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 cipher, Certificate or OCSP information.
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
37 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
38 than -2003 did; needs libidn2 in addition to libidn.
40 JH/02 The path option on a pipe transport is now expanded before use.
42 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
43 Patch provided by "Björn", documentation fix added too.
45 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
46 missing a wire-to-host endian conversion.
48 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
49 close after a BDAT command line could be taken as a following command,
50 giving a synch failure. Fix by only checking for synch immediately
51 before acknowledging the chunk.
53 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
54 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
55 macro. Patches provided by Josh Soref.
57 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
58 Previously we did not; the RFC seems ambiguous and VRFY is not listed
59 by IANA as a service extension. However, John Klensin suggests that we
62 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
63 the dkim code may be unix-mode line endings rather than smtp wire-format
64 CRLF, so prepend a CR to any bare LF.
66 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
67 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
69 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
70 the first were themselves being wrongly included in the feed into dkim
71 processing; with most chunk sizes in use this resulted in an incorrect
72 body hash calculated value.
74 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
75 DKIM signature block, for verification. Although advised against by
76 standards it is specifically not ruled illegal.
78 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
80 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
81 missing a body hash (the bh= tag).
83 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
84 It seems that HAProxy sends the Proxy Protocol information in clear and
85 only then does a TLS startup, so do the same.
87 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
88 TCP connections (such as for Spamd) unless the daemon successfully set
89 Fast Open mode on its listening sockets. This fixes breakage seen on
90 too-old kernels or those not configured for Fast Open, at the cost of
91 requiring both directions being enabled for TFO, and TFO never being used
92 by non-daemon-related Exim processes.
94 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
95 endings, at least on the first header line. Try to canonify any that get
96 past that check, despite the cost.
98 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
99 now limited to an arbitrary five deep, while parsing addresses with the
100 strip_excess_angle_brackets option enabled.
102 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
103 instead leave the unprompted TLS handshake in socket buffer for the
104 TLS library to consume.
106 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
108 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
110 JH/16 Drop variables when they go out of scope. Memory management drops a whole
111 region in one operation, for speed, and this leaves assigned pointers
112 dangling. Add checks run only under the testsuite which checks all
113 variables at a store-reset and panics on a dangling pointer; add code
114 explicitly nulling out all the variables discovered. Fixes one known
115 bug: a transport crash, where a dangling pointer for $sending_ip_address
116 originally assigned in a verify callout, is re-used.
118 PP/06 Drop '.' from @INC in various Perl scripts.
120 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
122 PP/08 Reduce a number of compilation warnings under clang; building with
123 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
124 should be warning-free.
126 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
128 HS/01 Fix portability problems introduced by PP/08 for platforms where
129 realloc(NULL) is not equivalent to malloc() [SunOS et al].
131 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
132 chunk. This allows us to accept broken chunked messages. We need a more
133 general solution here.
135 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
136 already-broken messages in the queue.
138 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
140 JH/19 Fix reference counting bug in routing-generated-address tracking.
146 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
147 supports it and a size is available (ie. the sending peer gave us one).
149 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
150 years of being deprecated). The replacements are the ACLs
151 acl_smtp_mime and acl_not_smtp_mime.
153 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
154 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
155 or even an in-clear connection were permitted. Now, if the host lookup
156 was dnssec and dane was requested then the host is only used if the
157 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
158 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
159 if one fails this test.
160 This means that a poorly-configured remote DNS will make it incommunicado;
161 but it protects against a DNS-interception attack on it.
163 JH/04 Bug 1810: make continued-use of an open smtp transport connection
164 non-noisy when a race steals the message being considered.
166 JH/05 If main configuration option tls_certificate is unset, generate a
167 self-signed certificate for inbound TLS connections.
169 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
170 in rewrites and routers.
172 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
173 and logged a warning sing 4.83; now they are a configuration file error.
175 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
176 (lacking @domain). Apply the same qualification processing as RCPT.
178 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
180 JH/10 Support ${sha256:} applied to a string (as well as the previous
183 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
184 a cutthrough deliver is pending, as we always want to make a connection.
185 This also avoids re-routing the message when later placing the cutthrough
186 connection after a verify cache hit.
187 Do not update it with the verify result either.
189 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
190 when routing results in more than one destination address.
192 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
193 signing (which inhibits the cutthrough capability). Previously only
194 the presence of an option was tested; now an expansion evaluating as
195 empty is permissible (obviously it should depend only on data available
196 when the cutthrough connection is made).
198 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
199 the relevant preceding SMTP command did not note the pipelining mode.
201 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
202 Previously they were not counted.
204 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
205 as one having no matching records. Previously we deferred the message
206 that needed the lookup.
208 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
209 distinguished as "(=".
211 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
212 for missing MX records. Previously it only worked for missing A records.
214 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
216 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
217 after the data-go-ahead and data-ack. Patch from Jason Betts.
219 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
220 even for a "none" policy. Patch from Tony Meyer.
222 JH/22 Fix continued use of a connection for further deliveries. If a port was
223 specified by a router, it must also match for the delivery to be
226 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
227 When one of the recipients of a message was unsuitable for the connection
228 (has no matching addresses), we lost track of needing to mark it
229 deferred. As a result mail would be lost.
231 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
233 JH/25 Decoding ACL controls is now done using a binary search; the source code
234 takes up less space and should be simpler to maintain. Merge the ACL
235 condition decode tables also, with similar effect.
237 JH/26 Fix problem with one_time used on a redirect router which returned the
238 parent address unchanged. A retry would see the parent address marked as
239 delivered, so not attempt the (identical) child. As a result mail would
242 JH/27 Fix a possible security hole, wherein a process operating with the Exim
243 UID can gain a root shell. Credit to http://www.halfdog.net/ for
244 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
247 JH/28 Enable {spool,log} filesystem space and inode checks as default.
248 Main config options check_{log,spool}_{inodes,space} are now
249 100 inodes, 10MB unless set otherwise in the configuration.
251 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
252 Previously it only applied to the main-section connection policy
255 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
257 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
258 by me. Added RFC7919 DH primes as an alternative.
260 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
261 are not in the system include path.
263 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
264 GnuTLS, when a session startup failed (eg because the client disconnected)
265 Exim did stdio operations after fclose. This was exposed by a recent
266 change which nulled out the file handle after the fclose.
268 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
269 signed directly by the cert-signing cert, rather than an intermediate
270 OCSP-signing cert. This is the model used by LetsEncrypt.
272 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
274 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
275 an incoming connection.
277 HS/02 Bug 1802: Do not half-close the connection after sending a request
280 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
281 fallback to "prime256v1".
283 JH/34 SECURITY: Use proper copy of DATA command in error message.
284 Could leak key material. Remotely exploitable. CVE-2016-9963.
290 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
291 and 3.4.4 - once the server is enabled to respond to an OCSP request
292 it does even when not requested, resulting in a stapling non-aware
293 client dropping the TLS connection.
295 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
296 support variable-length bit vectors. No functional change.
298 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
299 The I= interface field on outgoing lines is now after the H= remote
300 host field, same as incoming lines. There is a separate
301 outgoing_interface log selector which allows you to disable the
304 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
305 If not running log_selector +smtp_connection the mainlog would be held
306 open indefinitely after a "too many connections" event, including to a
307 deleted file after a log rotate. Leave the per net connection logging
308 leaving it open for efficiency as that will be quickly detected by the
309 check on the next write.
311 HS/01 Bug 1671: Fix post transport crash.
312 Processing the wait-<transport> messages could crash the delivery
313 process if the message IDs didn't exist for some reason. When
314 using 'split_spool_directory=yes' the construction of the spool
315 file name failed already, exposing the same netto behaviour.
317 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
318 mime_regex ACL conditions.
320 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
321 to DSN fail messages (bounces): remote IP, remote greeting, remote response
322 to HELO, local diagnostic string.
324 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
325 log line to debug. Even when configured with a tls authenticator many
326 client connections are expected to not authenticate in this way, so
327 an authenticate fail is not an error.
329 HS/02 Add the Exim version string to the process info. This way exiwhat
330 gives some more detail about the running daemon.
332 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
333 matter for fast-change records such as DNSBLs.
335 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
336 retry record, even if constant. There may be multiple transports with
337 different interface settings and the retry behaviour needs to be kept
340 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
342 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
344 JH/10 Bug 840: fix log_defer_output option of pipe transport
346 JH/11 Bug 830: use same host for all RCPTS of a message, even under
347 hosts_randomize. This matters a lot when combined with mua_wrapper.
349 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
350 ${quote_pgsql:<string>} operator.
352 JH/13 Bug 1708: avoid misaligned access in cached lookup.
354 JH/14 Change header file name for freeradius-client. Relevant if compiling
355 with Radius support; from the Gentoo tree and checked under Fedora.
357 JH/15 Bug 1712: Introduce $prdr_requested flag variable
359 JH/16 Bug 1714: Permit an empty string as expansion result for transport
360 option transport_filter, meaning no filtering.
362 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
364 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
365 defaults to "*" (all hosts). The variable is now available when not built
366 with TLS, default unset, mainly to enable keeping the testsuite sane.
367 If a server certificate is not supplied (via tls_certificate) an error is
368 logged, and clients will find TLS connections fail on startup. Presumably
369 they will retry in-clear.
370 Packagers of Exim are strongly encouraged to create a server certificate
371 at installation time.
373 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
374 with the $config_file variable.
376 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
377 in transport context, after the attempt, and per-recipient. The latter type
378 is per host attempted. The event data is the error message, and the errno
379 information encodes the lookup type (A vs. MX) used for the (first) host,
380 and the trailing two digits of the smtp 4xx response.
382 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
383 to write to mainlog (or rejectlog, paniclog) in the window between file
384 creation and permissions/ownership being changed. Particularly affects
385 installations where exicyclog is run as root, rather than exim user;
386 result is that the running daemon panics and dies.
388 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
390 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
391 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
392 "pri" and "weight". Note that the previous implicit priority given by the
393 list order is no longer honoured.
395 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
398 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
399 by defining SUPPORT_SOCKS.
401 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
402 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
403 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
404 variables are renamed to proxy_{local,external}_{address,port}.
406 JH/27 Move Internationalisation support from Experimental to mainline, enabled
407 for a build by defining SUPPORT_I18N
409 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
410 of the query string, and make ${quote_redis:} do that quoting.
412 JH/29 Move Events support from Experimental to mainline, enabled by default
413 and removable for a build by defining DISABLE_EVENT.
415 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
417 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
418 cached by the daemon.
420 JH/32 Move Redis support from Experimental to mainline, enabled for a build
421 by defining LOOKUP_REDIS. The libhiredis library is required.
423 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
424 keys are given for lookup.
426 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
427 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
428 only supported when built with TLS support. The PolarSSL SHA routines
429 are still used when the TLS library is too old for convenient support.
431 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
432 openssl_options), for security. OpenSSL forces this from version 1.1.0
433 server-side so match that on older versions.
435 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
436 allocation for $value could be released as the expansion processing
437 concluded, but leaving the global pointer active for it.
439 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
440 and to use the domains and local_parts ACL conditions.
442 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
443 incorrectly not doubled on cutthrough transmission, hence seen as a
444 body-termination at the receiving system - resulting in truncated mails.
445 Commonly the sender saw a TCP-level error, and retransmitted the message
446 via the normal store-and-forward channel. This could result in duplicates
447 received - but deduplicating mailstores were liable to retain only the
448 initial truncated version.
450 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
452 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
454 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
455 we're in there, support oversigning also; bug 1309.
457 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
459 HS/04 Add support for keep_environment and add_environment options.
461 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
462 either intentional arithmetic overflow during PRNG, or testing config-
465 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
466 delivery resulted in actual delivery. Cancel cutthrough before DATA
469 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
470 rejects a recipient: pass the reject to the originator.
472 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
473 Many were false-positives and ignorable, but it's worth fixing the
476 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
477 for the new environment-manipulation done at startup. Move the routines
478 from being local to tls.c to being global via the os.c file.
480 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
481 an extract embedded as result-arg for a map, the first arg for extract
482 is unavailable so we cannot tell if this is a numbered or keyed
483 extraction. Accept either.
489 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
492 JH/02 The smtp transport option "multi_domain" is now expanded.
494 JH/03 The smtp transport now requests PRDR by default, if the server offers
497 JH/04 Certificate name checking on server certificates, when exim is a client,
498 is now done by default. The transport option tls_verify_cert_hostnames
499 can be used to disable this per-host. The build option
500 EXPERIMENTAL_CERTNAMES is withdrawn.
502 JH/05 The value of the tls_verify_certificates smtp transport and main options
503 default to the word "system" to access the system default CA bundle.
504 For GnuTLS, only version 3.0.20 or later.
506 JH/06 Verification of the server certificate for a TLS connection is now tried
507 (but not required) by default. The verification status is now logged by
508 default, for both outbound TLS and client-certificate supplying inbound
511 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
514 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
515 Status Notification (bounce) messages are now MIME format per RFC 3464.
516 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
517 under the control of the dsn_advertise_hosts option, and routers may
518 have a dsn_lasthop option.
520 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
521 default, modifiable by a malware= option. The list separator for
522 the options can now be changed in the usual way. Bug 68.
524 JH/10 The smtp_receive_timeout main option is now expanded before use.
526 JH/11 The incoming_interface log option now also enables logging of the
527 local interface on delivery outgoing connections.
529 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
530 if the interface and destination host and port all match.
532 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
535 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
536 Patch from Andrew Lewis.
538 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
539 now supports optional time-restrictions, weighting, and priority
540 modifiers per server. Patch originally by <rommer@active.by>.
542 JH/16 The spamd_address main option now supports a mixed list of local
543 and remote servers. Remote servers can be IPv6 addresses, and
544 specify a port-range.
546 JH/17 Bug 68: The spamd_address main option now supports an optional
547 timeout value per server.
549 JH/18 Bug 1581: Router and transport options headers_add/remove can
550 now have the list separator specified.
552 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
555 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
558 JH/21 Support for the A6 type of dns record is withdrawn.
560 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
561 rather than the verbs used.
563 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
564 from 255 to 1024 chars.
566 JH/24 Verification callouts now attempt to use TLS by default.
568 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
569 are generic router options now. The defaults didn't change.
571 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
572 Original patch from Alexander Shikoff, worked over by JH.
574 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
575 files if ZCAT_COMMAND is not executable.
577 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
579 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
581 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
582 Normally benign, it bites when the pair was led to by a CNAME;
583 modern usage is to not canonicalize the domain to a CNAME target
584 (and we were inconsistent anyway for A-only vs AAAA+A).
586 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
588 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
589 when evaluating $sender_host_dnssec.
591 JH/31 Check the HELO verification lookup for DNSSEC, adding new
592 $sender_helo_dnssec variable.
594 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
596 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
598 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
600 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
601 documented as working, but never had. Support all but $spam_report.
603 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
604 added for tls authenticator.
606 HS/03 Add perl_taintmode main config option
612 TL/01 When running the test suite, the README says that variables such as
613 no_msglog_check are global and can be placed anywhere in a specific
614 test's script, however it was observed that placement needed to be near
615 the beginning for it to behave that way. Changed the runtest perl
616 script to read through the entire script once to detect and set these
617 variables, reset to the beginning of the script, and then run through
618 the script parsing/test process like normal.
620 TL/02 The BSD's have an arc4random API. One of the functions to induce
621 adding randomness was arc4random_stir(), but it has been removed in
622 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
623 function when detected.
625 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
626 cause callback expansion.
628 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
629 syntax errors in an expansion can be treated as a string instead of
630 logging or causing an error, due to the internal use of bool_lax
631 instead of bool when processing it.
633 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
634 server certificates when making smtp deliveries.
636 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
638 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
640 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
642 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
643 Merged patch from Sebastian Wiedenroth.
645 JH/05 Fix results-pipe from transport process. Several recipients, combined
646 with certificate use, exposed issues where response data items split
647 over buffer boundaries were not parsed properly. This eventually
648 resulted in duplicates being sent. This issue only became common enough
649 to notice due to the introduction of connection certificate information,
650 the item size being so much larger. Found and fixed by Wolfgang Breyha.
652 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
653 size buffer was used, resulting in syntax errors when an expansion
656 JH/07 Add support for directories of certificates when compiled with a GnuTLS
657 version 3.3.6 or later.
659 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
660 is EXPERIMENTAL_EVENT, the main-configuration and transport options
661 both become "event_action", the variables become $event_name, $event_data
662 and $event_defer_errno. There is a new variable $verify_mode, usable in
663 routers, transports and related events. The tls:cert event is now also
664 raised for inbound connections, if the main configuration event_action
667 TL/06 In test suite, disable OCSP for old versions of openssl which contained
668 early OCSP support, but no stapling (appears to be less than 1.0.0).
670 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
671 server certificate names available under the smtp transport option
672 "tls_verify_cert_hostname" now do not permit multi-component wildcard
675 JH/10 Time-related extraction expansions from certificates now use the main
676 option "timezone" setting for output formatting, and are consistent
677 between OpenSSL and GnuTLS compilations. Bug 1541.
679 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
680 encoded parameter in the incoming message. Bug 1558.
682 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
683 include certificate info, eximon was claiming there were spoolfile
686 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
688 JH/14 Log delivery-related information more consistently, using the sequence
689 "H=<name> [<ip>]" wherever possible.
691 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
692 are problematic for Debian distribution, omit them from the release
695 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
697 JH/16 Fix string representation of time values on 64bit time_t architectures.
700 JH/17 Fix a null-indirection in certextract expansions when a nondefault
701 output list separator was used.
706 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
707 checkers that were complaining about end of non-void function with no
710 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
711 This was a regression introduced in 4.83 by another bugfix.
713 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
715 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
716 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
722 TF/01 Correctly close the server side of TLS when forking for delivery.
724 When a message was received over SMTP with TLS, Exim failed to clear up
725 the incoming connection properly after forking off the child process to
726 deliver the message. In some situations the subsequent outgoing
727 delivery connection happened to have the same fd number as the incoming
728 connection previously had. Exim would try to use TLS and fail, logging
729 a "Bad file descriptor" error.
731 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
732 utilities have not been installed.
734 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
735 temporary space as the ACL may create new global variables.
737 TL/01 LDAP support uses per connection or global context settings, depending
738 upon the detected version of the libraries at build time.
740 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
741 to extract and use the src ip:port in logging and expansions as if it
742 were a direct connection from the outside internet. PPv2 support was
743 updated based on HAProxy spec change in May 2014.
745 JH/02 Add ${listextract {number}{list}{success}{fail}}.
747 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
748 Properly escape header and check for NULL return.
750 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
753 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
755 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
756 characters in header names, implemented as a verify condition.
757 Contributed by Michael Fischer v. Mollard.
759 TL/05 Rename SPF condition results err_perm and err_temp to standardized
760 results permerror and temperror. Previous values are deprecated but
761 still accepted. In a future release, err_perm and err_temp will be
762 completely removed, which will be a backward incompatibility if the
763 ACL tests for either of these two old results. Patch contributed by
764 user bes-internal on the mailing list.
766 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
768 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
769 selectors, in both main and reject logs.
771 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
774 JH/07 Add malware type "sock" for talking to simple daemon.
776 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
778 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
779 routers/transports under cutthrough routing.
781 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
782 numbers. Touch up "bool" conditional to keep the same definition.
784 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
786 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
788 JH/12 Expand items in router/transport headers_add or headers_remove lists
789 individually rather than the list as a whole. Bug 1452.
791 Required for reasonable handling of multiple headers_ options when
792 they may be empty; requires that headers_remove items with embedded
793 colons must have them doubled (or the list-separator changed).
795 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
796 view the policy declared in the DMARC record. Currently, $dmarc_status
797 is a combined value of both the record presence and the result of the
800 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
802 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
803 dnslookup router and the smtp transport (applying to the forward
806 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
807 of ldap servers used for a specific lookup. Patch provided by Heiko
810 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
811 New variable $lookup_dnssec_authenticated for observability.
813 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
814 Patch submitted by Lars Timman.
816 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
818 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
819 Requires trusted mode and valid format message id, aborts otherwise.
820 Patch contributed by Heiko Schlichting.
822 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
823 certextract with support for various fields. Bug 1358.
825 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
826 is requested by default, modifiable by smtp transport option
829 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
830 operate on certificate variables to give certificate fingerprints
831 Also new ${sha256:cert_variable}.
833 JH/23 The PRDR feature is moved from being Experimental into the mainline.
835 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
836 Christian Aistleitner.
838 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
840 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
841 file. Patch from Wolfgang Breyha.
843 JH/25 Expand the coverage of the delivery $host and $host_address to
844 client authenticators run in verify callout. Bug 1476.
846 JH/26 Port service names are now accepted for tls_on_connect_ports, to
847 align with daemon_smtp_ports. Bug 72.
849 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
850 support and error reporting did not work properly.
852 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
853 and is readable. Patch from Andrew Colin Kissa.
855 TL/14 Enhance documentation of ${run expansion and how it parses the
856 commandline after expansion, particularly in the case when an
857 unquoted variable expansion results in an empty value.
859 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
861 PP/02 Fix internal collision of T_APL on systems which support RFC3123
862 by renaming away from it. Addresses GH issue 15, reported by
865 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
867 TL/15 SECURITY: prevent double expansion in math comparison functions
868 (can expand unsanitized data). Not remotely exploitable.
875 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
877 PP/02 Make -n do something, by making it not do something.
878 When combined with -bP, the name of an option is not output.
880 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
883 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
884 $sender_host_name and config options to manage this, and basic check
887 PP/05 DSCP support for outbound connections and control modifier for inbound.
889 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
890 (Only plugin which currently uses this is kerberos4, which nobody should
891 be using, but we should make it available and other future plugins might
892 conceivably use it, even though it would break NAT; stuff *should* be
893 using channel bindings instead).
895 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
896 name; added for Sendmail compatibility; requires admin caller.
897 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
898 just ignore it); requires trusted caller.
899 Also parse but ignore: -Ac -Am -X<logfile>
902 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
904 TL/02 Add +smtp_confirmation as a default logging option.
906 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
907 Patch by Magnus Holmgren from 2007-02-20.
909 TL/04 Bugzilla 1281 - Spec typo.
910 Bugzilla 1283 - Spec typo.
911 Bugzilla 1290 - Spec grammar fixes.
913 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
915 TL/06 Add Experimental DMARC support using libopendmarc libraries.
917 TL/07 Fix an out of order global option causing a segfault. Reported to dev
918 mailing list by by Dmitry Isaikin.
920 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
922 JH/02 Support "G" suffix to numbers in ${if comparisons.
924 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
926 NM/01 Bugzilla 1197 - Spec typo
927 Bugzilla 1196 - Spec examples corrections
929 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
931 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
932 gnutls_enable_pkcs11, but renamed to more accurately indicate its
935 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
936 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
938 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
939 "acl {{name}{arg}...}", and optional args on acl condition
942 JH/05 Permit multiple router/transport headers_add/remove lines.
944 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
946 JH/07 Avoid using a waiting database for a single-message-only transport.
947 Performance patch from Paul Fisher. Bugzilla 1262.
949 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
952 JH/09 Add $headers_added variable, with content from use of ACL modifier
953 add_header (but not yet added to the message). Bugzilla 199.
955 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
956 Pulled from Bugzilla 817 by Wolfgang Breyha.
958 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
960 (nb: this is the same fix as in Exim 4.80.1)
962 JH/11 Add A= logging on delivery lines, and a client_set_id option on
965 JH/12 Add optional authenticated_sender logging to A= and a log_selector
968 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
970 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
971 advertise SMTP AUTH mechanism to us, instead of a generic
972 protocol violation error. Also, make Exim more robust to bad
973 data from the Dovecot auth socket.
975 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
977 When a queue runner is handling a message, Exim first routes the
978 recipient addresses, during which it prunes them based on the retry
979 hints database. After that it attempts to deliver the message to
980 any remaining recipients. It then updates the hints database using
983 So if a recipient address works intermittently, it can get repeatedly
984 deferred at routing time. The retry hints record remains fresh so the
985 address never reaches the final cutoff time.
987 This is a fairly common occurrence when a user is bumping up against
988 their storage quota. Exim had some logic in its local delivery code
989 to deal with this. However it did not apply to per-recipient defers
990 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
992 This change adds a proper retry rule check during routing so that the
993 final cutoff time is checked against the message's age. We only do
994 this check if there is an address retry record and there is not a
995 domain retry record; this implies that previous attempts to handle
996 the address had the retry_use_local_parts option turned on. We use
997 this as an approximation for the destination being like a local
998 delivery, as in LMTP.
1000 I suspect this new check makes the old local delivery cutoff check
1001 redundant, but I have not verified this so I left the code in place.
1003 TF/02 Correct gecos expansion when From: is a prefix of the username.
1005 Test 0254 submits a message to Exim with the header
1009 When I ran the test suite under the user fanf2, Exim expanded
1010 the header to contain my full name, whereas it should have added
1011 a Resent-Sender: header. It erroneously treats any prefix of the
1012 username as equal to the username.
1014 This change corrects that bug.
1016 GF/01 DCC debug and logging tidyup
1017 Error conditions log to paniclog rather than rejectlog.
1018 Debug lines prefixed by "DCC: " to remove any ambiguity.
1020 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1022 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1023 Bug spotted by Jeremy Harris; was flawed since initial commit.
1024 Would have resulted in OCSP responses post-SNI triggering an Exim
1025 NULL dereference and crash.
1027 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1029 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1030 Bug detection, analysis and fix by Samuel Thibault.
1031 Bugzilla 1331, Debian bug #698092.
1033 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1035 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1036 Server implementation by Todd Lyons, client by JH.
1037 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1038 config variable "prdr_enable" controls whether the server
1039 advertises the facility. If the client requests PRDR a new
1040 acl_data_smtp_prdr ACL is called once for each recipient, after
1041 the body content is received and before the acl_smtp_data ACL.
1042 The client is controlled by both of: a hosts_try_prdr option
1043 on the smtp transport, and the server advertisement.
1044 Default client logging of deliveries and rejections involving
1045 PRDR are flagged with the string "PRDR".
1047 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1048 fclose(). Diagnosis by Todd Lyons.
1050 PP/17 Update configure.default to handle IPv6 localhost better.
1051 Patch by Alain Williams (plus minor tweaks).
1054 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1055 This is now consistent with GnuTLS, and is now documented: the
1056 previous undocumented portable approach to treating the option as
1057 unset was to force an expansion failure. That still works, and
1058 an empty string is now equivalent.
1060 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1061 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1062 not performing validation itself.
1064 PP/20 Added force_command boolean option to pipe transport.
1065 Patch from Nick Koston, of cPanel Inc.
1067 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1070 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1072 PP/21 Fix eximon continuous updating with timestamped log-files.
1073 Broken in a format-string cleanup in 4.80, missed when I repaired the
1074 other false fix of the same issue.
1075 Report and fix from Heiko Schlichting.
1078 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1079 Report from Prashanth Katuri.
1081 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1082 It's SecureTransport, so affects any MacOS clients which use the
1083 system-integrated TLS libraries, including email clients.
1085 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1086 using a MIME ACL for non-SMTP local injection.
1087 Report and assistance in diagnosis by Warren Baker.
1089 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1091 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1093 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1094 last id that failed so it may be referenced in subsequent ACL's.
1096 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1099 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1100 ldap library initialization, allowing self-signed CA's to be
1101 used. Also properly sets require_cert option later in code by
1102 using NULL (global ldap config) instead of ldap handle (per
1103 session). Bug diagnosis and testing by alxgomz.
1105 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1106 the src/util/ subdirectory.
1108 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1109 renamed to Transport Post Delivery Action by Jeremy Harris, as
1112 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1113 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1114 redis_servers = needs to be configured which will be used by the redis
1115 lookup. Patch from Warren Baker, of The Packet Hub.
1117 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1119 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1120 hostname or reverse DNS when processing a host list. Used suggestions
1121 from multiple comments on this bug.
1123 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1125 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1126 Missed a few lines, added it to make the runtest require no keyboard
1129 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1130 contains upper case chars. Make router use caseful_local_part.
1132 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1133 support when GnuTLS has been built with p11-kit.
1139 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1141 This, or similar/improved, will also be change PP/11 of 4.82.
1147 PP/01 Handle short writes when writing local log-files.
1148 In practice, only affects FreeBSD (8 onwards).
1149 Bugzilla 1053, with thanks to Dmitry Isaikin.
1151 NM/01 Bugzilla 949 - Documentation tweak
1153 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1156 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1158 PP/02 Implemented gsasl authenticator.
1160 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1162 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1163 `pkg-config foo` for cflags/libs.
1165 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1166 with rest of GSASL and with heimdal_gssapi.
1168 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1169 `pkg-config foo` for cflags/libs for the TLS implementation.
1171 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1172 properties get this fed in as external SSF. A number of robustness
1173 and debugging improvements to the cyrus_sasl authenticator.
1175 PP/08 cyrus_sasl server now expands the server_realm option.
1177 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1178 Patch by Jeremy Harris.
1180 PP/10 Added dbmjz lookup type.
1182 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1184 PP/12 MAIL args handles TAB as well as SP, for better interop with
1185 non-compliant senders.
1186 Analysis and variant patch by Todd Lyons.
1188 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1189 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1190 Patch from Dirk Mueller <dmueller@suse.com>
1192 PP/13 tls_peerdn now print-escaped for spool files.
1193 Observed some $tls_peerdn in wild which contained \n, which resulted
1194 in spool file corruption.
1196 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1197 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1198 or write after TLS renegotiation, which otherwise led to messages
1201 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1202 as a tracking header (ie: a signed header comes before the signature).
1203 Patch from Wolfgang Breyha.
1205 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1206 comma-sep list; embedded commas doubled.
1208 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1210 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1212 Report and patch from Dmitry Banschikov.
1214 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1215 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1216 failures. We appear to now support TLS1.1+ with Exim.
1218 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1219 lets Exim select keys and certificates based upon TLS SNI from client.
1220 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1221 before an outbound SMTP session. New log_selector, +tls_sni.
1223 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1224 NULL dereference. Report and patch from Alun Jones.
1226 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1227 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1228 Not seeing resolver debug output on NetBSD, but suspect this is a
1229 resolver implementation change.
1231 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1232 Left warnings. Added "eximon gdb" invocation mode.
1234 PP/21 Defaulting "accept_8bitmime" to true, not false.
1236 PP/22 Added -bw for inetd wait mode support.
1238 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1239 locate the relevant includes and libraries. Made this the default.
1241 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1242 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1244 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1245 This may cause build issues on older platforms.
1247 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1248 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1249 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1250 Added SNI support via GnuTLS too.
1251 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1253 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1255 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1256 Applied second patch from Janne, implementing suggestion to default
1257 multiple-strings-in-record handling to match SPF spec.
1259 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1261 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1262 read-only, out of scope).
1263 Patch from Wolfgang Breyha, report from Stuart Northfield.
1265 PP/29 Fix three issues highlighted by clang analyser static analysis.
1266 Only crash-plausible issue would require the Cambridge-specific
1267 iplookup router and a misconfiguration.
1268 Report from Marcin Mirosław.
1270 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1272 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1273 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1274 As part of this, removing so much warning spew let me fix some minor
1275 real issues in debug logging.
1277 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1278 assignment on my part. Fixed.
1280 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1281 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1282 Janne Snabb (who went above and beyond: thank you).
1284 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1285 string otherwise requires a connection and a bunch more work and it's
1286 relatively easy to get wrong. Should also expose TLS library linkage
1289 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1290 64-bit ${eval} (JH/03).
1292 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1293 GNU libc to support some of the 64-bit stuff, should not lead to
1294 conflicts. Defined before os.h is pulled in, so if a given platform
1295 needs to override this, it can.
1297 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1298 protection layer was required, which is not implemented.
1299 Bugzilla 1254, patch from Wolfgang Breyha.
1301 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1302 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1303 tls_dhparam take prime identifiers. Also unbreak combination of
1304 OpenSSL+DH_params+TLSSNI.
1306 PP/39 Disable SSLv2 by default in OpenSSL support.
1312 PP/01 Solaris build fix for Oracle's LDAP libraries.
1313 Bugzilla 1109, patch from Stephen Usher.
1315 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1317 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1320 TF/02 Fix a couple more cases where we did not log the error message
1321 when unlink() failed. See also change 4.74-TF/03.
1323 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1324 lock up or crash if it happened to be inside a call to libc when it
1325 got a SIGUSR1 from exiwhat.
1327 The SIGUSR1 handler appends the current process status to the process
1328 log which is later printed by exiwhat. It used to use the general
1329 purpose logging code to do this, but several functions it calls are
1330 not safe for signals.
1332 The new output code in the SIGUSR1 handler is specific to the process
1333 log, and simple enough that it's easy to inspect for signal safety.
1334 Removing some special cases also simplifies the general logging code.
1335 Removing the spurious timestamps from the process log simplifies
1338 TF/04 Improved ratelimit ACL condition.
1340 The /noupdate option has been deprecated in favour of /readonly which
1341 has clearer semantics. The /leaky, /strict, and /readonly update modes
1342 are mutually exclusive. The update mode is no longer included in the
1343 database key; it just determines when the database is updated. (This
1344 means that when you upgrade Exim will forget old rate measurements.)
1346 Exim now checks that the per_* options are used with an update mode that
1347 makes sense for the current ACL. For example, when Exim is processing a
1348 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1349 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1350 must specify per_mail/readonly. If you omit the update mode it defaults to
1351 /leaky where that makes sense (as before) or /readonly where required.
1353 The /noupdate option is now undocumented but still supported for
1354 backwards compatibility. It is equivalent to /readonly except that in
1355 ACLs where /readonly is required you may specify /leaky/noupdate or
1356 /strict/noupdate which are treated the same as /readonly.
1358 A useful new feature is the /count= option. This is a generalization
1359 of the per_byte option, so that you can measure the throughput of other
1360 aggregate values. For example, the per_byte option is now equivalent
1361 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1363 The per_rcpt option has been generalized using the /count= mechanism
1364 (though it's more complicated than the per_byte equivalence). When it is
1365 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1366 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1367 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1368 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1369 Note that using per_rcpt with a non-readonly update mode in more than
1370 one ACL will cause the recipients to be double-counted. (The per_mail
1371 and per_byte options don't have this problem.)
1373 The handling of very low rates has changed slightly. If the computed rate
1374 is less than the event's count (usually one) then this event is the first
1375 after a long gap. In this case the rate is set to the same as this event's
1376 count, so that the first message of a spam run is counted properly.
1378 The major new feature is a mechanism for counting the rate of unique
1379 events. The new per_addr option counts the number of different
1380 recipients that someone has sent messages to in the last time period. It
1381 behaves like per_rcpt if all the recipient addresses are different, but
1382 duplicate recipient addresses do not increase the measured rate. Like
1383 the /count= option this is a general mechanism, so the per_addr option
1384 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1385 example, measure the rate that a client uses different sender addresses
1386 with the options per_mail/unique=$sender_address. There are further
1387 details in the main documentation.
1389 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1391 TF/06 Removed a few PCRE remnants.
1393 TF/07 Automatically extract Exim's version number from tags in the git
1394 repository when doing development or release builds.
1396 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1397 Bugzilla 879. Patch from Paul Fisher.
1399 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1400 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1403 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1405 PP/05 Variable $av_failed, true if the AV scanner deferred.
1406 Bugzilla 1078. Patch from John Horne.
1408 PP/06 Stop make process more reliably on build failure.
1409 Bugzilla 1087. Patch from Heiko Schlittermann.
1411 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1412 Bugzilla 1089. Patch from Heiko Schlittermann.
1414 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1415 Bugzilla 1131. Patch from Holger Weiß.
1417 PP/09 Handle IPv6 addresses with SPF.
1418 Bugzilla 860. Patch from Wolfgang Breyha.
1420 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1422 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1425 PP/11 match_* no longer expand right-hand-side by default.
1426 New compile-time build option, EXPAND_LISTMATCH_RHS.
1427 New expansion conditions, "inlist", "inlisti".
1429 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1431 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1433 PP/14 fix log_write() format string regression from TF/03.
1434 Bugzilla 1152. Patch from Dmitry Isaikin.
1440 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1442 PP/02 Harmonised TLS library version reporting; only show if debugging.
1443 Layout now matches that introduced for other libraries in 4.74 PP/03.
1445 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1447 PP/04 New "dns_use_edns0" global option.
1449 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1452 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1453 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1455 TK/01 Updated PolarSSL code to 0.14.2.
1456 Bugzilla 1097. Patch from Andreas Metzler.
1458 PP/07 Catch divide-by-zero in ${eval:...}.
1459 Fixes bugzilla 1102.
1461 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1464 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1465 format-string attack -- SECURITY: remote arbitrary code execution.
1467 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1468 time unintentionally subject to list matching rules, letting the header
1469 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1470 arbitrary string expansion). This allowed for information disclosure.
1472 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1473 INT_MIN/-1 -- value coerced to INT_MAX.
1479 NM/01 Workaround for PCRE version dependency in version reporting
1482 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1483 This fixes portability to compilers other than gcc, notably
1484 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1486 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1487 makefiles for portability to HP-UX and POSIX correctness.
1489 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1490 Also via indented variable definition in the Makefile.
1491 (Debugging by Oliver Heesakkers).
1493 PP/02 Restore caching of spamd results with expanded spamd_address.
1494 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1496 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1497 Improves build reliability. Fix from: Frank Elsner
1499 NM/02 Fix wide character breakage in the rfc2047 coding
1500 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1502 NM/03 Allow underscore in dnslist lookups
1503 Fixes bug 1026. Patch from Graeme Fowler
1505 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1506 Code patches from Adam Ciarcinski of NetBSD.
1508 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1511 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1512 is logged, to avoid truncation. Patch from John Horne.
1514 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1515 Patch from Jakob Hirsch.
1517 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1518 SQL string expansion failure details.
1519 Patch from Andrey Oktyabrski.
1521 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1522 Patch from Simon Arlott.
1524 PP/09 New lookups functionality failed to compile on old gcc which rejects
1525 extern declarations in function scope.
1526 Patch from Oliver Fleischmann
1528 PP/10 Use sig_atomic_t for flags set from signal handlers.
1529 Check getgroups() return and improve debugging.
1530 Fixed developed for diagnosis in bug 927 (which turned out to be
1533 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1534 Patch from Mark Zealey.
1536 PP/12 Bugzilla 1056: Improved spamd server selection.
1537 Patch from Mark Zealey.
1539 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1540 Based on patch from Heiko Schlittermann.
1542 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1543 Patch from Uwe Doering, sign-off by Michael Haardt.
1545 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1546 variable declaration deep within a block. Bug and patch from
1549 PP/15 lookups-Makefile IRIX compatibility coercion.
1551 PP/16 Make DISABLE_DKIM build knob functional.
1553 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1554 Patch by Simon Arlott
1556 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1557 variable argument macros. Our copy now differs from upstream.
1563 TF/01 Failure to get a lock on a hints database can have serious
1564 consequences so log it to the panic log.
1566 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1567 controlled using the smtp_confirmation log selector.
1569 TF/03 Include the error message when we fail to unlink a spool file.
1571 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1572 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1573 for maintaining out-of-tree patches for some time.
1575 PP/01 Bugzilla 139: Documentation and portability issues.
1576 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1577 Handle per-OS dynamic-module compilation flags.
1579 PP/02 Let /dev/null have normal permissions.
1580 The 4.73 fixes were a little too stringent and complained about the
1581 permissions on /dev/null. Exempt it from some checks.
1582 Reported by Andreas M. Kirchwitz.
1584 PP/03 Report version information for many libraries, including
1585 Exim version information for dynamically loaded libraries. Created
1586 version.h, now support a version extension string for distributors
1587 who patch heavily. Dynamic module ABI change.
1589 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1590 privilege escalation vulnerability whereby the Exim run-time user
1591 can cause root to append content of the attacker's choosing to
1594 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1597 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1598 If dropping privileges for untrusted macros, we disabled normal logging
1599 on the basis that it would fail; for the Exim run-time user, this is not
1600 the case, and it resulted in successful deliveries going unlogged.
1601 Fixed. Reported by Andreas Metzler.
1607 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1608 only prepend for the Resent-* case. Fixes regression introduced in
1609 Exim 4.70 by NM/22 for Bugzilla 607.
1611 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1612 increasing numbers of administrators be bitten by this.
1614 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1616 PP/03 Bugzilla 994: added openssl_options main configuration option.
1618 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1620 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1622 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1624 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1625 without a peer certificate, leading to a segfault because of an
1626 assumption that peers always have certificates. Be a little more
1627 paranoid. Problem reported by Martin Tscholak.
1629 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1630 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1631 NB: ClamAV planning to remove STREAM in "middle of 2010".
1632 CL also introduces -bmalware, various -d+acl logging additions and
1633 more caution in buffer sizes.
1635 PP/09 Implemented reverse_ip expansion operator.
1637 PP/10 Bugzilla 937: provide a "debug" ACL control.
1639 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1641 PP/12 Bugzilla 973: Implement --version.
1643 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1645 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1647 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1649 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1650 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1651 ignore trailing whitespace.
1653 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1655 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1658 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1659 Notification from Dr Andrew Aitchison.
1661 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1662 ExtendedDetectionInfo response format.
1663 Notification from John Horne.
1665 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1668 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1669 XSL and documented dependency on system catalogs, with examples of how
1672 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1675 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1676 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1677 configuration file which is writeable by the Exim user or group.
1679 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1680 of configuration files to cover files specified with the -C option if
1681 they are going to be used with root privileges, not just the default
1684 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1685 option (effectively making it always true).
1687 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1688 files to be used while preserving root privileges.
1690 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1691 that rogue child processes cannot use them.
1693 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1694 run-time user, instead of root.
1696 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1697 Exim run-time user without dropping privileges.
1699 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1700 result string, instead of calling string_vformat() twice with the same
1703 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1704 for other users. Others should always drop root privileges if they use
1705 -C on the command line, even for a whitelisted configure file.
1707 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1709 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1715 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1716 $data_path, and $header_path variables; fixed documentation bugs and
1719 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1720 exipick to access non-standard spools, including the "frozen" queue
1723 NM/01 Bugzilla 965: Support mysql stored procedures.
1724 Patch from Alain Williams
1726 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1728 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1729 Patch from Andreas Metzler
1731 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1732 Patch from Kirill Miazine
1734 NM/05 Bugzilla 671: Added umask to procmail example.
1736 JJ/03 installed exipick 20100323.0, fixing doc bug
1738 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1739 directory. Notification and patch from Dan Rosenberg.
1741 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1743 TK/02 Improve log output when DKIM signing operation fails.
1745 MH/01 Treat the transport option dkim_domain as a colon separated
1746 list, not as a single string, and sign the message with each element,
1747 omitting multiple occurences of the same signer.
1749 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1750 Bugzilla 985, 986. Patch by Simon Arlott
1752 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1753 Patch by Simon Arlott
1755 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1756 MBX locking. Notification from Dan Rosenberg.
1762 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1764 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1766 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1768 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1770 NM/04 Bugzilla 915: Flag broken perl installation during build.
1776 TK/01 Added patch by Johannes Berg that expands the main option
1777 "spamd_address" if it starts with a dollar sign.
1779 TK/02 Write list of recipients to X-Envelope-Sender header when building
1780 the mbox-format spool file for content scanning (suggested by Jakob
1783 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1784 (http://www.dcc-servers.net/) support via dccifd. Activated by
1785 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1787 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1788 by Mark Daniel Reidel <mr@df.eu>.
1790 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1791 When building exim an external PCRE library is now needed -
1792 PCRE is a system library on the majority of modern systems.
1793 See entry on PCRE_LIBS in EDITME file.
1795 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1796 conversation. Added nologin parameter to request.
1797 Patch contributed by Kirill Miazine.
1799 TF/01 Do not log submission mode rewrites if they do not change the address.
1801 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1803 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1804 log files in place. Contributed by Roberto Lima.
1806 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1808 TF/03 Bugzilla 615: When checking the local_parts router precondition
1809 after a local_part_suffix or local_part_prefix option, Exim now
1810 does not use the address's named list lookup cache, since this
1811 contains cached lookups for the whole local part.
1813 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1814 Robert Millan. Documentation is in experimental-spec.txt.
1816 TF/04 Bugzilla 668: Fix parallel build (make -j).
1818 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1820 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1821 Patch provided by Jan Srzednicki.
1823 TF/05 Leading white space used to be stripped from $spam_report which
1824 wrecked the formatting. Now it is preserved.
1826 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1827 that they are available at delivery time.
1829 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1831 TF/08 TLS error reporting now respects the incoming_interface and
1832 incoming_port log selectors.
1834 TF/09 Produce a more useful error message if an SMTP transport's hosts
1835 setting expands to an empty string.
1837 NM/06 Bugzilla 744: EXPN did not work under TLS.
1838 Patch provided by Phil Pennock.
1840 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1841 Patch provided by Richard Godbee.
1843 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1844 acl_smtp_notquit, added index entry.
1846 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1847 Patch provided by Eugene Bujak.
1849 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1850 accept(). Patch provided by Maxim Dounin.
1852 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1853 Patch provided by Phil Pennock.
1855 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1857 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1858 Patch provided by Brad "anomie" Jorsch.
1860 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1861 Patch provided by Dean Brooks.
1863 TK/05 Add native DKIM support (does not depend on external libraries).
1865 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1866 Patch provided by Graeme Fowler.
1868 NM/16 Bugzilla 851: Documentation example syntax fix.
1870 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1872 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1875 NM/19 Bugzilla 745: TLS version reporting.
1876 Patch provided by Phil Pennock.
1878 NM/20 Bugzilla 167: bool: condition support.
1879 Patch provided by Phil Pennock.
1881 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1882 clients. Patch provided by Phil Pennock.
1884 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1885 Patch provided by Brad "anomie" Jorsch.
1887 NM/23 Bugzilla 687: Fix misparses in eximstats.
1888 Patch provided by Heiko Schlittermann.
1890 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1891 Patch provided by Heiko Schlittermann.
1893 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1894 plus update to original patch.
1896 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1898 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1899 Patch provided by David Brownlee.
1901 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1903 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1905 NM/30 Bugzilla 888: TLS documentation bugfixes.
1907 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1909 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1910 Unlike the original bugzilla I have changed all shell scripts in src tree.
1912 NM/33 Bugzilla 898: Transport filter timeout fix.
1913 Patch by Todd Rinaldo.
1915 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1916 Patch by Serge Demonchaux.
1918 NM/35 Bugzilla 39: Base64 decode bug fixes.
1919 Patch by Jakob Hirsch.
1921 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1923 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1925 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1927 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1933 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1934 ALT-N's libdkim that I have put here:
1935 http://duncanthrax.net/exim-experimental/
1937 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1938 were called 'true' and it seems that C99 defines that as a reserved
1939 keyword to be used with 'bool' variable types. That means you could
1940 not include C99-style headers which use bools without triggering
1941 build errors in sieve.c.
1943 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1944 as mailq or other aliases. Changed the --help handling significantly
1945 to do whats expected. exim_usage() emits usage/help information.
1947 SC/01 Added the -bylocaldomain option to eximstats.
1949 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1951 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1953 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1959 PH/01 Another patch from the Sieve maintainer.
1961 PH/02 When an IPv6 address is converted to a string for single-key lookup
1962 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1963 dots are used instead of colons so that keys in lsearch files need not
1964 contain colons. This was done some time before quoting was made available
1965 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1966 (notated using the quote facility) so as to distinguish them from IPv4
1967 keys. This meant that lookups for IP addresses in host lists did not work
1968 for iplsearch lookups.
1970 This has been fixed by arranging for IPv6 addresses to be expressed with
1971 colons if the lookup type is iplsearch. This is not incompatible, because
1972 previously such lookups could never work.
1974 The situation is now rather anomalous, since one *can* have colons in
1975 ordinary lsearch keys. However, making the change in all cases is
1976 incompatible and would probably break a number of configurations.
1978 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1981 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1982 conversion specification without a maximum field width, thereby enabling
1983 a rogue spamd server to cause a buffer overflow. While nobody in their
1984 right mind would setup Exim to query an untrusted spamd server, an
1985 attacker that gains access to a server running spamd could potentially
1986 exploit this vulnerability to run arbitrary code as the Exim user.
1988 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1989 $primary_hostname instead of what libspf2 thinks the hosts name is.
1991 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1992 a directory entry by the name of the lookup key. Previously, if a
1993 symlink pointed to a non-existing file or a file in a directory that
1994 Exim lacked permissions to read, a lookup for a key matching that
1995 symlink would fail. Now it is enough that a matching directory entry
1996 exists, symlink or not. (Bugzilla 503.)
1998 PH/03 The body_linecount and body_zerocount variables are now exported in the
2001 PH/04 Added the $dnslist_matched variable.
2003 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2004 This means they are set thereafter only if the connection becomes
2007 PH/06 Added the client_condition to authenticators so that some can be skipped
2008 by clients under certain conditions.
2010 PH/07 The error message for a badly-placed control=no_multiline_responses left
2011 "_responses" off the end of the name.
2013 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2015 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2016 (without spaces) instead of just copying the configuration text.
2018 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2020 PH/11 Added $max_received_linelength.
2022 PH/12 Added +ignore_defer and +include_defer to host lists.
2024 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2025 way in which PCRE > 7.0 is built.
2027 PH/14 Implemented queue_only_load_latch.
2029 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2030 MAIL command. The effect was to mangle the value on 64-bit systems.
2032 PH/16 Another patch from the Sieve maintainer.
2034 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2036 PH/18 If a system quota error occurred while trying to create the file for
2037 a maildir delivery, the message "Mailbox is full" was not appended to the
2038 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2039 only to a quota excession during the actual writing of the file.
2041 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2042 characters?) which causes problems in log lines. The DN values are now
2043 passed through string_printing() before being added to log lines.
2045 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2046 and InterBase are left for another time.)
2048 PH/21 Added message_body_newlines option.
2050 PH/22 Guard against possible overflow in moan_check_errorcopy().
2052 PH/23 POSIX allows open() to be a macro; guard against that.
2054 PH/24 If the recipient of an error message contained an @ in the local part
2055 (suitably quoted, of course), incorrect values were put in $domain and
2056 $local_part during the evaluation of errors_copy.
2062 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2063 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2066 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2067 issue a MAIL command.
2069 PH/02 In an ACL statement such as
2071 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2073 if a client was not listed at all, or was listed with a value other than
2074 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2075 the condition was not true (as it should be), so access was not denied.
2076 The bug was that the ! inversion was incorrectly passed on to the second
2077 item. This has been fixed.
2079 PH/03 Added additional dnslists conditions == and =& which are different from
2080 = and & when the dns lookup returns more than one IP address.
2082 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2083 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2085 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2086 FSYNC, which compiles an option called disable_fsync that allows for
2087 bypassing fsync(). The documentation is heavily laced with warnings.
2089 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2091 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2092 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2093 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2094 including adding "make clean"; (3) Added -fPIC when compiling the test
2095 dynamically loaded module, to get rid of a warning.
2097 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2098 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2099 = 0s. The bug is otherwise harmless.
2101 PH/07 There was a bug in the dovecot authenticator such that the value of
2102 $auth1 could be overwritten, and so not correctly preserved, after a
2103 successful authentication. This usually meant that the value preserved by
2104 the server_setid option was incorrect.
2106 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2108 PH/09 Installed PCRE release 7.0.
2110 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2111 run for batched SMTP input. It is now run at the start of every message
2112 in the batch. While fixing this I discovered that the process information
2113 (output by running exiwhat) was not always getting set for -bs and -bS
2114 input. This is fixed, and it now also says "batched" for BSMTP.
2116 PH/11 Added control=no_pipelining.
2118 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2119 patch, slightly modified), and move the expansion of helo_data till after
2120 the connection is made in the smtp transport (so it can use these
2123 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2125 PH/14 Added log_selector = +pid.
2127 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2129 PH/16 Add ${if forany and ${if forall.
2131 PH/17 Added dsn_from option to vary the From: line in DSNs.
2133 PH/18 Flush SMTP output before performing a callout, unless control =
2134 no_callout_flush is set.
2136 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2137 was true (the default) a successful delivery failed to delete the retry
2138 item, thus causing premature timeout of the address. The bug is now
2141 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2143 PH/21 Long custom messages for fakedefer and fakereject are now split up
2144 into multiline responses in the same way that messages for "deny" and
2145 other ACL rejections are.
2147 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2148 with slight modification.
2150 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2151 draft, changing the syntax and factoring some duplicate code".
2153 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2154 for deliveries of the second and subsequent messages over the same SMTP
2157 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2158 ${reduce, with only minor "tidies".
2160 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2162 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2163 expansion side effects.
2165 PH/27 When a message times out after an over-quota error from an Exim-imposed
2166 quota, the bounce message says "mailbox is full". This message was not
2167 being given when it was a system quota that was exceeded. It now should
2170 MH/03 Made $recipients available in local_scan(). local_scan() already has
2171 better access to the recipient list through recipients_list[], but
2172 $recipients can be useful in postmaster-provided expansion strings.
2174 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2175 in the case of a MAIL command with additional options following the
2176 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2177 were accidentally chopped off.
2179 PH/29 SMTP synchronization checks are implemented when a command is read -
2180 there is a check that no more input is waiting when there shouldn't be
2181 any. However, for some commands, a delay in an ACL can mean that it is
2182 some time before the response is written. In this time, more input might
2183 arrive, invalidly. So now there are extra checks after an ACL has run for
2184 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2185 pipelining has not been advertised.
2187 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2189 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2190 returned in the SMTP response when smtp_return_error_details was false.
2191 This has been fixed.
2193 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2194 instead of the C I/O that was originally supplied, because problems were
2195 reported on Solaris.
2197 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2198 Exim which did not show up earlier: it was assuming that a call to
2199 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2200 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2201 was a macro that became an assignment, so it seemed to work. This has
2202 changed to a proper function call with a void return, hence the compile
2203 error. Exim's code has been fixed.
2205 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2208 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2210 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2212 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2213 because it turns out to be too big or there is a timeout) there is no
2214 "Completed" line in the log. When some messages of this type were
2215 selected by exigrep, they were listed as "not completed". Others were
2216 picked up by some special patterns. I have improved the selection
2217 criteria to be more general.
2219 PH/38 The host_find_failed option in the manualroute router can now be set
2220 to "ignore", to completely ignore a host whose IP address cannot be
2221 found. If all hosts are ignored, the behaviour is controlled by the new
2222 host_all_ignored option.
2224 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2225 homing or because of multiple MX records with /mx) generated more than
2226 one IP address, and the following item turned out to be the local host,
2227 all the secondary addresses of the first item were incorrectly removed
2228 from the list, along with the local host and any following hosts (which
2229 is what is supposed to happen).
2231 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2232 whoever called Exim into the -H file. In the case of the daemon it was
2233 behaving confusingly. When first started, it used values for whoever
2234 started the daemon, but after a SIGHUP it used the Exim user (because it
2235 calls itself on a restart). I have changed the code so that it now always
2238 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2239 message are rejected with the same error (e.g. no authentication or bad
2240 sender address), and a DATA command is nevertheless sent (as can happen
2241 with PIPELINING or a stupid MUA), the error message that was given to the
2242 RCPT commands is included in the rejection of the DATA command. This is
2243 intended to be helpful for MUAs that show only the final error to their
2246 PH/42 Another patch from the Sieve maintainer.
2248 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2249 Eximstats - Fixed some broken HTML links and added missing column headers
2251 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2252 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2254 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2256 PH/43 Yet another patch from the Sieve maintainer.
2258 PH/44 I found a way to check for a TCP/IP connection going away before sending
2259 the response to the final '.' that terminates a message, but only in the
2260 case where the client has not sent further data following the '.'
2261 (unfortunately, this is allowed). However, in many cases there won't be
2262 any further data because there won't be any more messages to send. A call
2263 to select() can be used: if it shows that the input is "ready", there is
2264 either input waiting, or the socket has been closed. An attempt to read
2265 the next input character can distinguish the two cases. Previously, Exim
2266 would have sent an OK response which the client would never have see.
2267 This could lead to message repetition. This fix should cure that, at
2268 least in a lot of common cases.
2270 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2271 advertised in response to EHLO.
2277 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2278 fixed by 4.65/MH/01 (is this a record?) are fixed:
2280 (i) An empty string was always treated as zero by the numeric comparison
2281 operators. This behaviour has been restored.
2283 (ii) It is documented that the numeric comparison operators always treat
2284 their arguments as decimal numbers. This was broken in that numbers
2285 starting with 0 were being interpreted as octal.
2287 While fixing these problems I realized that there was another issue that
2288 hadn't been noticed. Values of message_size_limit (both the global option
2289 and the transport option) were treated as octal if they started with 0.
2290 The documentation was vague. These values are now always treated as
2291 decimal, and I will make that clear in the documentation.
2297 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2298 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2301 MH/01 Don't check that the operands of numeric comparison operators are
2302 integers when their expansion is in "skipping" mode (fixes bug
2303 introduced by 4.64-PH/07).
2305 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2306 child addresses, Exim now panics and dies. Previously, because the count
2307 is held in a short int, deliveries were likely to be lost. As such a
2308 large number of recipients for a single message is ridiculous
2309 (performance will be very, very poor), I have chosen to impose a limit
2310 rather than extend the field.
2316 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2317 leftover -K file (the existence of which was triggered by #402).
2318 While we were at it, introduced process PID as part of the -K
2319 filename. This should rule out race conditions when creating
2322 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2323 processing considerably. Previous code took too long for large mails,
2324 triggering a timeout which in turn triggers #401.
2326 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2327 in the DK code in transports.c. sendfile() is not really portable,
2328 hence the _LINUX specificness.
2330 TF/01 In the add_headers option to the mail command in an Exim filter,
2331 there was a bug that Exim would claim a syntax error in any
2332 header after the first one which had an odd number of characters
2335 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2336 callout verification, Exim cached a "reject" for the entire domain. This
2337 is correct for most verifications, but it is not correct for a recipient
2338 verification with use_sender or use_postmaster set, because in that case
2339 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2340 case of MAIL FROM:<> rejection from other early rejections (e.g.
2341 rejection of HELO). When verifying a recipient using a non-null MAIL
2342 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2343 Whatever the result of the callout, the value of the domain cache is
2344 left unchanged (for any other kind of callout, getting as far as trying
2345 RCPT means that the domain itself is ok).
2347 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2350 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2351 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2352 session, the daemon ignores ECONNECT errors and logs others; it now
2353 ignores EPIPE as well.
2355 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2356 (quoted-printable decoding).
2358 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2359 later the small subsequent patch to fix an introduced bug.
2361 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2363 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2365 PH/08 An error is now given if message_size_limit is specified negative.
2367 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2368 to be given (somewhat) arbitrary names.
2370 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2373 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2374 miscellaneous code fixes
2376 PH/10 Added the log_reject_target ACL modifier to specify where to log
2379 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2380 hostname. This is wrong, because it relates to the incoming message (and
2381 probably the interface on which it is arriving) and not to the outgoing
2382 callout (which could be using a different interface). This has been
2383 changed to use the value of the helo_data option from the smtp transport
2384 instead - this is what is used when a message is actually being sent. If
2385 there is no remote transport (possible with a router that sets up host
2386 addresses), $smtp_active_hostname is used.
2388 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2389 tweaks were necessary in order to get it to work (see also 21 below):
2390 (a) The code assumed that strncpy() returns a negative number on buffer
2391 overflow, which isn't the case. Replaced with Exim's string_format()
2393 (b) There were several signed/unsigned issues. I just did the minimum
2394 hacking in of casts. There is scope for a larger refactoring.
2395 (c) The code used strcasecmp() which is not a standard C function.
2396 Replaced with Exim's strcmpic() function.
2397 (d) The code set only $1; it now sets $auth1 as well.
2398 (e) A simple test gave the error "authentication client didn't specify
2399 service in request". It would seem that Dovecot has changed its
2400 interface. Fortunately there's a specification; I followed it and
2401 changed what the client sends and it appears to be working now.
2403 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2406 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2407 address A is aliased to B and C, where B exists and C does not. Without
2408 -v the output is "A verified" because verification stops after a
2409 successful redirection if more than one address is generated. However,
2410 with -v the child addresses are also verified. Exim was outputting "A
2411 failed to verify" and then showing the successful verification for C,
2412 with its parentage. It now outputs "B failed to verify", showing B's
2413 parentage before showing the successful verification of C.
2415 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2416 look up a TXT record in a specific list after matching in a combined
2419 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2420 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2421 they consult the DNS. I had assumed they would set it the way they
2422 wanted; and indeed my experiments on Linux seem to show that in some
2423 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2424 To be on the safe side, however, I have now made the interface to
2425 host_find_byname() similar to host_find_bydns(), with an argument
2426 containing the DNS resolver options. The host_find_byname() function now
2427 sets these options at its start, just as host_find_bydns() does. The smtp
2428 transport options dns_qualify_single and dns_search_parents are passed to
2429 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2430 of host_find_byname() use the default settings of RES_DEFNAMES
2431 (qualify_single) but not RES_DNSRCH (search_parents).
2433 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2434 spool_read_header() do less string testing, by means of a preliminary
2435 switch on the second character of optional "-foo" lines. (This is
2436 overdue, caused by the large number of possibilities that now exist.
2437 Originally there were few.) While I was there, I also converted the
2438 str(n)cmp tests so they don't re-test the leading "-" and the first
2439 character, in the hope this might squeeze out yet more improvement.
2441 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2442 flag allowing group syntax was set by the header_syntax check but not
2443 turned off, possible causing trouble later; (2) The flag was not being
2444 set at all for the header_verify test, causing "group"-style headers to
2445 be rejected. I have now set it in this case, and also caused header_
2446 verify to ignore an empty address taken from a group. While doing this, I
2447 came across some other cases where the code for allowing group syntax
2448 while scanning a header line wasn't quite right (mostly, not resetting
2449 the flag correctly in the right place). These bugs could have caused
2450 trouble for malformed header lines. I hope it is now all correct.
2452 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2453 with the "reply" argument non-NULL. The code, however (which originally
2454 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2455 but it didn't always do it. This confused somebody who was copying the
2456 code for some other use. I have removed all the tests.
2458 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2459 feature that was used to support insecure browsers during the U.S. crypto
2460 embargo. It requires special client support, and Exim is probably the
2461 only MTA that supported it -- and would never use it because real RSA is
2462 always available. This code has been removed, because it had the bad
2463 effect of slowing Exim down by computing (never used) parameters for the
2464 RSA_EXPORT functionality.
2466 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2467 authenticator to fail if there's a tab character in the incoming data
2468 (there should never be unless someone is messing about, as it's supposed
2469 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2470 if the connection is using TLS or if the remote IP is the same as the
2471 local IP, and the "valid-client-cert option" if a client certificate has
2474 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2475 authenticators. This can be used for authorization after authentication
2476 succeeds. (In the case of plaintext, it servers for both authentication
2479 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2480 if any retry times were supplied.
2482 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2483 connection, where there is no client IP address to check. In this
2484 situation, the verify now always succeeds.
2486 PH/25 Applied John Jetmore's -Mset patch.
2488 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2490 PH/27 In a string expansion for a processed (not raw) header when multiple
2491 headers of the same name were present, leading whitespace was being
2492 removed from all of them, but trailing whitespace was being removed only
2493 from the last one. Now trailing whitespace is removed from each header
2494 before concatenation. Completely empty headers in a concatenation (as
2495 before) are ignored.
2497 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2498 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2500 PH/29 [Removed. This was a change that I later backed out, and forgot to
2501 correct the ChangeLog entry (that I had efficiently created) before
2502 committing the later change.]
2504 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2505 address errors (4xx response to RCPT) over the same connection as other
2506 messages routed to the same hosts. Such deliveries are always "forced",
2507 so retry times are not inspected. This resulted in far too many retries
2508 for the affected addresses. The effect occurred only when there were more
2509 hosts than the hosts_max_try setting in the smtp transport when it had
2510 the 4xx errors. Those hosts that it had tried were not added to the list
2511 of hosts for which the message was waiting, so if all were tried, there
2512 was no problem. Two fixes have been applied:
2514 (i) If there are any address or message errors in an SMTP delivery, none
2515 of the hosts (tried or untried) are now added to the list of hosts
2516 for which the message is waiting, so the message should not be a
2517 candidate for sending over the same connection that was used for a
2518 successful delivery of some other message. This seems entirely
2519 reasonable: after all the message is NOT "waiting for some host".
2520 This is so "obvious" that I'm not sure why it wasn't done
2521 previously. Hope I haven't missed anything, but it can't do any
2522 harm, as the worst effect is to miss an optimization.
2524 (ii) If, despite (i), such a delivery is accidentally attempted, the
2525 routing retry time is respected, so at least it doesn't keep
2526 hammering the server.
2528 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2529 in ${readsocket because some servers need this prod.
2531 PH/32 Added some extra debug output when updating a wait-xxx database.
2533 PH/33 The hint "could be header name not terminated by colon", which has been
2534 given for certain expansion errors for a long time, was not being given
2535 for the ${if def:h_colon_omitted{... case.
2537 PH/34 The spec says: "With one important exception, whenever a domain list is
2538 being scanned, $domain contains the subject domain." There was at least
2539 one case where this was not true.
2541 PH/35 The error "getsockname() failed: connection reset by peer" was being
2542 written to the panic log as well as the main log, but it isn't really
2543 panic-worthy as it just means the connection died rather early on. I have
2544 removed the panic log writing for the ECONNRESET error when getsockname()
2547 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2548 runs only) independently of the message's sender address. This meant
2549 that, if the 4xx error was in fact related to the sender, a different
2550 message to the same recipient with a different sender could confuse
2551 things. In particular, this can happen when sending to a greylisting
2552 server, but other circumstances could also provoke similar problems.
2553 I have changed the default so that the retry time for these errors is now
2554 based a combination of the sender and recipient addresses. This change
2555 can be overridden by setting address_retry_include_sender=false in the
2558 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2559 remote server are returned as part of bounce messages. This was not
2560 happening for LMTP over a pipe (the lmtp transport), but now it is the
2561 same for both kinds of LMTP.
2563 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2564 in header lines that were in fact CNAMEs. This is no longer the case.
2566 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2567 and queue runs started by the daemon processed all messages. This has
2568 been fixed so that -R and -S can now usefully be given with -q<time>.
2570 PH/40 Import PCRE release 6.7 (fixes some bugs).
2572 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2574 PH/42 Give an error if -q is specified more than once.
2576 PH/43 Renamed the variables $interface_address and $interface_port as
2577 $received_ip_address and $received_port, to make it clear that these
2578 values apply to message reception, and not to the outgoing interface when
2579 a message is delivered. (The old names remain recognized, of course.)
2581 PH/44 There was no timeout on the connect() call when using a Unix domain
2582 socket in the ${readsocket expansion. There now is.
2584 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2585 be meaningful with "accept".
2587 SC/01 Eximstats V1.43
2588 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2590 SC/02 Eximstats V1.44
2591 Use a glob alias rather than an array ref in the generated
2592 parser. This improves both readability and performance.
2594 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2595 Collect SpamAssassin and rejection statistics.
2596 Don't display local sender or destination tables unless
2597 there is data to show.
2598 Added average volumes into the top table text output.
2600 SC/04 Eximstats V1.46
2601 Collect data on the number of addresses (recipients)
2602 as well as the number of messages.
2604 SC/05 Eximstats V1.47
2605 Added 'Message too big' to the list of mail rejection
2606 reasons (thanks to Marco Gaiarin).
2608 SC/06 Eximstats V1.48
2609 Mainlog lines which have GMT offsets and are too short to
2610 have a flag are now skipped.
2612 SC/07 Eximstats V1.49 (Alain Williams)
2613 Added the -emptyok flag.
2615 SC/08 Eximstats V1.50
2616 Fixes for obtaining the IP address from reject messages.
2618 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2619 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2620 whitespace changes from 4.64-PH/27
2622 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2625 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2626 are found, allow negative numbers in numeric criteria)
2628 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2630 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2631 to match changes made in 4.64-PH/43
2633 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2635 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2636 "domain" is actually an IP address, and force a failure. This locks out
2637 those revolvers/nameservers that support "A-for-A" lookups, in
2638 contravention of the specifications.
2640 PH/48 When a host name was looked up from an IP address, and the subsequent
2641 forward lookup of the name timed out, the host name was left in
2642 $sender_host_name, contrary to the specification.
2644 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2645 restricted to single-key lookups, Exim was not diagnosing an error if
2646 * or *@ was used with a query-style lookup.
2648 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2650 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2651 long ago, but noone interested enough thought of it. Let's just say that
2652 the "1.1" means that there are some new functions that weren't there at
2653 some point in the past.
2655 PH/51 Error processing for expansion failure of helo_data from an smtp
2656 transport during callout processing was broken.
2658 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2659 tested/used via the -bh/-bhc/-bs options.
2661 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2662 bug, fixed in subsequent PCRE releases).
2664 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2665 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2667 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2673 SC/01 Use a glob alias rather than an array ref in eximstats generated
2674 parser. This improves both readability and performance.
2676 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2677 Don't display local sender or destination tables in eximstats unless
2678 there is data to show.
2679 Added average volumes into the eximstats top table text output.
2681 SC/03 Collect data on the number of addresses (recipients) as well
2682 as the number of messages in eximstats.
2684 TF/01 Correct an error in the documentation for the redirect router. Exim
2685 does (usually) call initgroups() when daemonizing.
2687 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2688 with consistent privilege compared to when running as a daemon.
2690 TF/03 Note in the spec that $authenticated_id is not set for local
2691 submissions from trusted users.
2693 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2694 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2696 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2697 by adding some example configuration directives to the default
2698 configuration file. A little bit of work is required to uncomment the
2699 directives and define how usernames and passwords are checked, but
2700 there is now a framework to start from.
2702 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2703 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2704 without this. I don't know how relevant this is to other LDAP libraries.
2706 PH/02 Add the verb name to the "unknown ACL verb" error.
2708 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2710 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2712 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2713 directory not expanded when it should be if an expanded home directory
2714 was set for the address (which is overridden by the transport).
2716 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2719 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2720 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2721 because it is too late at that time, and has no effect.
2723 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2724 security issue with \' (bugzilla #107). I could not use the
2725 PQescapeStringConn() function, because it needs a PGconn value as one of
2728 PH/08 When testing addresses using -bt, indicate those final addresses that
2729 are duplicates that would not cause an additional delivery. At least one
2730 person was confused, thinking that -bt output corresponded to deliveries.
2731 (Suppressing duplicates isn't a good idea as you lose the information
2732 about possibly different redirections that led to the duplicates.)
2734 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2735 systems where poll() doesn't work, in particular OS X.
2737 PH/10 Added more information to debugging output for retry time not reached.
2739 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2740 operations in malware.c.
2742 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2745 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2746 syslog_duplication set false, log lines that would normally be written
2747 both the the main log and to the reject log were not written to syslog at
2750 PH/14 In the default configuration, change the use of "message" in ACL warn
2751 statements to "add_header".
2753 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2754 not followed by a command (e.g. "seen endif").
2756 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2757 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2760 PH/17 Added extra conditions to the default value of delay_warning_condition
2764 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2765 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2766 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2769 The Auto-Submitted: and various List- headers are standardised, whereas I
2770 don't think Precedence: ever was.
2772 PH/18 Refactored debugging code in route_finduser() to show more information,
2773 in particular, the error code if getpwnam() issues one.
2775 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2776 This is apparently needed in addition to the PH/07 change above to avoid
2777 any possible encoding problems.
2779 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2780 but not after initializing Perl.
2782 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2783 output them only if debugging. By default they are written stderr,
2784 apparently, which is not desirable.
2786 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2789 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2792 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2794 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2795 authenticated or an ident call has been made. Suppress the default
2796 values for $authenticated_id and $authenticated_sender (but permit -oMai
2797 and -oMas) when testing with -bh.
2799 PH/24 Re-jigged the order of the tests in the default configuration so that the
2800 tests for valid domains and recipients precede the DNS black list and CSA
2801 tests, on the grounds that those ones are more expensive.
2803 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2804 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2805 This bug exists in every version of Exim that I still have, right back to
2808 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2809 However, an attempt to turn on case-sensitivity in a regex key by
2810 including (?-i) didn't work because the subject string was already
2811 lowercased, and the effects were non-intuitive. It turns out that a
2812 one-line patch can be used to allow (?-i) to work as expected.
2818 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2819 other effects) broke the use of negated acl sub-conditions.
2821 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2824 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2825 "Deny" causes Exim to reject the incoming connection with a 554 error.
2826 Unfortunately, if there is a major crisis, such as a disk failure,
2827 tcp-wrappers gives "deny", whereas what one would like would be some
2828 kind of temporary error. A kludge has been added to help with this.
2829 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2830 554 error is used if errno is still zero or contains ENOENT (which occurs
2831 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2834 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2836 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2837 errors. Otherwise a message that provokes a temporary error (when other
2838 messages do not) can cause a whole host to time out.
2840 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2841 addresses were routed directly to files or pipes from a redirect router.
2842 File deliveries just didn't batch; pipe deliveries might have suffered
2845 PH/06 A failure to get a lock for a hints database would erroneously always say
2846 "Failed to get write lock", even when it was really a read lock.
2848 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2849 of 0600. This has been changed to use the value of the lockfile_mode
2850 option (which defaults to 0600).
2852 PH/08 Applied small patch from the Sieve maintainer.
2854 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2855 folder from quota calculations, a direct delivery into this folder messed
2856 up the contents of the maildirsize file. This was because the regex was
2857 used only to exclude .Trash (or whatever) when the size of the mailbox
2858 was calculated. There was no check that a delivery was happening into an
2859 excluded directory. This bug has been fixed by ignoring all quota
2860 processing for deliveries into excluded directories.
2862 PH/10 Added the maildirfolder_create_regex option to appendfile.
2868 PH/01 The code for finding all the local interface addresses on a FreeBSD
2869 system running IPv6 was broken. This may well have applied to all BSD
2870 systems, as well as to others that have similar system calls. The broken
2871 code found IPv4 interfaces correctly, but gave incorrect values for the
2872 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2873 that it would not match correctly against @[] and not recognize the IPv6
2876 PH/02 The ipliteral router was not recognizing addresses of the form user@
2877 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2879 PH/03 Added disable_ipv6.
2881 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2882 decoded form, because it is most often used to construct To: headers
2883 lines in autoreplies, and the decoded form may well be syntactically
2884 invalid. However, $reply_address has leading white space removed, and all
2885 newlines turned into spaces so that the autoreply transport does not
2888 PH/05 If group was specified without a user on a router, and no group or user
2889 was specified on a transport, the group from the router was ignored.
2891 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2892 for visible compile-time settings that can be used to change these
2893 numbers, for those that want even more. Backwards compatibility with old
2894 spool files has been maintained. However, going back to a previous Exim
2895 release will lost any variables that are in spool files.
2897 PH/07 Two small changes when running in the test harness: increase delay when
2898 passing a TCP/IP connection to a new process, in case the original
2899 process has to generate a bounce, and remove special handling of
2900 127.0.0.2 (sic), which is no longer necessary.
2902 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2903 be the same on different OS.
2905 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2908 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2909 whether --show-vars was specified or not
2911 JJ/02 exipick: Added support for new ACL variable spool format introduced
2914 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2915 syntactically invalid From: or Reply-to: line, and a filter used this to
2916 generate an autoreply, and therefore failed to obtain an address for the
2917 autoreply, Exim could try to deliver to a non-existent relative file
2918 name, causing unrelated and misleading errors. What now happens is that
2919 it logs this as a hard delivery error, but does not attempt to create a
2922 PH/11 The exinext utility has a -C option for testing purposes, but although
2923 the given file was scanned by exinext itself; it wasn't being passed on
2924 when Exim was called.
2926 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2927 an end-of-file indication when reading a command response.
2929 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2930 compiled. In many other places in Exim, IPv6 addresses are always
2931 recognized, so I have changed this. It also means that IPv4 domain
2932 literals of the form [IPV4:n.n.n.n] are now always recognized.
2934 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2935 used if the router is not running as root, for example, when verifying at
2936 ACL time, or when using -bh. The debugging output from this situation was
2937 non-existent - all you got was a failure to exec. I have made two
2940 (a) Failures to set uid/gid, the current directory, or a process leader
2941 in a subprocess such as that created by queryprogram now generate
2942 suitable debugging output when -d is set.
2944 (b) The queryprogram router detects when it is not running as root,
2945 outputs suitable debugging information if -d is set, and then runs
2946 the subprocess without attempting to change uid/gid.
2948 PH/15 Minor change to Makefile for building test_host (undocumented testing
2951 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2952 additional section of a DNS packet that returns MX or SRV records.
2953 Instead, it always explicitly searches for A/AAAA records. This avoids
2954 major problems that occur when a DNS server includes only records of one
2955 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2956 fixed another bug: if SRV records were looked up and the corresponding
2957 address records were *not* found in the additional section, the port
2958 values from the SRV records were lost.
2960 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2961 using the correct key (the original address) when searching the retry
2962 rules in order to find which one to use for generating the retry hint.
2964 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2965 adding the default one. Similarly, if it contains a Reply-To: header, the
2966 errors_reply_to option, if set, is not used.
2968 PH/19 When calculating a retry time, Exim used to measure the "time since
2969 failure" by looking at the "first failed" field in the retry record. Now
2970 it does not use this if it is later than than the arrival time of the
2971 message. Instead it uses the arrival time. This makes for better
2972 behaviour in cases where some deliveries succeed, thus re-setting the
2973 "first failed" field. An example is a quota failure for a huge message
2974 when small messages continue to be delivered. Without this change, the
2975 "time since failure" will always be short, possible causing more frequent
2976 delivery attempts for the huge message than are intended.
2977 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2979 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2980 $1, $2, $3) because the numerical variables can be reset during some
2981 expansion items (e.g. "match"), thereby losing the authentication data.
2983 PH/21 Make -bV show the size of off_t variables so that the test suite can
2984 decide whether to run tests for quotas > 2G.
2986 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2987 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2988 filecount value is greater than 2G or if a quota value is greater than 2G
2989 on a system where the size of off_t is not greater than 4, a panic error
2992 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2993 never match. The debug and -bh output now contains an explicit error
2994 message indicating a malformed IPv4 address or mask.
2996 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2997 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2998 PH/23 above applies.
3000 PH/25 Do not write to syslog when running in the test harness. The only
3001 occasion when this arises is a failure to open the main or panic logs
3002 (for which there is an explicit test).
3004 PH/26 Added the /no_tell option to "control=freeze".
3006 PH/27 If a host name lookup failed very early in a connection, for example, if
3007 the IP address matched host_lookup and the reverse lookup yielded a name
3008 that did not have a forward lookup, an error message of the form "no IP
3009 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3010 could be logged. Now it outputs the IP address instead of "NULL".
3012 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3013 allows the sender and the authenticated sender to be set when
3014 submitting a message from within Exim. Since child_open_exim() is
3015 documented for local_scan(), the new function should be too.
3017 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3018 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3019 results in an empty string is now treated as unset.
3021 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3023 PH/31 Added sender_verify_fail logging option.
3025 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3026 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3027 tidied the source and removed it altogether.
3029 PH/33 When a queue run was abandoned because the load average was too high, a
3030 log line was always written; now it is written only if the queue_run log
3031 selector is set. In addition, the log line for abandonment now contains
3032 information about the queue run such as the pid. This is always present
3033 in "start" and "stop" lines but was omitted from the "abandon" line.
3035 PH/34 Omit spaces between a header name and the colon in the error message that
3036 is given when verify = headers_syntax fails (if there are lots of them,
3037 the message gets confusing).
3039 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3040 names, as there are now some PTR records that contain slashes. This check
3041 is only to protect against broken name servers that fall over on strange
3042 characters, so the fact that it applies to all lookups doesn't matter.
3044 PH/36 Now that the new test suite is complete, we can remove some of the
3045 special code in Exim that was needed for the old test suite. For example,
3046 sorting DNS records because real resolvers return them in an arbitrary
3047 order. The new test suite's fake resolver always returns records in the
3050 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3051 bounces) except when queue_only is set, to avoid logging races between
3052 the different processes.
3054 PH/38 Panic-die if .include specifies a non-absolute path.
3056 PH/39 A tweak to the "H" retry rule from its user.
3058 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3059 a label. They prevented compilation on older perls.
3061 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3062 a warning to be raised on newish perls.
3064 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3065 on queue. Changes to match documented behaviour of showing count of
3066 messages matching specified criteria.
3068 PH/40 Changed the default ident timeout from 30s to 5s.
3070 PH/41 Added support for the use of login_cap features, on those BSD systems
3071 that have them, for controlling the resources used by pipe deliveries.
3073 PH/42 The content-scanning code uses fopen() to create files in which to put
3074 message data. Previously it was not paying any attention to the mode of
3075 the files. Exim runs with umask(0) because the rest of the code creates
3076 files with open(), and sets the required mode explicitly. Thus, these
3077 files were ending up world-writeable. This was not a big issue, because,
3078 being within the spool directory, they were not world-accessible. I have
3079 created a function called modefopen, which takes an additional mode
3080 argument. It sets umask(777), creates the file, chmods it to the required
3081 mode, then resets the umask. All the relevant calls to fopen() in the
3082 content scanning code have been changed to use this function.
3084 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3085 to 24 hours. This avoids potential overflow problems when processing G
3086 and H retry rules. I suspect nobody ever tinkers with this value.
3088 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3090 PH/45 When the plaintext authenticator is running as a client, the server's
3091 challenges are checked to ensure they are valid base64 strings. By
3092 default, the authentication attempt is cancelled if an invalid string is
3093 received. Setting client_ignore_invalid_base64 true ignores these errors.
3094 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3095 they are received. Thus, the responses can be made to depend on the
3096 challenges. If an invalid string is ignored, an empty string is placed in
3099 PH/46 Messages that are created by the autoreply transport now contains a
3100 References: header, in accordance with RFCs 2822 and 3834.
3102 PH/47 Added authenticated_sender_force to the smtp transport.
3104 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3106 PH/49 Installed latest patch from the Sieve maintainer.
3108 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3109 was also set, the appendfile transport was unnecessarily scanning a
3110 directory of message files (e.g. for maildir delivery) to find the count
3111 of files (along with the size), even though it did not need this
3112 information. It now does the scan only if it needs to find either the
3113 size of the count of files.
3115 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3117 PH/52 Two bugs concerned with error handling when the smtp transport is
3120 (i) Exim was not creating retry information for temporary errors given
3121 for individual recipients after the DATA command when the smtp transport
3122 was used in LMTP mode. This meant that they could be retried too
3123 frequently, and not timed out correctly.
3125 (ii) Exim was setting the flag that allows error details to be returned
3126 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3127 recipients that were returned after the DATA command.
3129 PH/53 This is related to PH/52, but is more general: for any failing address,
3130 when detailed error information was permitted to be returned to the
3131 sender, but the error was temporary, then after the final timeout, only
3132 "retry timeout exceeded" was returned. Now it returns the full error as
3133 well as "retry timeout exceeded".
3135 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3136 do this, and (what is worse) MTAs that accept it.
3138 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3139 will now be deprecated.
3141 PH/56 New os.c-cygwin from the Cygwin maintainer.
3143 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3144 formats (previously only available in exim formats via -bpr, -bpru,
3145 and -bpra. Now also available in native and exiqgrep formats)
3147 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3148 with very large, slow to parse queues
3150 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3152 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3154 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3155 responses to authentication challenges, though it was showing the
3156 challenges; (ii) I've removed the CR characters from the debug output for
3159 PH/58 Allow for the insertion of a newline as well as a space when a string
3160 is turned into more than one encoded-word during RFC 2047 encoding. The
3161 Sieve code now uses this.
3163 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3164 data_4xx, lost_connection, tls_required.
3166 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3167 message was being sent as an SMTP response.
3169 PH/61 Add -l and -k options to exicyclog.
3171 PH/62 When verifying, if an address was redirected to one new address, so that
3172 verification continued, and the new address failed or deferred after
3173 having set something in $address_data, the value of $address_data was not
3174 passed back to the ACL. This was different to the case when no
3175 redirection occurred. The value is now passed back in both cases.
3177 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3178 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3179 use login_cap.h, so on its own it isn't the distinguishing feature. The
3180 new name refers directly to the setclassresources() function.
3182 PH/65 Added configuration files for NetBSD3.
3184 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3186 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3187 is preferred over IPv4.
3189 PH/68 The bounce_return_message and bounce_return_body options were not being
3190 honoured for bounces generated during the reception of non-SMTP messages.
3191 In particular, this applied to messages rejected by the ACL. This bug has
3192 been fixed. However, if bounce_return_message is true and bounce_return_
3193 body is false, the headers that are returned for a non-SMTP message
3194 include only those that have been read before the error was detected.
3195 (In the case of an ACL rejection, they have all been read.)
3197 PH/69 The HTML version of the specification is now built in a directory called
3198 spec_html instead of spec.html, because the latter looks like a path with
3199 a MIME-type, and this confuses some software.
3201 PH/70 Catch two compiler warnings in sieve.c.
3203 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3204 function verify_get_ident() calls ip_connect() to connect a socket, but
3205 if the "connect()" function timed out, ip_connect() used to close the
3206 socket. However, verify_get_ident() also closes the socket later, and in
3207 between Exim writes to the log, which may get opened at this point. When
3208 the socket was closed in ip_connect(), the log could get the same file
3209 descriptor number as the socket. This naturally causes chaos. The fix is
3210 not to close the socket in ip_connect(); the socket should be closed by
3211 the function that creates it. There was only one place in the code where
3212 this was missing, in the iplookup router, which I don't think anybody now
3213 uses, but I've fixed it anyway.
3215 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3216 well as to direct DNS lookups. Otherwise the handling of names in host
3217 lists is inconsistent and therefore confusing.
3223 PH/01 Two changes to the default runtime configuration:
3225 (1) Move the checks for relay_from_hosts and authenticated clients from
3226 after to before the (commented out) DNS black list checks.
3228 (2) Add control=submission to the relay_from_hosts and authenticated
3229 clients checks, on the grounds that messages accepted by these
3230 statements are most likely to be submissions.
3232 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3234 (1) Generate an error if the third argument for the ${prvs expansion is
3237 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3240 (3) Reset the variables that are obtained from the first argument of
3241 ${prvscheck and used in the second argument before leaving the code,
3242 because their memory is reclaimed, so using them afterwards may do
3245 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3246 one (it's much easier than Tom thought :-).
3248 (5) Because of (4), we can now allow for the use of $prvscheck_result
3249 inside the third argument.
3251 PH/03 For some reason, the default setting of PATH when running a command from
3252 a pipe transport was just "/usr/bin". I have changed it to
3255 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3256 anything to be listed in the output from -bV.
3258 PH/05 When a filter generated an autoreply, the entire To: header line was
3259 quoted in the delivery log line, like this:
3261 => >A.N.Other <ano@some.domain> <original@ddress> ...
3263 This has been changed so that it extracts the operative address. There
3264 may be more than one such address. If so, they are comma-separated, like
3267 => >ano@some.domain,ona@other.domain <original@ddress> ...
3269 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3270 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3271 not being looked up in the rDNS to get a host name, Exim was showing the
3272 IP address twice in Received: lines, even though the IP addresses were
3273 identical. For example:
3275 Received: from [1.2.3.4] (helo=[1.2.3.4])
3277 However, if the real host name was known, it was omitting the HELO data
3278 if it matched the actual IP address. This has been tidied up so that it
3279 doesn't show the same IP address twice.
3281 PH/07 When both +timestamp and +memory debugging was on, the value given by
3282 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3283 called by the debug printing, thereby overwriting the timestamp buffer.
3284 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3286 PH/08 When the original message was included in an autoreply transport, it
3287 always said "this is a copy of the message, including all the headers",
3288 even if body_only or headers_only was set. It now gives an appropriate
3291 PH/09 Applied a patch from the Sieve maintainer which:
3293 o fixes some comments
3294 o adds the (disabled) notify extension core
3295 o adds some debug output for the result of if/elsif tests
3296 o points to the current vacation draft in the documentation
3297 and documents the missing references header update
3301 o fixes a bug in processing the envelope test (when testing
3302 multiple envelope elements, the last element determined the
3305 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3306 Electronic Mail") by including:
3308 Auto-submitted: auto-generated
3310 in the messages that it generates (bounce messages and others, such as
3311 warnings). In the case of bounce messages for non-SMTP messages, there was
3312 also a typo: it was using "Auto_submitted" (underscore instead of
3313 hyphen). Since every message generated by Exim is necessarily in response
3314 to another message, thes have all been changed to:
3316 Auto-Submitted: auto-replied
3318 in accordance with these statements in the RFC:
3320 The auto-replied keyword:
3322 - SHOULD be used on messages sent in direct response to another
3323 message by an automatic process,
3325 - MUST NOT be used on manually-generated messages,
3327 - MAY be used on Delivery Status Notifications (DSNs) and Message
3328 Disposition Notifications (MDNs),
3330 - MUST NOT be used on messages generated by automatic or periodic
3331 processes, except for messages which are automatic responses to
3334 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3335 to the default Received: header definition.
3337 PH/12 Added log selector acl_warn_skipped (default on).
3339 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3340 variables because (a) they are in the wrong storage pool and (b) even if
3341 they were copied, it wouldn't work properly because of the caching.
3343 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3344 checking when decoding. Apparently there are clients that generate
3345 overlong encoded strings. Why am I not surprised?
3347 PH/15 If the first argument of "${if match_address" was not empty, but did not
3348 contain an "@" character, Exim crashed. Now it writes a panic log message
3349 and treats the condition as false.
3351 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3353 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3354 extension "envelope-auth". The code is finished and in agreement with
3355 other implementations, but there is no documentation so far and in fact,
3356 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3357 not changing the active code.
3359 Print executed "if" and "elsif" statements when debugging is used. This
3360 helps a great deal to understand what a filter does.
3362 Document more things not specified clearly in RFC3028. I had all this
3363 sorted out, when out of a sudden new issues came to my mind. Oops."
3365 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3368 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3369 canonical form (as documented). However, after a host name lookup from
3370 the IP address, check_host() was doing a simple string comparison with
3371 addresses acquired from the DNS when checking that the found name did
3372 have the original IP as one of its addresses. Since any found IPv6
3373 addresses are likely to be in abbreviated form, the comparison could
3374 fail. Luckily, there already exists a function for doing the comparison
3375 by converting both addresses to binary, so now that is used instead of
3376 the text comparison.
3378 PH/20 There was another similar case to PH/19, when a complete host name was
3379 given in a host list; looking up its IP address could give an abbreviated
3380 form, whereas the current host's name might or might not be abbreviated.
3381 The same fix has been applied.
3387 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3388 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3391 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3392 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3394 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3396 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3397 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3398 Also, for "Auto-Submitted", treat anything other than "no" as
3399 non-personal, in accordance with RFC 3834. (Previously it treated
3400 anything starting "auto-" as non-personal.)
3402 TF/01 The control=submission/name=... option had a problem with syntax
3403 errors if the name included a slash character. The /name= option
3404 now slurps the rest of the string, so it can include any characters
3405 but it must come last in the list of options (after /sender_retain
3408 PH/05 Some modifications to the interface to the fake nameserver for the new
3416 TK/01 Added the "success_on_redirect" address verification option. See
3417 NewStuff for rationale and an example.
3419 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3421 PH/02 Patch to exigrep to allow it to work on syslog lines.
3423 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3424 fread() to skip over the body file's header line, because in Cygwin the
3425 header line is locked and is inaccessible.
3427 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3428 co-exist for some time) to make it clear that it is the Exim ID that is
3429 referenced, not the Message-ID: header line.
3431 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3432 string_format() function, because snprintf() does not exist on all
3435 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3436 ${stat: expansion item.
3438 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3439 protocol synchronization error", to keep the pedants happy.
3441 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3442 well as for IRIX systems, when gcc is being used. See the host.c source
3445 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3447 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3450 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3451 transport and to the smtp transport in LMTP mode.
3453 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3455 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3456 run a filter in a subprocess. This could lead to confusion in subsequent
3457 lookups in the parent process. There should also be a search_tidyup() at
3458 the end of the subprocess.
3460 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3461 only if the host matched helo_try_verify_hosts, which caused the
3462 verification to occur when the EHLO/HELO command was issued. The ACL just
3463 tested the remembered result. Now, if a previous verification attempt has
3464 not happened, "verify = helo" does it there and then.
3466 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3468 TK/03 Fix log output including CR from clamd.
3470 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3471 exist provoked a memory error which could cause a segfault.
3473 PH/15 Installed PCRE 6.2
3475 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3477 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3478 of the problem. Specifically, suggested +O2 rather than +O1 for the
3481 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3483 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3484 record, the port was not correctly logged when the outgoing_port log
3485 selector was set (it logged the transort's default port).
3487 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3488 fallback_hosts, and "hosts" in the smtp transport.
3490 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3491 host errors such as "Connection refused".
3493 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3494 authentication with radiusclient 0.4.9:
3496 - Error returned from rc_read_config was caught wrongly
3497 - Username/password not passed on to radius server due to wrong length.
3499 The presumption is that some radiusclient API changes for 4.51/PH/17
3500 were not taken care of correctly. The code is still untested by me (my
3501 Linux distribution still has 0.3.2 of radiusclient), but it was
3502 contributed by a Radius user.
3504 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3505 expanding the "port" option of the smtp transport.
3507 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3508 while reading a MIME header. Thanks to Tom Hughes for a patch.
3510 PH/24 Include config.h inside local_scan.h so that configuration settings are
3513 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3514 that in an ACL for RCPT (for example), you can examine exactly what was
3517 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3518 commands, but it was not correctly comparing the address with the actual
3519 client host address. Thus, it would show the EHLO address in Received:
3520 header lines when this was not necessary.
3522 PH/27 Added the % operator to ${eval:}.
3524 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3525 it should be ignoring failures (because with -C, for example, it has lost
3526 privilege). It wasn't ignoring creation failures other than "already
3529 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3532 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3533 because an input line was too long, either on its own, or by virtue of
3534 too many continuations, the temporary file was not being removed, and the
3535 return code was incorrect.
3537 PH/31 Missing "BOOL" in function definition in filtertest.c.
3539 PH/32 Applied Sieve patches from the maintainer.
3541 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3543 PH/33 Added "verify = not_blind".
3545 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3546 Local/Makefile (with some defaults set). These are used in built scripts
3547 such as exicyclog, but they have never been used in the exim_install
3548 script (though there are many overriding facilities there). I have
3549 arranged that the exim_install script now takes note of these two
3552 PH/35 Installed configuration files for Dragonfly.
3554 PH/36 When a locally submitted message by a trusted user did not contain a
3555 From: header, and the sender address was obtained from -f or from an SMTP
3556 MAIL command, and the trusted user did not use -F to supply a sender
3557 name, $originator_name was incorrectly used when constructing a From:
3558 header. Furthermore, $originator_name was used for submission mode
3559 messages from external hosts without From: headers in a similar way,
3560 which is clearly wrong.
3562 PH/37 Added control=suppress_local_fixups.
3564 PH/38 When log_selector = +received_sender was set, and the addition of the
3565 sender made the log line's construction buffer exactly full, or one byte
3566 less than full, an overflow happened when the terminating "\n" was
3569 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3570 when the result of a list match is failure because a DNS lookup failed.
3572 PH/40 RM_COMMAND is now used in the building process.
3574 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3575 the "build-* directories that it finds.
3577 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3578 address in a domain literal was a prefix of an interface address.
3580 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3581 when verifying a sender address, unless rewrite_headers is false.
3583 PH/44 Wrote a long comment about why errors_to addresses are verified as
3584 recipients, not senders.
3586 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3587 the ratelimit ACL was added.
3589 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3591 PH/46 Added extra information about PostgreSQL errors to the error string.
3593 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3594 avoiding the need to install special zones in a real server. This is
3595 backwards compatible; if it can't find the fake resolver, it drops back.
3596 Thus, both old and new test suites can be run.
3598 TF/02 Added util/ratelimit.pl
3600 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3601 clock is set back in time.
3603 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3604 Candler <B.Candler@pobox.com>.
3606 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3607 OK for addresses that are the result of redirections.
3609 PH/48 A number of further additions for the benefit of the new test suite,
3610 including a fake gethostbyname() that interfaces to the fake DNS resolver
3613 TF/06 The fix for widen_domains has also been applied to qualify_single and
3614 search_parents which are the other dnslookup options that can cause
3617 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3620 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3622 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3623 service" if it encountered an MX record with an empty target hostname.
3624 The message is now "an MX or SRV record indicated no SMTP service".
3626 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3627 if the DNS of the sending site is misconfigured. This is quite a
3628 common situation. This change restores the behaviour of treating a
3629 helo verification defer as a failure.
3631 PH/51 If self=fail was set on a router, the bounce message did not include the
3632 actual error message.
3638 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3640 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3641 command itself ended in error, the underlying message about the transport
3642 filter timeout was being overwritten with the pipe command error. Now the
3643 underlying error message should be appended to the second error message.
3645 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3647 PH/02 Reduce the amount of output that "make" produces by default. Full output
3648 can still be requested.
3650 PH/03 The warning log line about a condition test deferring for a "warn" verb
3651 was being output only once per connection, rather than after each
3652 occurrence (because it was using the same function as for successful
3653 "warn" verbs). This seems wrong, so I have changed it.
3655 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3656 it should not have, which might have caused a crash in the right
3657 circumstances, but probably never did.
3659 PH/04 Installed a modified version of Tony Finch's patch to make submission
3660 mode fix the return path as well as the Sender: header line, and to
3661 add a /name= option so that you can make the user's friendly name appear
3664 TF/03 Added the control = fakedefer ACL modifier.
3666 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3667 Mark Lowes for thorough testing.
3669 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3671 TK/03 Merged latest SRS patch from Miles Wilton.
3673 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3674 with the definition in sysexits.h (which is #included earlier).
3675 Fortunately, Exim does not actually use EX_OK. The code used to try to
3676 preserve the sysexits.h value, by assuming that macro definitions were
3677 scanned for macro replacements. I have been disabused of this notion,
3678 so now the code just undefines EX_OK before #including unistd.h.
3680 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3681 in the smtp transport. When a block could not be written in a single
3682 write() function, the timeout was being re-applied to each part-write.
3683 This seems wrong - if the receiver was accepting one byte at a time it
3684 would take for ever. The timeout is now adjusted when this happens. It
3685 doesn't have to be particularly precise.
3687 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3688 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3690 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3691 without a domain if the check to <postmaster@domain> fails.
3693 SC/01 Eximstats: added -xls and the ability to specify output files
3694 (patch written by Frank Heydlauf).
3696 SC/02 Eximstats: use FileHandles for outputting results.
3698 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3700 SC/04 Eximstats: fixed display of large numbers with -nvr option
3702 SC/05 Eximstats: fixed merging of reports with empty tables.
3704 SC/06 Eximstats: added the -include_original_destination flag
3706 SC/07 Eximstats: removed tabs and trailing whitespace.
3708 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3710 TK/06 MBOX spool code: Add real "From " MBOX separator line
3711 so the .eml file is really in mbox format (even though
3712 most programs do not really care). Patch from Alex Miller.
3714 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3715 The latter is generated from $received_to and is only set if the
3716 message has one envelope recipient. SA can use these headers,
3717 obviously out-of-the-box. Patch from Alex Miller.
3719 PH/08 The ${def test on a variable was returning false if the variable's
3720 value was "0", contrary to what the specification has always said!
3721 The result should be true unless the variable is empty.
3723 PH/09 The syntax error of a character other than { following "${if
3724 def:variable_name" (after optional whitespace) was not being diagnosed.
3725 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3726 accidental colon was present, for example, could give incorrect results.
3728 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3729 result is used (not including appendfile, where other changes are about
3732 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3733 This involved changing a lot of size variables from int to off_t. It
3734 should work with maildirs and everything.
3736 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3737 spamd dying while we are connected to it.
3739 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3742 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3743 function for BDB 4.3.
3745 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3747 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3748 the sources for PCRE have been re-arranged and more files are now
3751 PH/15 The code I had for printing potentially long long variables in PH/11
3752 above was not the best (it lost precision). The length of off_t variables
3753 is now inspected at build time, and an appropriate printing format (%ld
3754 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3755 to be "long long int" or "long int". This is needed for the internal
3756 formatting function string_vformat().
3758 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3759 the configuration file to be ":syslog", then the script "guesses" where
3760 the logs files are, rather than using the compiled in default. In our
3761 case the guess is not the same as the compiled default, so the script
3762 suddenly stopped working when I started to use syslog. The patch checks
3763 to see if log_file_path is "". If so, it attempts to read it from exim
3764 with no configuration file to get the compiled in version, before it
3765 falls back to the previous guessing code."
3767 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3768 implementing BATV in an Exim configuration. See NewStuff for the gory
3771 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3772 Makefile that are specific to HP-UX.
3774 PH/18 If the "use_postmaster" option was set for a recipient callout together
3775 with the "random" option, the postmaster address was used as the MAIL
3776 FROM address for the random test, but not for the subsequent recipient
3777 test. It is now used for both.
3779 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3780 patch removes a few documentation additions to RFC 3028, because the
3781 latest draft now contains them. It adds the new en;ascii-case comparator
3782 and a new error check for 8bit text in MIME parts. Comparator and
3783 require names are now matched exactly. I enabled the subaddress
3784 extension, but it is not well tested yet (read: it works for me)."
3786 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3787 rework some of the code of TK/09 above to avoid the hardwired use of
3788 "%lld" and "long long". Replaced the call to snprintf() with a call to
3791 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3792 records point to non-existent hosts", "retry timeout exceeded", and
3793 "retry time not reached for any host after a long failure period".
3795 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3796 experimental DomainKeys support:
3798 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3799 (2) On an error such as an illegally used "control", the wrong name for
3800 the control was given.
3802 These problems did NOT occur unless DomainKeys support was compiled.
3804 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3806 PH/24 Added ${if match_ip condition.
3808 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3809 databases so that it will be absolutely obvious if a crash occurs in the
3810 DB library. This is a regular occurrence (often caused by mis-matched
3813 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3814 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3815 picked up on a user's system that detects such things. There doesn't seem
3816 to be a gcc warning option for this - only an attribute that has to be
3817 put on the function's prototype. It seems that in Fedora Core 4 they have
3818 set this on a number of new functions. No doubt there will be more in due
3821 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3822 specify a transport. However, if an address that was verified by such a
3823 router was the subject of a callout, Exim crashed because it tried to
3824 read the rcpt_include_affixes from the non-existent transport. Now it
3825 just assumes that the setting of that option is false. This bug was
3826 introduced by 4.51/PH/31.
3828 PH/28 Changed -d+all to exclude +memory, because that information is very
3829 rarely of interest, but it makes the output a lot bigger. People tend to
3830 do -d+all out of habit.
3832 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3833 code in os-type was giving problems when libc.so lives in lib64, like on
3836 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3837 aren't the modern standard, and it seems that some systems' include files
3838 don't always have them. Exim was already checking for some of the newer
3839 ones like T_AAAA, and defining it itself. I've added checks for all the
3840 record types that Exim uses.
3842 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3843 not automatically generating a new one, as it is supposed to. This
3844 prevented TLS from working. If the file did exist, but contained invalid
3845 data, a new version was generated, as expected. It was only the case of a
3846 non-existent file that was broken.
3848 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3849 with a change in libdomainkeys > 0.64.
3851 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3852 from DNS. If the selector record carries the flag, it now has
3853 precedence over the domain-wide flag.
3855 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3857 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3858 the use of a port name that isn't defined in /etc/services) occurred, the
3859 message was deferred as in a normal delivery, and thus remained on the
3860 spool, instead of being failed because of the mua_wrapper setting. This
3861 is now fixed, and I tidied up some of the mua_wrapper messages at the
3864 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3865 the messages in a hash of arrays rather than using individual hashes.
3866 This is a bit cleaner and results in dramatic memory savings, albeit
3867 at a slight CPU cost.
3869 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3870 as requested by Marc Sherman.
3872 SC/10 Eximstats: added histograms for user specified patterns as requested
3875 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3877 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3878 fopen() in the content-scanning modules that did not already have it.
3884 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3885 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3887 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3889 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3891 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3892 to test Sieve filters that use "vacation".
3894 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3895 that changes the way the GnuTLS parameters are stored in the cache file.
3896 The new format can be generated externally. For backward compatibility,
3897 if the data in the cache doesn't make sense, Exim assumes it has read an
3898 old-format file, and it generates new data and writes a new file. This
3899 means that you can't go back to an older release without removing the
3902 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3903 work if there are any delivery delays because "one_time" forces the
3904 parent to be marked "delivered", so its unseen clone is never tried
3905 again. For this reason, Exim now forbids the simultaneous setting of
3908 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3909 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3910 entry if you want to know the details. The fix, however, neglected to
3911 consider the case when local delivery batching is involved. The test for
3912 "previously delivered" was not happening when checking to see if an
3913 address could be batched with a previous (undelivered) one; under
3914 certain circumstances this could lead to multiple deliveries to the same
3917 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3918 in its include files, and this causes problems building Exim.
3920 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3921 header_syntax) but Exim was just ignoring anything given after a slash.
3922 In particular, this caused confusion with an attempt to use "verify =
3923 reverse_host_lookup/defer_ok". An error is now given when options are
3924 supplied for verify items that do not have them. (Maybe reverse_host_
3925 lookup should have a defer_ok option, but that's a different point.)
3927 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3928 defined by RFC 821) to 2048, because there were problems with some AUTH
3929 commands, and RFC 1869 says the size should be increased for extended
3930 SMTP commands that take arguments.
3932 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3935 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3936 "unknown" error; now it says that the functionality isn't in the binary.
3938 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3939 an address' error message when a string expansion fails (syntax or
3940 whatever). Otherwise the password may appear in the log. Following change
3941 PH/42 below, there is no longer a chance of it appearing in a bounce
3944 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3946 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3947 of its addresses was ever tried. (Bugzilla bug #2.)
3949 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3950 the result incorrectly in the debug output. (It correctly added a newline
3951 to what was transported.)
3953 TF/01 Added $received_time.
3955 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3956 commented out examples of how to interface to a virus scanner and to
3957 SpamAssassin. Also added commented examples of av_scanner and
3958 spamd_address settings.
3960 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3961 and controls are allowed in which ACLs. There were a couple of minor
3962 errors. Some of the entries in the conditions table (which is a table of
3963 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3964 negation of where the condition IS allowed.
3966 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3968 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3969 header file does not have a version number, so I've had to invent a new
3970 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3971 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3972 radiusclient), but it was contributed by a Radius user.
3974 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3975 files or directories, for OpenSSL.
3977 PH/19 When an Exim process that is running as root has to create an Exim log
3978 file, it does so in a subprocess that runs as exim:exim so as to get the
3979 ownership right at creation (otherwise, other Exim processes might see
3980 the file with the wrong ownership). There was no test for failure of this
3981 fork() call, which would lead to the process getting stuck as it waited
3982 for a non-existent subprocess. Forks do occasionally fail when resources
3983 run out. I reviewed all the other calls to fork(); they all seem to check
3986 PH/20 When checking for unexpected SMTP input at connect time (before writing
3987 the banner), Exim was not dealing correctly with a non-positive return
3988 from the read() function. If the client had disconnected by this time,
3989 the result was a log entry for a synchronization error with an empty
3990 string after "input=" when read() returned zero. If read() returned -1
3991 (an event I could not check), uninitialized data bytes were printed.
3992 There were reports of junk text (parts of files, etc) appearing after
3995 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3997 PH/22 Added support for macro redefinition, and (re)definition in between
3998 driver and ACL definitions.
4000 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4001 forgetting to use the resulting value; it was using the unexpanded value.
4003 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4004 hadn't been configured. The fix is from Juergen Kreileder, who
4005 understands it better than I do:
4007 "Here's what I see happening with three configured cyrus_sasl
4008 authenticators configured (plain, login, cram-md5):
4010 On startup auth_cyrus_sasl_init() gets called for each of these.
4011 This means three calls to sasl_listmech() without a specified mech_list.
4012 => SASL tests which mechs of all available mechs actually work
4013 => three warnings about OTP not working
4014 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4016 With the patch, sasl_listmech() also gets called three times. But now
4017 SASL's mech_list option is set to the server_mech specified in the the
4018 authenticator. Or in other words, the answer from sasl_listmech()
4019 gets limited to just the mech you're testing for (which is different
4021 => the return list contains just 'plain' or 'login', 'cram-md5' or
4022 nothing depending on the value of ob->server_mech.
4024 I've just tested the patch: Authentication still works fine,
4025 unavailable mechs specified in the exim configuration are still
4026 caught, and the auth.log warnings about OTP are gone."
4028 PH/25 When debugging is enabled, the contents of the command line are added
4029 to the debugging output, even when log_selector=+arguments is not
4032 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4033 answer is "GNU", and only if the return is "GNU/something" is the answer
4036 PH/27 $acl_verify_message is now set immediately after the failure of a
4037 verification in an ACL, and so is available in subsequent modifiers. In
4038 particular, the message can be preserved by coding like this:
4040 warn !verify = sender
4041 set acl_m0 = $acl_verify_message
4043 Previously, $acl_verify_message was set only while expanding "message"
4044 and "log_message" when a very denied access.
4046 PH/28 Modified OS/os.c-Linux with
4048 -#ifndef OS_LOAD_AVERAGE
4049 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4051 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4052 nomenclature these days.)
4054 PH/29 Installed patch from the Sieve maintainer that adds the options
4055 sieve_useraddress and sieve_subaddress to the redirect router.
4057 PH/30 In these circumstances:
4058 . Two addresses routed to the same list of hosts;
4059 . First host does not offer TLS;
4060 . First host accepts first address;
4061 . First host gives temporary error to second address;
4062 . Second host offers TLS and a TLS session is established;
4063 . Second host accepts second address.
4064 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4065 and peerdn, if requested) that were in fact used only for the second
4068 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4069 attention to any local part prefix or suffix that was matched by the
4070 router that accepted the address. It now behaves in the same way as it
4071 does for delivery: the affixes are removed from the local part unless
4072 rcpt_include_affixes is set on the transport.
4074 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4075 timeout during the DATA phase of an incoming message.
4077 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4078 applied a patch sanctioned by the Sieve maintainer.
4080 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4081 the uid or gid is negative. A case of a negative gid caused this to be
4082 noticed. The fix allows for either to be negative.
4084 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4085 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4088 PH/36 The change PH/12 above was broken. Fixed it.
4090 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4091 the grounds that routing the same address twice would always produce the
4092 same answer. This might have been true once, but it is certainly no
4093 longer true now. Routing a child address may depend on the previous
4094 routing that produced that child. Some complicated redirection strategies
4095 went wrong when messages had multiple recipients, and made Exim's
4096 behaviour dependent on the order in which the addresses were given.
4098 I have moved the duplicate checking until after the routing is complete.
4099 Exim scans the addresses that are assigned to local and remote
4100 transports, and removes any duplicates. This means that more work will be
4101 done, as duplicates will always all be routed, but duplicates are
4102 presumably rare, so I don't expect this is of any significance.
4104 For deliveries to pipes, files, and autoreplies, the duplicate checking
4105 still happens during the routing process, since they are not going to be
4108 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4109 It corrects a timeout issue with spamd. This is Ian's comment: "The
4110 background is that sometimes spamd either never reads data from a
4111 connection it has accepted, or it never writes response data. The exiscan
4112 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4113 blindly assumes that writes won't block so it may never time out."
4115 PH/39 Allow G after quota size as well as K and M.
4117 PH/40 The value set for $authenticated_id in an authenticator may not contain
4118 binary zeroes or newlines because the value is written to log lines and
4119 to spool files. There was no check on this. Now the value is run through
4120 the string_printing() function so that such characters are converted to
4121 printable escape sequences.
4123 PH/41 $message_linecount is a new variable that contains the total number of
4124 lines in the message. Compare $body_linecount, which is the count for the
4127 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4128 bounce and delay warning messages, except in certain special cases, which
4131 (a) An SMTP error message from a remote host;
4132 (b) A message specified in a :fail: redirection;
4133 (c) A message specified in a "fail" command in a system filter;
4134 (d) A message specified in a FAIL return from the queryprogram router;
4135 (e) A message specified by the cannot_route_message router option.
4137 In these cases only, Exim does include the error details in bounce and
4138 warning messages. There are also a few cases where bland messages such
4139 as "unrouteable address" or "local delivery error" are given.
4141 PH/43 $value is now also set for the "else" part of a ${run expansion.
4143 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4144 being worked on, but at least Exim now implements the latest version to
4147 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4148 process to complete was treated as a delivery failure, a timeout while
4149 writing the message to the pipe was logged, but erroneously treated as a
4150 successful delivery. Such timeouts include transport filter timeouts. For
4151 consistency with the overall process timeout, these timeouts are now
4152 treated as errors, giving rise to delivery failures by default. However,
4153 there is now a new Boolean option for the pipe transport called
4154 timeout_defer, which, if set TRUE, converts the failures into defers for
4155 both kinds of timeout. A transport filter timeout is now identified in
4158 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4159 systems where "make" and "gmake" are different, calling "gmake" at top
4160 level broke things. I've arranged for the value of $(MAKE) to be passed
4161 from the Makefile to this script so that it can call the same version of
4165 A note about Exim versions 4.44 and 4.50
4166 ----------------------------------------
4168 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4169 changes of various kinds. As a consequence, a big documentation update was
4170 needed. This delayed the release for rather longer than seemed good, especially
4171 in the light of a couple of (minor) security issues. Therefore, the changes
4172 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4173 release. So 4.44 and 4.50 are in effect two different branches that both start
4176 I have left the 4.50 change log unchanged; it contains all the changes since
4177 4.43. The change log for 4.44 is below; many of its items are identical to
4178 those for 4.50. This seems to be the most sensible way to preserve the
4179 historical information.
4185 1. Minor wording change to the doc/README.SIEVE file.
4187 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4188 computation of the current number of files was incorrect.
4190 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4191 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4194 4. Give more explanation in the error message when the command for a transport
4195 filter fails to execute.
4197 5. There are several places where Exim runs a non-Exim command in a
4198 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4199 was being done only for the command run by the queryprogram router. It is
4200 now done for all such subprocesses. The other cases are: ${run, transport
4201 filters, and the commands run by the lmtp and pipe transports.
4203 6. Added CONFIGURE_GROUP build-time option.
4205 7. Some older OS have a limit of 256 on the maximum number of file
4206 descriptors. Exim was using setrlimit() to set 1000 as a large value
4207 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4208 systems. I've change it so that if it can't get 1000, it tries for 256.
4210 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4211 was an oversight, and furthermore, ever since the addition of extra
4212 controls (e.g. 4.43/32), the checks on when to allow different forms of
4213 "control" were broken. There should now be diagnostics for all cases when a
4214 control that does not make sense is encountered.
4216 9. Added the /retain_sender option to "control=submission".
4218 10. $recipients is now available in the predata ACL (oversight).
4220 11. Tidy the search cache before the fork to do a delivery from a message
4221 received from the command line. Otherwise the child will trigger a lookup
4222 failure and thereby defer the delivery if it tries to use (for example) a
4223 cached ldap connection that the parent has called unbind on.
4225 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4226 of $address_data from the recipient verification was clobbered by the
4227 sender verification.
4229 13. The value of address_data from a sender verification is now available in
4230 $sender_address_data in subsequent conditions in the ACL statement.
4232 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4234 15. Added a new option "connect=<time>" to callout options, to set a different
4237 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4238 was its contents. (It was OK if the option was not defined at all.)
4240 17. A "Completed" log line is now written for messages that are removed from
4241 the spool by the -Mrm option.
4243 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4244 information about exactly what failed.
4246 19. Added -dd to debug only the daemon process.
4248 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4249 handles timeouts, both on the server side and network timeouts. Renamed the
4250 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4252 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4253 It is now set to "smtps".
4255 22. $host_address is now set to the target address during the checking of
4256 ignore_target_hosts.
4258 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4259 being passed; this would have caused $sender_host_name to have been used if
4260 matching the list had actually called for a host name (not very likely,
4261 since this list is usually IP addresses). A host name is now passed as
4264 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4265 code that specifies a non-restarting handler (typically sigaction() in
4266 modern systems) in an attempt to fix a rare and obscure crash bug.
4268 25. Narrowed the window for a race in the daemon that could cause it to ignore
4269 SIGCHLD signals. This is not a major problem, because they are used only to
4270 wake it up if nothing else does.
4272 26. A malformed maildirsize file could cause Exim to calculate negative values
4273 for the mailbox size or file count. Odd effects could occur as a result.
4274 The maildirsize information is now recalculated if the size or filecount
4277 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4278 support for a long time. Removed HAVE_SYS_VFS_H.
4280 28. Installed the latest version of exipick from John Jetmore.
4282 29. In an address list, if the pattern was not a regular expression, an empty
4283 subject address (from a bounce message) matched only if the pattern was an
4284 empty string. Non-empty patterns were not even tested. This was the wrong
4285 because it is perfectly reasonable to use an empty address as part of a
4286 database query. An empty address is now tested by patterns that are
4287 lookups. However, all the other forms of pattern expect the subject to
4288 contain a local part and a domain, and therefore, for them, an empty
4289 address still always fails if the pattern is not itself empty.
4291 30. Exim went into a mad DNS loop when attempting to do a callout where the
4292 host was specified on an smtp transport, and looking it up yielded more
4293 than one IP address.
4295 31. Re-factored the code for checking spool and log partition space into a
4296 function that finds that data and another that does the check. The former
4297 is then used to implement four new variables: $spool_space, $log_space,
4298 $spool_inodes, and $log_inodes.
4300 32. The RFC2047 encoding function was originally intended for short strings
4301 such as real names; it was not keeping to the 75-character limit for
4302 encoded words that the RFC imposes. It now respects the limit, and
4303 generates multiple encoded words if necessary. To be on the safe side, I
4304 have increased the buffer size for the ${rfc2047: expansion operator from
4307 33. It is now permitted to omit both strings after an "if" condition; if the
4308 condition is true, the result is "true". As before, when the second string
4309 is omitted, a false condition yields an empty string. This makes it less
4310 cumbersome to write custom ACL and router conditions.
4312 34. Failure to deliver a bounce message always caused it to be frozen, even if
4313 there was an errors_to setting on the router. The errors_to setting is now
4316 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4317 canonical form (fully expanded) before being placed in
4318 $sender_host_address.
4320 36. The table in the code that translates DNS record types into text (T_A to
4321 "A" for instance) was missing entries for NS and CNAME. It is just possible
4322 that this could have caused confusion if both these types were looked up
4323 for the same domain, because the text type is used as part of Exim's
4324 per-process caching. But the chance of anyone hitting this buglet seems
4327 37. The dnsdb lookup has been extended in a number of ways.
4329 (1) There is a new type, "zns", which walks up the domain tree until it
4330 finds some nameserver records. It should be used with care.
4332 (2) There is a new type, "mxh", which is like "mx" except that it returns
4333 just the host names, not the priorities.
4335 (3) It is now possible to give a list of domains (or IP addresses) to be
4336 looked up. The behaviour when one of the lookups defers can be
4337 controlled by a keyword.
4339 (4) It is now possible to specify the separator character for use when
4340 multiple records are returned.
4342 38. The dnslists ACL condition has been extended: it is now possible to supply
4343 a list of IP addresses and/or domains to be looked up in a particular DNS
4346 39. Added log_selector=+queue_time_overall.
4348 40. When running the queue in the test harness, wait just a tad after forking a
4349 delivery process, to get repeatability of debugging output.
4351 41. Include certificate and key file names in error message when GnuTLS fails
4352 to set them up, because the GnuTLS error message doesn't include the name
4353 of the failing file when there is a problem reading it.
4355 42. Allow both -bf and -bF in the same test run.
4357 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4359 44. The "Exiscan patch" is now merged into the mainline Exim source.
4361 45. Sometimes the final signoff response after QUIT could fail to get
4362 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4363 < 0 before doing a fflush(). This bug looks as though it goes back to the
4364 introduction of TLS in release 3.20, but "sometimes" must have been rare
4365 because the tests only now provoked it.
4367 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4368 (this can affect the format of dates).
4370 47. exim_tidydb, when checking for the continued existence of a message for
4371 which it has found a message-specific retry record, was not finding
4372 messages that were in split spool directories. Consequently, it was
4373 deleting retry records that should have stayed in existence.
4375 48. Steve fixed some bugs in eximstats.
4377 49. The SPA authentication driver was not abandoning authentication and moving
4378 on to the next authenticator when an expansion was forced to fail,
4379 contradicting the general specification for all authenticators. Instead it
4380 was generating a temporary error. It now behaves as specified.
4382 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4383 (the order specifies the preference for clients). The order is now AES256,
4384 AES128, 3DES, ARCFOUR128.
4386 51. Small patch to Sieve code - explicitly set From: when generating an
4389 52. Exim crashed if a remote delivery caused a very long error message to be
4390 recorded - for instance if somebody sent an entire SpamAssassin report back
4391 as a large number of 550 error lines. This bug was coincidentally fixed by
4392 increasing the size of one of Exim's internal buffers (big_buffer) that
4393 happened as part of the Exiscan merge. However, to be on the safe side, I
4394 have made the code more robust (and fixed the comments that describe what
4397 53. Now that there can be additional text after "Completed" in log lines (if
4398 the queue_time_overall log selector is set), a one-byte patch to exigrep
4399 was needed to allow it to recognize "Completed" as not the last thing in
4402 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4403 patch that reportedly fixes this has been added. I am not expert enough to
4404 create a test for it. This is what the patch creator wrote:
4406 "I found a little strange behaviour of ldap code when working with
4407 Windows 2003 AD Domain, where users was placed in more than one
4408 Organization Units. When I tried to give exim partial DN, the exit code
4409 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4410 But simultaneously result of request was absolutely normal ldap result,
4411 so I produce this patch..."
4413 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4414 REFERENCE, so I have modified the code to exclude the patch when that macro
4417 55. Some experimental protocols are using DNS PTR records for new purposes. The
4418 keys for these records are domain names, not reversed IP addresses. The
4419 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4420 leaves it alone. Component reversal etc. now happens only for IP addresses.
4423 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4425 57. Double the size of the debug message buffer (to 2048) so that more of very
4426 long debug lines gets shown.
4428 58. The exicyclog utility now does better if the number of log files to keep
4429 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4431 59. Two changes related to the smtp_active_hostname option:
4433 (1) $smtp_active_hostname is now available as a variable.
4434 (2) The default for smtp_banner uses $smtp_active_hostname instead
4435 of $primary_hostname.
4437 60. The host_aton() function is supposed to be passed a string that is known
4438 to be a valid IP address. However, in the case of IPv6 addresses, it was
4439 not checking this. This is a hostage to fortune. Exim now panics and dies
4440 if the condition is not met. A case was found where this could be provoked
4441 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4442 components; fortuitously, this particular loophole had already been fixed
4443 by change 4.50/55 above.
4445 If there are any other similar loopholes, the new check in host_aton()
4446 itself should stop them being exploited. The report I received stated that
4447 data on the command line could provoke the exploit when Exim was running as
4448 exim, but did not say which command line option was involved. All I could
4449 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4450 running as the user.
4453 61. There was a buffer overflow vulnerability in the SPA authentication code
4454 (which came originally from the Samba project). I have added a test to the
4455 spa_base64_to_bits() function which I hope fixes it.
4458 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4459 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4461 63. The daemon start-up calls getloadavg() while still root for those OS that
4462 need the first call to be done as root, but it missed one case: when
4463 deliver_queue_load_max is set with deliver_drop_privilege. This is
4464 necessary for the benefit of the queue runner, because there is no re-exec
4465 when deliver_drop_privilege is set.
4467 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4468 This has been fixed.
4470 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4471 was in use, was not putting the data itself into the right store pool;
4472 consequently, it could be overwritten for a subsequent message in the same
4473 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4476 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4478 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4479 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4480 the function were treating the return as a boolean value, which happened to
4481 work because 0=false and not-0=true, but is not correct code.
4483 68. The host_aton() function was not handling scoped IPv6 addresses (those
4484 with, for example, "%eth0" on the end) correctly.
4486 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4487 negated items (that is, ~something) in unsigned ints. Some compilers
4488 apparently mutter when there is no cast.
4490 70. If an address verification called from an ACL failed, and did not produce a
4491 user-specific message (i.e. there was only a "system" message), nothing was
4492 put in $acl_verify_message. In this situation, it now puts the system
4495 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4496 session; change 4.31/43 added the unwanted input to the log line - except
4497 that it did not do this in the start of session case. It now does.
4499 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4500 This is wrong and can cause the other end to generate a synchronization
4501 error if it is another Exim or anything else that does the synchronization
4502 check. A QUIT command is no longer sent after a timeout.
4504 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4505 during host lookups.
4507 74. The defer_ok option of callout verification was not working if it was used
4508 when verifying addresses in header lines, that is, for this case:
4510 verify = header_sender/callout=defer_ok
4512 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4513 those file descriptors could be used for SMTP connections. If anything
4514 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4515 could be sent to the SMTP client, causing chaos. The daemon now opens
4516 stdin, stdout, and stderr to /dev/null when it puts itself into the
4519 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4520 log by default. The user can override this with suitable Perl magic.
4522 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4523 the log message when discard triggers, was not working for the DATA ACL or
4524 for the non-SMTP ACL.
4526 78. Error message wording change in sieve.c.
4528 79. If smtp_accept_max_per_host was set, the number of connections could be
4529 restricted to fewer than expected, because the daemon was trying to set up
4530 a new connection before checking whether the processes handling previous
4531 connections had finished. The check for completed processes is now done
4532 earlier. On busy systems, this bug wouldn't be noticed because something
4533 else would have woken the daemon, and it would have reaped the completed
4536 80. If a message was submitted locally by a user whose login name contained one
4537 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4538 It caused a spool format error. I have fixed the spool reading code. A
4539 related problem was that the "from" clause in the Received: line became
4540 illegal because of the space(s). It is now covered by ${quote_local_part.
4542 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4545 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4547 83. Patch from Sieve maintainer to fix unterminated string problem in
4548 "vacation" handling.
4550 84. Some minor changes to the Linux configuration files to help with other
4551 OS variants using glibc.
4553 85. One more patch for Sieve to update vacation handling to latest spec.
4556 ----------------------------------------------------
4557 See the note above about the 4.44 and 4.50 releases.
4558 ----------------------------------------------------
4564 1. Change 4.43/35 introduced a bug that caused file counts to be
4565 incorrectly computed when quota_filecount was set in an appendfile
4568 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4569 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4572 3. Give more explanation in the error message when the command for a transport
4573 filter fails to execute.
4575 4. There are several places where Exim runs a non-Exim command in a
4576 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4577 was being done only for the command run by the queryprogram router. It is
4578 now done for all such subprocesses. The other cases are: ${run, transport
4579 filters, and the commands run by the lmtp and pipe transports.
4581 5. Some older OS have a limit of 256 on the maximum number of file
4582 descriptors. Exim was using setrlimit() to set 1000 as a large value
4583 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4584 systems. I've change it so that if it can't get 1000, it tries for 256.
4586 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4587 was an oversight, and furthermore, ever since the addition of extra
4588 controls (e.g. 4.43/32), the checks on when to allow different forms of
4589 "control" were broken. There should now be diagnostics for all cases when a
4590 control that does not make sense is encountered.
4592 7. $recipients is now available in the predata ACL (oversight).
4594 8. Tidy the search cache before the fork to do a delivery from a message
4595 received from the command line. Otherwise the child will trigger a lookup
4596 failure and thereby defer the delivery if it tries to use (for example) a
4597 cached ldap connection that the parent has called unbind on.
4599 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4600 of $address_data from the recipient verification was clobbered by the
4601 sender verification.
4603 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4604 was its contents. (It was OK if the option was not defined at all.)
4606 11. A "Completed" log line is now written for messages that are removed from
4607 the spool by the -Mrm option.
4609 12. $host_address is now set to the target address during the checking of
4610 ignore_target_hosts.
4612 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4613 being passed; this would have caused $sender_host_name to have been used if
4614 matching the list had actually called for a host name (not very likely,
4615 since this list is usually IP addresses). A host name is now passed as
4618 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4619 code that specifies a non-restarting handler (typically sigaction() in
4620 modern systems) in an attempt to fix a rare and obscure crash bug.
4622 15. Narrowed the window for a race in the daemon that could cause it to ignore
4623 SIGCHLD signals. This is not a major problem, because they are used only to
4624 wake it up if nothing else does.
4626 16. A malformed maildirsize file could cause Exim to calculate negative values
4627 for the mailbox size or file count. Odd effects could occur as a result.
4628 The maildirsize information is now recalculated if the size or filecount
4631 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4632 support for a long time. Removed HAVE_SYS_VFS_H.
4634 18. Updated exipick to current release from John Jetmore.
4636 19. Allow an empty sender to be matched against a lookup in an address list.
4637 Previously the only cases considered were a regular expression, or an
4640 20. Exim went into a mad DNS lookup loop when doing a callout where the
4641 host was specified on the transport, if the DNS lookup yielded more than
4644 21. The RFC2047 encoding function was originally intended for short strings
4645 such as real names; it was not keeping to the 75-character limit for
4646 encoded words that the RFC imposes. It now respects the limit, and
4647 generates multiple encoded words if necessary. To be on the safe side, I
4648 have increased the buffer size for the ${rfc2047: expansion operator from
4651 22. Failure to deliver a bounce message always caused it to be frozen, even if
4652 there was an errors_to setting on the router. The errors_to setting is now
4655 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4656 canonical form (fully expanded) before being placed in
4657 $sender_host_address.
4659 24. Updated eximstats to version 1.33
4661 25. Include certificate and key file names in error message when GnuTLS fails
4662 to set them up, because the GnuTLS error message doesn't include the name
4663 of the failing file when there is a problem reading it.
4665 26. Expand error message when OpenSSL has problems setting up cert/key files.
4668 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4669 (this can affect the format of dates).
4671 28. exim_tidydb, when checking for the continued existence of a message for
4672 which it has found a message-specific retry record, was not finding
4673 messages that were in split spool directories. Consequently, it was
4674 deleting retry records that should have stayed in existence.
4676 29. eximstats updated to version 1.35
4677 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4678 1.35 - bugfix such that pie charts by volume are generated correctly
4680 30. The SPA authentication driver was not abandoning authentication and moving
4681 on to the next authenticator when an expansion was forced to fail,
4682 contradicting the general specification for all authenticators. Instead it
4683 was generating a temporary error. It now behaves as specified.
4685 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4686 (the order specifies the preference for clients). The order is now AES256,
4687 AES128, 3DES, ARCFOUR128.
4689 31. Small patch to Sieve code - explicitly set From: when generating an
4692 32. Exim crashed if a remote delivery caused a very long error message to be
4693 recorded - for instance if somebody sent an entire SpamAssassin report back
4694 as a large number of 550 error lines. This bug was coincidentally fixed by
4695 increasing the size of one of Exim's internal buffers (big_buffer) that
4696 happened as part of the Exiscan merge. However, to be on the safe side, I
4697 have made the code more robust (and fixed the comments that describe what
4700 33. Some experimental protocols are using DNS PTR records for new purposes. The
4701 keys for these records are domain names, not reversed IP addresses. The
4702 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4703 leaves it alone. Component reversal etc. now happens only for IP addresses.
4706 34. The host_aton() function is supposed to be passed a string that is known
4707 to be a valid IP address. However, in the case of IPv6 addresses, it was
4708 not checking this. This is a hostage to fortune. Exim now panics and dies
4709 if the condition is not met. A case was found where this could be provoked
4710 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4711 components; fortuitously, this particular loophole had already been fixed
4712 by change 4.50/55 or 4.44/33 above.
4714 If there are any other similar loopholes, the new check in host_aton()
4715 itself should stop them being exploited. The report I received stated that
4716 data on the command line could provoke the exploit when Exim was running as
4717 exim, but did not say which command line option was involved. All I could
4718 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4719 running as the user.
4722 35. There was a buffer overflow vulnerability in the SPA authentication code
4723 (which came originally from the Samba project). I have added a test to the
4724 spa_base64_to_bits() function which I hope fixes it.
4727 36. The daemon start-up calls getloadavg() while still root for those OS that
4728 need the first call to be done as root, but it missed one case: when
4729 deliver_queue_load_max is set with deliver_drop_privilege. This is
4730 necessary for the benefit of the queue runner, because there is no re-exec
4731 when deliver_drop_privilege is set.
4733 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4734 was in use, was not putting the data itself into the right store pool;
4735 consequently, it could be overwritten for a subsequent message in the same
4736 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4739 38. Sometimes the final signoff response after QUIT could fail to get
4740 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4741 < 0 before doing a fflush(). This bug looks as though it goes back to the
4742 introduction of TLS in release 3.20, but "sometimes" must have been rare
4743 because the tests only now provoked it.
4749 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4750 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4751 Now it returns an integer. A number of calls were still expecting a T/F
4752 return. Fortuitously, in all cases, the tests worked in OK situations,
4753 which is the norm. However, things would have gone wrong on any write
4754 failures on the smtp file descriptor. This function is used when sending
4755 messages over SMTP and also when doing verify callouts.
4757 2. When Exim is called to do synchronous delivery of a locally submitted
4758 message (the -odf or -odi options), it no longer closes stderr before doing
4761 3. Implemented the mua_wrapper option.
4763 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4765 5. Implemented the functions header_remove(), header_testname(),
4766 header_add_at_position(), and receive_remove_recipient(), and exported them
4769 6. If an ACL "warn" statement specified the addition of headers, Exim already
4770 inserted X-ACL-Warn: at the start if there was no header name. However, it
4771 was not making this test for the second and subsequent header lines if
4772 there were newlines in the string. This meant that an invalid header could
4773 be inserted if Exim was badly configured.
4775 7. Allow an ACL "warn" statement to add header lines at the start or after all
4776 the Received: headers, as well as at the end.
4778 8. Added the rcpt_4xx retry error code.
4780 9. Added postmaster_mailfrom=xxx to callout verification option.
4782 10. Added mailfrom=xxxx to the callout verification option, for verify=
4785 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4786 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4788 12. Inserted some casts to stop certain compilers warning when using pointer
4789 differences as field lengths or precisions in printf-type calls (mostly
4790 affecting debugging statements).
4792 13. Added optional readline() support for -be (dynamically loaded).
4794 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4795 same clock tick as a message's arrival, so that its received time was the
4796 same as the "first fail" time on the retry record, and that message
4797 remained on the queue past the ultimate address timeout, every queue runner
4798 would try a delivery (because it was past the ultimate address timeout) but
4799 after another failure, the ultimate address timeout, which should have then
4800 bounced the address, did not kick in. This was a "< instead of <=" error;
4801 in most cases the first failure would have been in the next clock tick
4802 after the received time, and all would be well.
4804 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4805 being recognized when the domain list was tested by the match_domain
4806 condition in an expansion string.
4808 16. Added the ${str2b64: operator.
4810 17. Exim was always calling setrlimit() to set a large limit for the number of
4811 processes, without checking whether the existing limit was already
4812 adequate. (It did check for the limit on file descriptors.) Furthermore,
4813 errors from getrlimit() and setrlimit() were being ignored. Now they are
4814 logged to the main and panic logs, but Exim does carry on, to try to do its
4815 job under whatever limits there are.
4817 18. Imported PCRE 5.0.
4819 19. Trivial typo in log message " temporarily refused connection" (the leading
4822 20. If the log selector return_path_on_delivery was set and an address was
4823 redirected to /dev/null, the delivery process crashed because it assumed
4824 that a return path would always be set for a "successful" delivery. In this
4825 case, the whole delivery is bypassed as an optimization, and therefore no
4828 21. Internal re-arrangement: the function for sending a challenge and reading
4829 a response while authentication was assuming a zero-terminated challenge
4830 string. It's now changed to take a pointer and a length, to allow for
4831 binary data in such strings.
4833 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4835 23. Exim was not respecting finduser_retries when seeking the login of the
4836 uid under which it was called; it was always trying 10 times. (The default
4837 setting of finduser_retries is zero.) Also, it was sleeping after the final
4838 failure, which is pointless.
4840 24. Implemented tls_on_connect_ports.
4842 25. Implemented acl_smtp_predata.
4844 26. If the domain in control=submission is set empty, Exim assumes that the
4845 authenticated id is a complete email address when it generates From: or
4846 Sender: header lines.
4848 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4849 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4850 chown and chgrp in /bin and hostname in /usr/bin.
4852 28. Exim was keeping the "process log" file open after each use, just as it
4853 does for the main log. This opens the possibility of it remaining open for
4854 long periods when the USR1 signal hits a daemon. Occasional processlog
4855 errors were reported, that could have been caused by this. Anyway, it seems
4856 much more sensible not to leave this file open at all, so that is what now
4859 29. The long-running daemon process does not normally write to the log once it
4860 has entered its main loop, and it closes the log before doing so. This is
4861 so that log files can straightforwardly be renamed and moved. However,
4862 there are a couple of unusual error situations where the daemon does write
4863 log entries, and I had neglected to close the log afterwards.
4865 30. The text of an SMTP error response that was received during a remote
4866 delivery was being truncated at 512 bytes. This is too short for some of
4867 the long messages that one sometimes sees. I've increased the limit to
4870 31. It is now possible to make retry rules that apply only when a message has a
4871 specific sender, in particular, an empty sender.
4873 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4874 it possible to be selective about when SMTP synchronization is enforced.
4876 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4878 32. Implemented hosts_connection_nolog.
4880 33. Added an ACL for QUIT.
4882 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4885 35. Added mailbox_size and mailbox_filecount to appendfile.
4887 36. Added control = no_multiline_responses to ACLs.
4889 37. There was a bug in the logic of the code that waits for the clock to tick
4890 in the case where the clock went backwards by a substantial amount such
4891 that the microsecond fraction of "now" was more than the microsecond
4892 fraction of "then" (but the whole seconds number was less).
4894 38. Added support for the libradius Radius client library this is found on
4895 FreeBSD (previously only the radiusclient library was supported).
4901 1. When certain lookups returned multiple values in the form name=value, the
4902 quoting of the values was not always being done properly. Specifically:
4903 (a) If the value started with a double quote, but contained no whitespace,
4905 (b) If the value contained whitespace other than a space character (i.e.
4906 tabs or newlines or carriage returns) it was not quoted.
4907 This fix has been applied to the mysql and pgsql lookups by writing a
4908 separate quoting function and calling it from the lookup code. The fix
4909 should probably also be applied to nisplus, ibase and oracle lookups, but
4910 since I cannot test any of those, I have not disturbed their existing code.
4912 2. A hit in the callout cache for a specific address caused a log line with no
4913 reason for rejecting RCPT. Now it says "Previous (cached) callout
4914 verification failure".
4916 3. There was an off-by-one bug in the queryprogram router. An over-long
4917 return line was truncated at 256 instead of 255 characters, thereby
4918 overflowing its buffer with the terminating zero. As well as fixing this, I
4919 have increased the buffer size to 1024 (and made a note to document this).
4921 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4922 when Exim is waiting for an SMTP response from a remote server, Exim
4923 restarts its select() call on the socket, thereby resetting its timeout.
4924 This is not a problem when such interrupts are rare. Somebody set up a cron
4925 job to run exiwhat every 2 minutes, which is less than the normal select()
4926 timeout (5 or 10 minutes). This meant that the select() timeout never
4927 kicked in because it was always reset. I have fixed this by comparing the
4928 time when an interrupt arrives with the time at the start of the first call
4929 to select(). If more time than the timeout has elapsed, the interrupt is
4930 treated as a timeout.
4932 5. Some internal re-factoring in preparation for the addition of Sieve
4933 extensions (by MH). In particular, the "personal" test is moved to a
4934 separate function, and given an option for scanning Cc: and Bcc: (which is
4935 not set for Exim filters).
4937 6. When Exim created an email address using the login of the caller as the
4938 local part (e.g. when creating a From: or Sender: header line), it was not
4939 quoting the local part when it contained special characters such as @.
4941 7. Installed new OpenBSD configuration files.
4943 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4944 try to make them clearer.
4946 9. Callout options, other than the timeout value, were being ignored when
4947 verifying sender addresses in header lines. For example, when using
4949 verify = header_sender/callout=no_cache
4951 the cache was (incorrectly) being used.
4953 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4954 only the Cygwin environment.
4956 11. When return_path_on_delivery was set as a log selector, if different remote
4957 addresses in the same message used different return paths and parallel
4958 remote delivery occurred, the wrong values would sometimes be logged.
4959 (Whenever a remote delivery process finished, the return path value from
4960 the most recently started remote delivery process was logged.)
4962 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4963 lines when AUTH and/or TLS are in use. This is the "received protocol"
4964 field. Exim used to use "asmtp" for authenticated SMTP, without any
4965 indication (in the protocol name) for TLS use. Now it follows the RFC and
4966 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4967 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4968 names appear in log lines as well as in Received: header lines.
4970 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4971 extensions, and comparison tests, and to fix some bugs.
4973 14. Changes to the "personal" filter test:
4975 (1) The test was buggy in that it was just doing the equivalent of
4976 "contains" tests on header lines. For example, if a user's address was
4977 anne@some.where, the "personal" test would incorrectly be true for
4979 To: susanne@some.where
4981 This test is now done by extracting each address from the header in turn,
4982 and checking the entire address. Other tests that are part of "personal"
4983 are now done using regular expressions (for example, to check local parts
4984 of addresses in From: header lines).
4986 (2) The list of non-personal local parts in From: addresses has been
4987 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4988 taken from the Sieve specification recommendations.
4990 (3) If the message contains any header line starting with "List-" it is
4991 treated as non-personal.
4993 (4) The test for "circular" in the Subject: header line has been removed
4994 because it now seems ill-conceived.
4996 15. Minor typos in src/EDITME comments corrected.
4998 16. Installed latest exipick from John Jetmore.
5000 17. If headers_add on a router specified a text string that was too long for
5001 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5002 of string_sprintf() is now avoided.
5004 18. $message_body_size was not set (it was always zero) when running the DATA
5005 ACL and the local_scan() function.
5007 19. For the "mail" command in an Exim filter, no default was being set for
5008 the once_repeat time, causing a random time value to be used if "once" was
5009 specified. (If the value happened to be <= 0, no repeat happened.) The
5010 default is now 0s, meaning "never repeat". The "vacation" command was OK
5011 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5012 (I found it when inspecting the code).
5014 20. There is now an overall timeout for performing a callout verification. It
5015 defaults to 4 times the callout timeout, which applies to individual SMTP
5016 commands during the callout. The overall timeout applies when there is more
5017 than one host that can be tried. The timeout is checked before trying the
5018 next host. This prevents very long delays if there are a large number of
5019 hosts and all are timing out (e.g. when the network connections are timing
5020 out). The value of the overall timeout can be changed by specifying an
5021 additional sub-option for "callout", called "maxwait". For example:
5023 verify = sender/callout=5s,maxwait=20s
5025 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5026 to the end before writing, but this should make it even safer).
5028 22. Exim was forgetting that it had advertised PIPELINING for the second and
5029 subsequent messages on an SMTP connection. It was also not resetting its
5030 memory on STARTTLS and an internal HELO.
5032 23. When Exim logs an SMTP synchronization error within a session, it now
5033 records whether PIPELINING has been advertised or not.
5035 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5036 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5037 rather than long int.
5039 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5041 26. Added the never_mail option to autoreply.
5047 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5048 crash if the getsockname() call failed; this can happen if a connection is
5049 closed very soon after it is established. The problem was simply in the
5050 order in which certain operations were done, causing Exim to try to write
5051 to the SMTP stream before it had set up the file descriptor. The bug has
5052 been fixed by making things happen in the correct order.
5058 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5059 before the connection was closed, thus losing the rejection response.
5061 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5062 some early Solaris releases, but causes trouble in current releases where
5063 socklen_t is defined.
5065 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5068 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5071 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5072 that releasing the top of it at the end releases what was used for sub-
5073 expansions (unless the block got too big). However, discard this block if
5074 the first thing is a variable or header, so that we can use its block when
5075 it is dynamic (useful for very large $message_headers, for example).
5077 6. Lookups now cache *every* query, not just the most recent. A new, separate
5078 store pool is used for this. It can be recovered when all lookup caches are
5079 flushed. Lookups now release memory at the end of their result strings.
5080 This has involved some general refactoring of the lookup sources.
5082 7. Some code has been added to the store_xxx() functions to reduce the amount
5083 of flapping under certain conditions.
5085 8. log_incoming_interface used to affect only the <= reception log lines. Now
5086 it causes the local interface and port to be added to several more SMTP log
5087 lines, for example "SMTP connection from", and rejection lines.
5089 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5091 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5093 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5094 could be overwritten at the end of the current message (or the start of a
5095 new message if it was set in a HELO ACL). The value is now preserved for
5096 the duration of the SMTP connection.
5098 12. If a transport had a headers_rewrite setting, and a matching header line
5099 contained an unqualified address, that address was qualified, even if it
5100 did not match any rewriting rules. The underlying bug was that the values
5101 of the flags that permit the existence of unqualified sender and recipient
5102 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5103 non-local messages, and by -bnq for local messages) were not being
5104 preserved with the message after it was received.
5106 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5107 "next input=" as part of the text comprising the host identity instead of
5108 the correct text. The code was using the same buffer for two different
5109 strings. However, depending on which order the printing function evaluated
5110 its arguments, the bug did not always show up. Under Linux, for example, my
5111 test suite worked just fine.
5113 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5114 doesn't work with some older versions of Perl. It has been changed to "my",
5115 which in any case is probably the better facility to use.
5117 15. A really picky compiler found some instances of statements for creating
5118 error messages that either had too many or two few arguments for the format
5121 16. The size of the buffer for calls to the DNS resolver has been increased
5122 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5123 for addresses that have a lot of PTR records. This alleviates a problem; it
5124 does not fully solve it.
5126 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5127 buffer now truncates the list and logs the incident, which is the same
5128 action as happens when Exim is looking up a host name and its aliases.
5129 Previously in this situation something unpredictable would happen;
5130 sometimes it was "internal error: store_reset failed".
5132 18. If a server dropped the connection unexpectedly when an Exim client was
5133 using GnuTLS and trying to read a response, the client delivery process
5134 crashed while trying to generate an error log message.
5136 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5137 string, for example:
5139 warn message = H1: something\nH2: something
5141 the text was added as a single header line from Exim's point of view
5142 though it ended up OK in the delivered message. However, searching for the
5143 second and subsequent header lines using $h_h2: did not work. This has been
5144 fixed. Similarly, if a system filter added multiple headers in this way,
5145 the routers could not see them.
5147 20. Expanded the error message when iplsearch is called with an invalid key to
5148 suggest using net-iplsearch in a host list.
5150 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5151 ACLs are no longer actually imposed (and a message to that effect is
5154 22. If a "gecos" field in a passwd entry contained escaped characters, in
5155 particular, if it contained a \" sequence, Exim got it wrong when building
5156 a From: or a Sender: header from that name. A second bug also caused
5157 incorrect handling when an unquoted " was present following a character
5158 that needed quoting.
5160 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5161 was not being matched caselessly.
5163 24. Arranged for all hyphens in the exim.8 source to be escaped with
5166 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5167 properly. Recipient callout cache records were still being keyed to include
5168 the sender, even when use_sender was set false. This led to far more
5169 callouts that were necessary. The sender is no longer included in the key
5170 when use_sender is false.
5172 26. Added "control = submission" modifier to ACLs.
5174 27. Added the ${base62d: operator to decode base 62 numbers.
5176 28. dnsdb lookups can now access SRV records.
5178 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5179 the configuration file.
5181 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5182 (-v) mode. This makes the output for a verbose queue run more intelligible.
5184 31. Added a use_postmaster feature to recipient callouts.
5186 32. Added the $body_zerocount variable, containing the number of binary zero
5187 bytes in the message body.
5189 33. The time of last modification of the "new" subdirectory is now used as the
5190 "mailbox time last read" when there is a quota error for a maildir
5193 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5195 35. Added +ignore_unknown as a special item in host lists.
5197 36. Code for decoding IPv6 addresses in host lists is now included, even if
5198 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5199 address was recognized as an IP address, but was then not correctly decoded
5200 into binary, causing unexpected and incorrect effects when compared with
5207 1. Very minor rewording of debugging text in manualroute to say "list of
5208 hosts" instead of "hostlist".
5210 2. If verify=header_syntax was set, and a header line with an unqualified
5211 address (no domain) and a large number of spaces between the end of the
5212 name and the colon was received, the reception process suffered a buffer
5213 overflow, and (when I tested it) crashed. This was caused by some obsolete
5214 code that should have been removed. The fix is to remove it!
5216 3. When running in the test harness, delay a bit after writing a bounce
5217 message to get a bit more predictability in the log output.
5219 4. Added a call to search_tidyup() just before forking a reception process. In
5220 theory, someone could use a lookup in the expansion of smtp_accept_max_
5221 per_host which, without the tidyup, could leave open a database connection.
5223 5. Added the variables $recipient_data and $sender_data which get set from a
5224 lookup success in an ACL "recipients" or "senders" condition, or a router
5225 "senders" option, similar to $domain_data and $local_part_data.
5227 6. Moved the writing of debug_print from before to after the "senders" test
5230 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5231 problems for message scanning, either using a data ACL, or using
5232 local_scan() because the Received: header was not generated till after they
5233 were called (in order to set the time as the time of reception completion).
5234 I have revised the way this works. The header is now generated after the
5235 body is received, but before the ACL or local_scan() are called. After they
5236 are run, the timestamp in the header is updated.
5242 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5243 before starting a queue runner without re-exec. This happened only when
5244 deliver_drop_privilege was set or when the Exim user was set to root. The
5245 effect of the bug was that timeouts during subsequent deliveries caused
5246 crashes instead of being properly handled. The handler is now left at its
5247 default (and expected) setting.
5249 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5250 message, again when deliver_drop_privilege is set or Exim is run as root.
5251 The bug described in (1) was not present in this case, but the tidying up
5252 of the other signals was missing. I have made the two cases consistent.
5254 3. The ignore_target_hosts setting on a manualroute router was being ignored
5255 for hosts that were looked up using the /MX notation.
5257 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5260 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5261 operated on the sender address. After changing the $sender_address to <>
5262 for the sender address verify, Exim was re-instated it as the original
5263 (before rewriting) address, but remembering that it had rewritten it, so it
5264 wasn't rewriting it again. This bug also had the effect of breaking the
5265 sender address verification caching when the sender address was rewritten.
5267 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5268 This has been changed so that if the ip literal address matches
5269 ignore_target_hosts, the router declines.
5271 7. Added expansion conditions match_domain, match_address, and match_local_
5272 part (NOT match_host).
5274 8. The placeholder for the Received: header didn't have a length field set.
5276 9. Added code to Exim itself and to exim_lock to test for a specific race
5277 condition that could lead to file corruption when using MBX delivery. The
5278 issue is with the lockfile that is created in /tmp. If this file is removed
5279 after a process has opened it but before that process has acquired a lock,
5280 there is the potential for a second process to recreate the file and also
5281 acquire a lock. This could lead to two Exim processes writing to the file
5282 at the same time. The added code performs the same test as UW imapd; it
5283 checks after acquiring the lock that its file descriptor still refers to
5284 the same named file.
5286 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5287 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5290 11. Added the smtp_active_hostname option. If used, this will typically be made
5291 to depend on the incoming interface address. Because $interface_address is
5292 not set up until the daemon has forked a reception process, error responses
5293 that can happen earlier (such as "too many connections") no longer contain
5296 12. If an expansion in a condition on a "warn" statement fails because a lookup
5297 defers, the "warn" statement is abandoned, and the next ACL statement is
5298 processed. Previously this caused the whole ACL to be aborted.
5300 13. Added the iplsearch lookup type.
5302 14. Added ident_timeout as a log selector.
5304 15. Added tls_certificate_verified as a log selector.
5306 16. Added a global option tls_require_ciphers (compare the smtp transport
5307 option of the same name). This controls incoming TLS connections.
5309 17. I finally figured out how to make tls_require_ciphers do a similar thing
5310 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5311 before starting the TLS session.
5313 18. Tabs are now shown as \t in -bP output.
5315 19. If the log selector return_path_on_delivery was set, Exim crashed when
5316 bouncing a message because it had too many Received: header lines.
5318 20. If two routers both had headers_remove settings, and the first one included
5319 a superfluous trailing colon, the final name in the first list and the
5320 first name in the second list were incorrectly joined into one item (with a
5321 colon in the middle).
5327 1. Added -C and -D options to the exinext utility, mainly to make it easier
5328 to include in the automated testing, but these could be helpful when
5329 multiple configurations are in use.
5331 2. The exinext utility was not formatting the output nicely when there was
5332 an alternate port involved in the retry record key, nor when there was a
5333 message id as well (for retries that were specific to a specific message
5334 and a specific host). It was also confused by IPv6 addresses, because of
5335 the additional colons they contain. I have fixed the IPv4 problem, and
5336 patched it up to do a reasonable job for IPv6.
5338 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5339 delivery, the log line now contains "pipelined" if PIPELINING was used.
5341 4. An SMTP transport process used to panic and die if the bind() call to set
5342 an explicit outgoing interface failed. This has been changed; it is now
5343 treated in the same way as a connect() failure.
5345 5. A reference to $sender_host_name in the part of a conditional expansion
5346 that was being skipped was still causing a DNS lookup. This no longer
5349 6. The def: expansion condition was not recognizing references to header lines
5350 that used bh_ and bheader_.
5352 7. Added the _cache feature to named lists.
5354 8. The code for checking quota_filecount in the appendfile transport was
5355 allowing one more file than it should have been.
5357 9. For compatibility with Sendmail, the command line option
5365 and sets the incoming protocol and host name (for trusted callers). The
5366 host name and its colon can be omitted when only the protocol is to be set.
5367 Note the Exim already has two private options, -pd and -ps, that refer to
5368 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5369 "s", but I don't think that's a major issue.
5371 10. A number of refactoring changes to the code, none of which should affect
5374 (a) The number of logging options was getting close to filling up the
5375 32-bit word that was used as a bit map. I have split them into two classes:
5376 those that are passed in the argument to log_write(), and those that are
5377 only ever tested independently outside of that function. These are now in
5378 separate 32-bit words, so there is plenty of room for expansion again.
5379 There is no change in the user interface or the logging behaviour.
5381 (b) When building, for example, log lines, the code previously used a
5382 macro that called string_cat() twice, in order to add two strings. This is
5383 not really sufficiently general. Furthermore, there was one instance where
5384 it was actually wrong because one of the argument was used twice, and in
5385 one call a function was used. (As it happened, calling the function twice
5386 did not affect the overall behaviour.) The macro has been replaced by a
5387 function that can join an arbitrary number of extra strings onto a growing
5390 (c) The code for expansion conditions now uses a table and a binary chop
5391 instead of a serial search (which was left over from when there were very
5392 few conditions). Also, it now recognizes conditions like "pam" even when
5393 the relevant support is not compiled in: a suitably worded error message is
5394 given if an attempt is made to use such a condition.
5396 11. Added ${time_interval:xxxxx}.
5398 12. A bug was causing one of the ddress fields not to be passed back correctly
5399 from remote delivery subprocesses. The field in question was not being
5400 subsequently used, so this caused to problems in practice.
5402 13. Added new log selectors queue_time and deliver_time.
5404 14. Might have fixed a bug in maildirsizefile handling that threw up
5405 "unexpected character" debug warnings, and recalculated the data
5406 unnecessarily. In any case, I expanded the warning message to give more
5409 15. Added the message "Restricted characters in address" to the statements in
5410 the default ACL that block characters like @ and % in local parts.
5412 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5413 Three changes have been made:
5415 (a) There was a serious bug; a negative response to MAIL caused the whole
5416 recipient domain to be cached as invalid, thereby blocking all messages
5417 to all local parts at the same domain, from all senders. This bug has
5418 been fixed. The domain is no longer cached after a negative response to
5419 MAIL if the sender used is not empty.
5421 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5424 (c) A new callout option, "use_sender" has been added for people who want
5425 the modified behaviour.
5431 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5434 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5435 indeed breaks things for older releases.
5437 3. Added additional logging to the case where there is a problem reading data
5438 from a filter that is running in a subprocess using a pipe, in order to
5439 try to track down a specific problem.
5441 4. Testing facility fudge: when running in the test harness and attempting
5442 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5443 getting "No route to host". Convert this to a timeout.
5445 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5448 6. Some OS don't have socklen_t but use size_t instead. This affects the
5449 fifth argument of getsockopt() amongst other things. This is now
5450 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5451 can be set for individual OS. I have set it for SunOS5, OSF1, and
5452 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5453 some earlier ones do not.
5455 7. Change 4.30/15 was not doing the test caselessly.
5457 8. The standard form for an IPv6 address literal was being rejected by address
5458 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5459 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5460 this, as well as the form without the "IPv6" on the front (but only when
5461 address literals are enabled, of course).
5463 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5465 10. Exim crashed if a message with an empty sender address specified by -f
5466 encountered a router with an errors_to setting. This could be provoked only
5467 by a command such as
5471 where an empty string was supplied; "<>" did not hit this bug.
5473 11. Installed PCRE release 4.5.
5475 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5476 remained set. It is now erased.
5478 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5479 times from message ids (which are base 36 rather than the normal 62).
5481 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5482 were being counted as actual protocol errors, and logged if the log
5483 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5484 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5485 and DATA following a set of rejected RCPTs do not count as protocol errors.
5486 In other words, Exim assumes they were pipelined, though this may not
5487 actually be the case. Of course, in all cases the client gets an
5488 appropriate error code.
5490 15. If a lookup fails in an ACL condition, a message about the failure may
5491 be available; it is used if testing the ACL cannot continue, because most
5492 such messages specify what the cause of the deferral is. However, some
5493 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5494 that caused an old message to be retained and used if a later statement
5495 caused a defer, replacing the real cause of the deferral.
5497 16. If an IP address had so many PTR records that the DNS lookup buffer
5498 was not large enough to hold them, Exim could crash while trying to process
5499 the truncated data. It now detects and logs this case.
5501 17. Further to 4.21/58, another change has been made: if (and only if) the
5502 first line of a message (the first header line) ends with CRLF, a bare LF
5503 in a subsequent header line has a space inserted after it, so as not to
5504 terminate the header.
5506 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5507 unnecessarily, used atoi() instead of strtol(), and didn't check the
5508 termination when getting file sizes from file names by regex.
5510 19. Completely re-implemented the support for maildirsize files, in the light
5511 of a number of problems with the previous contributed implementation
5512 (4.30/29). In particular:
5514 . If the quota is zero, the maildirsize file is maintained, but no quota is
5517 . If the maildir directory does not exist, it is created before any attempt
5518 to write a maildirsize file.
5520 . The quota value in the file is just a cache; if the quota is changed in
5521 the transport, the new value overrides.
5523 . A regular expression is available for excluding directories from the
5526 20. The autoreply transport checks the characters in options that define the
5527 message's headers; it allows continued headers, but it was checking with
5528 isspace() after an embedded newline instead of explicitly looking for a
5531 21. If all the "regular" hosts to which an address was routed had passed their
5532 expiry times, and had not reached their retry times, the address was
5533 bounced, even if fallback hosts were defined. Now Exim should go on to try
5536 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5537 equivalent code in the SMTP transport. Some hosts send humungous responses
5538 to HELO/EHLO, more than 1024 it seems.
5540 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5541 gives compiler warnings in some environments. I've now done it "properly",
5544 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5545 (because of problems with the built-in one) was declared to return uschar *
5546 instead of char *, causing compiler failure.
5548 25. Fixed a file descriptor leak when processing alias/forward files.
5550 26. Fixed a minor format string issue in dbfn.c.
5552 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5554 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5555 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5558 29. When a sender address is verified, it is cached, to save repeating the test
5559 when there is more than one recipient in a message. However, when the
5560 verification involves a callout, it is possible for different callout
5561 options to be set for different recipients. It is too complicated to keep
5562 track of this in the cache, so now Exim always runs a verification when a
5563 callout is required, relying on the callout cache for the optimization.
5564 The overhead is duplication of the address routing, but this should not be
5567 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5568 to be verified with callout=postmaster, and the main callout worked but the
5569 postmaster check failed, the verification correctly failed. However, if a
5570 subsequent RCPT command asked for sender verification *without* the
5571 postmaster check, incorrect caching caused this verification also to fail,
5574 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5575 it was not caching the DNS options (qualify_single, search_parents) that
5576 were used when the lookup failed. A subsequent lookup with different
5577 options therefore always gave the same answer, though there were cases
5578 where it should not have. (Example: a "domains = !$mx_any" option on a
5579 dnslookup router: the "domains" option is always processed without any
5580 widening, but the router might have qualify_single set.) Now Exim uses the
5581 cached value only when the same options are set.
5583 32. Added John Jetmore's "exipick" utility to the distribution.
5585 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5586 than a timeout (e.g. a certificate is required, and is not provided), an
5587 Exim server now closes the connection immediately. Previously it waited for
5588 the client to close - but if the client is SSL, it seems that they each
5589 wait for each other, leading to a delay before one of them times out.
5591 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5592 maintained 0.8.x compatibility because I don't think many are using it, and
5593 it is clearly obsolete.
5595 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5598 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5599 client certificate was expired. A simple patch fixes this, though I don't
5600 understand the full logic of why the verify callback is called multiple
5603 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5604 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5605 which causes problems with some clients (such as the Certicom SSL Plus
5606 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5607 disables the coutermeasure allowing Eudora to connect."
5609 38. Exim was not checking that a write() to a log file succeeded. This could
5610 lead to Bad Things if a log got too big, in particular if it hit a file
5611 size limit. Exim now panics and dies if it cannot write to a log file, just
5612 as it does if it cannot open a log file.
5614 39. Modified OS/Makefile-Linux so that it now contains
5616 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5618 The two -D definitions ensure that Exim is compiled with large file
5619 support, which makes it possible to handle log files that are bigger than
5622 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5623 instance) a domain was checked against a named list that involved a lookup,
5624 causing $domain_data to be set, then another domain was checked against the
5625 same list, then the first domain was re-checked, the value of $domain_data
5626 after the final check could be wrong. In particular, if the second check
5627 failed, it could be set empty. This bug probably also applied to
5630 41. The strip_trailing_dot option was not being applied to the address given
5631 with the -f command-line option.
5633 42. The code for reading a message's header from the spool was incrementing
5634 $received_count, but never initializing it. This meant that the value was
5635 incorrect (doubled) while delivering a message in the same process in which
5636 it was received. In the most common configuration of Exim, this never
5637 happens - a fresh exec is done - but it can happen when
5638 deliver_drop_privilege is set.
5640 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5641 it now includes up to 150 characters of the unexpected data in the log
5644 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5645 and building data strings. The size of both of these buffers was 10 000
5646 bytes - far larger than anybody would *ever* want, thought I. Needless to
5647 say, somebody hit the limit. I have increased the maximum line length to
5648 20 000 and the maximum data length of concatenated lines to 100 000. I have
5649 also fixed two bugs, because there was no checking on these buffers. Tsk,
5650 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5651 buffer is too small.
5653 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5654 lsearch lookups. Now it does.
5656 46. When parsing a route_list item in a manualroute router, a fixed-length
5657 buffer was used for the list of hosts. I made this 1024 bytes long,
5658 thinking that nobody would ever have a list of hosts that long. Wrong.
5659 Somebody had a whole pile of complicated expansion conditions, and the
5660 string was silently truncated, leading to an expansion error. It turns out
5661 that it is easier to change to an unlimited length (owing to other changes
5662 that have happened since this code was originally written) than to build
5663 structure for giving a limitation error. The length of the item that
5664 expands into the list of hosts is now unlimited.
5666 47. The lsearch lookup could not handle data where the length of text line was
5667 more than 4095 characters. Such lines were truncated, leading to shortened
5668 data being returned. It should now handle lines of any length.
5670 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5671 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5674 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5675 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5676 no longer happens in comment lines. A list of replacements is now placed
5677 at the head of all of the source files, except those whose only change is
5678 to replace PERL_COMMAND in the very first #! line.
5680 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5681 messages on the queue, with a bottom-up merge sort, using code contributed
5682 by Michael Haardt. This should make operations like -bp somewhat faster on
5683 large queues. It won't affect queue runners, except when queue_run_in_order
5686 51. Installed eximstats 1.31 in the distribution.
5688 52. Added support for SRV lookups to the dnslookup router.
5690 53. If an ACL referred to $message_body or $message_body_end, the value was not
5691 reset for any messages that followed in the same SMTP session.
5693 54. The store-handling optimization for building very long strings was not
5694 differentiating between the different store pools. I don't think this
5695 actually made any difference in practice, but I've tidied it.
5697 55. While running the routers to verify a sender address, $sender_address
5698 was still set to the sender address. This is wrong, because when routing to
5699 send a bounce to the sender, it would be empty. Therefore, I have changed
5700 it so that, while verifying a sender address, $sender_address is set to <>.
5701 (There is no change to what happens when verifying a recipient address.)
5703 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5704 target A or AAAA records (if not already returned) without resetting the
5705 qualify_single or search_parents options of the DNS resolver. These are
5706 inappropriate in this case because the targets of MX and SRV records must
5707 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5708 target that, when qualified, matched something in the local domain. These
5709 two options are now turned off when doing these lookups.
5711 57. It seems that at least some releases of Reiserfs (which does not have the
5712 concept of a fixed number of inodes) returns zero and not -1 for the
5713 number of available inodes. This interacted badly with check_spool_inodes,
5714 which assumed that -1 was the "no such thing" setting. What I have done is
5715 to check that the total number of inodes is greater than zero before doing
5716 the test of how many are available.
5718 58. When a "warn" ACL statement has a log_message modifier, the message is
5719 remembered, and not repeated. This is to avoid a lot of repetition when a
5720 message has many recipients that cause the same warning to be written.
5721 However, Exim was preserving the list of already written lines for an
5722 entire SMTP session, which doesn't seem right. The memory is now reset if a
5723 new message is started.
5725 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5726 debugging output unless log_rewrite was also set.
5728 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5729 of (int)(handle) when we know that handle contains (void *)(-1).
5731 61. The Exim daemon panic-logs an error return when it closes the incoming
5732 connection. However "connection reset by peer" seems to be common, and
5733 isn't really an error worthy of noting specially, so that particular error
5736 62. When Exim is trying to find all the local interfaces, it used to panic and
5737 die if the ioctl to get the interface flags failed. However, it seems that
5738 on at least one OS (Solaris 9) it is possible to have an interface that is
5739 included in the list of interfaces, but for which you get a failure error
5740 for this call. This happens when the interface is not "plumbed" into a
5741 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5742 failure of the "get flags" call assumes that the interface is down.
5744 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5745 makes life easier for people who are doing arithmetic on fields extracted
5746 from dates, where you often get leading zeros that should not be
5747 interpreted as octal.
5749 64. Added qualify_domain to the redirect router, to override the global
5752 65. If a pathologically long header line contained very many addresses (the
5753 report of this problem mentioned 10 000) and each of them was rewritten,
5754 Exim could use up a very large amount of memory. (It kept on making new
5755 copies of the header line as it rewrote, and never released the old ones.)
5756 At the expense of a bit more processing, the header rewriting function has
5757 been changed so that it no longer eats memory in this way.
5759 66. The generation of the Received: header has been moved from the time that a
5760 message starts to be received, to the time that it finishes. The timestamp
5761 in the Received: header should now be very close to that of the <= log
5762 line. There are two side-effects of this change:
5764 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5765 logged header lines no longer include the local Received: line, because
5766 it has not yet been created. The same applies to a copy of the message
5767 that is returned to a non-SMTP sender when a message is rejected.
5769 (b) When a filter file is tested using -bf, no additional Received: header
5770 is added to the test message. After some thought, I decided that this
5773 This change does not affect the value of $received_for. It is still set
5774 after address rewriting, but before local_scan() is called.
5776 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5778 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5779 gave an unhelpful panic error message, and a defer error. I have managed to
5780 change this behaviour so that it now rejects any supplied certificate,
5781 which seems right, as the list of acceptable certificates is empty.
5783 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5784 gave an unhelpful defer error. I have not managed to make this reject any
5785 supplied certificates, but the error message it gives is "no certificate
5786 supplied", which is not helpful.
5788 70. exigrep's output now also includes lines that are not associated with any
5789 message, but which match the given pattern. Implemented by a patch from
5790 Martin Sluka, which also tidied up the Perl a bit.
5792 71. Recipient callout verification, like sender verification, was using <> in
5793 the MAIL FROM command. This isn't really the right thing, since the actual
5794 sender may affect whether the remote host accepts the recipient or not. I
5795 have changed it to use the actual sender in the callout; this means that
5796 the cache record is now keyed on a recipient/sender pair, not just the
5797 recipient address. There doesn't seem to be a real danger of callout loops,
5798 since a callout by the remote host to check the sender would use <>.
5799 [SEE ABOVE: changed after hitting problems.]
5801 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5802 temporary errors. However, in the case of such a code being given after
5803 the end of a data transmission (i.e. after ".") Exim was failing to write
5804 a retry record for the message. (Yes, there was some broken host that was
5805 actually sending 8xx at this point.)
5807 73. An unknown lookup type in a host list could cause Exim to panic-die when
5808 the list was checked. (An example that provoked this was putting <; in the
5809 middle of a list instead of at the start.) If this happened during a DATA
5810 ACL check, a -D file could be left lying around. This kind of configuration
5811 error no longer causes Exim to die; instead it causes a defer error. The
5812 incident is still logged to the main and panic logs.
5814 74. Buglet left over from Exim 3 conversion. The message "too many messages
5815 in one connection" was written to the rejectlog but not the mainlog, except
5816 when address rewriting (yes!) was being logged.
5818 75. Added write_rejectlog option.
5820 76. When a system filter was run not as root (that is, when system_filter_user
5821 was set), the values of the $n variables were not being returned to the
5822 main process; thus, they were not subsequently available in the $sn
5825 77. Added +return_path_on_delivery log selector.
5827 78. A connection timeout was being treated differently from recipients deferred
5828 when testing hosts_max_try with a message that was older than the host's
5829 retry timeout. (The host should not be counted, thus allowing all hosts to
5830 be tried at least once before bouncing.) This may have been the cause of an
5831 occasionally reported bug whereby a message would remain on the queue
5832 longer than the retry timeout, but would be bounced if a delivery was
5833 forced. I say "may" because I never totally pinned down the problem;
5834 setting up timeout/retry tests is difficult. See also the next item.
5836 79. The ultimate address timeout was not being applied to errors that involved
5837 a combination of host plus message (for example, a timeout on a MAIL
5838 command). When an address resolved to a number of possible hosts, and they
5839 were not all tried for each delivery (e.g. because of hosts_max_try), a
5840 message could remain on the queue longer than the retry timeout.
5842 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5845 81. Fixed an obscure SMTP outgoing bug which required at least the following
5846 conditions: (a) there was another message waiting for the same server;
5847 (b) the server returned 5xx to all RCPT commands in the first message so
5848 that the message was not completed; (c) the server dropped the connection
5849 or gave a negative response to the RSET that Exim sends to abort the
5850 transaction. The observed case was a dropped connection after DATA that had
5851 been sent in pipelining mode. That is, the server had advertised PIPELINING
5852 but was not implementing it correctly. The effect of the bug was incorrect
5853 behaviour, such as trying another host, and this could lead to a crash.
5859 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5860 and daemon.c were passed as pointers to ints; they should have been
5861 pointers to socklen_t variables (which are typically unsigned ints).
5863 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5866 3. Fixed a really odd bug that affected only the testing scheme; patching a
5867 certain fixed string in the binary changed the value of another string that
5868 happened to be identical to the end of the original first string.
5870 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5871 name", it returns that address as the IP address. On some operating
5872 systems (e.g. Solaris), it also passes back the IP address string as the
5873 "host name". However, on others (e.g. Linux), it passes back an empty
5874 string. Exim wasn't checking for this, and was changing the host name to an
5875 empty string, assuming it had been canonicalized.
5877 5. Although rare, it is permitted to have more than one PTR record for a given
5878 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5879 all the names associated with an address, because they do in Solaris.
5880 However, it seems that they do not in Linux for data that comes from the
5881 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5882 I found this out when I moved to a new Linux workstation and tried to run
5883 the Exim test suite.
5885 To get round this problem I have changed the code so that it now does its
5886 own call to the DNS to look up PTR records when searching for a host name.
5887 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5888 addresses that are only in /etc/hosts are still found.
5890 This behaviour is, however, controlled by an option called host_lookup_
5891 order, which defaults to "bydns:byaddr". If people want to use the other
5892 order, or indeed, just use one or the other means of lookup, they can
5893 specify it in this variable.
5895 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5896 some operating systems, this comes back from gethostbyaddr() as an empty
5897 string, and this is what Exim used to test for. However, it seems that in
5898 other systems, "." is yielded. Exim now tests for this case too.
5900 7. The values of check_spool_space and check_log_space are now held internally
5901 as a number of kilobytes instead of an absolute number of bytes. If a
5902 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5903 kilobyte. This means that much larger values can be stored.
5905 8. Exim monitor: an attempt to get the action menu when not actually pointing
5906 at a message produces an empty menu entitled "No message selected". This
5907 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5908 no entries in it ("Shell widget menu has zero width and/or height"). So I
5909 have added a single, blank menu entry in this case.
5911 9. Added ${quote_local_part.
5913 10. MIME decoding is now applied to the contents of Subject: header lines when
5916 11. Now that a reference to $sender_host_address automatically causes a reverse
5917 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5918 host lookup before query-style lookups in lists that might use this
5919 variable. This has therefore been abolished, and the "net-" prefix is no
5920 longer necessary for query-style lookups.
5922 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5923 have been a typo for LFLAGS, so it has been changed.
5925 13. The install script calls Exim with "-C /dev/null" in order to find the
5926 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5927 to be output. However, since Exim outputs its version number before the
5928 error, it didn't break the script. It just looked ugly. I fixed this by
5929 always allowing "-C /dev/null" if the caller is root.
5931 14. Ignore overlarge ACL variable number when reading spool file - insurance
5932 against a later release with more variables having written the file.
5934 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5935 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5936 this, as well as the form without the "IPv6" on the front.
5938 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5939 OS/Makefile-Darwin file.
5941 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5942 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5944 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5945 variable. However, some OS use ints for inodes. I've added cast to long int
5946 to get rid of the compiler warning.
5948 19. I had forgotten to lock out "/../" in configuration file names when
5949 ALT_CONFIG_PREFIX was set.
5951 20. Routers used for verification do not need to specify transports. However,
5952 if such a router generated a host list, and callout was configured, Exim
5953 crashed, because it could not find a port number from the (non-existent)
5954 transport. It now assumes port 25 in this circumstance.
5956 21. Added the -t option to exigrep.
5958 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5959 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5960 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5961 them is set, LOOKUP_LSEARCH is forced.
5963 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5965 24. Added sender and host information to the "rejected by local_scan()" log
5966 line; previously there was no indication of these.
5968 25. Added .include_if_exists.
5970 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5971 Linux. It turns out that not all file systems support this. Apparently some
5972 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5973 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5974 not supported on the file descriptor, is now ignored when Exim is trying to
5975 sync a directory. This applies only to Linux.
5977 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5979 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5980 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5983 29. Added support for maildirsize files from supplied patch (modified a bit).
5985 30. The use of :fail: followed by an empty string could lead Exim to respond to
5986 sender verification failures with (e.g.):
5988 550 Verification failed for <xxx>
5989 550 Sender verify failed
5991 where the first response line was missing the '-' that indicates it is not
5992 the final line of the response.
5994 31. The loop for finding the name of the user that called Exim had a hardwired
5995 limit of 10; it now uses the value of finduser_retries, which is used for
5996 all other user lookups.
5998 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6001 33. Exim was neglecting to zero errno before one call of strtol() when
6002 expanding a string and expecting an integer value. On some systems this
6003 resulted in spurious "integer overflow" errors. Also, it was casting the
6004 result into an int without checking.
6006 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6007 did not work. The code looks as if it has *never* worked, though it appears
6008 to have been documented since at least release 1.62. I have made it work.
6010 35. The "timeout_DNS" error in retry rules, also documented since at least
6011 1.62, also never worked. As it isn't clear exactly what this means, and
6012 clearly it isn't a major issue, I have abolished the feature by treating it
6013 as "timeout", and writing a warning to the main and panic logs.
6015 36. The display of retry rules for -brt wasn't always showing the error code
6018 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6019 timeout_connect_A, timeout_connect_MX.
6021 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6022 to the empty sender.
6024 39. The daemon was not analysing the content of -oX till after it had closed
6025 stderr and disconnected from the controlling terminal. This meant that any
6026 syntax errors were only noted on the panic log, and the return code from
6027 the command was 0. By re-arranging the code a little, I've made the
6028 decoding happen first, so such errors now appear on stderr, and the return
6029 code is 1. However, the actual setting up of the sockets still happens in
6030 the disconnected process, so errors there are still only recorded on the
6033 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6034 connections (as happens on some IP stacks) was logged at start up time as
6035 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6036 it from "IPv6 and IPv4", which means that two separate sockets are being
6039 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6040 says whether AF_INET or AF_INET6 was passed as an argument.
6042 42. Exiwhat output was messed up when time zones were included in log
6045 43. Exiwhat now gives more information about the daemon's listening ports,
6046 and whether -tls-on-connect was used.
6048 44. The "port" option of the smtp transport is now expanded.
6050 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6051 silently ignored. Now an error message is written to the main and panic
6054 46. There's a new ACL modifier called "logwrite" which writes to a log file
6055 as soon as it is encountered.
6057 47. Added $local_user_uid and $local_user_gid at routing time.
6059 48. Exim crashed when trying to verify a sender address that was being
6062 49. Exim was recognizing only a space character after ".include". It now also
6063 recognizes a tab character.
6065 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6066 extracting the relevant information from the specification. The man page no
6067 longer contains scrambled data for the -d option, and I've added a section
6068 at the front about calling Exim under different names.
6070 51. Added "extra_headers" argument to the "mail" command in filter files.
6072 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6075 53. Installed eximstats 1.29.
6077 54. Added transport_filter_timeout as a generic transport option.
6079 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6080 Cc: header lines. This was required by RFC 822, but it not required by RFC
6083 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6084 incoming messages that did not have them. Now it does so only if the
6085 message originates locally, that is, if there is no associated remote host
6086 address. When Resent- header lines are present, this applies to the Resent-
6087 lines rather than the non-Resent- lines.
6089 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6090 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6092 58. Messages for configuration errors now include the name of the main
6093 configuration file - useful now that there may be more than one file in a
6094 list (.included file names were always shown).
6096 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6097 for those rare installations that do not start the daemon as root or run it
6098 setuid root. I've cut out the call to initgroups() if the daemon is not
6101 60. The Exim user and group can now be bound into the binary as text strings
6102 that are looked up at the start of Exim's processing.
6104 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6106 62. Added $mailstore_basename variable.
6108 63. Installed patch to sieve.c from Michael Haardt.
6110 64. When Exim failed to open the panic log after failing to open the main log,
6111 the original message it was trying to log was written to stderr and debug
6112 output, but if they were not available (the usual case in production), it
6113 was lost. Now it is written to syslog before the two lines that record the
6114 failures to open the logs.
6116 65. Users' Exim filters run in subprocesses under the user's uid. It is
6117 possible for a "deliver" command or an alias in a "personal" command to
6118 provoke an address rewrite. If logging of address rewriting is configured,
6119 this fails because the process is not running as root or exim. There may be
6120 a better way of dealing with this, but for the moment (because 4.30 needs
6121 to be released), I have disabled address rewrite logging when running a
6122 filter in a non-root, non-exim process.
6128 1. The buildconfig auxiliary program wasn't quoting the value set for
6129 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6130 not defined. This bug was masked in 4.22 by the effect that was fixed in
6133 2. Some messages that were rejected after a message id was allocated were
6134 shown as "incomplete" by exigrep. It no longer does this for messages that
6135 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6137 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6138 have allow_domain_literals set, the ID did not get logged in the <= line.
6139 Domain literals are now always recognized in Message-ID: header lines.
6141 4. The first argument for a ${extract expansion item is the key name or field
6142 number. Leading and trailing spaces in this item were not being ignored,
6143 causing some misleading effects.
6145 5. When deliver_drop_privilege was set, single queue runner processes started
6146 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6147 same command in the process it spins off) were not dropping privilege.
6149 6. When the daemon running as "exim" started a queue runner, it always
6150 re-executed Exim in the spun-off process. This is a waste of effort when
6151 deliver_drop_privilege is set. The new process now just calls the
6152 queue-runner function directly.
6158 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6161 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6162 ignored. Though the use of -oP was forcing the writing of a pid file, it
6163 was always written to the default place.
6165 3. If the message "no IP address found for host xxxx" is generated during
6166 incoming verification, it is now followed by identification of the incoming
6167 connection (so you can more easily find what provoked it).
6169 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6171 5. Added some features to "harden" Exim a bit more against certain attacks:
6173 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6174 be put in Local/Makefile. This is like the never_users runtime option,
6175 but it cannot be overridden. The default setting is "root".
6177 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6178 prefix string with which any file named in a -C command line option
6181 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6182 is retained for -C and -D only if the caller of Exim is root. Without
6183 it, the exim user may also use -C and -D and retain privilege.
6185 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6186 command line option is disabled.
6188 6. Macro names set by the -D option must start with an upper case letter, just
6189 like macro names defined in the configuration file.
6191 7. Added "dereference=" facility to LDAP.
6193 8. Two instances of the typo "uknown" in the source files are fixed.
6195 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6196 the Configure-Makefile script screwed up while processing it.
6198 10. Incorporated PCRE 4.4.
6200 11. The SMTP synchronization check was not operating right at the start of an
6201 SMTP session. For example, it could not catch a HELO sent before the client
6202 waited for the greeting. There is now a check for outstanding input at the
6203 point when the greeting is written. Because of the duplex, asynchronous
6204 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6205 way, but not yet received, when the check is performed.
6207 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6208 on TCP/IP sockets, because this apparently causes some broken clients to
6211 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6212 unchanged) from the Cygwin maintainer.
6214 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6215 was supported instead of testing for mbx. Effectively a typo.
6217 15. The spa authenticator server code was not checking that the input it
6218 received was valid base64.
6220 16. The debug output line for the "set" modifier in ACLs was not showing the
6221 name of the variable that was being set.
6223 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6225 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6226 lookup on its own. Something else was needed to trigger the lookup. For
6227 example, a match in host_lookup or the need for a host name in a host list.
6228 Now, if $sender_host_name is referenced and the host name has not yet been
6229 looked up, a lookup is performed. If the lookup fails, the variable remains
6230 empty, and $host_lookup_failed is set to "1".
6232 19. Added "eqi" as a case-independent comparison operator.
6234 20. The saslauthd authentication condition could segfault if neither service
6235 nor realm was specified.
6237 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6238 error message that was logged was misleading, and incoming SMTP
6239 connections were dropped. The message is now more accurate, and temporary
6240 errors are given to SMTP connections.
6242 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6243 (except RSET) with a 421 error, until QUIT is received. However, it was
6244 failing to send a response to QUIT.
6246 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6247 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6248 ACL is now run after the helo_try_verify_hosts code.
6250 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6251 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6252 case-independent, so other case variants are also recognized). Apparently
6253 some systems use these upper case variants.
6255 25. If more than two messages were waiting for the same host, and a transport
6256 filter was specified for the transport, Exim sent two messages over the
6257 same TCP/IP connection, and then failed with "socket operation on non-
6258 socket" when it tried to send the third.
6260 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6262 27. The extern definition of crypt16() in expand.c was not being excluded when
6263 the OS had its own crypt16() function.
6265 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6266 as a preferred synonym for return_size_limit, both as an option and as an
6269 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6271 30. Changed the default configuration ACL to relax the local part checking rule
6272 for addresses that are not in any local domains. For these addresses,
6273 slashes and pipe symbols are allowed within local parts, but the sequence
6274 /../ is explicitly forbidden.
6276 31. SPA server authentication was not clearing the challenge buffer before
6279 32. log_message in a "warn" ACL statement was writing to the reject log as
6280 well as to the main log, which contradicts the documentation and doesn't
6281 seem right (because no rejection is happening). So I have stopped it.
6283 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6284 However, I am unable to do any testing of this.
6286 34. Fixed an infelicity in the appendfile transport. When checking directories
6287 for a mailbox, to see if any needed to be created, it was accidentally
6288 using path names with one or more superfluous leading slashes; tracing
6289 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6291 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6292 contents are added to the log line that is written for every discarded
6293 recipient. (Previously a log_message setting was ignored.)
6295 36. The ${quote: operator now quotes the string if it is empty.
6297 37. The install script runs exim in order to find its version number. If for
6298 some reason other than non-existence or emptiness, which it checks, it
6299 could not run './exim', it was installing it with an empty version number,
6300 i.e. as "exim-". This error state is now caught, and the installation is
6303 38. An argument was missing from the function that creates an error message
6304 when Exim fails to connect to the socket for saslauthd authentication.
6305 This could cause Exim to crash, or give a corrupted message.
6307 39. Added isip, isip4, and isip6 to ${if conditions.
6309 40. The ACL variables $acl_xx are now saved with the message, and can be
6310 accessed later in routers, transports, and filters.
6312 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6313 strings in the file are not string-expanded.
6315 42. If a MAIL command specified a SIZE value that was too large to fit into an
6316 int variable, the check against message_size_limit failed. Such values are
6317 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6318 day this will have to be increased, but I don't think I want to be around
6319 when emails are that large.
6326 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6327 iconv() is not standard in FreeBSD.
6329 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6330 IPv6 enabled. The observed symptom was a segmentation fault on return from
6331 the function os_common_find_running_interfaces() in src/os.c.
6333 3. In the check_special_case() function in daemon.c I had used "errno" as an
6334 argument name, which causes warnings on some systems. This was basically a
6335 typo, since it was named "eno" in the comments!
6337 4. The code that waits for the clock to tick (at a resolution of some fraction
6338 of a second) so as to ensure message-id uniqueness was always waiting for
6339 at least one whole tick, when it could have waited for less. [This is
6340 almost certainly not relevant at current processor speeds, where it is
6341 unlikely to ever wait at all. But we try to future-proof.]
6343 5. The function that sleeps for a time interval that includes fractions of a
6344 second contained a race. It did not block SIGALRM between setting the
6345 timer, and suspending (a couple of lines later). If the interval was short
6346 and the sigsuspend() was delayed until after it had expired, the suspension
6347 never ended. On busy systems this could lead to processes getting stuck for
6350 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6351 process, before it forks any delivery processes. The open lookup caching
6352 mechanism meant that the open file or database connection was passed into
6353 the delivery process. The problem was that delivery processes always tidy
6354 up cached lookup data. This could cause a problem for the next delivery
6355 process started by the queue runner, because the external queue runner
6356 process does not know about the closure. So the next delivery process
6357 still has data in the lookup cache. In the case of a file lookup, there was
6358 no problem because closing a file descriptor in a subprocess doesn't affect
6359 the parent. However, if the lookup was caching a connection to a database,
6360 the connection was closed, and the second delivery process was likely to
6361 see errors such as "PGSQL: query failed: server closed the connection
6362 unexpectedly". The problem has been fixed by closing all cached lookups
6363 in a queue runner before running a delivery process.
6365 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6366 seem to have the "const" qualifier which it has on other OS. I've
6369 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6370 *of the same type* (client or server) with the same public name that an
6371 error should be diagnosed.
6373 9. When Exim looked up a host name for an IP address, but failed to find the
6374 original IP address when looking up the host name (a safety check), it
6375 output the message "<ip address> does not match any IP for NULL", which was
6376 confusing, to say the least. The bug was that the host name should have
6377 appeared instead of "NULL".
6379 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6380 user that is built into the binary, and the -C or -D options is used, root
6381 privilege is dropped before the configuration file is read. In addition,
6382 logging is switched to stderr instead of the normal log files. If the
6383 configuration then re-defines the Exim user, the unprivileged environment
6384 is probably not what is expected, so Exim logs a panic warning message (but
6387 However, if deliver_drop_privilege is set, the unprivileged state may well
6388 be exactly what is intended, so the warning has been cut out in that case,
6389 and Exim is allowed to try to write to its normal log files.
6395 1. smtp_return_error_details was not giving details for temporary sender
6396 or receiver verification errors.
6398 2. Diagnose a configuration error if two authenticators have the same public
6401 3. Exim used not to create the message log file for a message until the first
6402 delivery attempt. This could be confusing when incoming messages were held
6403 for policy or load reasons. The message log file is now created at the time
6404 the message is received, and an initial "Received" line is written to it.
6406 4. The automatically generated man page for command line options had a minor
6407 bug that caused no ill effects; however, a more serious problem was that
6408 the procedure for building the man page automatically didn't always
6409 operate. Consequently, release 4.20 contains an out-of-date version. This
6410 shouldn't happen again.
6412 5. When building Exim with embedded Perl support, the script that builds the
6413 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6414 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6416 6. The freeze_tell option was not being used for messages that were frozen on
6417 arrival, either by an ACL or by local_scan().
6419 7. Added the smtp_incomplete_transaction log selector.
6421 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6422 was accepting AUTH without a new EHLO.
6424 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6425 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6428 10. Logging of TCP/IP connections (when configured) now happens in the main
6429 daemon process instead of the child process, so that the TCP/IP connection
6430 count is more accurate (but it can never be perfect).
6432 11. The use of "drop" in a nested ACL was not being handled correctly in the
6433 outer ACL. Now, if condition failure induced by the nested "drop" causes
6434 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6435 or "require"), the connection is dropped.
6437 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6438 that yield "discard" can now be used with an "accept" or a "discard" verb,
6439 but an error is generated for any others (because I can't see a useful way
6440 to define what should happen).
6442 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6443 are now processed in the same way as lines in the Exim configuration file.
6444 In particular, continuation lines are supported.
6446 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6448 15. Added -ti meaning -t -i.
6450 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6451 domains, and warn by logging if others are found.
6453 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6454 returned from GIFCONF when Exim is trying to find the list of interfaces on
6455 a host. The code in os.c has been modified to copy each ifreq to an aligned
6456 structure in all cases.
6458 Also, in some cases, the returned ifreq's were being copied to a 'struct
6459 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6460 means the last couple of bytes of an IPv6 address could be chopped if the
6461 ifreq contained only a normal sockaddr (14 bytes storage).
6463 18. Named domain lists were not supported in the hosts_treat_as_local option.
6464 An entry such as +xxxx was not recognized, and was treated as a literal
6467 19. Ensure that header lines added by a DATA ACL are included in the reject log
6468 if the ACL subsequently rejects the message.
6470 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6471 MD5 (which is deprecated).
6473 21. When testing a filter file using -bf, Exim was writing a message when it
6474 took the sender from a "From " line in the message, but it was not doing so
6475 when it took $return_path from a Return-Path: header line. It now does.
6477 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6478 with a valid header line field name (a series of printing characters
6479 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6481 23. Changed "disc" in the source to "disk" to conform to the documentation and
6482 the book and for uniformity.
6484 24. Ignore Sendmail's -Ooption=value command line item.
6486 25. When execve() failed while trying to run a command in a pipe transport,
6487 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6488 could be confused with a return value of 69 from the command itself. This
6489 has been changed to 127, the value the shell returns if it is asked to run
6490 a non-existent command. The wording for the related log line suggests a
6491 non-existent command as the problem.
6493 26. If received_header_text expands to an empty string, do not add a Received:
6494 header line to the message. (Well, it adds a token one on the spool, but
6495 marks it "old" so that it doesn't get used or transmitted.)
6497 27. Installed eximstats 1.28 (addition of -nt option).
6499 28. There was no check for failure on the call to getsockname() in the daemon
6500 code. This can fail if there is a shortage of resources on the system, with
6501 ENOMEM, for example. A temporary error is now given on failure.
6503 29. Contrary to the C standard, it seems that in some environments, the
6504 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6505 program. Exim now does this explicitly; it affects the formatting of
6506 timestamps using strftime().
6508 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6509 complaints. I've now initialized all the variables, to avoid this.
6511 32. Header lines added by a system filter were not being "seen" during
6512 transport-time rewrites.
6514 33. The info_callback() function passed to OpenSSL is set up with type void
6515 (*)(SSL *, int, int), as described somewhere. However, when calling the
6516 function (actually a macro) that sets it up, the type void(*)() is
6517 expected. I've put in a cast to prevent warnings from picky compilers.
6519 34. If a DNS black list lookup found a CNAME record, but there were no A
6520 records associated with the domain it pointed at, Exim crashed.
6522 35. If a DNS black list lookup returned more than one A record, Exim ignored
6523 all but the first. It now scans all returned addresses if a particular IP
6524 value is being sought. In this situation, the contents of the
6525 $dnslist_value variable are a list of all the addresses, separated by a
6528 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6529 to accept a host name and all its aliases if the forward lookup for any of
6530 them yielded the IP address of the incoming connection. Now it accepts only
6531 those names whose forward lookup yields the correct IP address. Any other
6532 names are discarded. This closes a loophole whereby a rogue DNS
6533 administrator could create reverse DNS records to break through a
6534 wildcarded host restriction in an ACL.
6536 37. If a user filter or a system filter that ran in a subprocess used any of
6537 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6538 the wrong values were passed to the pipe command ($thisaddress had the
6539 value of $0, $0 had the value of $1, etc). This bug was introduced by
6540 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6542 38. Improved the line breaking for long SMTP error messages from ACLs.
6543 Previously, if there was no break point between 40 and 75 characters, Exim
6544 left the rest of the message alone. Two changes have been made: (a) I've
6545 reduced the minimum length to 35 characters; (b) if it can't find a break
6546 point between 35 and 75 characters, it looks ahead and uses the first one
6547 that it finds. This may give the occasional overlong line, but at least the
6548 remaining text gets split now.
6550 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6551 file descriptors might be low, and that setting 1000 would always raise it.
6552 It turns out that in some environments, the limit is already over 1000 and
6553 that lowering it causes trouble. So now Exim takes care not to decrease it.
6555 40. When delivering a message, the value of $return_path is set to $sender_
6556 address at the start of routing (routers may change the value). By an
6557 oversight, this default was not being set up when an address was tested by
6558 -bt or -bv, which affected the outcome if any router or filter referred to
6561 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6562 log only when "message" or "log_message" are set. However, if one of the
6563 conditions was an address verification, or a call to a nested ACL, the
6564 messages generated by the underlying test were being passed through. This
6565 no longer happens. The underlying message is available in $acl_verify_
6566 message for both "message" and "log_message" expansions, so it can be
6567 passed through if needed.
6569 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6570 new expansion $bh_ to give the encoded byte string without charset
6571 translation. Translation happens only if iconv() is available; HAVE_ICONV
6572 indicates this at build time. HEADERS_CHARSET gives the charset to
6573 translate to; headers_charset can change it in the configuration, and
6574 "headers charset" can change it in an individual filter file.
6576 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6577 that creates RFC 2047 encoded "words" labels them as that charset instead
6578 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6579 expansion operator; (ii) when Exim creates a From: line for a local
6580 message; (iii) when a header line is rewritten to include a "phrase" part.
6582 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6583 buggy, causing it to skip the first lines of messages whose message ID
6584 ended in 'D'. This would not have bitten before Exim release 4.14, because
6585 message IDs were unlikely to end in 'D' before then. The effect was to have
6586 incorrect size information for certain domains.
6588 45. #include "config.h" was missing at the start of the crypt16.c module. This
6589 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6592 46. If there was a timeout during a "random" callout check, Exim treated it as
6593 a failure of the random address, and carried on sending RSET and the real
6594 address. If the delay was just some slowness somewhere, the response to the
6595 original RCPT would be taken as a response to RSET and so on, causing
6596 mayhem of various kinds.
6598 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6599 when I implemented it. It didn't allow for the fact that some option values
6600 may legitimately be negative (e.g. size_addition), and it didn't even do
6601 the right test for positive values.
6603 48. Domain names in DNS records are case-independent. Exim always looks them up
6604 in lower case. Some resolvers return domain names in exactly the case they
6605 appear in the zone file, that is, they may contain uppercase letters. Not
6606 all resolvers do this - some return always lower case. Exim was treating a
6607 change of case by a resolver as a change of domain, similar to a widening
6608 of a domain abbreviation. This triggered its re-routing code and so it was
6609 trying to route what was effectively the same domain again. This normally
6610 caused routing to fail (because the router wouldn't handle the domain
6611 twice). Now Exim checks for this case specially, and just changes the
6612 casing of the domain that it ultimately uses when it transmits the message
6615 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6618 50. If a filter generated a file delivery with a non-absolute name (possible if
6619 no home directory exists for the router), the forbid_file option was not
6622 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6623 the existing equality matching.
6625 52. Exim was using ints instead of ino_t variables in some places where it was
6626 dealing with inode numbers.
6628 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6629 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6630 called TMPDIR, and if it finds it is different, it changes its value.
6632 54. The smtp_printf() function is now made available to local_scan() so
6633 additional output lines can be written before returning. There is also an
6634 smtp_fflush() function to enable the detection of a dropped connection.
6635 The variables smtp_input and smtp_batched_input are exported to
6638 55. Changed the default runtime configuration: the message "Unknown user"
6639 has been removed from the ACL, and instead placed on the localuser router,
6640 using the cannot_route_message feature. This means that any verification
6641 failures that generate their own messages won't get overridden. Similarly,
6642 the "Unrouteable address" message that was in the ACL for unverifiable
6643 relay addresses has also been removed.
6645 56. Added hosts_avoid_esmtp to the smtp transport.
6647 57. The exicyclog script was not checking for the esoteric option
6648 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6649 will work only if exicyclog is run under the appropriate euid.
6651 58. Following a discussion on the list, the rules by which Exim recognises line
6652 endings on incoming messages have been changed. The -dropcr and drop_cr
6653 options are now no-ops, retained only for backwards compatibility. The
6654 following line terminators are recognized: LF CRLF CR. However, special
6655 processing applies to CR:
6657 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6658 nor a local message in the state where . is a terminator.
6660 (ii) If a bare CR is encountered in a header line, an extra space is added
6661 after the line terminator so as not to end the header. The reasoning
6662 behind this is that bare CRs in header lines are most likely either
6663 to be mistakes, or people trying to play silly games.
6665 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6666 was being incorrectly given as 18 bytes larger than it should have been.
6667 This is a VOB (very old bug).
6669 60. This may never have affected anything current, but just in case it has:
6670 When the local host is found other than at the start of a list of hosts,
6671 the local host, those with the same MX, and any that follow, are discarded.
6672 When the list in question was part of a longer list of hosts, the following
6673 hosts (not currently being processed) were also being discarded. This no
6674 longer happens. I'm not sure if this situation could ever has previously
6677 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6680 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6681 header. This is something that is recommended in a new Internet Draft, and
6682 is something that is documented as being done by Sendmail. There are two
6683 possible values. For messages generated by the autoreply transport, Exim
6686 Auto-Submitted: auto-replied
6688 whereas for all other generated messages (e.g. bounces) it adds
6690 Auto-Submitted: auto-generated
6692 63. The "personal" condition in filters now includes a test for the
6693 Auto-Submitted: header. If it contains the string "auto-" the message it
6694 not considered personal.
6696 64. Added rcpt_include_affixes as a generic transport option.
6698 65. Added queue_only_override (default true).
6700 66. Added the syslog_duplication option.
6702 67. If what should have been the first header line of a message consisted of
6703 a space followed by a colon, Exim was mis-interpreting it as a header line.
6704 It isn't of course - it is syntactically invalid and should therefore be
6705 treated as the start of the message body. The misbehaviour could have
6706 caused a number of strange effects, including loss of data in subsequent
6707 header lines, and spool format errors.
6709 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6710 client host had authenticated. This control can now be exercised by an ACL
6711 for more flexibility.
6713 69. By default, callouts do not happen when testing with -bh. There is now a
6714 variant, -bhc, which does actually run the callout code, including
6715 consulting and updating the callout cache.
6717 70. Added support for saslauthd authentication, courtesy of Alexander
6720 71. If statvfs() failed on the spool or log directories while checking their
6721 size for availability, Exim confusingly gave the error "space shortage".
6722 Furthermore, in debugging mode it crashed with a floating point exception.
6723 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6724 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6725 serious problem, Exim now writes to the main and panic logs when this
6726 happens, with details of the failure. It then refuses to accept the
6727 incoming message, giving the message "spool directory problem" or "log
6728 directory problem" with a 421 code for SMTP messages.
6730 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6731 0, 1, and 2 exist, because some OS complain for execs without them (see
6732 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6733 descriptors. However, the code omitted to check that the open succeeded,
6734 causing mysterious errors if for some reason the permissions on /dev/null
6735 got screwed. Now Exim writes a message to the main and panic logs, and
6736 bombs out if it can't open /dev/null.
6738 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6739 interact so that it is all more flexible. It is supposed to remain
6740 backwards compatible. Also added extra_local_interfaces.
6742 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6743 to bomb out with an assertion failure - to the client this appears as a
6744 connection drop. This problem occurs in the part of the code that was taken
6745 from the Samba project. Fortunately, the assertion is in a very simple
6746 function, so I have fixed this by reproducing the function inline in the
6747 one place where it is called, and arranging for authentication to fail
6748 instead of killing the process with assert().
6750 75. The SPA client code was not working when the server requested OEM rather
6751 than Unicode encoding.
6753 76. Added code to make require_files with a specific uid setting more usable in
6754 the case where statting the file as root fails - usually a non-root-mounted
6755 NFS file system. When this happens and the failure is EACCES, Exim now
6756 forks a subprocess and does the per-uid checking as the relevant uid.
6758 77. Added process_log_path.
6760 78. If log_file_path was not explicitly set, a setting of check_log_space or
6761 check_log_inodes was ignored.
6763 79. If a space check for the spool or log partitions fails, the incident is now
6764 logged. Of course, in the latter case the data may get lost...
6766 80. Added the %p formatting code to string_format() so that it can be used to
6767 print addresses in debug_print(). Adjusted all the address printing in the
6768 debugging in store.c to use %p rather than %d.
6770 81. There was a concern that a line of code in smtp_in.c could overflow a
6771 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6772 initially expressed, the concern was not well-founded, because trailing
6773 spaces are removed early. However, if the trailing spaces were followed by
6774 a NULL, they did not get removed, so the overflow was possible. Two fixes
6777 (a) I re-wrote the offending code in a cleaner fashion.
6778 (b) If an incoming SMTP command contains a NULL character, it is rejected
6781 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6782 runs initgroups(), so that if the Exim user is in any additional groups,
6783 they will be used during message reception.
6789 The change log for 4.20 and earlier releases has been archived.