
4988 5. Some internal re-factoring in preparation for the addition of Sieve
4989 extensions (by MH). In particular, the "personal" test is moved to a
4990 separate function, and given an option for scanning Cc: and Bcc: (which is
4991 not set for Exim filters).
4993 6. When Exim created an email address using the login of the caller as the
4994 local part (e.g. when creating a From: or Sender: header line), it was not
4995 quoting the local part when it contained special characters such as @.
4997 7. Installed new OpenBSD configuration files.
4999 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5000 try to make them clearer.
5002 9. Callout options, other than the timeout value, were being ignored when
5003 verifying sender addresses in header lines. For example, when using
5005 verify = header_sender/callout=no_cache
5007 the cache was (incorrectly) being used.
5009 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5010 only the Cygwin environment.
5012 11. When return_path_on_delivery was set as a log selector, if different remote
5013 addresses in the same message used different return paths and parallel
5014 remote delivery occurred, the wrong values would sometimes be logged.
5015 (Whenever a remote delivery process finished, the return path value from
5016 the most recently started remote delivery process was logged.)
5018 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5019 lines when AUTH and/or TLS are in use. This is the "received protocol"
5020 field. Exim used to use "asmtp" for authenticated SMTP, without any
5021 indication (in the protocol name) for TLS use. Now it follows the RFC and
5022 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5023 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5024 names appear in log lines as well as in Received: header lines.
5026 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5027 extensions, and comparison tests, and to fix some bugs.
5029 14. Changes to the "personal" filter test:
5031 (1) The test was buggy in that it was just doing the equivalent of
5032 "contains" tests on header lines. For example, if a user's address was
5033 anne@some.where, the "personal" test would incorrectly be true for
5035 To: susanne@some.where
5037 This test is now done by extracting each address from the header in turn,
5038 and checking the entire address. Other tests that are part of "personal"
5039 are now done using regular expressions (for example, to check local parts
5040 of addresses in From: header lines).
5042 (2) The list of non-personal local parts in From: addresses has been
5043 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5044 taken from the Sieve specification recommendations.
5046 (3) If the message contains any header line starting with "List-" it is
5047 treated as non-personal.
5049 (4) The test for "circular" in the Subject: header line has been removed
5050 because it now seems ill-conceived.
5052 15. Minor typos in src/EDITME comments corrected.
5054 16. Installed latest exipick from John Jetmore.
5056 17. If headers_add on a router specified a text string that was too long for
5057 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5058 of string_sprintf() is now avoided.
5060 18. $message_body_size was not set (it was always zero) when running the DATA
5061 ACL and the local_scan() function.
5063 19. For the "mail" command in an Exim filter, no default was being set for
5064 the once_repeat time, causing a random time value to be used if "once" was
5065 specified. (If the value happened to be <= 0, no repeat happened.) The
5066 default is now 0s, meaning "never repeat". The "vacation" command was OK
5067 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5068 (I found it when inspecting the code).
5070 20. There is now an overall timeout for performing a callout verification. It
5071 defaults to 4 times the callout timeout, which applies to individual SMTP
5072 commands during the callout. The overall timeout applies when there is more
5073 than one host that can be tried. The timeout is checked before trying the
5074 next host. This prevents very long delays if there are a large number of
5075 hosts and all are timing out (e.g. when the network connections are timing
5076 out). The value of the overall timeout can be changed by specifying an
5077 additional sub-option for "callout", called "maxwait". For example:
5079 verify = sender/callout=5s,maxwait=20s
5081 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5082 to the end before writing, but this should make it even safer).
5084 22. Exim was forgetting that it had advertised PIPELINING for the second and
5085 subsequent messages on an SMTP connection. It was also not resetting its
5086 memory on STARTTLS and an internal HELO.
5088 23. When Exim logs an SMTP synchronization error within a session, it now
5089 records whether PIPELINING has been advertised or not.
5091 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5092 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5093 rather than long int.
5095 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5097 26. Added the never_mail option to autoreply.
5103 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5104 crash if the getsockname() call failed; this can happen if a connection is
5105 closed very soon after it is established. The problem was simply in the
5106 order in which certain operations were done, causing Exim to try to write
5107 to the SMTP stream before it had set up the file descriptor. The bug has
5108 been fixed by making things happen in the correct order.
5114 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5115 before the connection was closed, thus losing the rejection response.
5117 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5118 some early Solaris releases, but causes trouble in current releases where
5119 socklen_t is defined.
5121 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5124 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5127 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5128 that releasing the top of it at the end releases what was used for sub-
5129 expansions (unless the block got too big). However, discard this block if
5130 the first thing is a variable or header, so that we can use its block when
5131 it is dynamic (useful for very large $message_headers, for example).
5133 6. Lookups now cache *every* query, not just the most recent. A new, separate
5134 store pool is used for this. It can be recovered when all lookup caches are
5135 flushed. Lookups now release memory at the end of their result strings.
5136 This has involved some general refactoring of the lookup sources.
5138 7. Some code has been added to the store_xxx() functions to reduce the amount
5139 of flapping under certain conditions.
5141 8. log_incoming_interface used to affect only the <= reception log lines. Now
5142 it causes the local interface and port to be added to several more SMTP log
5143 lines, for example "SMTP connection from", and rejection lines.
5145 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5147 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5149 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5150 could be overwritten at the end of the current message (or the start of a
5151 new message if it was set in a HELO ACL). The value is now preserved for
5152 the duration of the SMTP connection.
5154 12. If a transport had a headers_rewrite setting, and a matching header line
5155 contained an unqualified address, that address was qualified, even if it
5156 did not match any rewriting rules. The underlying bug was that the values
5157 of the flags that permit the existence of unqualified sender and recipient
5158 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5159 non-local messages, and by -bnq for local messages) were not being
5160 preserved with the message after it was received.
5162 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5163 "next input=" as part of the text comprising the host identity instead of
5164 the correct text. The code was using the same buffer for two different
5165 strings. However, depending on which order the printing function evaluated
5166 its arguments, the bug did not always show up. Under Linux, for example, my
5167 test suite worked just fine.
5169 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5170 doesn't work with some older versions of Perl. It has been changed to "my",
5171 which in any case is probably the better facility to use.
5173 15. A really picky compiler found some instances of statements for creating
5174 error messages that either had too many or two few arguments for the format
5177 16. The size of the buffer for calls to the DNS resolver has been increased
5178 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5179 for addresses that have a lot of PTR records. This alleviates a problem; it
5180 does not fully solve it.
5182 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5183 buffer now truncates the list and logs the incident, which is the same
5184 action as happens when Exim is looking up a host name and its aliases.
5185 Previously in this situation something unpredictable would happen;
5186 sometimes it was "internal error: store_reset failed".
5188 18. If a server dropped the connection unexpectedly when an Exim client was
5189 using GnuTLS and trying to read a response, the client delivery process
5190 crashed while trying to generate an error log message.
5192 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5193 string, for example:
5195 warn message = H1: something\nH2: something
5197 the text was added as a single header line from Exim's point of view
5198 though it ended up OK in the delivered message. However, searching for the
5199 second and subsequent header lines using $h_h2: did not work. This has been
5200 fixed. Similarly, if a system filter added multiple headers in this way,
5201 the routers could not see them.
5203 20. Expanded the error message when iplsearch is called with an invalid key to
5204 suggest using net-iplsearch in a host list.
5206 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5207 ACLs are no longer actually imposed (and a message to that effect is
5210 22. If a "gecos" field in a passwd entry contained escaped characters, in
5211 particular, if it contained a \" sequence, Exim got it wrong when building
5212 a From: or a Sender: header from that name. A second bug also caused
5213 incorrect handling when an unquoted " was present following a character
5214 that needed quoting.
5216 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5217 was not being matched caselessly.
5219 24. Arranged for all hyphens in the exim.8 source to be escaped with
5222 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5223 properly. Recipient callout cache records were still being keyed to include
5224 the sender, even when use_sender was set false. This led to far more
5225 callouts that were necessary. The sender is no longer included in the key
5226 when use_sender is false.
5228 26. Added "control = submission" modifier to ACLs.
5230 27. Added the ${base62d: operator to decode base 62 numbers.
5232 28. dnsdb lookups can now access SRV records.
5234 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5235 the configuration file.
5237 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5238 (-v) mode. This makes the output for a verbose queue run more intelligible.
5240 31. Added a use_postmaster feature to recipient callouts.
5242 32. Added the $body_zerocount variable, containing the number of binary zero
5243 bytes in the message body.
5245 33. The time of last modification of the "new" subdirectory is now used as the
5246 "mailbox time last read" when there is a quota error for a maildir
5249 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5251 35. Added +ignore_unknown as a special item in host lists.
5253 36. Code for decoding IPv6 addresses in host lists is now included, even if
5254 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5255 address was recognized as an IP address, but was then not correctly decoded
5256 into binary, causing unexpected and incorrect effects when compared with
5263 1. Very minor rewording of debugging text in manualroute to say "list of
5264 hosts" instead of "hostlist".
5266 2. If verify=header_syntax was set, and a header line with an unqualified
5267 address (no domain) and a large number of spaces between the end of the
5268 name and the colon was received, the reception process suffered a buffer
5269 overflow, and (when I tested it) crashed. This was caused by some obsolete
5270 code that should have been removed. The fix is to remove it!
5272 3. When running in the test harness, delay a bit after writing a bounce
5273 message to get a bit more predictability in the log output.
5275 4. Added a call to search_tidyup() just before forking a reception process. In
5276 theory, someone could use a lookup in the expansion of smtp_accept_max_
5277 per_host which, without the tidyup, could leave open a database connection.
5279 5. Added the variables $recipient_data and $sender_data which get set from a
5280 lookup success in an ACL "recipients" or "senders" condition, or a router
5281 "senders" option, similar to $domain_data and $local_part_data.
5283 6. Moved the writing of debug_print from before to after the "senders" test
5286 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5287 problems for message scanning, either using a data ACL, or using
5288 local_scan() because the Received: header was not generated till after they
5289 were called (in order to set the time as the time of reception completion).
5290 I have revised the way this works. The header is now generated after the
5291 body is received, but before the ACL or local_scan() are called. After they
5292 are run, the timestamp in the header is updated.
5298 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5299 before starting a queue runner without re-exec. This happened only when
5300 deliver_drop_privilege was set or when the Exim user was set to root. The
5301 effect of the bug was that timeouts during subsequent deliveries caused
5302 crashes instead of being properly handled. The handler is now left at its
5303 default (and expected) setting.
5305 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5306 message, again when deliver_drop_privilege is set or Exim is run as root.
5307 The bug described in (1) was not present in this case, but the tidying up
5308 of the other signals was missing. I have made the two cases consistent.
5310 3. The ignore_target_hosts setting on a manualroute router was being ignored
5311 for hosts that were looked up using the /MX notation.
5313 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5316 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5317 operated on the sender address. After changing the $sender_address to <>
5318 for the sender address verify, Exim was re-instated it as the original
5319 (before rewriting) address, but remembering that it had rewritten it, so it
5320 wasn't rewriting it again. This bug also had the effect of breaking the
5321 sender address verification caching when the sender address was rewritten.
5323 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5324 This has been changed so that if the ip literal address matches
5325 ignore_target_hosts, the router declines.
5327 7. Added expansion conditions match_domain, match_address, and match_local_
5328 part (NOT match_host).
5330 8. The placeholder for the Received: header didn't have a length field set.
5332 9. Added code to Exim itself and to exim_lock to test for a specific race
5333 condition that could lead to file corruption when using MBX delivery. The
5334 issue is with the lockfile that is created in /tmp. If this file is removed
5335 after a process has opened it but before that process has acquired a lock,
5336 there is the potential for a second process to recreate the file and also
5337 acquire a lock. This could lead to two Exim processes writing to the file
5338 at the same time. The added code performs the same test as UW imapd; it
5339 checks after acquiring the lock that its file descriptor still refers to
5340 the same named file.
5342 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5343 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5346 11. Added the smtp_active_hostname option. If used, this will typically be made
5347 to depend on the incoming interface address. Because $interface_address is
5348 not set up until the daemon has forked a reception process, error responses
5349 that can happen earlier (such as "too many connections") no longer contain
5352 12. If an expansion in a condition on a "warn" statement fails because a lookup
5353 defers, the "warn" statement is abandoned, and the next ACL statement is
5354 processed. Previously this caused the whole ACL to be aborted.
5356 13. Added the iplsearch lookup type.
5358 14. Added ident_timeout as a log selector.
5360 15. Added tls_certificate_verified as a log selector.
5362 16. Added a global option tls_require_ciphers (compare the smtp transport
5363 option of the same name). This controls incoming TLS connections.
5365 17. I finally figured out how to make tls_require_ciphers do a similar thing
5366 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5367 before starting the TLS session.
5369 18. Tabs are now shown as \t in -bP output.
5371 19. If the log selector return_path_on_delivery was set, Exim crashed when
5372 bouncing a message because it had too many Received: header lines.
5374 20. If two routers both had headers_remove settings, and the first one included
5375 a superfluous trailing colon, the final name in the first list and the
5376 first name in the second list were incorrectly joined into one item (with a
5377 colon in the middle).
5383 1. Added -C and -D options to the exinext utility, mainly to make it easier
5384 to include in the automated testing, but these could be helpful when
5385 multiple configurations are in use.
5387 2. The exinext utility was not formatting the output nicely when there was
5388 an alternate port involved in the retry record key, nor when there was a
5389 message id as well (for retries that were specific to a specific message
5390 and a specific host). It was also confused by IPv6 addresses, because of
5391 the additional colons they contain. I have fixed the IPv4 problem, and
5392 patched it up to do a reasonable job for IPv6.
5394 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5395 delivery, the log line now contains "pipelined" if PIPELINING was used.
5397 4. An SMTP transport process used to panic and die if the bind() call to set
5398 an explicit outgoing interface failed. This has been changed; it is now
5399 treated in the same way as a connect() failure.
5401 5. A reference to $sender_host_name in the part of a conditional expansion
5402 that was being skipped was still causing a DNS lookup. This no longer
5405 6. The def: expansion condition was not recognizing references to header lines
5406 that used bh_ and bheader_.
5408 7. Added the _cache feature to named lists.
5410 8. The code for checking quota_filecount in the appendfile transport was
5411 allowing one more file than it should have been.
5413 9. For compatibility with Sendmail, the command line option
5421 and sets the incoming protocol and host name (for trusted callers). The
5422 host name and its colon can be omitted when only the protocol is to be set.
5423 Note the Exim already has two private options, -pd and -ps, that refer to
5424 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5425 "s", but I don't think that's a major issue.
5427 10. A number of refactoring changes to the code, none of which should affect
5430 (a) The number of logging options was getting close to filling up the
5431 32-bit word that was used as a bit map. I have split them into two classes:
5432 those that are passed in the argument to log_write(), and those that are
5433 only ever tested independently outside of that function. These are now in
5434 separate 32-bit words, so there is plenty of room for expansion again.
5435 There is no change in the user interface or the logging behaviour.
5437 (b) When building, for example, log lines, the code previously used a
5438 macro that called string_cat() twice, in order to add two strings. This is
5439 not really sufficiently general. Furthermore, there was one instance where
5440 it was actually wrong because one of the argument was used twice, and in
5441 one call a function was used. (As it happened, calling the function twice
5442 did not affect the overall behaviour.) The macro has been replaced by a
5443 function that can join an arbitrary number of extra strings onto a growing
5446 (c) The code for expansion conditions now uses a table and a binary chop
5447 instead of a serial search (which was left over from when there were very
5448 few conditions). Also, it now recognizes conditions like "pam" even when
5449 the relevant support is not compiled in: a suitably worded error message is
5450 given if an attempt is made to use such a condition.
5452 11. Added ${time_interval:xxxxx}.
5454 12. A bug was causing one of the ddress fields not to be passed back correctly
5455 from remote delivery subprocesses. The field in question was not being
5456 subsequently used, so this caused to problems in practice.
5458 13. Added new log selectors queue_time and deliver_time.
5460 14. Might have fixed a bug in maildirsizefile handling that threw up
5461 "unexpected character" debug warnings, and recalculated the data
5462 unnecessarily. In any case, I expanded the warning message to give more
5465 15. Added the message "Restricted characters in address" to the statements in
5466 the default ACL that block characters like @ and % in local parts.
5468 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5469 Three changes have been made:
5471 (a) There was a serious bug; a negative response to MAIL caused the whole
5472 recipient domain to be cached as invalid, thereby blocking all messages
5473 to all local parts at the same domain, from all senders. This bug has
5474 been fixed. The domain is no longer cached after a negative response to
5475 MAIL if the sender used is not empty.
5477 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5480 (c) A new callout option, "use_sender" has been added for people who want
5481 the modified behaviour.
5487 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5490 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5491 indeed breaks things for older releases.
5493 3. Added additional logging to the case where there is a problem reading data
5494 from a filter that is running in a subprocess using a pipe, in order to
5495 try to track down a specific problem.
5497 4. Testing facility fudge: when running in the test harness and attempting
5498 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5499 getting "No route to host". Convert this to a timeout.
5501 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5504 6. Some OS don't have socklen_t but use size_t instead. This affects the
5505 fifth argument of getsockopt() amongst other things. This is now
5506 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5507 can be set for individual OS. I have set it for SunOS5, OSF1, and
5508 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5509 some earlier ones do not.
5511 7. Change 4.30/15 was not doing the test caselessly.
5513 8. The standard form for an IPv6 address literal was being rejected by address
5514 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5515 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5516 this, as well as the form without the "IPv6" on the front (but only when
5517 address literals are enabled, of course).
5519 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5521 10. Exim crashed if a message with an empty sender address specified by -f
5522 encountered a router with an errors_to setting. This could be provoked only
5523 by a command such as
5527 where an empty string was supplied; "<>" did not hit this bug.
5529 11. Installed PCRE release 4.5.
5531 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5532 remained set. It is now erased.
5534 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5535 times from message ids (which are base 36 rather than the normal 62).
5537 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5538 were being counted as actual protocol errors, and logged if the log
5539 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5540 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5541 and DATA following a set of rejected RCPTs do not count as protocol errors.
5542 In other words, Exim assumes they were pipelined, though this may not
5543 actually be the case. Of course, in all cases the client gets an
5544 appropriate error code.
5546 15. If a lookup fails in an ACL condition, a message about the failure may
5547 be available; it is used if testing the ACL cannot continue, because most
5548 such messages specify what the cause of the deferral is. However, some
5549 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5550 that caused an old message to be retained and used if a later statement
5551 caused a defer, replacing the real cause of the deferral.
5553 16. If an IP address had so many PTR records that the DNS lookup buffer
5554 was not large enough to hold them, Exim could crash while trying to process
5555 the truncated data. It now detects and logs this case.
5557 17. Further to 4.21/58, another change has been made: if (and only if) the
5558 first line of a message (the first header line) ends with CRLF, a bare LF
5559 in a subsequent header line has a space inserted after it, so as not to
5560 terminate the header.
5562 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5563 unnecessarily, used atoi() instead of strtol(), and didn't check the
5564 termination when getting file sizes from file names by regex.
5566 19. Completely re-implemented the support for maildirsize files, in the light
5567 of a number of problems with the previous contributed implementation
5568 (4.30/29). In particular:
5570 . If the quota is zero, the maildirsize file is maintained, but no quota is
5573 . If the maildir directory does not exist, it is created before any attempt
5574 to write a maildirsize file.
5576 . The quota value in the file is just a cache; if the quota is changed in
5577 the transport, the new value overrides.
5579 . A regular expression is available for excluding directories from the
5582 20. The autoreply transport checks the characters in options that define the
5583 message's headers; it allows continued headers, but it was checking with
5584 isspace() after an embedded newline instead of explicitly looking for a
5587 21. If all the "regular" hosts to which an address was routed had passed their
5588 expiry times, and had not reached their retry times, the address was
5589 bounced, even if fallback hosts were defined. Now Exim should go on to try
5592 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5593 equivalent code in the SMTP transport. Some hosts send humungous responses
5594 to HELO/EHLO, more than 1024 it seems.
5596 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5597 gives compiler warnings in some environments. I've now done it "properly",
5600 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5601 (because of problems with the built-in one) was declared to return uschar *
5602 instead of char *, causing compiler failure.
5604 25. Fixed a file descriptor leak when processing alias/forward files.
5606 26. Fixed a minor format string issue in dbfn.c.
5608 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5610 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5611 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5614 29. When a sender address is verified, it is cached, to save repeating the test
5615 when there is more than one recipient in a message. However, when the
5616 verification involves a callout, it is possible for different callout
5617 options to be set for different recipients. It is too complicated to keep
5618 track of this in the cache, so now Exim always runs a verification when a
5619 callout is required, relying on the callout cache for the optimization.
5620 The overhead is duplication of the address routing, but this should not be
5623 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5624 to be verified with callout=postmaster, and the main callout worked but the
5625 postmaster check failed, the verification correctly failed. However, if a
5626 subsequent RCPT command asked for sender verification *without* the
5627 postmaster check, incorrect caching caused this verification also to fail,
5630 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5631 it was not caching the DNS options (qualify_single, search_parents) that
5632 were used when the lookup failed. A subsequent lookup with different
5633 options therefore always gave the same answer, though there were cases
5634 where it should not have. (Example: a "domains = !$mx_any" option on a
5635 dnslookup router: the "domains" option is always processed without any
5636 widening, but the router might have qualify_single set.) Now Exim uses the
5637 cached value only when the same options are set.
5639 32. Added John Jetmore's "exipick" utility to the distribution.
5641 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5642 than a timeout (e.g. a certificate is required, and is not provided), an
5643 Exim server now closes the connection immediately. Previously it waited for
5644 the client to close - but if the client is SSL, it seems that they each
5645 wait for each other, leading to a delay before one of them times out.
5647 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5648 maintained 0.8.x compatibility because I don't think many are using it, and
5649 it is clearly obsolete.
5651 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5654 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5655 client certificate was expired. A simple patch fixes this, though I don't
5656 understand the full logic of why the verify callback is called multiple
5659 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5660 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5661 which causes problems with some clients (such as the Certicom SSL Plus
5662 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5663 disables the coutermeasure allowing Eudora to connect."
5665 38. Exim was not checking that a write() to a log file succeeded. This could
5666 lead to Bad Things if a log got too big, in particular if it hit a file
5667 size limit. Exim now panics and dies if it cannot write to a log file, just
5668 as it does if it cannot open a log file.
5670 39. Modified OS/Makefile-Linux so that it now contains
5672 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5674 The two -D definitions ensure that Exim is compiled with large file
5675 support, which makes it possible to handle log files that are bigger than
5678 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5679 instance) a domain was checked against a named list that involved a lookup,
5680 causing $domain_data to be set, then another domain was checked against the
5681 same list, then the first domain was re-checked, the value of $domain_data
5682 after the final check could be wrong. In particular, if the second check
5683 failed, it could be set empty. This bug probably also applied to
5686 41. The strip_trailing_dot option was not being applied to the address given
5687 with the -f command-line option.
5689 42. The code for reading a message's header from the spool was incrementing
5690 $received_count, but never initializing it. This meant that the value was
5691 incorrect (doubled) while delivering a message in the same process in which
5692 it was received. In the most common configuration of Exim, this never
5693 happens - a fresh exec is done - but it can happen when
5694 deliver_drop_privilege is set.
5696 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5697 it now includes up to 150 characters of the unexpected data in the log
5700 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5701 and building data strings. The size of both of these buffers was 10 000
5702 bytes - far larger than anybody would *ever* want, thought I. Needless to
5703 say, somebody hit the limit. I have increased the maximum line length to
5704 20 000 and the maximum data length of concatenated lines to 100 000. I have
5705 also fixed two bugs, because there was no checking on these buffers. Tsk,
5706 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5707 buffer is too small.
5709 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5710 lsearch lookups. Now it does.
5712 46. When parsing a route_list item in a manualroute router, a fixed-length
5713 buffer was used for the list of hosts. I made this 1024 bytes long,
5714 thinking that nobody would ever have a list of hosts that long. Wrong.
5715 Somebody had a whole pile of complicated expansion conditions, and the
5716 string was silently truncated, leading to an expansion error. It turns out
5717 that it is easier to change to an unlimited length (owing to other changes
5718 that have happened since this code was originally written) than to build
5719 structure for giving a limitation error. The length of the item that
5720 expands into the list of hosts is now unlimited.
5722 47. The lsearch lookup could not handle data where the length of text line was
5723 more than 4095 characters. Such lines were truncated, leading to shortened
5724 data being returned. It should now handle lines of any length.
5726 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5727 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5730 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5731 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5732 no longer happens in comment lines. A list of replacements is now placed
5733 at the head of all of the source files, except those whose only change is
5734 to replace PERL_COMMAND in the very first #! line.
5736 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5737 messages on the queue, with a bottom-up merge sort, using code contributed
5738 by Michael Haardt. This should make operations like -bp somewhat faster on
5739 large queues. It won't affect queue runners, except when queue_run_in_order
5742 51. Installed eximstats 1.31 in the distribution.
5744 52. Added support for SRV lookups to the dnslookup router.
5746 53. If an ACL referred to $message_body or $message_body_end, the value was not
5747 reset for any messages that followed in the same SMTP session.
5749 54. The store-handling optimization for building very long strings was not
5750 differentiating between the different store pools. I don't think this
5751 actually made any difference in practice, but I've tidied it.
5753 55. While running the routers to verify a sender address, $sender_address
5754 was still set to the sender address. This is wrong, because when routing to
5755 send a bounce to the sender, it would be empty. Therefore, I have changed
5756 it so that, while verifying a sender address, $sender_address is set to <>.
5757 (There is no change to what happens when verifying a recipient address.)
5759 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5760 target A or AAAA records (if not already returned) without resetting the
5761 qualify_single or search_parents options of the DNS resolver. These are
5762 inappropriate in this case because the targets of MX and SRV records must
5763 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5764 target that, when qualified, matched something in the local domain. These
5765 two options are now turned off when doing these lookups.
5767 57. It seems that at least some releases of Reiserfs (which does not have the
5768 concept of a fixed number of inodes) returns zero and not -1 for the
5769 number of available inodes. This interacted badly with check_spool_inodes,
5770 which assumed that -1 was the "no such thing" setting. What I have done is
5771 to check that the total number of inodes is greater than zero before doing
5772 the test of how many are available.
5774 58. When a "warn" ACL statement has a log_message modifier, the message is
5775 remembered, and not repeated. This is to avoid a lot of repetition when a
5776 message has many recipients that cause the same warning to be written.
5777 However, Exim was preserving the list of already written lines for an
5778 entire SMTP session, which doesn't seem right. The memory is now reset if a
5779 new message is started.
5781 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5782 debugging output unless log_rewrite was also set.
5784 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5785 of (int)(handle) when we know that handle contains (void *)(-1).
5787 61. The Exim daemon panic-logs an error return when it closes the incoming
5788 connection. However "connection reset by peer" seems to be common, and
5789 isn't really an error worthy of noting specially, so that particular error
5792 62. When Exim is trying to find all the local interfaces, it used to panic and
5793 die if the ioctl to get the interface flags failed. However, it seems that
5794 on at least one OS (Solaris 9) it is possible to have an interface that is
5795 included in the list of interfaces, but for which you get a failure error
5796 for this call. This happens when the interface is not "plumbed" into a
5797 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5798 failure of the "get flags" call assumes that the interface is down.
5800 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5801 makes life easier for people who are doing arithmetic on fields extracted
5802 from dates, where you often get leading zeros that should not be
5803 interpreted as octal.
5805 64. Added qualify_domain to the redirect router, to override the global
5808 65. If a pathologically long header line contained very many addresses (the
5809 report of this problem mentioned 10 000) and each of them was rewritten,
5810 Exim could use up a very large amount of memory. (It kept on making new
5811 copies of the header line as it rewrote, and never released the old ones.)
5812 At the expense of a bit more processing, the header rewriting function has
5813 been changed so that it no longer eats memory in this way.
5815 66. The generation of the Received: header has been moved from the time that a
5816 message starts to be received, to the time that it finishes. The timestamp
5817 in the Received: header should now be very close to that of the <= log
5818 line. There are two side-effects of this change:
5820 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5821 logged header lines no longer include the local Received: line, because
5822 it has not yet been created. The same applies to a copy of the message
5823 that is returned to a non-SMTP sender when a message is rejected.
5825 (b) When a filter file is tested using -bf, no additional Received: header
5826 is added to the test message. After some thought, I decided that this
5829 This change does not affect the value of $received_for. It is still set
5830 after address rewriting, but before local_scan() is called.
5832 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5834 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5835 gave an unhelpful panic error message, and a defer error. I have managed to
5836 change this behaviour so that it now rejects any supplied certificate,
5837 which seems right, as the list of acceptable certificates is empty.
5839 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5840 gave an unhelpful defer error. I have not managed to make this reject any
5841 supplied certificates, but the error message it gives is "no certificate
5842 supplied", which is not helpful.
5844 70. exigrep's output now also includes lines that are not associated with any
5845 message, but which match the given pattern. Implemented by a patch from
5846 Martin Sluka, which also tidied up the Perl a bit.
5848 71. Recipient callout verification, like sender verification, was using <> in
5849 the MAIL FROM command. This isn't really the right thing, since the actual
5850 sender may affect whether the remote host accepts the recipient or not. I
5851 have changed it to use the actual sender in the callout; this means that
5852 the cache record is now keyed on a recipient/sender pair, not just the
5853 recipient address. There doesn't seem to be a real danger of callout loops,
5854 since a callout by the remote host to check the sender would use <>.
5855 [SEE ABOVE: changed after hitting problems.]
5857 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5858 temporary errors. However, in the case of such a code being given after
5859 the end of a data transmission (i.e. after ".") Exim was failing to write
5860 a retry record for the message. (Yes, there was some broken host that was
5861 actually sending 8xx at this point.)
5863 73. An unknown lookup type in a host list could cause Exim to panic-die when
5864 the list was checked. (An example that provoked this was putting <; in the
5865 middle of a list instead of at the start.) If this happened during a DATA
5866 ACL check, a -D file could be left lying around. This kind of configuration
5867 error no longer causes Exim to die; instead it causes a defer error. The
5868 incident is still logged to the main and panic logs.
5870 74. Buglet left over from Exim 3 conversion. The message "too many messages
5871 in one connection" was written to the rejectlog but not the mainlog, except
5872 when address rewriting (yes!) was being logged.
5874 75. Added write_rejectlog option.
5876 76. When a system filter was run not as root (that is, when system_filter_user
5877 was set), the values of the $n variables were not being returned to the
5878 main process; thus, they were not subsequently available in the $sn
5881 77. Added +return_path_on_delivery log selector.
5883 78. A connection timeout was being treated differently from recipients deferred
5884 when testing hosts_max_try with a message that was older than the host's
5885 retry timeout. (The host should not be counted, thus allowing all hosts to
5886 be tried at least once before bouncing.) This may have been the cause of an
5887 occasionally reported bug whereby a message would remain on the queue
5888 longer than the retry timeout, but would be bounced if a delivery was
5889 forced. I say "may" because I never totally pinned down the problem;
5890 setting up timeout/retry tests is difficult. See also the next item.
5892 79. The ultimate address timeout was not being applied to errors that involved
5893 a combination of host plus message (for example, a timeout on a MAIL
5894 command). When an address resolved to a number of possible hosts, and they
5895 were not all tried for each delivery (e.g. because of hosts_max_try), a
5896 message could remain on the queue longer than the retry timeout.
5898 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5901 81. Fixed an obscure SMTP outgoing bug which required at least the following
5902 conditions: (a) there was another message waiting for the same server;
5903 (b) the server returned 5xx to all RCPT commands in the first message so
5904 that the message was not completed; (c) the server dropped the connection
5905 or gave a negative response to the RSET that Exim sends to abort the
5906 transaction. The observed case was a dropped connection after DATA that had
5907 been sent in pipelining mode. That is, the server had advertised PIPELINING
5908 but was not implementing it correctly. The effect of the bug was incorrect
5909 behaviour, such as trying another host, and this could lead to a crash.
5915 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5916 and daemon.c were passed as pointers to ints; they should have been
5917 pointers to socklen_t variables (which are typically unsigned ints).
5919 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5922 3. Fixed a really odd bug that affected only the testing scheme; patching a
5923 certain fixed string in the binary changed the value of another string that
5924 happened to be identical to the end of the original first string.
5926 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5927 name", it returns that address as the IP address. On some operating
5928 systems (e.g. Solaris), it also passes back the IP address string as the
5929 "host name". However, on others (e.g. Linux), it passes back an empty
5930 string. Exim wasn't checking for this, and was changing the host name to an
5931 empty string, assuming it had been canonicalized.
5933 5. Although rare, it is permitted to have more than one PTR record for a given
5934 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5935 all the names associated with an address, because they do in Solaris.
5936 However, it seems that they do not in Linux for data that comes from the
5937 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5938 I found this out when I moved to a new Linux workstation and tried to run
5939 the Exim test suite.
5941 To get round this problem I have changed the code so that it now does its
5942 own call to the DNS to look up PTR records when searching for a host name.
5943 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5944 addresses that are only in /etc/hosts are still found.
5946 This behaviour is, however, controlled by an option called host_lookup_
5947 order, which defaults to "bydns:byaddr". If people want to use the other
5948 order, or indeed, just use one or the other means of lookup, they can
5949 specify it in this variable.
5951 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5952 some operating systems, this comes back from gethostbyaddr() as an empty
5953 string, and this is what Exim used to test for. However, it seems that in
5954 other systems, "." is yielded. Exim now tests for this case too.
5956 7. The values of check_spool_space and check_log_space are now held internally
5957 as a number of kilobytes instead of an absolute number of bytes. If a
5958 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5959 kilobyte. This means that much larger values can be stored.
5961 8. Exim monitor: an attempt to get the action menu when not actually pointing
5962 at a message produces an empty menu entitled "No message selected". This
5963 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5964 no entries in it ("Shell widget menu has zero width and/or height"). So I
5965 have added a single, blank menu entry in this case.
5967 9. Added ${quote_local_part.
5969 10. MIME decoding is now applied to the contents of Subject: header lines when
5972 11. Now that a reference to $sender_host_address automatically causes a reverse
5973 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5974 host lookup before query-style lookups in lists that might use this
5975 variable. This has therefore been abolished, and the "net-" prefix is no
5976 longer necessary for query-style lookups.
5978 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5979 have been a typo for LFLAGS, so it has been changed.
5981 13. The install script calls Exim with "-C /dev/null" in order to find the
5982 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5983 to be output. However, since Exim outputs its version number before the
5984 error, it didn't break the script. It just looked ugly. I fixed this by
5985 always allowing "-C /dev/null" if the caller is root.
5987 14. Ignore overlarge ACL variable number when reading spool file - insurance
5988 against a later release with more variables having written the file.
5990 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5991 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5992 this, as well as the form without the "IPv6" on the front.
5994 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5995 OS/Makefile-Darwin file.
5997 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5998 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6000 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6001 variable. However, some OS use ints for inodes. I've added cast to long int
6002 to get rid of the compiler warning.
6004 19. I had forgotten to lock out "/../" in configuration file names when
6005 ALT_CONFIG_PREFIX was set.
6007 20. Routers used for verification do not need to specify transports. However,
6008 if such a router generated a host list, and callout was configured, Exim
6009 crashed, because it could not find a port number from the (non-existent)
6010 transport. It now assumes port 25 in this circumstance.
6012 21. Added the -t option to exigrep.
6014 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6015 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6016 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6017 them is set, LOOKUP_LSEARCH is forced.
6019 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6021 24. Added sender and host information to the "rejected by local_scan()" log
6022 line; previously there was no indication of these.
6024 25. Added .include_if_exists.
6026 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6027 Linux. It turns out that not all file systems support this. Apparently some
6028 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6029 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6030 not supported on the file descriptor, is now ignored when Exim is trying to
6031 sync a directory. This applies only to Linux.
6033 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6035 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6036 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6039 29. Added support for maildirsize files from supplied patch (modified a bit).
6041 30. The use of :fail: followed by an empty string could lead Exim to respond to
6042 sender verification failures with (e.g.):
6044 550 Verification failed for <xxx>
6045 550 Sender verify failed
6047 where the first response line was missing the '-' that indicates it is not
6048 the final line of the response.
6050 31. The loop for finding the name of the user that called Exim had a hardwired
6051 limit of 10; it now uses the value of finduser_retries, which is used for
6052 all other user lookups.
6054 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6057 33. Exim was neglecting to zero errno before one call of strtol() when
6058 expanding a string and expecting an integer value. On some systems this
6059 resulted in spurious "integer overflow" errors. Also, it was casting the
6060 result into an int without checking.
6062 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6063 did not work. The code looks as if it has *never* worked, though it appears
6064 to have been documented since at least release 1.62. I have made it work.
6066 35. The "timeout_DNS" error in retry rules, also documented since at least
6067 1.62, also never worked. As it isn't clear exactly what this means, and
6068 clearly it isn't a major issue, I have abolished the feature by treating it
6069 as "timeout", and writing a warning to the main and panic logs.
6071 36. The display of retry rules for -brt wasn't always showing the error code
6074 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6075 timeout_connect_A, timeout_connect_MX.
6077 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6078 to the empty sender.
6080 39. The daemon was not analysing the content of -oX till after it had closed
6081 stderr and disconnected from the controlling terminal. This meant that any
6082 syntax errors were only noted on the panic log, and the return code from
6083 the command was 0. By re-arranging the code a little, I've made the
6084 decoding happen first, so such errors now appear on stderr, and the return
6085 code is 1. However, the actual setting up of the sockets still happens in
6086 the disconnected process, so errors there are still only recorded on the
6089 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6090 connections (as happens on some IP stacks) was logged at start up time as
6091 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6092 it from "IPv6 and IPv4", which means that two separate sockets are being
6095 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6096 says whether AF_INET or AF_INET6 was passed as an argument.
6098 42. Exiwhat output was messed up when time zones were included in log
6101 43. Exiwhat now gives more information about the daemon's listening ports,
6102 and whether -tls-on-connect was used.
6104 44. The "port" option of the smtp transport is now expanded.
6106 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6107 silently ignored. Now an error message is written to the main and panic
6110 46. There's a new ACL modifier called "logwrite" which writes to a log file
6111 as soon as it is encountered.
6113 47. Added $local_user_uid and $local_user_gid at routing time.
6115 48. Exim crashed when trying to verify a sender address that was being
6118 49. Exim was recognizing only a space character after ".include". It now also
6119 recognizes a tab character.
6121 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6122 extracting the relevant information from the specification. The man page no
6123 longer contains scrambled data for the -d option, and I've added a section
6124 at the front about calling Exim under different names.
6126 51. Added "extra_headers" argument to the "mail" command in filter files.
6128 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6131 53. Installed eximstats 1.29.
6133 54. Added transport_filter_timeout as a generic transport option.
6135 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6136 Cc: header lines. This was required by RFC 822, but it not required by RFC
6139 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6140 incoming messages that did not have them. Now it does so only if the
6141 message originates locally, that is, if there is no associated remote host
6142 address. When Resent- header lines are present, this applies to the Resent-
6143 lines rather than the non-Resent- lines.
6145 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6146 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6148 58. Messages for configuration errors now include the name of the main
6149 configuration file - useful now that there may be more than one file in a
6150 list (.included file names were always shown).
6152 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6153 for those rare installations that do not start the daemon as root or run it
6154 setuid root. I've cut out the call to initgroups() if the daemon is not
6157 60. The Exim user and group can now be bound into the binary as text strings
6158 that are looked up at the start of Exim's processing.
6160 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6162 62. Added $mailstore_basename variable.
6164 63. Installed patch to sieve.c from Michael Haardt.
6166 64. When Exim failed to open the panic log after failing to open the main log,
6167 the original message it was trying to log was written to stderr and debug
6168 output, but if they were not available (the usual case in production), it
6169 was lost. Now it is written to syslog before the two lines that record the
6170 failures to open the logs.
6172 65. Users' Exim filters run in subprocesses under the user's uid. It is
6173 possible for a "deliver" command or an alias in a "personal" command to
6174 provoke an address rewrite. If logging of address rewriting is configured,
6175 this fails because the process is not running as root or exim. There may be
6176 a better way of dealing with this, but for the moment (because 4.30 needs
6177 to be released), I have disabled address rewrite logging when running a
6178 filter in a non-root, non-exim process.
6184 1. The buildconfig auxiliary program wasn't quoting the value set for
6185 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6186 not defined. This bug was masked in 4.22 by the effect that was fixed in
6189 2. Some messages that were rejected after a message id was allocated were
6190 shown as "incomplete" by exigrep. It no longer does this for messages that
6191 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6193 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6194 have allow_domain_literals set, the ID did not get logged in the <= line.
6195 Domain literals are now always recognized in Message-ID: header lines.
6197 4. The first argument for a ${extract expansion item is the key name or field
6198 number. Leading and trailing spaces in this item were not being ignored,
6199 causing some misleading effects.
6201 5. When deliver_drop_privilege was set, single queue runner processes started
6202 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6203 same command in the process it spins off) were not dropping privilege.
6205 6. When the daemon running as "exim" started a queue runner, it always
6206 re-executed Exim in the spun-off process. This is a waste of effort when
6207 deliver_drop_privilege is set. The new process now just calls the
6208 queue-runner function directly.
6214 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6217 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6218 ignored. Though the use of -oP was forcing the writing of a pid file, it
6219 was always written to the default place.
6221 3. If the message "no IP address found for host xxxx" is generated during
6222 incoming verification, it is now followed by identification of the incoming
6223 connection (so you can more easily find what provoked it).
6225 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6227 5. Added some features to "harden" Exim a bit more against certain attacks:
6229 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6230 be put in Local/Makefile. This is like the never_users runtime option,
6231 but it cannot be overridden. The default setting is "root".
6233 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6234 prefix string with which any file named in a -C command line option
6237 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6238 is retained for -C and -D only if the caller of Exim is root. Without
6239 it, the exim user may also use -C and -D and retain privilege.
6241 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6242 command line option is disabled.
6244 6. Macro names set by the -D option must start with an upper case letter, just
6245 like macro names defined in the configuration file.
6247 7. Added "dereference=" facility to LDAP.
6249 8. Two instances of the typo "uknown" in the source files are fixed.
6251 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6252 the Configure-Makefile script screwed up while processing it.
6254 10. Incorporated PCRE 4.4.
6256 11. The SMTP synchronization check was not operating right at the start of an
6257 SMTP session. For example, it could not catch a HELO sent before the client
6258 waited for the greeting. There is now a check for outstanding input at the
6259 point when the greeting is written. Because of the duplex, asynchronous
6260 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6261 way, but not yet received, when the check is performed.
6263 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6264 on TCP/IP sockets, because this apparently causes some broken clients to
6267 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6268 unchanged) from the Cygwin maintainer.
6270 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6271 was supported instead of testing for mbx. Effectively a typo.
6273 15. The spa authenticator server code was not checking that the input it
6274 received was valid base64.
6276 16. The debug output line for the "set" modifier in ACLs was not showing the
6277 name of the variable that was being set.
6279 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6281 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6282 lookup on its own. Something else was needed to trigger the lookup. For
6283 example, a match in host_lookup or the need for a host name in a host list.
6284 Now, if $sender_host_name is referenced and the host name has not yet been
6285 looked up, a lookup is performed. If the lookup fails, the variable remains
6286 empty, and $host_lookup_failed is set to "1".
6288 19. Added "eqi" as a case-independent comparison operator.
6290 20. The saslauthd authentication condition could segfault if neither service
6291 nor realm was specified.
6293 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6294 error message that was logged was misleading, and incoming SMTP
6295 connections were dropped. The message is now more accurate, and temporary
6296 errors are given to SMTP connections.
6298 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6299 (except RSET) with a 421 error, until QUIT is received. However, it was
6300 failing to send a response to QUIT.
6302 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6303 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6304 ACL is now run after the helo_try_verify_hosts code.
6306 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6307 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6308 case-independent, so other case variants are also recognized). Apparently
6309 some systems use these upper case variants.
6311 25. If more than two messages were waiting for the same host, and a transport
6312 filter was specified for the transport, Exim sent two messages over the
6313 same TCP/IP connection, and then failed with "socket operation on non-
6314 socket" when it tried to send the third.
6316 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6318 27. The extern definition of crypt16() in expand.c was not being excluded when
6319 the OS had its own crypt16() function.
6321 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6322 as a preferred synonym for return_size_limit, both as an option and as an
6325 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6327 30. Changed the default configuration ACL to relax the local part checking rule
6328 for addresses that are not in any local domains. For these addresses,
6329 slashes and pipe symbols are allowed within local parts, but the sequence
6330 /../ is explicitly forbidden.
6332 31. SPA server authentication was not clearing the challenge buffer before
6335 32. log_message in a "warn" ACL statement was writing to the reject log as
6336 well as to the main log, which contradicts the documentation and doesn't
6337 seem right (because no rejection is happening). So I have stopped it.
6339 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6340 However, I am unable to do any testing of this.
6342 34. Fixed an infelicity in the appendfile transport. When checking directories
6343 for a mailbox, to see if any needed to be created, it was accidentally
6344 using path names with one or more superfluous leading slashes; tracing
6345 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6347 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6348 contents are added to the log line that is written for every discarded
6349 recipient. (Previously a log_message setting was ignored.)
6351 36. The ${quote: operator now quotes the string if it is empty.
6353 37. The install script runs exim in order to find its version number. If for
6354 some reason other than non-existence or emptiness, which it checks, it
6355 could not run './exim', it was installing it with an empty version number,
6356 i.e. as "exim-". This error state is now caught, and the installation is
6359 38. An argument was missing from the function that creates an error message
6360 when Exim fails to connect to the socket for saslauthd authentication.
6361 This could cause Exim to crash, or give a corrupted message.
6363 39. Added isip, isip4, and isip6 to ${if conditions.
6365 40. The ACL variables $acl_xx are now saved with the message, and can be
6366 accessed later in routers, transports, and filters.
6368 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6369 strings in the file are not string-expanded.
6371 42. If a MAIL command specified a SIZE value that was too large to fit into an
6372 int variable, the check against message_size_limit failed. Such values are
6373 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6374 day this will have to be increased, but I don't think I want to be around
6375 when emails are that large.
6382 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6383 iconv() is not standard in FreeBSD.
6385 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6386 IPv6 enabled. The observed symptom was a segmentation fault on return from
6387 the function os_common_find_running_interfaces() in src/os.c.
6389 3. In the check_special_case() function in daemon.c I had used "errno" as an
6390 argument name, which causes warnings on some systems. This was basically a
6391 typo, since it was named "eno" in the comments!
6393 4. The code that waits for the clock to tick (at a resolution of some fraction
6394 of a second) so as to ensure message-id uniqueness was always waiting for
6395 at least one whole tick, when it could have waited for less. [This is
6396 almost certainly not relevant at current processor speeds, where it is
6397 unlikely to ever wait at all. But we try to future-proof.]
6399 5. The function that sleeps for a time interval that includes fractions of a
6400 second contained a race. It did not block SIGALRM between setting the
6401 timer, and suspending (a couple of lines later). If the interval was short
6402 and the sigsuspend() was delayed until after it had expired, the suspension
6403 never ended. On busy systems this could lead to processes getting stuck for
6406 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6407 process, before it forks any delivery processes. The open lookup caching
6408 mechanism meant that the open file or database connection was passed into
6409 the delivery process. The problem was that delivery processes always tidy
6410 up cached lookup data. This could cause a problem for the next delivery
6411 process started by the queue runner, because the external queue runner
6412 process does not know about the closure. So the next delivery process
6413 still has data in the lookup cache. In the case of a file lookup, there was
6414 no problem because closing a file descriptor in a subprocess doesn't affect
6415 the parent. However, if the lookup was caching a connection to a database,
6416 the connection was closed, and the second delivery process was likely to
6417 see errors such as "PGSQL: query failed: server closed the connection
6418 unexpectedly". The problem has been fixed by closing all cached lookups
6419 in a queue runner before running a delivery process.
6421 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6422 seem to have the "const" qualifier which it has on other OS. I've
6425 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6426 *of the same type* (client or server) with the same public name that an
6427 error should be diagnosed.
6429 9. When Exim looked up a host name for an IP address, but failed to find the
6430 original IP address when looking up the host name (a safety check), it
6431 output the message "<ip address> does not match any IP for NULL", which was
6432 confusing, to say the least. The bug was that the host name should have
6433 appeared instead of "NULL".
6435 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6436 user that is built into the binary, and the -C or -D options is used, root
6437 privilege is dropped before the configuration file is read. In addition,
6438 logging is switched to stderr instead of the normal log files. If the
6439 configuration then re-defines the Exim user, the unprivileged environment
6440 is probably not what is expected, so Exim logs a panic warning message (but
6443 However, if deliver_drop_privilege is set, the unprivileged state may well
6444 be exactly what is intended, so the warning has been cut out in that case,
6445 and Exim is allowed to try to write to its normal log files.
6451 1. smtp_return_error_details was not giving details for temporary sender
6452 or receiver verification errors.
6454 2. Diagnose a configuration error if two authenticators have the same public
6457 3. Exim used not to create the message log file for a message until the first
6458 delivery attempt. This could be confusing when incoming messages were held
6459 for policy or load reasons. The message log file is now created at the time
6460 the message is received, and an initial "Received" line is written to it.
6462 4. The automatically generated man page for command line options had a minor
6463 bug that caused no ill effects; however, a more serious problem was that
6464 the procedure for building the man page automatically didn't always
6465 operate. Consequently, release 4.20 contains an out-of-date version. This
6466 shouldn't happen again.
6468 5. When building Exim with embedded Perl support, the script that builds the
6469 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6470 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6472 6. The freeze_tell option was not being used for messages that were frozen on
6473 arrival, either by an ACL or by local_scan().
6475 7. Added the smtp_incomplete_transaction log selector.
6477 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6478 was accepting AUTH without a new EHLO.
6480 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6481 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6484 10. Logging of TCP/IP connections (when configured) now happens in the main
6485 daemon process instead of the child process, so that the TCP/IP connection
6486 count is more accurate (but it can never be perfect).
6488 11. The use of "drop" in a nested ACL was not being handled correctly in the
6489 outer ACL. Now, if condition failure induced by the nested "drop" causes
6490 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6491 or "require"), the connection is dropped.
6493 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6494 that yield "discard" can now be used with an "accept" or a "discard" verb,
6495 but an error is generated for any others (because I can't see a useful way
6496 to define what should happen).
6498 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6499 are now processed in the same way as lines in the Exim configuration file.
6500 In particular, continuation lines are supported.
6502 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6504 15. Added -ti meaning -t -i.
6506 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6507 domains, and warn by logging if others are found.
6509 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6510 returned from GIFCONF when Exim is trying to find the list of interfaces on
6511 a host. The code in os.c has been modified to copy each ifreq to an aligned
6512 structure in all cases.
6514 Also, in some cases, the returned ifreq's were being copied to a 'struct
6515 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6516 means the last couple of bytes of an IPv6 address could be chopped if the
6517 ifreq contained only a normal sockaddr (14 bytes storage).
6519 18. Named domain lists were not supported in the hosts_treat_as_local option.
6520 An entry such as +xxxx was not recognized, and was treated as a literal
6523 19. Ensure that header lines added by a DATA ACL are included in the reject log
6524 if the ACL subsequently rejects the message.
6526 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6527 MD5 (which is deprecated).
6529 21. When testing a filter file using -bf, Exim was writing a message when it
6530 took the sender from a "From " line in the message, but it was not doing so
6531 when it took $return_path from a Return-Path: header line. It now does.
6533 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6534 with a valid header line field name (a series of printing characters
6535 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6537 23. Changed "disc" in the source to "disk" to conform to the documentation and
6538 the book and for uniformity.
6540 24. Ignore Sendmail's -Ooption=value command line item.
6542 25. When execve() failed while trying to run a command in a pipe transport,
6543 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6544 could be confused with a return value of 69 from the command itself. This
6545 has been changed to 127, the value the shell returns if it is asked to run
6546 a non-existent command. The wording for the related log line suggests a
6547 non-existent command as the problem.
6549 26. If received_header_text expands to an empty string, do not add a Received:
6550 header line to the message. (Well, it adds a token one on the spool, but
6551 marks it "old" so that it doesn't get used or transmitted.)
6553 27. Installed eximstats 1.28 (addition of -nt option).
6555 28. There was no check for failure on the call to getsockname() in the daemon
6556 code. This can fail if there is a shortage of resources on the system, with
6557 ENOMEM, for example. A temporary error is now given on failure.
6559 29. Contrary to the C standard, it seems that in some environments, the
6560 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6561 program. Exim now does this explicitly; it affects the formatting of
6562 timestamps using strftime().
6564 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6565 complaints. I've now initialized all the variables, to avoid this.
6567 32. Header lines added by a system filter were not being "seen" during
6568 transport-time rewrites.
6570 33. The info_callback() function passed to OpenSSL is set up with type void
6571 (*)(SSL *, int, int), as described somewhere. However, when calling the
6572 function (actually a macro) that sets it up, the type void(*)() is
6573 expected. I've put in a cast to prevent warnings from picky compilers.
6575 34. If a DNS black list lookup found a CNAME record, but there were no A
6576 records associated with the domain it pointed at, Exim crashed.
6578 35. If a DNS black list lookup returned more than one A record, Exim ignored
6579 all but the first. It now scans all returned addresses if a particular IP
6580 value is being sought. In this situation, the contents of the
6581 $dnslist_value variable are a list of all the addresses, separated by a
6584 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6585 to accept a host name and all its aliases if the forward lookup for any of
6586 them yielded the IP address of the incoming connection. Now it accepts only
6587 those names whose forward lookup yields the correct IP address. Any other
6588 names are discarded. This closes a loophole whereby a rogue DNS
6589 administrator could create reverse DNS records to break through a
6590 wildcarded host restriction in an ACL.
6592 37. If a user filter or a system filter that ran in a subprocess used any of
6593 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6594 the wrong values were passed to the pipe command ($thisaddress had the
6595 value of $0, $0 had the value of $1, etc). This bug was introduced by
6596 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6598 38. Improved the line breaking for long SMTP error messages from ACLs.
6599 Previously, if there was no break point between 40 and 75 characters, Exim
6600 left the rest of the message alone. Two changes have been made: (a) I've
6601 reduced the minimum length to 35 characters; (b) if it can't find a break
6602 point between 35 and 75 characters, it looks ahead and uses the first one
6603 that it finds. This may give the occasional overlong line, but at least the
6604 remaining text gets split now.
6606 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6607 file descriptors might be low, and that setting 1000 would always raise it.
6608 It turns out that in some environments, the limit is already over 1000 and
6609 that lowering it causes trouble. So now Exim takes care not to decrease it.
6611 40. When delivering a message, the value of $return_path is set to $sender_
6612 address at the start of routing (routers may change the value). By an
6613 oversight, this default was not being set up when an address was tested by
6614 -bt or -bv, which affected the outcome if any router or filter referred to
6617 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6618 log only when "message" or "log_message" are set. However, if one of the
6619 conditions was an address verification, or a call to a nested ACL, the
6620 messages generated by the underlying test were being passed through. This
6621 no longer happens. The underlying message is available in $acl_verify_
6622 message for both "message" and "log_message" expansions, so it can be
6623 passed through if needed.
6625 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6626 new expansion $bh_ to give the encoded byte string without charset
6627 translation. Translation happens only if iconv() is available; HAVE_ICONV
6628 indicates this at build time. HEADERS_CHARSET gives the charset to
6629 translate to; headers_charset can change it in the configuration, and
6630 "headers charset" can change it in an individual filter file.
6632 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6633 that creates RFC 2047 encoded "words" labels them as that charset instead
6634 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6635 expansion operator; (ii) when Exim creates a From: line for a local
6636 message; (iii) when a header line is rewritten to include a "phrase" part.
6638 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6639 buggy, causing it to skip the first lines of messages whose message ID
6640 ended in 'D'. This would not have bitten before Exim release 4.14, because
6641 message IDs were unlikely to end in 'D' before then. The effect was to have
6642 incorrect size information for certain domains.
6644 45. #include "config.h" was missing at the start of the crypt16.c module. This
6645 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6648 46. If there was a timeout during a "random" callout check, Exim treated it as
6649 a failure of the random address, and carried on sending RSET and the real
6650 address. If the delay was just some slowness somewhere, the response to the
6651 original RCPT would be taken as a response to RSET and so on, causing
6652 mayhem of various kinds.
6654 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6655 when I implemented it. It didn't allow for the fact that some option values
6656 may legitimately be negative (e.g. size_addition), and it didn't even do
6657 the right test for positive values.
6659 48. Domain names in DNS records are case-independent. Exim always looks them up
6660 in lower case. Some resolvers return domain names in exactly the case they
6661 appear in the zone file, that is, they may contain uppercase letters. Not
6662 all resolvers do this - some return always lower case. Exim was treating a
6663 change of case by a resolver as a change of domain, similar to a widening
6664 of a domain abbreviation. This triggered its re-routing code and so it was
6665 trying to route what was effectively the same domain again. This normally
6666 caused routing to fail (because the router wouldn't handle the domain
6667 twice). Now Exim checks for this case specially, and just changes the
6668 casing of the domain that it ultimately uses when it transmits the message
6671 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6674 50. If a filter generated a file delivery with a non-absolute name (possible if
6675 no home directory exists for the router), the forbid_file option was not
6678 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6679 the existing equality matching.
6681 52. Exim was using ints instead of ino_t variables in some places where it was
6682 dealing with inode numbers.
6684 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6685 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6686 called TMPDIR, and if it finds it is different, it changes its value.
6688 54. The smtp_printf() function is now made available to local_scan() so
6689 additional output lines can be written before returning. There is also an
6690 smtp_fflush() function to enable the detection of a dropped connection.
6691 The variables smtp_input and smtp_batched_input are exported to
6694 55. Changed the default runtime configuration: the message "Unknown user"
6695 has been removed from the ACL, and instead placed on the localuser router,
6696 using the cannot_route_message feature. This means that any verification
6697 failures that generate their own messages won't get overridden. Similarly,
6698 the "Unrouteable address" message that was in the ACL for unverifiable
6699 relay addresses has also been removed.
6701 56. Added hosts_avoid_esmtp to the smtp transport.
6703 57. The exicyclog script was not checking for the esoteric option
6704 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6705 will work only if exicyclog is run under the appropriate euid.
6707 58. Following a discussion on the list, the rules by which Exim recognises line
6708 endings on incoming messages have been changed. The -dropcr and drop_cr
6709 options are now no-ops, retained only for backwards compatibility. The
6710 following line terminators are recognized: LF CRLF CR. However, special
6711 processing applies to CR:
6713 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6714 nor a local message in the state where . is a terminator.
6716 (ii) If a bare CR is encountered in a header line, an extra space is added
6717 after the line terminator so as not to end the header. The reasoning
6718 behind this is that bare CRs in header lines are most likely either
6719 to be mistakes, or people trying to play silly games.
6721 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6722 was being incorrectly given as 18 bytes larger than it should have been.
6723 This is a VOB (very old bug).
6725 60. This may never have affected anything current, but just in case it has:
6726 When the local host is found other than at the start of a list of hosts,
6727 the local host, those with the same MX, and any that follow, are discarded.
6728 When the list in question was part of a longer list of hosts, the following
6729 hosts (not currently being processed) were also being discarded. This no
6730 longer happens. I'm not sure if this situation could ever has previously
6733 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6736 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6737 header. This is something that is recommended in a new Internet Draft, and
6738 is something that is documented as being done by Sendmail. There are two
6739 possible values. For messages generated by the autoreply transport, Exim
6742 Auto-Submitted: auto-replied
6744 whereas for all other generated messages (e.g. bounces) it adds
6746 Auto-Submitted: auto-generated
6748 63. The "personal" condition in filters now includes a test for the
6749 Auto-Submitted: header. If it contains the string "auto-" the message it
6750 not considered personal.
6752 64. Added rcpt_include_affixes as a generic transport option.
6754 65. Added queue_only_override (default true).
6756 66. Added the syslog_duplication option.
6758 67. If what should have been the first header line of a message consisted of
6759 a space followed by a colon, Exim was mis-interpreting it as a header line.
6760 It isn't of course - it is syntactically invalid and should therefore be
6761 treated as the start of the message body. The misbehaviour could have
6762 caused a number of strange effects, including loss of data in subsequent
6763 header lines, and spool format errors.
6765 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6766 client host had authenticated. This control can now be exercised by an ACL
6767 for more flexibility.
6769 69. By default, callouts do not happen when testing with -bh. There is now a
6770 variant, -bhc, which does actually run the callout code, including
6771 consulting and updating the callout cache.
6773 70. Added support for saslauthd authentication, courtesy of Alexander
6776 71. If statvfs() failed on the spool or log directories while checking their
6777 size for availability, Exim confusingly gave the error "space shortage".
6778 Furthermore, in debugging mode it crashed with a floating point exception.
6779 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6780 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6781 serious problem, Exim now writes to the main and panic logs when this
6782 happens, with details of the failure. It then refuses to accept the
6783 incoming message, giving the message "spool directory problem" or "log
6784 directory problem" with a 421 code for SMTP messages.
6786 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6787 0, 1, and 2 exist, because some OS complain for execs without them (see
6788 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6789 descriptors. However, the code omitted to check that the open succeeded,
6790 causing mysterious errors if for some reason the permissions on /dev/null
6791 got screwed. Now Exim writes a message to the main and panic logs, and
6792 bombs out if it can't open /dev/null.
6794 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6795 interact so that it is all more flexible. It is supposed to remain
6796 backwards compatible. Also added extra_local_interfaces.
6798 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6799 to bomb out with an assertion failure - to the client this appears as a
6800 connection drop. This problem occurs in the part of the code that was taken
6801 from the Samba project. Fortunately, the assertion is in a very simple
6802 function, so I have fixed this by reproducing the function inline in the
6803 one place where it is called, and arranging for authentication to fail
6804 instead of killing the process with assert().
6806 75. The SPA client code was not working when the server requested OEM rather
6807 than Unicode encoding.
6809 76. Added code to make require_files with a specific uid setting more usable in
6810 the case where statting the file as root fails - usually a non-root-mounted
6811 NFS file system. When this happens and the failure is EACCES, Exim now
6812 forks a subprocess and does the per-uid checking as the relevant uid.
6814 77. Added process_log_path.
6816 78. If log_file_path was not explicitly set, a setting of check_log_space or
6817 check_log_inodes was ignored.
6819 79. If a space check for the spool or log partitions fails, the incident is now
6820 logged. Of course, in the latter case the data may get lost...
6822 80. Added the %p formatting code to string_format() so that it can be used to
6823 print addresses in debug_print(). Adjusted all the address printing in the
6824 debugging in store.c to use %p rather than %d.
6826 81. There was a concern that a line of code in smtp_in.c could overflow a
6827 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6828 initially expressed, the concern was not well-founded, because trailing
6829 spaces are removed early. However, if the trailing spaces were followed by
6830 a NULL, they did not get removed, so the overflow was possible. Two fixes
6833 (a) I re-wrote the offending code in a cleaner fashion.
6834 (b) If an incoming SMTP command contains a NULL character, it is rejected
6837 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6838 runs initgroups(), so that if the Exim user is in any additional groups,
6839 they will be used during message reception.
6845 The change log for 4.20 and earlier releases has been archived.