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.
8 Exim version 4.89+fixes
9 -----------------------
10 Cherry-Picked from the master development branch
11 -------------------------------------------------
13 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
15 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
16 Disable the session-cache too, which might reduce our load. Since we
17 currrectly use a new context for every connection, both as server and
18 client, there is no benefit for these.
19 GnuTLS appears to not support tickets server-side by default (we don't
20 call gnutls_session_ticket_enable_server()) but client side is enabled
21 by default on recent versions (3.1.3 +) unless the PFS priority string
24 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
25 the check for any unsuccessful recipients did not notice the limit, and
26 erroneously found still-pending ones.
28 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
29 compatible one, to avoid the (poorly documented) possibility of a config
30 file in the working directory redirecting the DB files, possibly correpting
31 some existing file. CVE-2017-10140 assigned for BDB.
33 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
34 cache-hot. Previously, although the result was properly cached, the
35 initial verify call returned a defer.
37 JH/21 Bug 2151 (partial):
38 Avoid using SIZE on the MAIL for a callout verify, on any but
39 the main verify for receipient in uncached-mode.
41 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
42 the right size to place the terminating semicolon on its own folded
43 line, the header hash was calculated to an incorrect value thanks to
44 the (relaxed) space the fold became.
46 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
47 which could crash as a result. This could lead to undeliverable messages.
49 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
50 are unsuable for non-message-specific reasons - eg. connection timeout,
51 banner-time rejection.
53 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
54 matters on fast-turnover and PID-randomising systems, which were getting
55 out-of-order delivery.
57 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
58 a possibly-overlapping copy. The symptom was that "Remote host closed
59 connection in response to HELO" was logged instead of the actual 4xx
62 JH/34 Bug 2199: fix a use-after-free while reading smtp input for header lines.
63 A crafted sequence of BDAT commands could result in in-use memory beeing
64 freed. CVE-2017-16943.
66 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
67 from SMTP input. Previously it was always done; now only done for DATA
68 and not BDAT commands. CVE-2017-16944.
74 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
75 than -2003 did; needs libidn2 in addition to libidn.
77 JH/02 The path option on a pipe transport is now expanded before use.
79 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
80 Patch provided by "Björn", documentation fix added too.
82 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
83 missing a wire-to-host endian conversion.
85 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
86 close after a BDAT command line could be taken as a following command,
87 giving a synch failure. Fix by only checking for synch immediately
88 before acknowledging the chunk.
90 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
91 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
92 macro. Patches provided by Josh Soref.
94 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
95 Previously we did not; the RFC seems ambiguous and VRFY is not listed
96 by IANA as a service extension. However, John Klensin suggests that we
99 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
100 the dkim code may be unix-mode line endings rather than smtp wire-format
101 CRLF, so prepend a CR to any bare LF.
103 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
104 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
106 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
107 the first were themselves being wrongly included in the feed into dkim
108 processing; with most chunk sizes in use this resulted in an incorrect
109 body hash calculated value.
111 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
112 DKIM signature block, for verification. Although advised against by
113 standards it is specifically not ruled illegal.
115 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
117 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
118 missing a body hash (the bh= tag).
120 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
121 It seems that HAProxy sends the Proxy Protocol information in clear and
122 only then does a TLS startup, so do the same.
124 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
125 TCP connections (such as for Spamd) unless the daemon successfully set
126 Fast Open mode on its listening sockets. This fixes breakage seen on
127 too-old kernels or those not configured for Fast Open, at the cost of
128 requiring both directions being enabled for TFO, and TFO never being used
129 by non-daemon-related Exim processes.
131 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
132 endings, at least on the first header line. Try to canonify any that get
133 past that check, despite the cost.
135 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
136 now limited to an arbitrary five deep, while parsing addresses with the
137 strip_excess_angle_brackets option enabled.
139 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
140 instead leave the unprompted TLS handshake in socket buffer for the
141 TLS library to consume.
143 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
145 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
147 JH/16 Drop variables when they go out of scope. Memory management drops a whole
148 region in one operation, for speed, and this leaves assigned pointers
149 dangling. Add checks run only under the testsuite which checks all
150 variables at a store-reset and panics on a dangling pointer; add code
151 explicitly nulling out all the variables discovered. Fixes one known
152 bug: a transport crash, where a dangling pointer for $sending_ip_address
153 originally assigned in a verify callout, is re-used.
155 PP/06 Drop '.' from @INC in various Perl scripts.
157 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
159 PP/08 Reduce a number of compilation warnings under clang; building with
160 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
161 should be warning-free.
163 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
165 HS/01 Fix portability problems introduced by PP/08 for platforms where
166 realloc(NULL) is not equivalent to malloc() [SunOS et al].
168 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
169 chunk. This allows us to accept broken chunked messages. We need a more
170 general solution here.
172 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
173 already-broken messages in the queue.
175 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
177 JH/19 Fix reference counting bug in routing-generated-address tracking.
183 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
184 supports it and a size is available (ie. the sending peer gave us one).
186 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
187 years of being deprecated). The replacements are the ACLs
188 acl_smtp_mime and acl_not_smtp_mime.
190 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
191 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
192 or even an in-clear connection were permitted. Now, if the host lookup
193 was dnssec and dane was requested then the host is only used if the
194 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
195 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
196 if one fails this test.
197 This means that a poorly-configured remote DNS will make it incommunicado;
198 but it protects against a DNS-interception attack on it.
200 JH/04 Bug 1810: make continued-use of an open smtp transport connection
201 non-noisy when a race steals the message being considered.
203 JH/05 If main configuration option tls_certificate is unset, generate a
204 self-signed certificate for inbound TLS connections.
206 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
207 in rewrites and routers.
209 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
210 and logged a warning sing 4.83; now they are a configuration file error.
212 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
213 (lacking @domain). Apply the same qualification processing as RCPT.
215 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
217 JH/10 Support ${sha256:} applied to a string (as well as the previous
220 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
221 a cutthrough deliver is pending, as we always want to make a connection.
222 This also avoids re-routing the message when later placing the cutthrough
223 connection after a verify cache hit.
224 Do not update it with the verify result either.
226 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
227 when routing results in more than one destination address.
229 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
230 signing (which inhibits the cutthrough capability). Previously only
231 the presence of an option was tested; now an expansion evaluating as
232 empty is permissible (obviously it should depend only on data available
233 when the cutthrough connection is made).
235 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
236 the relevant preceding SMTP command did not note the pipelining mode.
238 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
239 Previously they were not counted.
241 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
242 as one having no matching records. Previously we deferred the message
243 that needed the lookup.
245 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
246 distinguished as "(=".
248 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
249 for missing MX records. Previously it only worked for missing A records.
251 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
253 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
254 after the data-go-ahead and data-ack. Patch from Jason Betts.
256 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
257 even for a "none" policy. Patch from Tony Meyer.
259 JH/22 Fix continued use of a connection for further deliveries. If a port was
260 specified by a router, it must also match for the delivery to be
263 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
264 When one of the recipients of a message was unsuitable for the connection
265 (has no matching addresses), we lost track of needing to mark it
266 deferred. As a result mail would be lost.
268 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
270 JH/25 Decoding ACL controls is now done using a binary search; the source code
271 takes up less space and should be simpler to maintain. Merge the ACL
272 condition decode tables also, with similar effect.
274 JH/26 Fix problem with one_time used on a redirect router which returned the
275 parent address unchanged. A retry would see the parent address marked as
276 delivered, so not attempt the (identical) child. As a result mail would
279 JH/27 Fix a possible security hole, wherein a process operating with the Exim
280 UID can gain a root shell. Credit to http://www.halfdog.net/ for
281 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
284 JH/28 Enable {spool,log} filesystem space and inode checks as default.
285 Main config options check_{log,spool}_{inodes,space} are now
286 100 inodes, 10MB unless set otherwise in the configuration.
288 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
289 Previously it only applied to the main-section connection policy
292 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
294 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
295 by me. Added RFC7919 DH primes as an alternative.
297 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
298 are not in the system include path.
300 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
301 GnuTLS, when a session startup failed (eg because the client disconnected)
302 Exim did stdio operations after fclose. This was exposed by a recent
303 change which nulled out the file handle after the fclose.
305 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
306 signed directly by the cert-signing cert, rather than an intermediate
307 OCSP-signing cert. This is the model used by LetsEncrypt.
309 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
311 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
312 an incoming connection.
314 HS/02 Bug 1802: Do not half-close the connection after sending a request
317 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
318 fallback to "prime256v1".
320 JH/34 SECURITY: Use proper copy of DATA command in error message.
321 Could leak key material. Remotely exploitable. CVE-2016-9963.
327 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
328 and 3.4.4 - once the server is enabled to respond to an OCSP request
329 it does even when not requested, resulting in a stapling non-aware
330 client dropping the TLS connection.
332 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
333 support variable-length bit vectors. No functional change.
335 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
336 The I= interface field on outgoing lines is now after the H= remote
337 host field, same as incoming lines. There is a separate
338 outgoing_interface log selector which allows you to disable the
341 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
342 If not running log_selector +smtp_connection the mainlog would be held
343 open indefinitely after a "too many connections" event, including to a
344 deleted file after a log rotate. Leave the per net connection logging
345 leaving it open for efficiency as that will be quickly detected by the
346 check on the next write.
348 HS/01 Bug 1671: Fix post transport crash.
349 Processing the wait-<transport> messages could crash the delivery
350 process if the message IDs didn't exist for some reason. When
351 using 'split_spool_directory=yes' the construction of the spool
352 file name failed already, exposing the same netto behaviour.
354 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
355 mime_regex ACL conditions.
357 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
358 to DSN fail messages (bounces): remote IP, remote greeting, remote response
359 to HELO, local diagnostic string.
361 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
362 log line to debug. Even when configured with a tls authenticator many
363 client connections are expected to not authenticate in this way, so
364 an authenticate fail is not an error.
366 HS/02 Add the Exim version string to the process info. This way exiwhat
367 gives some more detail about the running daemon.
369 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
370 matter for fast-change records such as DNSBLs.
372 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
373 retry record, even if constant. There may be multiple transports with
374 different interface settings and the retry behaviour needs to be kept
377 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
379 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
381 JH/10 Bug 840: fix log_defer_output option of pipe transport
383 JH/11 Bug 830: use same host for all RCPTS of a message, even under
384 hosts_randomize. This matters a lot when combined with mua_wrapper.
386 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
387 ${quote_pgsql:<string>} operator.
389 JH/13 Bug 1708: avoid misaligned access in cached lookup.
391 JH/14 Change header file name for freeradius-client. Relevant if compiling
392 with Radius support; from the Gentoo tree and checked under Fedora.
394 JH/15 Bug 1712: Introduce $prdr_requested flag variable
396 JH/16 Bug 1714: Permit an empty string as expansion result for transport
397 option transport_filter, meaning no filtering.
399 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
401 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
402 defaults to "*" (all hosts). The variable is now available when not built
403 with TLS, default unset, mainly to enable keeping the testsuite sane.
404 If a server certificate is not supplied (via tls_certificate) an error is
405 logged, and clients will find TLS connections fail on startup. Presumably
406 they will retry in-clear.
407 Packagers of Exim are strongly encouraged to create a server certificate
408 at installation time.
410 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
411 with the $config_file variable.
413 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
414 in transport context, after the attempt, and per-recipient. The latter type
415 is per host attempted. The event data is the error message, and the errno
416 information encodes the lookup type (A vs. MX) used for the (first) host,
417 and the trailing two digits of the smtp 4xx response.
419 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
420 to write to mainlog (or rejectlog, paniclog) in the window between file
421 creation and permissions/ownership being changed. Particularly affects
422 installations where exicyclog is run as root, rather than exim user;
423 result is that the running daemon panics and dies.
425 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
427 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
428 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
429 "pri" and "weight". Note that the previous implicit priority given by the
430 list order is no longer honoured.
432 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
435 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
436 by defining SUPPORT_SOCKS.
438 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
439 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
440 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
441 variables are renamed to proxy_{local,external}_{address,port}.
443 JH/27 Move Internationalisation support from Experimental to mainline, enabled
444 for a build by defining SUPPORT_I18N
446 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
447 of the query string, and make ${quote_redis:} do that quoting.
449 JH/29 Move Events support from Experimental to mainline, enabled by default
450 and removable for a build by defining DISABLE_EVENT.
452 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
454 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
455 cached by the daemon.
457 JH/32 Move Redis support from Experimental to mainline, enabled for a build
458 by defining LOOKUP_REDIS. The libhiredis library is required.
460 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
461 keys are given for lookup.
463 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
464 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
465 only supported when built with TLS support. The PolarSSL SHA routines
466 are still used when the TLS library is too old for convenient support.
468 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
469 openssl_options), for security. OpenSSL forces this from version 1.1.0
470 server-side so match that on older versions.
472 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
473 allocation for $value could be released as the expansion processing
474 concluded, but leaving the global pointer active for it.
476 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
477 and to use the domains and local_parts ACL conditions.
479 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
480 incorrectly not doubled on cutthrough transmission, hence seen as a
481 body-termination at the receiving system - resulting in truncated mails.
482 Commonly the sender saw a TCP-level error, and retransmitted the message
483 via the normal store-and-forward channel. This could result in duplicates
484 received - but deduplicating mailstores were liable to retain only the
485 initial truncated version.
487 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
489 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
491 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
492 we're in there, support oversigning also; bug 1309.
494 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
496 HS/04 Add support for keep_environment and add_environment options.
498 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
499 either intentional arithmetic overflow during PRNG, or testing config-
502 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
503 delivery resulted in actual delivery. Cancel cutthrough before DATA
506 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
507 rejects a recipient: pass the reject to the originator.
509 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
510 Many were false-positives and ignorable, but it's worth fixing the
513 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
514 for the new environment-manipulation done at startup. Move the routines
515 from being local to tls.c to being global via the os.c file.
517 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
518 an extract embedded as result-arg for a map, the first arg for extract
519 is unavailable so we cannot tell if this is a numbered or keyed
520 extraction. Accept either.
526 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
529 JH/02 The smtp transport option "multi_domain" is now expanded.
531 JH/03 The smtp transport now requests PRDR by default, if the server offers
534 JH/04 Certificate name checking on server certificates, when exim is a client,
535 is now done by default. The transport option tls_verify_cert_hostnames
536 can be used to disable this per-host. The build option
537 EXPERIMENTAL_CERTNAMES is withdrawn.
539 JH/05 The value of the tls_verify_certificates smtp transport and main options
540 default to the word "system" to access the system default CA bundle.
541 For GnuTLS, only version 3.0.20 or later.
543 JH/06 Verification of the server certificate for a TLS connection is now tried
544 (but not required) by default. The verification status is now logged by
545 default, for both outbound TLS and client-certificate supplying inbound
548 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
551 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
552 Status Notification (bounce) messages are now MIME format per RFC 3464.
553 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
554 under the control of the dsn_advertise_hosts option, and routers may
555 have a dsn_lasthop option.
557 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
558 default, modifiable by a malware= option. The list separator for
559 the options can now be changed in the usual way. Bug 68.
561 JH/10 The smtp_receive_timeout main option is now expanded before use.
563 JH/11 The incoming_interface log option now also enables logging of the
564 local interface on delivery outgoing connections.
566 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
567 if the interface and destination host and port all match.
569 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
572 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
573 Patch from Andrew Lewis.
575 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
576 now supports optional time-restrictions, weighting, and priority
577 modifiers per server. Patch originally by <rommer@active.by>.
579 JH/16 The spamd_address main option now supports a mixed list of local
580 and remote servers. Remote servers can be IPv6 addresses, and
581 specify a port-range.
583 JH/17 Bug 68: The spamd_address main option now supports an optional
584 timeout value per server.
586 JH/18 Bug 1581: Router and transport options headers_add/remove can
587 now have the list separator specified.
589 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
592 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
595 JH/21 Support for the A6 type of dns record is withdrawn.
597 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
598 rather than the verbs used.
600 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
601 from 255 to 1024 chars.
603 JH/24 Verification callouts now attempt to use TLS by default.
605 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
606 are generic router options now. The defaults didn't change.
608 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
609 Original patch from Alexander Shikoff, worked over by JH.
611 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
612 files if ZCAT_COMMAND is not executable.
614 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
616 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
618 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
619 Normally benign, it bites when the pair was led to by a CNAME;
620 modern usage is to not canonicalize the domain to a CNAME target
621 (and we were inconsistent anyway for A-only vs AAAA+A).
623 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
625 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
626 when evaluating $sender_host_dnssec.
628 JH/31 Check the HELO verification lookup for DNSSEC, adding new
629 $sender_helo_dnssec variable.
631 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
633 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
635 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
637 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
638 documented as working, but never had. Support all but $spam_report.
640 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
641 added for tls authenticator.
643 HS/03 Add perl_taintmode main config option
649 TL/01 When running the test suite, the README says that variables such as
650 no_msglog_check are global and can be placed anywhere in a specific
651 test's script, however it was observed that placement needed to be near
652 the beginning for it to behave that way. Changed the runtest perl
653 script to read through the entire script once to detect and set these
654 variables, reset to the beginning of the script, and then run through
655 the script parsing/test process like normal.
657 TL/02 The BSD's have an arc4random API. One of the functions to induce
658 adding randomness was arc4random_stir(), but it has been removed in
659 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
660 function when detected.
662 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
663 cause callback expansion.
665 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
666 syntax errors in an expansion can be treated as a string instead of
667 logging or causing an error, due to the internal use of bool_lax
668 instead of bool when processing it.
670 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
671 server certificates when making smtp deliveries.
673 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
675 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
677 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
679 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
680 Merged patch from Sebastian Wiedenroth.
682 JH/05 Fix results-pipe from transport process. Several recipients, combined
683 with certificate use, exposed issues where response data items split
684 over buffer boundaries were not parsed properly. This eventually
685 resulted in duplicates being sent. This issue only became common enough
686 to notice due to the introduction of connection certificate information,
687 the item size being so much larger. Found and fixed by Wolfgang Breyha.
689 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
690 size buffer was used, resulting in syntax errors when an expansion
693 JH/07 Add support for directories of certificates when compiled with a GnuTLS
694 version 3.3.6 or later.
696 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
697 is EXPERIMENTAL_EVENT, the main-configuration and transport options
698 both become "event_action", the variables become $event_name, $event_data
699 and $event_defer_errno. There is a new variable $verify_mode, usable in
700 routers, transports and related events. The tls:cert event is now also
701 raised for inbound connections, if the main configuration event_action
704 TL/06 In test suite, disable OCSP for old versions of openssl which contained
705 early OCSP support, but no stapling (appears to be less than 1.0.0).
707 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
708 server certificate names available under the smtp transport option
709 "tls_verify_cert_hostname" now do not permit multi-component wildcard
712 JH/10 Time-related extraction expansions from certificates now use the main
713 option "timezone" setting for output formatting, and are consistent
714 between OpenSSL and GnuTLS compilations. Bug 1541.
716 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
717 encoded parameter in the incoming message. Bug 1558.
719 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
720 include certificate info, eximon was claiming there were spoolfile
723 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
725 JH/14 Log delivery-related information more consistently, using the sequence
726 "H=<name> [<ip>]" wherever possible.
728 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
729 are problematic for Debian distribution, omit them from the release
732 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
734 JH/16 Fix string representation of time values on 64bit time_t architectures.
737 JH/17 Fix a null-indirection in certextract expansions when a nondefault
738 output list separator was used.
743 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
744 checkers that were complaining about end of non-void function with no
747 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
748 This was a regression introduced in 4.83 by another bugfix.
750 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
752 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
753 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
759 TF/01 Correctly close the server side of TLS when forking for delivery.
761 When a message was received over SMTP with TLS, Exim failed to clear up
762 the incoming connection properly after forking off the child process to
763 deliver the message. In some situations the subsequent outgoing
764 delivery connection happened to have the same fd number as the incoming
765 connection previously had. Exim would try to use TLS and fail, logging
766 a "Bad file descriptor" error.
768 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
769 utilities have not been installed.
771 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
772 temporary space as the ACL may create new global variables.
774 TL/01 LDAP support uses per connection or global context settings, depending
775 upon the detected version of the libraries at build time.
777 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
778 to extract and use the src ip:port in logging and expansions as if it
779 were a direct connection from the outside internet. PPv2 support was
780 updated based on HAProxy spec change in May 2014.
782 JH/02 Add ${listextract {number}{list}{success}{fail}}.
784 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
785 Properly escape header and check for NULL return.
787 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
790 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
792 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
793 characters in header names, implemented as a verify condition.
794 Contributed by Michael Fischer v. Mollard.
796 TL/05 Rename SPF condition results err_perm and err_temp to standardized
797 results permerror and temperror. Previous values are deprecated but
798 still accepted. In a future release, err_perm and err_temp will be
799 completely removed, which will be a backward incompatibility if the
800 ACL tests for either of these two old results. Patch contributed by
801 user bes-internal on the mailing list.
803 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
805 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
806 selectors, in both main and reject logs.
808 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
811 JH/07 Add malware type "sock" for talking to simple daemon.
813 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
815 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
816 routers/transports under cutthrough routing.
818 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
819 numbers. Touch up "bool" conditional to keep the same definition.
821 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
823 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
825 JH/12 Expand items in router/transport headers_add or headers_remove lists
826 individually rather than the list as a whole. Bug 1452.
828 Required for reasonable handling of multiple headers_ options when
829 they may be empty; requires that headers_remove items with embedded
830 colons must have them doubled (or the list-separator changed).
832 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
833 view the policy declared in the DMARC record. Currently, $dmarc_status
834 is a combined value of both the record presence and the result of the
837 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
839 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
840 dnslookup router and the smtp transport (applying to the forward
843 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
844 of ldap servers used for a specific lookup. Patch provided by Heiko
847 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
848 New variable $lookup_dnssec_authenticated for observability.
850 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
851 Patch submitted by Lars Timman.
853 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
855 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
856 Requires trusted mode and valid format message id, aborts otherwise.
857 Patch contributed by Heiko Schlichting.
859 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
860 certextract with support for various fields. Bug 1358.
862 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
863 is requested by default, modifiable by smtp transport option
866 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
867 operate on certificate variables to give certificate fingerprints
868 Also new ${sha256:cert_variable}.
870 JH/23 The PRDR feature is moved from being Experimental into the mainline.
872 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
873 Christian Aistleitner.
875 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
877 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
878 file. Patch from Wolfgang Breyha.
880 JH/25 Expand the coverage of the delivery $host and $host_address to
881 client authenticators run in verify callout. Bug 1476.
883 JH/26 Port service names are now accepted for tls_on_connect_ports, to
884 align with daemon_smtp_ports. Bug 72.
886 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
887 support and error reporting did not work properly.
889 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
890 and is readable. Patch from Andrew Colin Kissa.
892 TL/14 Enhance documentation of ${run expansion and how it parses the
893 commandline after expansion, particularly in the case when an
894 unquoted variable expansion results in an empty value.
896 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
898 PP/02 Fix internal collision of T_APL on systems which support RFC3123
899 by renaming away from it. Addresses GH issue 15, reported by
902 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
904 TL/15 SECURITY: prevent double expansion in math comparison functions
905 (can expand unsanitized data). Not remotely exploitable.
912 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
914 PP/02 Make -n do something, by making it not do something.
915 When combined with -bP, the name of an option is not output.
917 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
920 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
921 $sender_host_name and config options to manage this, and basic check
924 PP/05 DSCP support for outbound connections and control modifier for inbound.
926 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
927 (Only plugin which currently uses this is kerberos4, which nobody should
928 be using, but we should make it available and other future plugins might
929 conceivably use it, even though it would break NAT; stuff *should* be
930 using channel bindings instead).
932 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
933 name; added for Sendmail compatibility; requires admin caller.
934 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
935 just ignore it); requires trusted caller.
936 Also parse but ignore: -Ac -Am -X<logfile>
939 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
941 TL/02 Add +smtp_confirmation as a default logging option.
943 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
944 Patch by Magnus Holmgren from 2007-02-20.
946 TL/04 Bugzilla 1281 - Spec typo.
947 Bugzilla 1283 - Spec typo.
948 Bugzilla 1290 - Spec grammar fixes.
950 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
952 TL/06 Add Experimental DMARC support using libopendmarc libraries.
954 TL/07 Fix an out of order global option causing a segfault. Reported to dev
955 mailing list by by Dmitry Isaikin.
957 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
959 JH/02 Support "G" suffix to numbers in ${if comparisons.
961 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
963 NM/01 Bugzilla 1197 - Spec typo
964 Bugzilla 1196 - Spec examples corrections
966 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
968 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
969 gnutls_enable_pkcs11, but renamed to more accurately indicate its
972 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
973 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
975 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
976 "acl {{name}{arg}...}", and optional args on acl condition
979 JH/05 Permit multiple router/transport headers_add/remove lines.
981 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
983 JH/07 Avoid using a waiting database for a single-message-only transport.
984 Performance patch from Paul Fisher. Bugzilla 1262.
986 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
989 JH/09 Add $headers_added variable, with content from use of ACL modifier
990 add_header (but not yet added to the message). Bugzilla 199.
992 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
993 Pulled from Bugzilla 817 by Wolfgang Breyha.
995 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
997 (nb: this is the same fix as in Exim 4.80.1)
999 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1002 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1005 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1007 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1008 advertise SMTP AUTH mechanism to us, instead of a generic
1009 protocol violation error. Also, make Exim more robust to bad
1010 data from the Dovecot auth socket.
1012 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1014 When a queue runner is handling a message, Exim first routes the
1015 recipient addresses, during which it prunes them based on the retry
1016 hints database. After that it attempts to deliver the message to
1017 any remaining recipients. It then updates the hints database using
1020 So if a recipient address works intermittently, it can get repeatedly
1021 deferred at routing time. The retry hints record remains fresh so the
1022 address never reaches the final cutoff time.
1024 This is a fairly common occurrence when a user is bumping up against
1025 their storage quota. Exim had some logic in its local delivery code
1026 to deal with this. However it did not apply to per-recipient defers
1027 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1029 This change adds a proper retry rule check during routing so that the
1030 final cutoff time is checked against the message's age. We only do
1031 this check if there is an address retry record and there is not a
1032 domain retry record; this implies that previous attempts to handle
1033 the address had the retry_use_local_parts option turned on. We use
1034 this as an approximation for the destination being like a local
1035 delivery, as in LMTP.
1037 I suspect this new check makes the old local delivery cutoff check
1038 redundant, but I have not verified this so I left the code in place.
1040 TF/02 Correct gecos expansion when From: is a prefix of the username.
1042 Test 0254 submits a message to Exim with the header
1046 When I ran the test suite under the user fanf2, Exim expanded
1047 the header to contain my full name, whereas it should have added
1048 a Resent-Sender: header. It erroneously treats any prefix of the
1049 username as equal to the username.
1051 This change corrects that bug.
1053 GF/01 DCC debug and logging tidyup
1054 Error conditions log to paniclog rather than rejectlog.
1055 Debug lines prefixed by "DCC: " to remove any ambiguity.
1057 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1059 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1060 Bug spotted by Jeremy Harris; was flawed since initial commit.
1061 Would have resulted in OCSP responses post-SNI triggering an Exim
1062 NULL dereference and crash.
1064 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1066 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1067 Bug detection, analysis and fix by Samuel Thibault.
1068 Bugzilla 1331, Debian bug #698092.
1070 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1072 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1073 Server implementation by Todd Lyons, client by JH.
1074 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1075 config variable "prdr_enable" controls whether the server
1076 advertises the facility. If the client requests PRDR a new
1077 acl_data_smtp_prdr ACL is called once for each recipient, after
1078 the body content is received and before the acl_smtp_data ACL.
1079 The client is controlled by both of: a hosts_try_prdr option
1080 on the smtp transport, and the server advertisement.
1081 Default client logging of deliveries and rejections involving
1082 PRDR are flagged with the string "PRDR".
1084 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1085 fclose(). Diagnosis by Todd Lyons.
1087 PP/17 Update configure.default to handle IPv6 localhost better.
1088 Patch by Alain Williams (plus minor tweaks).
1091 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1092 This is now consistent with GnuTLS, and is now documented: the
1093 previous undocumented portable approach to treating the option as
1094 unset was to force an expansion failure. That still works, and
1095 an empty string is now equivalent.
1097 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1098 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1099 not performing validation itself.
1101 PP/20 Added force_command boolean option to pipe transport.
1102 Patch from Nick Koston, of cPanel Inc.
1104 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1107 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1109 PP/21 Fix eximon continuous updating with timestamped log-files.
1110 Broken in a format-string cleanup in 4.80, missed when I repaired the
1111 other false fix of the same issue.
1112 Report and fix from Heiko Schlichting.
1115 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1116 Report from Prashanth Katuri.
1118 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1119 It's SecureTransport, so affects any MacOS clients which use the
1120 system-integrated TLS libraries, including email clients.
1122 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1123 using a MIME ACL for non-SMTP local injection.
1124 Report and assistance in diagnosis by Warren Baker.
1126 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1128 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1130 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1131 last id that failed so it may be referenced in subsequent ACL's.
1133 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1136 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1137 ldap library initialization, allowing self-signed CA's to be
1138 used. Also properly sets require_cert option later in code by
1139 using NULL (global ldap config) instead of ldap handle (per
1140 session). Bug diagnosis and testing by alxgomz.
1142 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1143 the src/util/ subdirectory.
1145 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1146 renamed to Transport Post Delivery Action by Jeremy Harris, as
1149 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1150 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1151 redis_servers = needs to be configured which will be used by the redis
1152 lookup. Patch from Warren Baker, of The Packet Hub.
1154 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1156 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1157 hostname or reverse DNS when processing a host list. Used suggestions
1158 from multiple comments on this bug.
1160 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1162 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1163 Missed a few lines, added it to make the runtest require no keyboard
1166 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1167 contains upper case chars. Make router use caseful_local_part.
1169 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1170 support when GnuTLS has been built with p11-kit.
1176 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1178 This, or similar/improved, will also be change PP/11 of 4.82.
1184 PP/01 Handle short writes when writing local log-files.
1185 In practice, only affects FreeBSD (8 onwards).
1186 Bugzilla 1053, with thanks to Dmitry Isaikin.
1188 NM/01 Bugzilla 949 - Documentation tweak
1190 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1193 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1195 PP/02 Implemented gsasl authenticator.
1197 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1199 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1200 `pkg-config foo` for cflags/libs.
1202 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1203 with rest of GSASL and with heimdal_gssapi.
1205 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1206 `pkg-config foo` for cflags/libs for the TLS implementation.
1208 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1209 properties get this fed in as external SSF. A number of robustness
1210 and debugging improvements to the cyrus_sasl authenticator.
1212 PP/08 cyrus_sasl server now expands the server_realm option.
1214 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1215 Patch by Jeremy Harris.
1217 PP/10 Added dbmjz lookup type.
1219 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1221 PP/12 MAIL args handles TAB as well as SP, for better interop with
1222 non-compliant senders.
1223 Analysis and variant patch by Todd Lyons.
1225 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1226 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1227 Patch from Dirk Mueller <dmueller@suse.com>
1229 PP/13 tls_peerdn now print-escaped for spool files.
1230 Observed some $tls_peerdn in wild which contained \n, which resulted
1231 in spool file corruption.
1233 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1234 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1235 or write after TLS renegotiation, which otherwise led to messages
1238 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1239 as a tracking header (ie: a signed header comes before the signature).
1240 Patch from Wolfgang Breyha.
1242 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1243 comma-sep list; embedded commas doubled.
1245 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1247 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1249 Report and patch from Dmitry Banschikov.
1251 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1252 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1253 failures. We appear to now support TLS1.1+ with Exim.
1255 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1256 lets Exim select keys and certificates based upon TLS SNI from client.
1257 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1258 before an outbound SMTP session. New log_selector, +tls_sni.
1260 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1261 NULL dereference. Report and patch from Alun Jones.
1263 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1264 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1265 Not seeing resolver debug output on NetBSD, but suspect this is a
1266 resolver implementation change.
1268 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1269 Left warnings. Added "eximon gdb" invocation mode.
1271 PP/21 Defaulting "accept_8bitmime" to true, not false.
1273 PP/22 Added -bw for inetd wait mode support.
1275 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1276 locate the relevant includes and libraries. Made this the default.
1278 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1279 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1281 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1282 This may cause build issues on older platforms.
1284 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1285 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1286 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1287 Added SNI support via GnuTLS too.
1288 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1290 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1292 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1293 Applied second patch from Janne, implementing suggestion to default
1294 multiple-strings-in-record handling to match SPF spec.
1296 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1298 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1299 read-only, out of scope).
1300 Patch from Wolfgang Breyha, report from Stuart Northfield.
1302 PP/29 Fix three issues highlighted by clang analyser static analysis.
1303 Only crash-plausible issue would require the Cambridge-specific
1304 iplookup router and a misconfiguration.
1305 Report from Marcin Mirosław.
1307 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1309 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1310 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1311 As part of this, removing so much warning spew let me fix some minor
1312 real issues in debug logging.
1314 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1315 assignment on my part. Fixed.
1317 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1318 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1319 Janne Snabb (who went above and beyond: thank you).
1321 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1322 string otherwise requires a connection and a bunch more work and it's
1323 relatively easy to get wrong. Should also expose TLS library linkage
1326 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1327 64-bit ${eval} (JH/03).
1329 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1330 GNU libc to support some of the 64-bit stuff, should not lead to
1331 conflicts. Defined before os.h is pulled in, so if a given platform
1332 needs to override this, it can.
1334 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1335 protection layer was required, which is not implemented.
1336 Bugzilla 1254, patch from Wolfgang Breyha.
1338 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1339 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1340 tls_dhparam take prime identifiers. Also unbreak combination of
1341 OpenSSL+DH_params+TLSSNI.
1343 PP/39 Disable SSLv2 by default in OpenSSL support.
1349 PP/01 Solaris build fix for Oracle's LDAP libraries.
1350 Bugzilla 1109, patch from Stephen Usher.
1352 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1354 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1357 TF/02 Fix a couple more cases where we did not log the error message
1358 when unlink() failed. See also change 4.74-TF/03.
1360 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1361 lock up or crash if it happened to be inside a call to libc when it
1362 got a SIGUSR1 from exiwhat.
1364 The SIGUSR1 handler appends the current process status to the process
1365 log which is later printed by exiwhat. It used to use the general
1366 purpose logging code to do this, but several functions it calls are
1367 not safe for signals.
1369 The new output code in the SIGUSR1 handler is specific to the process
1370 log, and simple enough that it's easy to inspect for signal safety.
1371 Removing some special cases also simplifies the general logging code.
1372 Removing the spurious timestamps from the process log simplifies
1375 TF/04 Improved ratelimit ACL condition.
1377 The /noupdate option has been deprecated in favour of /readonly which
1378 has clearer semantics. The /leaky, /strict, and /readonly update modes
1379 are mutually exclusive. The update mode is no longer included in the
1380 database key; it just determines when the database is updated. (This
1381 means that when you upgrade Exim will forget old rate measurements.)
1383 Exim now checks that the per_* options are used with an update mode that
1384 makes sense for the current ACL. For example, when Exim is processing a
1385 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1386 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1387 must specify per_mail/readonly. If you omit the update mode it defaults to
1388 /leaky where that makes sense (as before) or /readonly where required.
1390 The /noupdate option is now undocumented but still supported for
1391 backwards compatibility. It is equivalent to /readonly except that in
1392 ACLs where /readonly is required you may specify /leaky/noupdate or
1393 /strict/noupdate which are treated the same as /readonly.
1395 A useful new feature is the /count= option. This is a generalization
1396 of the per_byte option, so that you can measure the throughput of other
1397 aggregate values. For example, the per_byte option is now equivalent
1398 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1400 The per_rcpt option has been generalized using the /count= mechanism
1401 (though it's more complicated than the per_byte equivalence). When it is
1402 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1403 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1404 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1405 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1406 Note that using per_rcpt with a non-readonly update mode in more than
1407 one ACL will cause the recipients to be double-counted. (The per_mail
1408 and per_byte options don't have this problem.)
1410 The handling of very low rates has changed slightly. If the computed rate
1411 is less than the event's count (usually one) then this event is the first
1412 after a long gap. In this case the rate is set to the same as this event's
1413 count, so that the first message of a spam run is counted properly.
1415 The major new feature is a mechanism for counting the rate of unique
1416 events. The new per_addr option counts the number of different
1417 recipients that someone has sent messages to in the last time period. It
1418 behaves like per_rcpt if all the recipient addresses are different, but
1419 duplicate recipient addresses do not increase the measured rate. Like
1420 the /count= option this is a general mechanism, so the per_addr option
1421 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1422 example, measure the rate that a client uses different sender addresses
1423 with the options per_mail/unique=$sender_address. There are further
1424 details in the main documentation.
1426 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1428 TF/06 Removed a few PCRE remnants.
1430 TF/07 Automatically extract Exim's version number from tags in the git
1431 repository when doing development or release builds.
1433 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1434 Bugzilla 879. Patch from Paul Fisher.
1436 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1437 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1440 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1442 PP/05 Variable $av_failed, true if the AV scanner deferred.
1443 Bugzilla 1078. Patch from John Horne.
1445 PP/06 Stop make process more reliably on build failure.
1446 Bugzilla 1087. Patch from Heiko Schlittermann.
1448 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1449 Bugzilla 1089. Patch from Heiko Schlittermann.
1451 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1452 Bugzilla 1131. Patch from Holger Weiß.
1454 PP/09 Handle IPv6 addresses with SPF.
1455 Bugzilla 860. Patch from Wolfgang Breyha.
1457 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1459 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1462 PP/11 match_* no longer expand right-hand-side by default.
1463 New compile-time build option, EXPAND_LISTMATCH_RHS.
1464 New expansion conditions, "inlist", "inlisti".
1466 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1468 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1470 PP/14 fix log_write() format string regression from TF/03.
1471 Bugzilla 1152. Patch from Dmitry Isaikin.
1477 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1479 PP/02 Harmonised TLS library version reporting; only show if debugging.
1480 Layout now matches that introduced for other libraries in 4.74 PP/03.
1482 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1484 PP/04 New "dns_use_edns0" global option.
1486 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1489 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1490 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1492 TK/01 Updated PolarSSL code to 0.14.2.
1493 Bugzilla 1097. Patch from Andreas Metzler.
1495 PP/07 Catch divide-by-zero in ${eval:...}.
1496 Fixes bugzilla 1102.
1498 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1501 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1502 format-string attack -- SECURITY: remote arbitrary code execution.
1504 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1505 time unintentionally subject to list matching rules, letting the header
1506 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1507 arbitrary string expansion). This allowed for information disclosure.
1509 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1510 INT_MIN/-1 -- value coerced to INT_MAX.
1516 NM/01 Workaround for PCRE version dependency in version reporting
1519 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1520 This fixes portability to compilers other than gcc, notably
1521 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1523 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1524 makefiles for portability to HP-UX and POSIX correctness.
1526 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1527 Also via indented variable definition in the Makefile.
1528 (Debugging by Oliver Heesakkers).
1530 PP/02 Restore caching of spamd results with expanded spamd_address.
1531 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1533 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1534 Improves build reliability. Fix from: Frank Elsner
1536 NM/02 Fix wide character breakage in the rfc2047 coding
1537 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1539 NM/03 Allow underscore in dnslist lookups
1540 Fixes bug 1026. Patch from Graeme Fowler
1542 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1543 Code patches from Adam Ciarcinski of NetBSD.
1545 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1548 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1549 is logged, to avoid truncation. Patch from John Horne.
1551 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1552 Patch from Jakob Hirsch.
1554 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1555 SQL string expansion failure details.
1556 Patch from Andrey Oktyabrski.
1558 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1559 Patch from Simon Arlott.
1561 PP/09 New lookups functionality failed to compile on old gcc which rejects
1562 extern declarations in function scope.
1563 Patch from Oliver Fleischmann
1565 PP/10 Use sig_atomic_t for flags set from signal handlers.
1566 Check getgroups() return and improve debugging.
1567 Fixed developed for diagnosis in bug 927 (which turned out to be
1570 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1571 Patch from Mark Zealey.
1573 PP/12 Bugzilla 1056: Improved spamd server selection.
1574 Patch from Mark Zealey.
1576 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1577 Based on patch from Heiko Schlittermann.
1579 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1580 Patch from Uwe Doering, sign-off by Michael Haardt.
1582 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1583 variable declaration deep within a block. Bug and patch from
1586 PP/15 lookups-Makefile IRIX compatibility coercion.
1588 PP/16 Make DISABLE_DKIM build knob functional.
1590 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1591 Patch by Simon Arlott
1593 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1594 variable argument macros. Our copy now differs from upstream.
1600 TF/01 Failure to get a lock on a hints database can have serious
1601 consequences so log it to the panic log.
1603 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1604 controlled using the smtp_confirmation log selector.
1606 TF/03 Include the error message when we fail to unlink a spool file.
1608 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1609 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1610 for maintaining out-of-tree patches for some time.
1612 PP/01 Bugzilla 139: Documentation and portability issues.
1613 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1614 Handle per-OS dynamic-module compilation flags.
1616 PP/02 Let /dev/null have normal permissions.
1617 The 4.73 fixes were a little too stringent and complained about the
1618 permissions on /dev/null. Exempt it from some checks.
1619 Reported by Andreas M. Kirchwitz.
1621 PP/03 Report version information for many libraries, including
1622 Exim version information for dynamically loaded libraries. Created
1623 version.h, now support a version extension string for distributors
1624 who patch heavily. Dynamic module ABI change.
1626 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1627 privilege escalation vulnerability whereby the Exim run-time user
1628 can cause root to append content of the attacker's choosing to
1631 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1634 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1635 If dropping privileges for untrusted macros, we disabled normal logging
1636 on the basis that it would fail; for the Exim run-time user, this is not
1637 the case, and it resulted in successful deliveries going unlogged.
1638 Fixed. Reported by Andreas Metzler.
1644 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1645 only prepend for the Resent-* case. Fixes regression introduced in
1646 Exim 4.70 by NM/22 for Bugzilla 607.
1648 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1649 increasing numbers of administrators be bitten by this.
1651 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1653 PP/03 Bugzilla 994: added openssl_options main configuration option.
1655 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1657 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1659 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1661 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1662 without a peer certificate, leading to a segfault because of an
1663 assumption that peers always have certificates. Be a little more
1664 paranoid. Problem reported by Martin Tscholak.
1666 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1667 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1668 NB: ClamAV planning to remove STREAM in "middle of 2010".
1669 CL also introduces -bmalware, various -d+acl logging additions and
1670 more caution in buffer sizes.
1672 PP/09 Implemented reverse_ip expansion operator.
1674 PP/10 Bugzilla 937: provide a "debug" ACL control.
1676 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1678 PP/12 Bugzilla 973: Implement --version.
1680 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1682 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1684 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1686 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1687 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1688 ignore trailing whitespace.
1690 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1692 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1695 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1696 Notification from Dr Andrew Aitchison.
1698 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1699 ExtendedDetectionInfo response format.
1700 Notification from John Horne.
1702 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1705 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1706 XSL and documented dependency on system catalogs, with examples of how
1709 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1712 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1713 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1714 configuration file which is writeable by the Exim user or group.
1716 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1717 of configuration files to cover files specified with the -C option if
1718 they are going to be used with root privileges, not just the default
1721 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1722 option (effectively making it always true).
1724 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1725 files to be used while preserving root privileges.
1727 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1728 that rogue child processes cannot use them.
1730 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1731 run-time user, instead of root.
1733 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1734 Exim run-time user without dropping privileges.
1736 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1737 result string, instead of calling string_vformat() twice with the same
1740 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1741 for other users. Others should always drop root privileges if they use
1742 -C on the command line, even for a whitelisted configure file.
1744 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1746 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1752 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1753 $data_path, and $header_path variables; fixed documentation bugs and
1756 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1757 exipick to access non-standard spools, including the "frozen" queue
1760 NM/01 Bugzilla 965: Support mysql stored procedures.
1761 Patch from Alain Williams
1763 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1765 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1766 Patch from Andreas Metzler
1768 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1769 Patch from Kirill Miazine
1771 NM/05 Bugzilla 671: Added umask to procmail example.
1773 JJ/03 installed exipick 20100323.0, fixing doc bug
1775 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1776 directory. Notification and patch from Dan Rosenberg.
1778 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1780 TK/02 Improve log output when DKIM signing operation fails.
1782 MH/01 Treat the transport option dkim_domain as a colon separated
1783 list, not as a single string, and sign the message with each element,
1784 omitting multiple occurences of the same signer.
1786 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1787 Bugzilla 985, 986. Patch by Simon Arlott
1789 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1790 Patch by Simon Arlott
1792 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1793 MBX locking. Notification from Dan Rosenberg.
1799 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1801 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1803 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1805 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1807 NM/04 Bugzilla 915: Flag broken perl installation during build.
1813 TK/01 Added patch by Johannes Berg that expands the main option
1814 "spamd_address" if it starts with a dollar sign.
1816 TK/02 Write list of recipients to X-Envelope-Sender header when building
1817 the mbox-format spool file for content scanning (suggested by Jakob
1820 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1821 (http://www.dcc-servers.net/) support via dccifd. Activated by
1822 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1824 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1825 by Mark Daniel Reidel <mr@df.eu>.
1827 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1828 When building exim an external PCRE library is now needed -
1829 PCRE is a system library on the majority of modern systems.
1830 See entry on PCRE_LIBS in EDITME file.
1832 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1833 conversation. Added nologin parameter to request.
1834 Patch contributed by Kirill Miazine.
1836 TF/01 Do not log submission mode rewrites if they do not change the address.
1838 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1840 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1841 log files in place. Contributed by Roberto Lima.
1843 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1845 TF/03 Bugzilla 615: When checking the local_parts router precondition
1846 after a local_part_suffix or local_part_prefix option, Exim now
1847 does not use the address's named list lookup cache, since this
1848 contains cached lookups for the whole local part.
1850 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1851 Robert Millan. Documentation is in experimental-spec.txt.
1853 TF/04 Bugzilla 668: Fix parallel build (make -j).
1855 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1857 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1858 Patch provided by Jan Srzednicki.
1860 TF/05 Leading white space used to be stripped from $spam_report which
1861 wrecked the formatting. Now it is preserved.
1863 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1864 that they are available at delivery time.
1866 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1868 TF/08 TLS error reporting now respects the incoming_interface and
1869 incoming_port log selectors.
1871 TF/09 Produce a more useful error message if an SMTP transport's hosts
1872 setting expands to an empty string.
1874 NM/06 Bugzilla 744: EXPN did not work under TLS.
1875 Patch provided by Phil Pennock.
1877 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1878 Patch provided by Richard Godbee.
1880 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1881 acl_smtp_notquit, added index entry.
1883 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1884 Patch provided by Eugene Bujak.
1886 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1887 accept(). Patch provided by Maxim Dounin.
1889 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1890 Patch provided by Phil Pennock.
1892 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1894 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1895 Patch provided by Brad "anomie" Jorsch.
1897 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1898 Patch provided by Dean Brooks.
1900 TK/05 Add native DKIM support (does not depend on external libraries).
1902 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1903 Patch provided by Graeme Fowler.
1905 NM/16 Bugzilla 851: Documentation example syntax fix.
1907 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1909 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1912 NM/19 Bugzilla 745: TLS version reporting.
1913 Patch provided by Phil Pennock.
1915 NM/20 Bugzilla 167: bool: condition support.
1916 Patch provided by Phil Pennock.
1918 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1919 clients. Patch provided by Phil Pennock.
1921 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1922 Patch provided by Brad "anomie" Jorsch.
1924 NM/23 Bugzilla 687: Fix misparses in eximstats.
1925 Patch provided by Heiko Schlittermann.
1927 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1928 Patch provided by Heiko Schlittermann.
1930 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1931 plus update to original patch.
1933 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1935 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1936 Patch provided by David Brownlee.
1938 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1940 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1942 NM/30 Bugzilla 888: TLS documentation bugfixes.
1944 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1946 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1947 Unlike the original bugzilla I have changed all shell scripts in src tree.
1949 NM/33 Bugzilla 898: Transport filter timeout fix.
1950 Patch by Todd Rinaldo.
1952 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1953 Patch by Serge Demonchaux.
1955 NM/35 Bugzilla 39: Base64 decode bug fixes.
1956 Patch by Jakob Hirsch.
1958 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1960 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1962 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1964 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1970 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1971 ALT-N's libdkim that I have put here:
1972 http://duncanthrax.net/exim-experimental/
1974 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1975 were called 'true' and it seems that C99 defines that as a reserved
1976 keyword to be used with 'bool' variable types. That means you could
1977 not include C99-style headers which use bools without triggering
1978 build errors in sieve.c.
1980 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1981 as mailq or other aliases. Changed the --help handling significantly
1982 to do whats expected. exim_usage() emits usage/help information.
1984 SC/01 Added the -bylocaldomain option to eximstats.
1986 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1988 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1990 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1996 PH/01 Another patch from the Sieve maintainer.
1998 PH/02 When an IPv6 address is converted to a string for single-key lookup
1999 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2000 dots are used instead of colons so that keys in lsearch files need not
2001 contain colons. This was done some time before quoting was made available
2002 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2003 (notated using the quote facility) so as to distinguish them from IPv4
2004 keys. This meant that lookups for IP addresses in host lists did not work
2005 for iplsearch lookups.
2007 This has been fixed by arranging for IPv6 addresses to be expressed with
2008 colons if the lookup type is iplsearch. This is not incompatible, because
2009 previously such lookups could never work.
2011 The situation is now rather anomalous, since one *can* have colons in
2012 ordinary lsearch keys. However, making the change in all cases is
2013 incompatible and would probably break a number of configurations.
2015 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2018 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2019 conversion specification without a maximum field width, thereby enabling
2020 a rogue spamd server to cause a buffer overflow. While nobody in their
2021 right mind would setup Exim to query an untrusted spamd server, an
2022 attacker that gains access to a server running spamd could potentially
2023 exploit this vulnerability to run arbitrary code as the Exim user.
2025 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2026 $primary_hostname instead of what libspf2 thinks the hosts name is.
2028 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2029 a directory entry by the name of the lookup key. Previously, if a
2030 symlink pointed to a non-existing file or a file in a directory that
2031 Exim lacked permissions to read, a lookup for a key matching that
2032 symlink would fail. Now it is enough that a matching directory entry
2033 exists, symlink or not. (Bugzilla 503.)
2035 PH/03 The body_linecount and body_zerocount variables are now exported in the
2038 PH/04 Added the $dnslist_matched variable.
2040 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2041 This means they are set thereafter only if the connection becomes
2044 PH/06 Added the client_condition to authenticators so that some can be skipped
2045 by clients under certain conditions.
2047 PH/07 The error message for a badly-placed control=no_multiline_responses left
2048 "_responses" off the end of the name.
2050 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2052 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2053 (without spaces) instead of just copying the configuration text.
2055 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2057 PH/11 Added $max_received_linelength.
2059 PH/12 Added +ignore_defer and +include_defer to host lists.
2061 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2062 way in which PCRE > 7.0 is built.
2064 PH/14 Implemented queue_only_load_latch.
2066 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2067 MAIL command. The effect was to mangle the value on 64-bit systems.
2069 PH/16 Another patch from the Sieve maintainer.
2071 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2073 PH/18 If a system quota error occurred while trying to create the file for
2074 a maildir delivery, the message "Mailbox is full" was not appended to the
2075 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2076 only to a quota excession during the actual writing of the file.
2078 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2079 characters?) which causes problems in log lines. The DN values are now
2080 passed through string_printing() before being added to log lines.
2082 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2083 and InterBase are left for another time.)
2085 PH/21 Added message_body_newlines option.
2087 PH/22 Guard against possible overflow in moan_check_errorcopy().
2089 PH/23 POSIX allows open() to be a macro; guard against that.
2091 PH/24 If the recipient of an error message contained an @ in the local part
2092 (suitably quoted, of course), incorrect values were put in $domain and
2093 $local_part during the evaluation of errors_copy.
2099 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2100 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2103 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2104 issue a MAIL command.
2106 PH/02 In an ACL statement such as
2108 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2110 if a client was not listed at all, or was listed with a value other than
2111 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2112 the condition was not true (as it should be), so access was not denied.
2113 The bug was that the ! inversion was incorrectly passed on to the second
2114 item. This has been fixed.
2116 PH/03 Added additional dnslists conditions == and =& which are different from
2117 = and & when the dns lookup returns more than one IP address.
2119 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2120 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2122 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2123 FSYNC, which compiles an option called disable_fsync that allows for
2124 bypassing fsync(). The documentation is heavily laced with warnings.
2126 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2128 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2129 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2130 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2131 including adding "make clean"; (3) Added -fPIC when compiling the test
2132 dynamically loaded module, to get rid of a warning.
2134 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2135 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2136 = 0s. The bug is otherwise harmless.
2138 PH/07 There was a bug in the dovecot authenticator such that the value of
2139 $auth1 could be overwritten, and so not correctly preserved, after a
2140 successful authentication. This usually meant that the value preserved by
2141 the server_setid option was incorrect.
2143 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2145 PH/09 Installed PCRE release 7.0.
2147 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2148 run for batched SMTP input. It is now run at the start of every message
2149 in the batch. While fixing this I discovered that the process information
2150 (output by running exiwhat) was not always getting set for -bs and -bS
2151 input. This is fixed, and it now also says "batched" for BSMTP.
2153 PH/11 Added control=no_pipelining.
2155 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2156 patch, slightly modified), and move the expansion of helo_data till after
2157 the connection is made in the smtp transport (so it can use these
2160 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2162 PH/14 Added log_selector = +pid.
2164 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2166 PH/16 Add ${if forany and ${if forall.
2168 PH/17 Added dsn_from option to vary the From: line in DSNs.
2170 PH/18 Flush SMTP output before performing a callout, unless control =
2171 no_callout_flush is set.
2173 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2174 was true (the default) a successful delivery failed to delete the retry
2175 item, thus causing premature timeout of the address. The bug is now
2178 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2180 PH/21 Long custom messages for fakedefer and fakereject are now split up
2181 into multiline responses in the same way that messages for "deny" and
2182 other ACL rejections are.
2184 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2185 with slight modification.
2187 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2188 draft, changing the syntax and factoring some duplicate code".
2190 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2191 for deliveries of the second and subsequent messages over the same SMTP
2194 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2195 ${reduce, with only minor "tidies".
2197 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2199 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2200 expansion side effects.
2202 PH/27 When a message times out after an over-quota error from an Exim-imposed
2203 quota, the bounce message says "mailbox is full". This message was not
2204 being given when it was a system quota that was exceeded. It now should
2207 MH/03 Made $recipients available in local_scan(). local_scan() already has
2208 better access to the recipient list through recipients_list[], but
2209 $recipients can be useful in postmaster-provided expansion strings.
2211 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2212 in the case of a MAIL command with additional options following the
2213 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2214 were accidentally chopped off.
2216 PH/29 SMTP synchronization checks are implemented when a command is read -
2217 there is a check that no more input is waiting when there shouldn't be
2218 any. However, for some commands, a delay in an ACL can mean that it is
2219 some time before the response is written. In this time, more input might
2220 arrive, invalidly. So now there are extra checks after an ACL has run for
2221 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2222 pipelining has not been advertised.
2224 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2226 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2227 returned in the SMTP response when smtp_return_error_details was false.
2228 This has been fixed.
2230 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2231 instead of the C I/O that was originally supplied, because problems were
2232 reported on Solaris.
2234 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2235 Exim which did not show up earlier: it was assuming that a call to
2236 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2237 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2238 was a macro that became an assignment, so it seemed to work. This has
2239 changed to a proper function call with a void return, hence the compile
2240 error. Exim's code has been fixed.
2242 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2245 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2247 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2249 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2250 because it turns out to be too big or there is a timeout) there is no
2251 "Completed" line in the log. When some messages of this type were
2252 selected by exigrep, they were listed as "not completed". Others were
2253 picked up by some special patterns. I have improved the selection
2254 criteria to be more general.
2256 PH/38 The host_find_failed option in the manualroute router can now be set
2257 to "ignore", to completely ignore a host whose IP address cannot be
2258 found. If all hosts are ignored, the behaviour is controlled by the new
2259 host_all_ignored option.
2261 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2262 homing or because of multiple MX records with /mx) generated more than
2263 one IP address, and the following item turned out to be the local host,
2264 all the secondary addresses of the first item were incorrectly removed
2265 from the list, along with the local host and any following hosts (which
2266 is what is supposed to happen).
2268 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2269 whoever called Exim into the -H file. In the case of the daemon it was
2270 behaving confusingly. When first started, it used values for whoever
2271 started the daemon, but after a SIGHUP it used the Exim user (because it
2272 calls itself on a restart). I have changed the code so that it now always
2275 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2276 message are rejected with the same error (e.g. no authentication or bad
2277 sender address), and a DATA command is nevertheless sent (as can happen
2278 with PIPELINING or a stupid MUA), the error message that was given to the
2279 RCPT commands is included in the rejection of the DATA command. This is
2280 intended to be helpful for MUAs that show only the final error to their
2283 PH/42 Another patch from the Sieve maintainer.
2285 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2286 Eximstats - Fixed some broken HTML links and added missing column headers
2288 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2289 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2291 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2293 PH/43 Yet another patch from the Sieve maintainer.
2295 PH/44 I found a way to check for a TCP/IP connection going away before sending
2296 the response to the final '.' that terminates a message, but only in the
2297 case where the client has not sent further data following the '.'
2298 (unfortunately, this is allowed). However, in many cases there won't be
2299 any further data because there won't be any more messages to send. A call
2300 to select() can be used: if it shows that the input is "ready", there is
2301 either input waiting, or the socket has been closed. An attempt to read
2302 the next input character can distinguish the two cases. Previously, Exim
2303 would have sent an OK response which the client would never have see.
2304 This could lead to message repetition. This fix should cure that, at
2305 least in a lot of common cases.
2307 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2308 advertised in response to EHLO.
2314 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2315 fixed by 4.65/MH/01 (is this a record?) are fixed:
2317 (i) An empty string was always treated as zero by the numeric comparison
2318 operators. This behaviour has been restored.
2320 (ii) It is documented that the numeric comparison operators always treat
2321 their arguments as decimal numbers. This was broken in that numbers
2322 starting with 0 were being interpreted as octal.
2324 While fixing these problems I realized that there was another issue that
2325 hadn't been noticed. Values of message_size_limit (both the global option
2326 and the transport option) were treated as octal if they started with 0.
2327 The documentation was vague. These values are now always treated as
2328 decimal, and I will make that clear in the documentation.
2334 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2335 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2338 MH/01 Don't check that the operands of numeric comparison operators are
2339 integers when their expansion is in "skipping" mode (fixes bug
2340 introduced by 4.64-PH/07).
2342 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2343 child addresses, Exim now panics and dies. Previously, because the count
2344 is held in a short int, deliveries were likely to be lost. As such a
2345 large number of recipients for a single message is ridiculous
2346 (performance will be very, very poor), I have chosen to impose a limit
2347 rather than extend the field.
2353 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2354 leftover -K file (the existence of which was triggered by #402).
2355 While we were at it, introduced process PID as part of the -K
2356 filename. This should rule out race conditions when creating
2359 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2360 processing considerably. Previous code took too long for large mails,
2361 triggering a timeout which in turn triggers #401.
2363 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2364 in the DK code in transports.c. sendfile() is not really portable,
2365 hence the _LINUX specificness.
2367 TF/01 In the add_headers option to the mail command in an Exim filter,
2368 there was a bug that Exim would claim a syntax error in any
2369 header after the first one which had an odd number of characters
2372 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2373 callout verification, Exim cached a "reject" for the entire domain. This
2374 is correct for most verifications, but it is not correct for a recipient
2375 verification with use_sender or use_postmaster set, because in that case
2376 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2377 case of MAIL FROM:<> rejection from other early rejections (e.g.
2378 rejection of HELO). When verifying a recipient using a non-null MAIL
2379 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2380 Whatever the result of the callout, the value of the domain cache is
2381 left unchanged (for any other kind of callout, getting as far as trying
2382 RCPT means that the domain itself is ok).
2384 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2387 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2388 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2389 session, the daemon ignores ECONNECT errors and logs others; it now
2390 ignores EPIPE as well.
2392 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2393 (quoted-printable decoding).
2395 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2396 later the small subsequent patch to fix an introduced bug.
2398 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2400 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2402 PH/08 An error is now given if message_size_limit is specified negative.
2404 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2405 to be given (somewhat) arbitrary names.
2407 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2410 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2411 miscellaneous code fixes
2413 PH/10 Added the log_reject_target ACL modifier to specify where to log
2416 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2417 hostname. This is wrong, because it relates to the incoming message (and
2418 probably the interface on which it is arriving) and not to the outgoing
2419 callout (which could be using a different interface). This has been
2420 changed to use the value of the helo_data option from the smtp transport
2421 instead - this is what is used when a message is actually being sent. If
2422 there is no remote transport (possible with a router that sets up host
2423 addresses), $smtp_active_hostname is used.
2425 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2426 tweaks were necessary in order to get it to work (see also 21 below):
2427 (a) The code assumed that strncpy() returns a negative number on buffer
2428 overflow, which isn't the case. Replaced with Exim's string_format()
2430 (b) There were several signed/unsigned issues. I just did the minimum
2431 hacking in of casts. There is scope for a larger refactoring.
2432 (c) The code used strcasecmp() which is not a standard C function.
2433 Replaced with Exim's strcmpic() function.
2434 (d) The code set only $1; it now sets $auth1 as well.
2435 (e) A simple test gave the error "authentication client didn't specify
2436 service in request". It would seem that Dovecot has changed its
2437 interface. Fortunately there's a specification; I followed it and
2438 changed what the client sends and it appears to be working now.
2440 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2443 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2444 address A is aliased to B and C, where B exists and C does not. Without
2445 -v the output is "A verified" because verification stops after a
2446 successful redirection if more than one address is generated. However,
2447 with -v the child addresses are also verified. Exim was outputting "A
2448 failed to verify" and then showing the successful verification for C,
2449 with its parentage. It now outputs "B failed to verify", showing B's
2450 parentage before showing the successful verification of C.
2452 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2453 look up a TXT record in a specific list after matching in a combined
2456 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2457 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2458 they consult the DNS. I had assumed they would set it the way they
2459 wanted; and indeed my experiments on Linux seem to show that in some
2460 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2461 To be on the safe side, however, I have now made the interface to
2462 host_find_byname() similar to host_find_bydns(), with an argument
2463 containing the DNS resolver options. The host_find_byname() function now
2464 sets these options at its start, just as host_find_bydns() does. The smtp
2465 transport options dns_qualify_single and dns_search_parents are passed to
2466 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2467 of host_find_byname() use the default settings of RES_DEFNAMES
2468 (qualify_single) but not RES_DNSRCH (search_parents).
2470 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2471 spool_read_header() do less string testing, by means of a preliminary
2472 switch on the second character of optional "-foo" lines. (This is
2473 overdue, caused by the large number of possibilities that now exist.
2474 Originally there were few.) While I was there, I also converted the
2475 str(n)cmp tests so they don't re-test the leading "-" and the first
2476 character, in the hope this might squeeze out yet more improvement.
2478 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2479 flag allowing group syntax was set by the header_syntax check but not
2480 turned off, possible causing trouble later; (2) The flag was not being
2481 set at all for the header_verify test, causing "group"-style headers to
2482 be rejected. I have now set it in this case, and also caused header_
2483 verify to ignore an empty address taken from a group. While doing this, I
2484 came across some other cases where the code for allowing group syntax
2485 while scanning a header line wasn't quite right (mostly, not resetting
2486 the flag correctly in the right place). These bugs could have caused
2487 trouble for malformed header lines. I hope it is now all correct.
2489 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2490 with the "reply" argument non-NULL. The code, however (which originally
2491 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2492 but it didn't always do it. This confused somebody who was copying the
2493 code for some other use. I have removed all the tests.
2495 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2496 feature that was used to support insecure browsers during the U.S. crypto
2497 embargo. It requires special client support, and Exim is probably the
2498 only MTA that supported it -- and would never use it because real RSA is
2499 always available. This code has been removed, because it had the bad
2500 effect of slowing Exim down by computing (never used) parameters for the
2501 RSA_EXPORT functionality.
2503 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2504 authenticator to fail if there's a tab character in the incoming data
2505 (there should never be unless someone is messing about, as it's supposed
2506 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2507 if the connection is using TLS or if the remote IP is the same as the
2508 local IP, and the "valid-client-cert option" if a client certificate has
2511 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2512 authenticators. This can be used for authorization after authentication
2513 succeeds. (In the case of plaintext, it servers for both authentication
2516 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2517 if any retry times were supplied.
2519 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2520 connection, where there is no client IP address to check. In this
2521 situation, the verify now always succeeds.
2523 PH/25 Applied John Jetmore's -Mset patch.
2525 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2527 PH/27 In a string expansion for a processed (not raw) header when multiple
2528 headers of the same name were present, leading whitespace was being
2529 removed from all of them, but trailing whitespace was being removed only
2530 from the last one. Now trailing whitespace is removed from each header
2531 before concatenation. Completely empty headers in a concatenation (as
2532 before) are ignored.
2534 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2535 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2537 PH/29 [Removed. This was a change that I later backed out, and forgot to
2538 correct the ChangeLog entry (that I had efficiently created) before
2539 committing the later change.]
2541 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2542 address errors (4xx response to RCPT) over the same connection as other
2543 messages routed to the same hosts. Such deliveries are always "forced",
2544 so retry times are not inspected. This resulted in far too many retries
2545 for the affected addresses. The effect occurred only when there were more
2546 hosts than the hosts_max_try setting in the smtp transport when it had
2547 the 4xx errors. Those hosts that it had tried were not added to the list
2548 of hosts for which the message was waiting, so if all were tried, there
2549 was no problem. Two fixes have been applied:
2551 (i) If there are any address or message errors in an SMTP delivery, none
2552 of the hosts (tried or untried) are now added to the list of hosts
2553 for which the message is waiting, so the message should not be a
2554 candidate for sending over the same connection that was used for a
2555 successful delivery of some other message. This seems entirely
2556 reasonable: after all the message is NOT "waiting for some host".
2557 This is so "obvious" that I'm not sure why it wasn't done
2558 previously. Hope I haven't missed anything, but it can't do any
2559 harm, as the worst effect is to miss an optimization.
2561 (ii) If, despite (i), such a delivery is accidentally attempted, the
2562 routing retry time is respected, so at least it doesn't keep
2563 hammering the server.
2565 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2566 in ${readsocket because some servers need this prod.
2568 PH/32 Added some extra debug output when updating a wait-xxx database.
2570 PH/33 The hint "could be header name not terminated by colon", which has been
2571 given for certain expansion errors for a long time, was not being given
2572 for the ${if def:h_colon_omitted{... case.
2574 PH/34 The spec says: "With one important exception, whenever a domain list is
2575 being scanned, $domain contains the subject domain." There was at least
2576 one case where this was not true.
2578 PH/35 The error "getsockname() failed: connection reset by peer" was being
2579 written to the panic log as well as the main log, but it isn't really
2580 panic-worthy as it just means the connection died rather early on. I have
2581 removed the panic log writing for the ECONNRESET error when getsockname()
2584 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2585 runs only) independently of the message's sender address. This meant
2586 that, if the 4xx error was in fact related to the sender, a different
2587 message to the same recipient with a different sender could confuse
2588 things. In particular, this can happen when sending to a greylisting
2589 server, but other circumstances could also provoke similar problems.
2590 I have changed the default so that the retry time for these errors is now
2591 based a combination of the sender and recipient addresses. This change
2592 can be overridden by setting address_retry_include_sender=false in the
2595 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2596 remote server are returned as part of bounce messages. This was not
2597 happening for LMTP over a pipe (the lmtp transport), but now it is the
2598 same for both kinds of LMTP.
2600 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2601 in header lines that were in fact CNAMEs. This is no longer the case.
2603 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2604 and queue runs started by the daemon processed all messages. This has
2605 been fixed so that -R and -S can now usefully be given with -q<time>.
2607 PH/40 Import PCRE release 6.7 (fixes some bugs).
2609 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2611 PH/42 Give an error if -q is specified more than once.
2613 PH/43 Renamed the variables $interface_address and $interface_port as
2614 $received_ip_address and $received_port, to make it clear that these
2615 values apply to message reception, and not to the outgoing interface when
2616 a message is delivered. (The old names remain recognized, of course.)
2618 PH/44 There was no timeout on the connect() call when using a Unix domain
2619 socket in the ${readsocket expansion. There now is.
2621 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2622 be meaningful with "accept".
2624 SC/01 Eximstats V1.43
2625 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2627 SC/02 Eximstats V1.44
2628 Use a glob alias rather than an array ref in the generated
2629 parser. This improves both readability and performance.
2631 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2632 Collect SpamAssassin and rejection statistics.
2633 Don't display local sender or destination tables unless
2634 there is data to show.
2635 Added average volumes into the top table text output.
2637 SC/04 Eximstats V1.46
2638 Collect data on the number of addresses (recipients)
2639 as well as the number of messages.
2641 SC/05 Eximstats V1.47
2642 Added 'Message too big' to the list of mail rejection
2643 reasons (thanks to Marco Gaiarin).
2645 SC/06 Eximstats V1.48
2646 Mainlog lines which have GMT offsets and are too short to
2647 have a flag are now skipped.
2649 SC/07 Eximstats V1.49 (Alain Williams)
2650 Added the -emptyok flag.
2652 SC/08 Eximstats V1.50
2653 Fixes for obtaining the IP address from reject messages.
2655 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2656 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2657 whitespace changes from 4.64-PH/27
2659 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2662 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2663 are found, allow negative numbers in numeric criteria)
2665 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2667 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2668 to match changes made in 4.64-PH/43
2670 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2672 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2673 "domain" is actually an IP address, and force a failure. This locks out
2674 those revolvers/nameservers that support "A-for-A" lookups, in
2675 contravention of the specifications.
2677 PH/48 When a host name was looked up from an IP address, and the subsequent
2678 forward lookup of the name timed out, the host name was left in
2679 $sender_host_name, contrary to the specification.
2681 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2682 restricted to single-key lookups, Exim was not diagnosing an error if
2683 * or *@ was used with a query-style lookup.
2685 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2687 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2688 long ago, but noone interested enough thought of it. Let's just say that
2689 the "1.1" means that there are some new functions that weren't there at
2690 some point in the past.
2692 PH/51 Error processing for expansion failure of helo_data from an smtp
2693 transport during callout processing was broken.
2695 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2696 tested/used via the -bh/-bhc/-bs options.
2698 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2699 bug, fixed in subsequent PCRE releases).
2701 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2702 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2704 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2710 SC/01 Use a glob alias rather than an array ref in eximstats generated
2711 parser. This improves both readability and performance.
2713 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2714 Don't display local sender or destination tables in eximstats unless
2715 there is data to show.
2716 Added average volumes into the eximstats top table text output.
2718 SC/03 Collect data on the number of addresses (recipients) as well
2719 as the number of messages in eximstats.
2721 TF/01 Correct an error in the documentation for the redirect router. Exim
2722 does (usually) call initgroups() when daemonizing.
2724 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2725 with consistent privilege compared to when running as a daemon.
2727 TF/03 Note in the spec that $authenticated_id is not set for local
2728 submissions from trusted users.
2730 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2731 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2733 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2734 by adding some example configuration directives to the default
2735 configuration file. A little bit of work is required to uncomment the
2736 directives and define how usernames and passwords are checked, but
2737 there is now a framework to start from.
2739 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2740 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2741 without this. I don't know how relevant this is to other LDAP libraries.
2743 PH/02 Add the verb name to the "unknown ACL verb" error.
2745 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2747 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2749 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2750 directory not expanded when it should be if an expanded home directory
2751 was set for the address (which is overridden by the transport).
2753 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2756 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2757 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2758 because it is too late at that time, and has no effect.
2760 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2761 security issue with \' (bugzilla #107). I could not use the
2762 PQescapeStringConn() function, because it needs a PGconn value as one of
2765 PH/08 When testing addresses using -bt, indicate those final addresses that
2766 are duplicates that would not cause an additional delivery. At least one
2767 person was confused, thinking that -bt output corresponded to deliveries.
2768 (Suppressing duplicates isn't a good idea as you lose the information
2769 about possibly different redirections that led to the duplicates.)
2771 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2772 systems where poll() doesn't work, in particular OS X.
2774 PH/10 Added more information to debugging output for retry time not reached.
2776 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2777 operations in malware.c.
2779 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2782 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2783 syslog_duplication set false, log lines that would normally be written
2784 both the the main log and to the reject log were not written to syslog at
2787 PH/14 In the default configuration, change the use of "message" in ACL warn
2788 statements to "add_header".
2790 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2791 not followed by a command (e.g. "seen endif").
2793 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2794 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2797 PH/17 Added extra conditions to the default value of delay_warning_condition
2801 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2802 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2803 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2806 The Auto-Submitted: and various List- headers are standardised, whereas I
2807 don't think Precedence: ever was.
2809 PH/18 Refactored debugging code in route_finduser() to show more information,
2810 in particular, the error code if getpwnam() issues one.
2812 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2813 This is apparently needed in addition to the PH/07 change above to avoid
2814 any possible encoding problems.
2816 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2817 but not after initializing Perl.
2819 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2820 output them only if debugging. By default they are written stderr,
2821 apparently, which is not desirable.
2823 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2826 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2829 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2831 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2832 authenticated or an ident call has been made. Suppress the default
2833 values for $authenticated_id and $authenticated_sender (but permit -oMai
2834 and -oMas) when testing with -bh.
2836 PH/24 Re-jigged the order of the tests in the default configuration so that the
2837 tests for valid domains and recipients precede the DNS black list and CSA
2838 tests, on the grounds that those ones are more expensive.
2840 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2841 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2842 This bug exists in every version of Exim that I still have, right back to
2845 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2846 However, an attempt to turn on case-sensitivity in a regex key by
2847 including (?-i) didn't work because the subject string was already
2848 lowercased, and the effects were non-intuitive. It turns out that a
2849 one-line patch can be used to allow (?-i) to work as expected.
2855 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2856 other effects) broke the use of negated acl sub-conditions.
2858 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2861 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2862 "Deny" causes Exim to reject the incoming connection with a 554 error.
2863 Unfortunately, if there is a major crisis, such as a disk failure,
2864 tcp-wrappers gives "deny", whereas what one would like would be some
2865 kind of temporary error. A kludge has been added to help with this.
2866 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2867 554 error is used if errno is still zero or contains ENOENT (which occurs
2868 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2871 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2873 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2874 errors. Otherwise a message that provokes a temporary error (when other
2875 messages do not) can cause a whole host to time out.
2877 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2878 addresses were routed directly to files or pipes from a redirect router.
2879 File deliveries just didn't batch; pipe deliveries might have suffered
2882 PH/06 A failure to get a lock for a hints database would erroneously always say
2883 "Failed to get write lock", even when it was really a read lock.
2885 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2886 of 0600. This has been changed to use the value of the lockfile_mode
2887 option (which defaults to 0600).
2889 PH/08 Applied small patch from the Sieve maintainer.
2891 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2892 folder from quota calculations, a direct delivery into this folder messed
2893 up the contents of the maildirsize file. This was because the regex was
2894 used only to exclude .Trash (or whatever) when the size of the mailbox
2895 was calculated. There was no check that a delivery was happening into an
2896 excluded directory. This bug has been fixed by ignoring all quota
2897 processing for deliveries into excluded directories.
2899 PH/10 Added the maildirfolder_create_regex option to appendfile.
2905 PH/01 The code for finding all the local interface addresses on a FreeBSD
2906 system running IPv6 was broken. This may well have applied to all BSD
2907 systems, as well as to others that have similar system calls. The broken
2908 code found IPv4 interfaces correctly, but gave incorrect values for the
2909 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2910 that it would not match correctly against @[] and not recognize the IPv6
2913 PH/02 The ipliteral router was not recognizing addresses of the form user@
2914 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2916 PH/03 Added disable_ipv6.
2918 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2919 decoded form, because it is most often used to construct To: headers
2920 lines in autoreplies, and the decoded form may well be syntactically
2921 invalid. However, $reply_address has leading white space removed, and all
2922 newlines turned into spaces so that the autoreply transport does not
2925 PH/05 If group was specified without a user on a router, and no group or user
2926 was specified on a transport, the group from the router was ignored.
2928 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2929 for visible compile-time settings that can be used to change these
2930 numbers, for those that want even more. Backwards compatibility with old
2931 spool files has been maintained. However, going back to a previous Exim
2932 release will lost any variables that are in spool files.
2934 PH/07 Two small changes when running in the test harness: increase delay when
2935 passing a TCP/IP connection to a new process, in case the original
2936 process has to generate a bounce, and remove special handling of
2937 127.0.0.2 (sic), which is no longer necessary.
2939 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2940 be the same on different OS.
2942 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2945 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2946 whether --show-vars was specified or not
2948 JJ/02 exipick: Added support for new ACL variable spool format introduced
2951 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2952 syntactically invalid From: or Reply-to: line, and a filter used this to
2953 generate an autoreply, and therefore failed to obtain an address for the
2954 autoreply, Exim could try to deliver to a non-existent relative file
2955 name, causing unrelated and misleading errors. What now happens is that
2956 it logs this as a hard delivery error, but does not attempt to create a
2959 PH/11 The exinext utility has a -C option for testing purposes, but although
2960 the given file was scanned by exinext itself; it wasn't being passed on
2961 when Exim was called.
2963 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2964 an end-of-file indication when reading a command response.
2966 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2967 compiled. In many other places in Exim, IPv6 addresses are always
2968 recognized, so I have changed this. It also means that IPv4 domain
2969 literals of the form [IPV4:n.n.n.n] are now always recognized.
2971 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2972 used if the router is not running as root, for example, when verifying at
2973 ACL time, or when using -bh. The debugging output from this situation was
2974 non-existent - all you got was a failure to exec. I have made two
2977 (a) Failures to set uid/gid, the current directory, or a process leader
2978 in a subprocess such as that created by queryprogram now generate
2979 suitable debugging output when -d is set.
2981 (b) The queryprogram router detects when it is not running as root,
2982 outputs suitable debugging information if -d is set, and then runs
2983 the subprocess without attempting to change uid/gid.
2985 PH/15 Minor change to Makefile for building test_host (undocumented testing
2988 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2989 additional section of a DNS packet that returns MX or SRV records.
2990 Instead, it always explicitly searches for A/AAAA records. This avoids
2991 major problems that occur when a DNS server includes only records of one
2992 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2993 fixed another bug: if SRV records were looked up and the corresponding
2994 address records were *not* found in the additional section, the port
2995 values from the SRV records were lost.
2997 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2998 using the correct key (the original address) when searching the retry
2999 rules in order to find which one to use for generating the retry hint.
3001 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3002 adding the default one. Similarly, if it contains a Reply-To: header, the
3003 errors_reply_to option, if set, is not used.
3005 PH/19 When calculating a retry time, Exim used to measure the "time since
3006 failure" by looking at the "first failed" field in the retry record. Now
3007 it does not use this if it is later than than the arrival time of the
3008 message. Instead it uses the arrival time. This makes for better
3009 behaviour in cases where some deliveries succeed, thus re-setting the
3010 "first failed" field. An example is a quota failure for a huge message
3011 when small messages continue to be delivered. Without this change, the
3012 "time since failure" will always be short, possible causing more frequent
3013 delivery attempts for the huge message than are intended.
3014 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3016 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3017 $1, $2, $3) because the numerical variables can be reset during some
3018 expansion items (e.g. "match"), thereby losing the authentication data.
3020 PH/21 Make -bV show the size of off_t variables so that the test suite can
3021 decide whether to run tests for quotas > 2G.
3023 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3024 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3025 filecount value is greater than 2G or if a quota value is greater than 2G
3026 on a system where the size of off_t is not greater than 4, a panic error
3029 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3030 never match. The debug and -bh output now contains an explicit error
3031 message indicating a malformed IPv4 address or mask.
3033 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3034 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3035 PH/23 above applies.
3037 PH/25 Do not write to syslog when running in the test harness. The only
3038 occasion when this arises is a failure to open the main or panic logs
3039 (for which there is an explicit test).
3041 PH/26 Added the /no_tell option to "control=freeze".
3043 PH/27 If a host name lookup failed very early in a connection, for example, if
3044 the IP address matched host_lookup and the reverse lookup yielded a name
3045 that did not have a forward lookup, an error message of the form "no IP
3046 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3047 could be logged. Now it outputs the IP address instead of "NULL".
3049 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3050 allows the sender and the authenticated sender to be set when
3051 submitting a message from within Exim. Since child_open_exim() is
3052 documented for local_scan(), the new function should be too.
3054 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3055 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3056 results in an empty string is now treated as unset.
3058 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3060 PH/31 Added sender_verify_fail logging option.
3062 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3063 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3064 tidied the source and removed it altogether.
3066 PH/33 When a queue run was abandoned because the load average was too high, a
3067 log line was always written; now it is written only if the queue_run log
3068 selector is set. In addition, the log line for abandonment now contains
3069 information about the queue run such as the pid. This is always present
3070 in "start" and "stop" lines but was omitted from the "abandon" line.
3072 PH/34 Omit spaces between a header name and the colon in the error message that
3073 is given when verify = headers_syntax fails (if there are lots of them,
3074 the message gets confusing).
3076 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3077 names, as there are now some PTR records that contain slashes. This check
3078 is only to protect against broken name servers that fall over on strange
3079 characters, so the fact that it applies to all lookups doesn't matter.
3081 PH/36 Now that the new test suite is complete, we can remove some of the
3082 special code in Exim that was needed for the old test suite. For example,
3083 sorting DNS records because real resolvers return them in an arbitrary
3084 order. The new test suite's fake resolver always returns records in the
3087 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3088 bounces) except when queue_only is set, to avoid logging races between
3089 the different processes.
3091 PH/38 Panic-die if .include specifies a non-absolute path.
3093 PH/39 A tweak to the "H" retry rule from its user.
3095 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3096 a label. They prevented compilation on older perls.
3098 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3099 a warning to be raised on newish perls.
3101 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3102 on queue. Changes to match documented behaviour of showing count of
3103 messages matching specified criteria.
3105 PH/40 Changed the default ident timeout from 30s to 5s.
3107 PH/41 Added support for the use of login_cap features, on those BSD systems
3108 that have them, for controlling the resources used by pipe deliveries.
3110 PH/42 The content-scanning code uses fopen() to create files in which to put
3111 message data. Previously it was not paying any attention to the mode of
3112 the files. Exim runs with umask(0) because the rest of the code creates
3113 files with open(), and sets the required mode explicitly. Thus, these
3114 files were ending up world-writeable. This was not a big issue, because,
3115 being within the spool directory, they were not world-accessible. I have
3116 created a function called modefopen, which takes an additional mode
3117 argument. It sets umask(777), creates the file, chmods it to the required
3118 mode, then resets the umask. All the relevant calls to fopen() in the
3119 content scanning code have been changed to use this function.
3121 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3122 to 24 hours. This avoids potential overflow problems when processing G
3123 and H retry rules. I suspect nobody ever tinkers with this value.
3125 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3127 PH/45 When the plaintext authenticator is running as a client, the server's
3128 challenges are checked to ensure they are valid base64 strings. By
3129 default, the authentication attempt is cancelled if an invalid string is
3130 received. Setting client_ignore_invalid_base64 true ignores these errors.
3131 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3132 they are received. Thus, the responses can be made to depend on the
3133 challenges. If an invalid string is ignored, an empty string is placed in
3136 PH/46 Messages that are created by the autoreply transport now contains a
3137 References: header, in accordance with RFCs 2822 and 3834.
3139 PH/47 Added authenticated_sender_force to the smtp transport.
3141 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3143 PH/49 Installed latest patch from the Sieve maintainer.
3145 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3146 was also set, the appendfile transport was unnecessarily scanning a
3147 directory of message files (e.g. for maildir delivery) to find the count
3148 of files (along with the size), even though it did not need this
3149 information. It now does the scan only if it needs to find either the
3150 size of the count of files.
3152 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3154 PH/52 Two bugs concerned with error handling when the smtp transport is
3157 (i) Exim was not creating retry information for temporary errors given
3158 for individual recipients after the DATA command when the smtp transport
3159 was used in LMTP mode. This meant that they could be retried too
3160 frequently, and not timed out correctly.
3162 (ii) Exim was setting the flag that allows error details to be returned
3163 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3164 recipients that were returned after the DATA command.
3166 PH/53 This is related to PH/52, but is more general: for any failing address,
3167 when detailed error information was permitted to be returned to the
3168 sender, but the error was temporary, then after the final timeout, only
3169 "retry timeout exceeded" was returned. Now it returns the full error as
3170 well as "retry timeout exceeded".
3172 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3173 do this, and (what is worse) MTAs that accept it.
3175 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3176 will now be deprecated.
3178 PH/56 New os.c-cygwin from the Cygwin maintainer.
3180 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3181 formats (previously only available in exim formats via -bpr, -bpru,
3182 and -bpra. Now also available in native and exiqgrep formats)
3184 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3185 with very large, slow to parse queues
3187 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3189 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3191 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3192 responses to authentication challenges, though it was showing the
3193 challenges; (ii) I've removed the CR characters from the debug output for
3196 PH/58 Allow for the insertion of a newline as well as a space when a string
3197 is turned into more than one encoded-word during RFC 2047 encoding. The
3198 Sieve code now uses this.
3200 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3201 data_4xx, lost_connection, tls_required.
3203 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3204 message was being sent as an SMTP response.
3206 PH/61 Add -l and -k options to exicyclog.
3208 PH/62 When verifying, if an address was redirected to one new address, so that
3209 verification continued, and the new address failed or deferred after
3210 having set something in $address_data, the value of $address_data was not
3211 passed back to the ACL. This was different to the case when no
3212 redirection occurred. The value is now passed back in both cases.
3214 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3215 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3216 use login_cap.h, so on its own it isn't the distinguishing feature. The
3217 new name refers directly to the setclassresources() function.
3219 PH/65 Added configuration files for NetBSD3.
3221 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3223 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3224 is preferred over IPv4.
3226 PH/68 The bounce_return_message and bounce_return_body options were not being
3227 honoured for bounces generated during the reception of non-SMTP messages.
3228 In particular, this applied to messages rejected by the ACL. This bug has
3229 been fixed. However, if bounce_return_message is true and bounce_return_
3230 body is false, the headers that are returned for a non-SMTP message
3231 include only those that have been read before the error was detected.
3232 (In the case of an ACL rejection, they have all been read.)
3234 PH/69 The HTML version of the specification is now built in a directory called
3235 spec_html instead of spec.html, because the latter looks like a path with
3236 a MIME-type, and this confuses some software.
3238 PH/70 Catch two compiler warnings in sieve.c.
3240 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3241 function verify_get_ident() calls ip_connect() to connect a socket, but
3242 if the "connect()" function timed out, ip_connect() used to close the
3243 socket. However, verify_get_ident() also closes the socket later, and in
3244 between Exim writes to the log, which may get opened at this point. When
3245 the socket was closed in ip_connect(), the log could get the same file
3246 descriptor number as the socket. This naturally causes chaos. The fix is
3247 not to close the socket in ip_connect(); the socket should be closed by
3248 the function that creates it. There was only one place in the code where
3249 this was missing, in the iplookup router, which I don't think anybody now
3250 uses, but I've fixed it anyway.
3252 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3253 well as to direct DNS lookups. Otherwise the handling of names in host
3254 lists is inconsistent and therefore confusing.
3260 PH/01 Two changes to the default runtime configuration:
3262 (1) Move the checks for relay_from_hosts and authenticated clients from
3263 after to before the (commented out) DNS black list checks.
3265 (2) Add control=submission to the relay_from_hosts and authenticated
3266 clients checks, on the grounds that messages accepted by these
3267 statements are most likely to be submissions.
3269 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3271 (1) Generate an error if the third argument for the ${prvs expansion is
3274 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3277 (3) Reset the variables that are obtained from the first argument of
3278 ${prvscheck and used in the second argument before leaving the code,
3279 because their memory is reclaimed, so using them afterwards may do
3282 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3283 one (it's much easier than Tom thought :-).
3285 (5) Because of (4), we can now allow for the use of $prvscheck_result
3286 inside the third argument.
3288 PH/03 For some reason, the default setting of PATH when running a command from
3289 a pipe transport was just "/usr/bin". I have changed it to
3292 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3293 anything to be listed in the output from -bV.
3295 PH/05 When a filter generated an autoreply, the entire To: header line was
3296 quoted in the delivery log line, like this:
3298 => >A.N.Other <ano@some.domain> <original@ddress> ...
3300 This has been changed so that it extracts the operative address. There
3301 may be more than one such address. If so, they are comma-separated, like
3304 => >ano@some.domain,ona@other.domain <original@ddress> ...
3306 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3307 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3308 not being looked up in the rDNS to get a host name, Exim was showing the
3309 IP address twice in Received: lines, even though the IP addresses were
3310 identical. For example:
3312 Received: from [1.2.3.4] (helo=[1.2.3.4])
3314 However, if the real host name was known, it was omitting the HELO data
3315 if it matched the actual IP address. This has been tidied up so that it
3316 doesn't show the same IP address twice.
3318 PH/07 When both +timestamp and +memory debugging was on, the value given by
3319 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3320 called by the debug printing, thereby overwriting the timestamp buffer.
3321 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3323 PH/08 When the original message was included in an autoreply transport, it
3324 always said "this is a copy of the message, including all the headers",
3325 even if body_only or headers_only was set. It now gives an appropriate
3328 PH/09 Applied a patch from the Sieve maintainer which:
3330 o fixes some comments
3331 o adds the (disabled) notify extension core
3332 o adds some debug output for the result of if/elsif tests
3333 o points to the current vacation draft in the documentation
3334 and documents the missing references header update
3338 o fixes a bug in processing the envelope test (when testing
3339 multiple envelope elements, the last element determined the
3342 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3343 Electronic Mail") by including:
3345 Auto-submitted: auto-generated
3347 in the messages that it generates (bounce messages and others, such as
3348 warnings). In the case of bounce messages for non-SMTP messages, there was
3349 also a typo: it was using "Auto_submitted" (underscore instead of
3350 hyphen). Since every message generated by Exim is necessarily in response
3351 to another message, thes have all been changed to:
3353 Auto-Submitted: auto-replied
3355 in accordance with these statements in the RFC:
3357 The auto-replied keyword:
3359 - SHOULD be used on messages sent in direct response to another
3360 message by an automatic process,
3362 - MUST NOT be used on manually-generated messages,
3364 - MAY be used on Delivery Status Notifications (DSNs) and Message
3365 Disposition Notifications (MDNs),
3367 - MUST NOT be used on messages generated by automatic or periodic
3368 processes, except for messages which are automatic responses to
3371 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3372 to the default Received: header definition.
3374 PH/12 Added log selector acl_warn_skipped (default on).
3376 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3377 variables because (a) they are in the wrong storage pool and (b) even if
3378 they were copied, it wouldn't work properly because of the caching.
3380 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3381 checking when decoding. Apparently there are clients that generate
3382 overlong encoded strings. Why am I not surprised?
3384 PH/15 If the first argument of "${if match_address" was not empty, but did not
3385 contain an "@" character, Exim crashed. Now it writes a panic log message
3386 and treats the condition as false.
3388 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3390 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3391 extension "envelope-auth". The code is finished and in agreement with
3392 other implementations, but there is no documentation so far and in fact,
3393 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3394 not changing the active code.
3396 Print executed "if" and "elsif" statements when debugging is used. This
3397 helps a great deal to understand what a filter does.
3399 Document more things not specified clearly in RFC3028. I had all this
3400 sorted out, when out of a sudden new issues came to my mind. Oops."
3402 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3405 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3406 canonical form (as documented). However, after a host name lookup from
3407 the IP address, check_host() was doing a simple string comparison with
3408 addresses acquired from the DNS when checking that the found name did
3409 have the original IP as one of its addresses. Since any found IPv6
3410 addresses are likely to be in abbreviated form, the comparison could
3411 fail. Luckily, there already exists a function for doing the comparison
3412 by converting both addresses to binary, so now that is used instead of
3413 the text comparison.
3415 PH/20 There was another similar case to PH/19, when a complete host name was
3416 given in a host list; looking up its IP address could give an abbreviated
3417 form, whereas the current host's name might or might not be abbreviated.
3418 The same fix has been applied.
3424 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3425 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3428 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3429 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3431 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3433 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3434 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3435 Also, for "Auto-Submitted", treat anything other than "no" as
3436 non-personal, in accordance with RFC 3834. (Previously it treated
3437 anything starting "auto-" as non-personal.)
3439 TF/01 The control=submission/name=... option had a problem with syntax
3440 errors if the name included a slash character. The /name= option
3441 now slurps the rest of the string, so it can include any characters
3442 but it must come last in the list of options (after /sender_retain
3445 PH/05 Some modifications to the interface to the fake nameserver for the new
3453 TK/01 Added the "success_on_redirect" address verification option. See
3454 NewStuff for rationale and an example.
3456 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3458 PH/02 Patch to exigrep to allow it to work on syslog lines.
3460 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3461 fread() to skip over the body file's header line, because in Cygwin the
3462 header line is locked and is inaccessible.
3464 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3465 co-exist for some time) to make it clear that it is the Exim ID that is
3466 referenced, not the Message-ID: header line.
3468 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3469 string_format() function, because snprintf() does not exist on all
3472 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3473 ${stat: expansion item.
3475 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3476 protocol synchronization error", to keep the pedants happy.
3478 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3479 well as for IRIX systems, when gcc is being used. See the host.c source
3482 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3484 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3487 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3488 transport and to the smtp transport in LMTP mode.
3490 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3492 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3493 run a filter in a subprocess. This could lead to confusion in subsequent
3494 lookups in the parent process. There should also be a search_tidyup() at
3495 the end of the subprocess.
3497 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3498 only if the host matched helo_try_verify_hosts, which caused the
3499 verification to occur when the EHLO/HELO command was issued. The ACL just
3500 tested the remembered result. Now, if a previous verification attempt has
3501 not happened, "verify = helo" does it there and then.
3503 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3505 TK/03 Fix log output including CR from clamd.
3507 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3508 exist provoked a memory error which could cause a segfault.
3510 PH/15 Installed PCRE 6.2
3512 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3514 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3515 of the problem. Specifically, suggested +O2 rather than +O1 for the
3518 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3520 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3521 record, the port was not correctly logged when the outgoing_port log
3522 selector was set (it logged the transort's default port).
3524 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3525 fallback_hosts, and "hosts" in the smtp transport.
3527 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3528 host errors such as "Connection refused".
3530 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3531 authentication with radiusclient 0.4.9:
3533 - Error returned from rc_read_config was caught wrongly
3534 - Username/password not passed on to radius server due to wrong length.
3536 The presumption is that some radiusclient API changes for 4.51/PH/17
3537 were not taken care of correctly. The code is still untested by me (my
3538 Linux distribution still has 0.3.2 of radiusclient), but it was
3539 contributed by a Radius user.
3541 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3542 expanding the "port" option of the smtp transport.
3544 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3545 while reading a MIME header. Thanks to Tom Hughes for a patch.
3547 PH/24 Include config.h inside local_scan.h so that configuration settings are
3550 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3551 that in an ACL for RCPT (for example), you can examine exactly what was
3554 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3555 commands, but it was not correctly comparing the address with the actual
3556 client host address. Thus, it would show the EHLO address in Received:
3557 header lines when this was not necessary.
3559 PH/27 Added the % operator to ${eval:}.
3561 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3562 it should be ignoring failures (because with -C, for example, it has lost
3563 privilege). It wasn't ignoring creation failures other than "already
3566 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3569 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3570 because an input line was too long, either on its own, or by virtue of
3571 too many continuations, the temporary file was not being removed, and the
3572 return code was incorrect.
3574 PH/31 Missing "BOOL" in function definition in filtertest.c.
3576 PH/32 Applied Sieve patches from the maintainer.
3578 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3580 PH/33 Added "verify = not_blind".
3582 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3583 Local/Makefile (with some defaults set). These are used in built scripts
3584 such as exicyclog, but they have never been used in the exim_install
3585 script (though there are many overriding facilities there). I have
3586 arranged that the exim_install script now takes note of these two
3589 PH/35 Installed configuration files for Dragonfly.
3591 PH/36 When a locally submitted message by a trusted user did not contain a
3592 From: header, and the sender address was obtained from -f or from an SMTP
3593 MAIL command, and the trusted user did not use -F to supply a sender
3594 name, $originator_name was incorrectly used when constructing a From:
3595 header. Furthermore, $originator_name was used for submission mode
3596 messages from external hosts without From: headers in a similar way,
3597 which is clearly wrong.
3599 PH/37 Added control=suppress_local_fixups.
3601 PH/38 When log_selector = +received_sender was set, and the addition of the
3602 sender made the log line's construction buffer exactly full, or one byte
3603 less than full, an overflow happened when the terminating "\n" was
3606 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3607 when the result of a list match is failure because a DNS lookup failed.
3609 PH/40 RM_COMMAND is now used in the building process.
3611 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3612 the "build-* directories that it finds.
3614 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3615 address in a domain literal was a prefix of an interface address.
3617 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3618 when verifying a sender address, unless rewrite_headers is false.
3620 PH/44 Wrote a long comment about why errors_to addresses are verified as
3621 recipients, not senders.
3623 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3624 the ratelimit ACL was added.
3626 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3628 PH/46 Added extra information about PostgreSQL errors to the error string.
3630 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3631 avoiding the need to install special zones in a real server. This is
3632 backwards compatible; if it can't find the fake resolver, it drops back.
3633 Thus, both old and new test suites can be run.
3635 TF/02 Added util/ratelimit.pl
3637 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3638 clock is set back in time.
3640 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3641 Candler <B.Candler@pobox.com>.
3643 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3644 OK for addresses that are the result of redirections.
3646 PH/48 A number of further additions for the benefit of the new test suite,
3647 including a fake gethostbyname() that interfaces to the fake DNS resolver
3650 TF/06 The fix for widen_domains has also been applied to qualify_single and
3651 search_parents which are the other dnslookup options that can cause
3654 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3657 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3659 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3660 service" if it encountered an MX record with an empty target hostname.
3661 The message is now "an MX or SRV record indicated no SMTP service".
3663 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3664 if the DNS of the sending site is misconfigured. This is quite a
3665 common situation. This change restores the behaviour of treating a
3666 helo verification defer as a failure.
3668 PH/51 If self=fail was set on a router, the bounce message did not include the
3669 actual error message.
3675 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3677 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3678 command itself ended in error, the underlying message about the transport
3679 filter timeout was being overwritten with the pipe command error. Now the
3680 underlying error message should be appended to the second error message.
3682 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3684 PH/02 Reduce the amount of output that "make" produces by default. Full output
3685 can still be requested.
3687 PH/03 The warning log line about a condition test deferring for a "warn" verb
3688 was being output only once per connection, rather than after each
3689 occurrence (because it was using the same function as for successful
3690 "warn" verbs). This seems wrong, so I have changed it.
3692 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3693 it should not have, which might have caused a crash in the right
3694 circumstances, but probably never did.
3696 PH/04 Installed a modified version of Tony Finch's patch to make submission
3697 mode fix the return path as well as the Sender: header line, and to
3698 add a /name= option so that you can make the user's friendly name appear
3701 TF/03 Added the control = fakedefer ACL modifier.
3703 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3704 Mark Lowes for thorough testing.
3706 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3708 TK/03 Merged latest SRS patch from Miles Wilton.
3710 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3711 with the definition in sysexits.h (which is #included earlier).
3712 Fortunately, Exim does not actually use EX_OK. The code used to try to
3713 preserve the sysexits.h value, by assuming that macro definitions were
3714 scanned for macro replacements. I have been disabused of this notion,
3715 so now the code just undefines EX_OK before #including unistd.h.
3717 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3718 in the smtp transport. When a block could not be written in a single
3719 write() function, the timeout was being re-applied to each part-write.
3720 This seems wrong - if the receiver was accepting one byte at a time it
3721 would take for ever. The timeout is now adjusted when this happens. It
3722 doesn't have to be particularly precise.
3724 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3725 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3727 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3728 without a domain if the check to <postmaster@domain> fails.
3730 SC/01 Eximstats: added -xls and the ability to specify output files
3731 (patch written by Frank Heydlauf).
3733 SC/02 Eximstats: use FileHandles for outputting results.
3735 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3737 SC/04 Eximstats: fixed display of large numbers with -nvr option
3739 SC/05 Eximstats: fixed merging of reports with empty tables.
3741 SC/06 Eximstats: added the -include_original_destination flag
3743 SC/07 Eximstats: removed tabs and trailing whitespace.
3745 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3747 TK/06 MBOX spool code: Add real "From " MBOX separator line
3748 so the .eml file is really in mbox format (even though
3749 most programs do not really care). Patch from Alex Miller.
3751 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3752 The latter is generated from $received_to and is only set if the
3753 message has one envelope recipient. SA can use these headers,
3754 obviously out-of-the-box. Patch from Alex Miller.
3756 PH/08 The ${def test on a variable was returning false if the variable's
3757 value was "0", contrary to what the specification has always said!
3758 The result should be true unless the variable is empty.
3760 PH/09 The syntax error of a character other than { following "${if
3761 def:variable_name" (after optional whitespace) was not being diagnosed.
3762 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3763 accidental colon was present, for example, could give incorrect results.
3765 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3766 result is used (not including appendfile, where other changes are about
3769 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3770 This involved changing a lot of size variables from int to off_t. It
3771 should work with maildirs and everything.
3773 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3774 spamd dying while we are connected to it.
3776 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3779 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3780 function for BDB 4.3.
3782 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3784 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3785 the sources for PCRE have been re-arranged and more files are now
3788 PH/15 The code I had for printing potentially long long variables in PH/11
3789 above was not the best (it lost precision). The length of off_t variables
3790 is now inspected at build time, and an appropriate printing format (%ld
3791 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3792 to be "long long int" or "long int". This is needed for the internal
3793 formatting function string_vformat().
3795 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3796 the configuration file to be ":syslog", then the script "guesses" where
3797 the logs files are, rather than using the compiled in default. In our
3798 case the guess is not the same as the compiled default, so the script
3799 suddenly stopped working when I started to use syslog. The patch checks
3800 to see if log_file_path is "". If so, it attempts to read it from exim
3801 with no configuration file to get the compiled in version, before it
3802 falls back to the previous guessing code."
3804 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3805 implementing BATV in an Exim configuration. See NewStuff for the gory
3808 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3809 Makefile that are specific to HP-UX.
3811 PH/18 If the "use_postmaster" option was set for a recipient callout together
3812 with the "random" option, the postmaster address was used as the MAIL
3813 FROM address for the random test, but not for the subsequent recipient
3814 test. It is now used for both.
3816 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3817 patch removes a few documentation additions to RFC 3028, because the
3818 latest draft now contains them. It adds the new en;ascii-case comparator
3819 and a new error check for 8bit text in MIME parts. Comparator and
3820 require names are now matched exactly. I enabled the subaddress
3821 extension, but it is not well tested yet (read: it works for me)."
3823 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3824 rework some of the code of TK/09 above to avoid the hardwired use of
3825 "%lld" and "long long". Replaced the call to snprintf() with a call to
3828 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3829 records point to non-existent hosts", "retry timeout exceeded", and
3830 "retry time not reached for any host after a long failure period".
3832 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3833 experimental DomainKeys support:
3835 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3836 (2) On an error such as an illegally used "control", the wrong name for
3837 the control was given.
3839 These problems did NOT occur unless DomainKeys support was compiled.
3841 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3843 PH/24 Added ${if match_ip condition.
3845 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3846 databases so that it will be absolutely obvious if a crash occurs in the
3847 DB library. This is a regular occurrence (often caused by mis-matched
3850 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3851 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3852 picked up on a user's system that detects such things. There doesn't seem
3853 to be a gcc warning option for this - only an attribute that has to be
3854 put on the function's prototype. It seems that in Fedora Core 4 they have
3855 set this on a number of new functions. No doubt there will be more in due
3858 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3859 specify a transport. However, if an address that was verified by such a
3860 router was the subject of a callout, Exim crashed because it tried to
3861 read the rcpt_include_affixes from the non-existent transport. Now it
3862 just assumes that the setting of that option is false. This bug was
3863 introduced by 4.51/PH/31.
3865 PH/28 Changed -d+all to exclude +memory, because that information is very
3866 rarely of interest, but it makes the output a lot bigger. People tend to
3867 do -d+all out of habit.
3869 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3870 code in os-type was giving problems when libc.so lives in lib64, like on
3873 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3874 aren't the modern standard, and it seems that some systems' include files
3875 don't always have them. Exim was already checking for some of the newer
3876 ones like T_AAAA, and defining it itself. I've added checks for all the
3877 record types that Exim uses.
3879 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3880 not automatically generating a new one, as it is supposed to. This
3881 prevented TLS from working. If the file did exist, but contained invalid
3882 data, a new version was generated, as expected. It was only the case of a
3883 non-existent file that was broken.
3885 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3886 with a change in libdomainkeys > 0.64.
3888 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3889 from DNS. If the selector record carries the flag, it now has
3890 precedence over the domain-wide flag.
3892 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3894 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3895 the use of a port name that isn't defined in /etc/services) occurred, the
3896 message was deferred as in a normal delivery, and thus remained on the
3897 spool, instead of being failed because of the mua_wrapper setting. This
3898 is now fixed, and I tidied up some of the mua_wrapper messages at the
3901 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3902 the messages in a hash of arrays rather than using individual hashes.
3903 This is a bit cleaner and results in dramatic memory savings, albeit
3904 at a slight CPU cost.
3906 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3907 as requested by Marc Sherman.
3909 SC/10 Eximstats: added histograms for user specified patterns as requested
3912 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3914 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3915 fopen() in the content-scanning modules that did not already have it.
3921 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3922 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3924 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3926 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3928 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3929 to test Sieve filters that use "vacation".
3931 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3932 that changes the way the GnuTLS parameters are stored in the cache file.
3933 The new format can be generated externally. For backward compatibility,
3934 if the data in the cache doesn't make sense, Exim assumes it has read an
3935 old-format file, and it generates new data and writes a new file. This
3936 means that you can't go back to an older release without removing the
3939 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3940 work if there are any delivery delays because "one_time" forces the
3941 parent to be marked "delivered", so its unseen clone is never tried
3942 again. For this reason, Exim now forbids the simultaneous setting of
3945 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3946 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3947 entry if you want to know the details. The fix, however, neglected to
3948 consider the case when local delivery batching is involved. The test for
3949 "previously delivered" was not happening when checking to see if an
3950 address could be batched with a previous (undelivered) one; under
3951 certain circumstances this could lead to multiple deliveries to the same
3954 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3955 in its include files, and this causes problems building Exim.
3957 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3958 header_syntax) but Exim was just ignoring anything given after a slash.
3959 In particular, this caused confusion with an attempt to use "verify =
3960 reverse_host_lookup/defer_ok". An error is now given when options are
3961 supplied for verify items that do not have them. (Maybe reverse_host_
3962 lookup should have a defer_ok option, but that's a different point.)
3964 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3965 defined by RFC 821) to 2048, because there were problems with some AUTH
3966 commands, and RFC 1869 says the size should be increased for extended
3967 SMTP commands that take arguments.
3969 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3972 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3973 "unknown" error; now it says that the functionality isn't in the binary.
3975 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3976 an address' error message when a string expansion fails (syntax or
3977 whatever). Otherwise the password may appear in the log. Following change
3978 PH/42 below, there is no longer a chance of it appearing in a bounce
3981 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3983 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3984 of its addresses was ever tried. (Bugzilla bug #2.)
3986 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3987 the result incorrectly in the debug output. (It correctly added a newline
3988 to what was transported.)
3990 TF/01 Added $received_time.
3992 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3993 commented out examples of how to interface to a virus scanner and to
3994 SpamAssassin. Also added commented examples of av_scanner and
3995 spamd_address settings.
3997 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3998 and controls are allowed in which ACLs. There were a couple of minor
3999 errors. Some of the entries in the conditions table (which is a table of
4000 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4001 negation of where the condition IS allowed.
4003 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4005 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4006 header file does not have a version number, so I've had to invent a new
4007 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4008 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4009 radiusclient), but it was contributed by a Radius user.
4011 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4012 files or directories, for OpenSSL.
4014 PH/19 When an Exim process that is running as root has to create an Exim log
4015 file, it does so in a subprocess that runs as exim:exim so as to get the
4016 ownership right at creation (otherwise, other Exim processes might see
4017 the file with the wrong ownership). There was no test for failure of this
4018 fork() call, which would lead to the process getting stuck as it waited
4019 for a non-existent subprocess. Forks do occasionally fail when resources
4020 run out. I reviewed all the other calls to fork(); they all seem to check
4023 PH/20 When checking for unexpected SMTP input at connect time (before writing
4024 the banner), Exim was not dealing correctly with a non-positive return
4025 from the read() function. If the client had disconnected by this time,
4026 the result was a log entry for a synchronization error with an empty
4027 string after "input=" when read() returned zero. If read() returned -1
4028 (an event I could not check), uninitialized data bytes were printed.
4029 There were reports of junk text (parts of files, etc) appearing after
4032 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4034 PH/22 Added support for macro redefinition, and (re)definition in between
4035 driver and ACL definitions.
4037 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4038 forgetting to use the resulting value; it was using the unexpanded value.
4040 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4041 hadn't been configured. The fix is from Juergen Kreileder, who
4042 understands it better than I do:
4044 "Here's what I see happening with three configured cyrus_sasl
4045 authenticators configured (plain, login, cram-md5):
4047 On startup auth_cyrus_sasl_init() gets called for each of these.
4048 This means three calls to sasl_listmech() without a specified mech_list.
4049 => SASL tests which mechs of all available mechs actually work
4050 => three warnings about OTP not working
4051 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4053 With the patch, sasl_listmech() also gets called three times. But now
4054 SASL's mech_list option is set to the server_mech specified in the the
4055 authenticator. Or in other words, the answer from sasl_listmech()
4056 gets limited to just the mech you're testing for (which is different
4058 => the return list contains just 'plain' or 'login', 'cram-md5' or
4059 nothing depending on the value of ob->server_mech.
4061 I've just tested the patch: Authentication still works fine,
4062 unavailable mechs specified in the exim configuration are still
4063 caught, and the auth.log warnings about OTP are gone."
4065 PH/25 When debugging is enabled, the contents of the command line are added
4066 to the debugging output, even when log_selector=+arguments is not
4069 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4070 answer is "GNU", and only if the return is "GNU/something" is the answer
4073 PH/27 $acl_verify_message is now set immediately after the failure of a
4074 verification in an ACL, and so is available in subsequent modifiers. In
4075 particular, the message can be preserved by coding like this:
4077 warn !verify = sender
4078 set acl_m0 = $acl_verify_message
4080 Previously, $acl_verify_message was set only while expanding "message"
4081 and "log_message" when a very denied access.
4083 PH/28 Modified OS/os.c-Linux with
4085 -#ifndef OS_LOAD_AVERAGE
4086 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4088 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4089 nomenclature these days.)
4091 PH/29 Installed patch from the Sieve maintainer that adds the options
4092 sieve_useraddress and sieve_subaddress to the redirect router.
4094 PH/30 In these circumstances:
4095 . Two addresses routed to the same list of hosts;
4096 . First host does not offer TLS;
4097 . First host accepts first address;
4098 . First host gives temporary error to second address;
4099 . Second host offers TLS and a TLS session is established;
4100 . Second host accepts second address.
4101 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4102 and peerdn, if requested) that were in fact used only for the second
4105 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4106 attention to any local part prefix or suffix that was matched by the
4107 router that accepted the address. It now behaves in the same way as it
4108 does for delivery: the affixes are removed from the local part unless
4109 rcpt_include_affixes is set on the transport.
4111 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4112 timeout during the DATA phase of an incoming message.
4114 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4115 applied a patch sanctioned by the Sieve maintainer.
4117 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4118 the uid or gid is negative. A case of a negative gid caused this to be
4119 noticed. The fix allows for either to be negative.
4121 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4122 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4125 PH/36 The change PH/12 above was broken. Fixed it.
4127 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4128 the grounds that routing the same address twice would always produce the
4129 same answer. This might have been true once, but it is certainly no
4130 longer true now. Routing a child address may depend on the previous
4131 routing that produced that child. Some complicated redirection strategies
4132 went wrong when messages had multiple recipients, and made Exim's
4133 behaviour dependent on the order in which the addresses were given.
4135 I have moved the duplicate checking until after the routing is complete.
4136 Exim scans the addresses that are assigned to local and remote
4137 transports, and removes any duplicates. This means that more work will be
4138 done, as duplicates will always all be routed, but duplicates are
4139 presumably rare, so I don't expect this is of any significance.
4141 For deliveries to pipes, files, and autoreplies, the duplicate checking
4142 still happens during the routing process, since they are not going to be
4145 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4146 It corrects a timeout issue with spamd. This is Ian's comment: "The
4147 background is that sometimes spamd either never reads data from a
4148 connection it has accepted, or it never writes response data. The exiscan
4149 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4150 blindly assumes that writes won't block so it may never time out."
4152 PH/39 Allow G after quota size as well as K and M.
4154 PH/40 The value set for $authenticated_id in an authenticator may not contain
4155 binary zeroes or newlines because the value is written to log lines and
4156 to spool files. There was no check on this. Now the value is run through
4157 the string_printing() function so that such characters are converted to
4158 printable escape sequences.
4160 PH/41 $message_linecount is a new variable that contains the total number of
4161 lines in the message. Compare $body_linecount, which is the count for the
4164 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4165 bounce and delay warning messages, except in certain special cases, which
4168 (a) An SMTP error message from a remote host;
4169 (b) A message specified in a :fail: redirection;
4170 (c) A message specified in a "fail" command in a system filter;
4171 (d) A message specified in a FAIL return from the queryprogram router;
4172 (e) A message specified by the cannot_route_message router option.
4174 In these cases only, Exim does include the error details in bounce and
4175 warning messages. There are also a few cases where bland messages such
4176 as "unrouteable address" or "local delivery error" are given.
4178 PH/43 $value is now also set for the "else" part of a ${run expansion.
4180 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4181 being worked on, but at least Exim now implements the latest version to
4184 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4185 process to complete was treated as a delivery failure, a timeout while
4186 writing the message to the pipe was logged, but erroneously treated as a
4187 successful delivery. Such timeouts include transport filter timeouts. For
4188 consistency with the overall process timeout, these timeouts are now
4189 treated as errors, giving rise to delivery failures by default. However,
4190 there is now a new Boolean option for the pipe transport called
4191 timeout_defer, which, if set TRUE, converts the failures into defers for
4192 both kinds of timeout. A transport filter timeout is now identified in
4195 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4196 systems where "make" and "gmake" are different, calling "gmake" at top
4197 level broke things. I've arranged for the value of $(MAKE) to be passed
4198 from the Makefile to this script so that it can call the same version of
4202 A note about Exim versions 4.44 and 4.50
4203 ----------------------------------------
4205 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4206 changes of various kinds. As a consequence, a big documentation update was
4207 needed. This delayed the release for rather longer than seemed good, especially
4208 in the light of a couple of (minor) security issues. Therefore, the changes
4209 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4210 release. So 4.44 and 4.50 are in effect two different branches that both start
4213 I have left the 4.50 change log unchanged; it contains all the changes since
4214 4.43. The change log for 4.44 is below; many of its items are identical to
4215 those for 4.50. This seems to be the most sensible way to preserve the
4216 historical information.
4222 1. Minor wording change to the doc/README.SIEVE file.
4224 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4225 computation of the current number of files was incorrect.
4227 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4228 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4231 4. Give more explanation in the error message when the command for a transport
4232 filter fails to execute.
4234 5. There are several places where Exim runs a non-Exim command in a
4235 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4236 was being done only for the command run by the queryprogram router. It is
4237 now done for all such subprocesses. The other cases are: ${run, transport
4238 filters, and the commands run by the lmtp and pipe transports.
4240 6. Added CONFIGURE_GROUP build-time option.
4242 7. Some older OS have a limit of 256 on the maximum number of file
4243 descriptors. Exim was using setrlimit() to set 1000 as a large value
4244 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4245 systems. I've change it so that if it can't get 1000, it tries for 256.
4247 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4248 was an oversight, and furthermore, ever since the addition of extra
4249 controls (e.g. 4.43/32), the checks on when to allow different forms of
4250 "control" were broken. There should now be diagnostics for all cases when a
4251 control that does not make sense is encountered.
4253 9. Added the /retain_sender option to "control=submission".
4255 10. $recipients is now available in the predata ACL (oversight).
4257 11. Tidy the search cache before the fork to do a delivery from a message
4258 received from the command line. Otherwise the child will trigger a lookup
4259 failure and thereby defer the delivery if it tries to use (for example) a
4260 cached ldap connection that the parent has called unbind on.
4262 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4263 of $address_data from the recipient verification was clobbered by the
4264 sender verification.
4266 13. The value of address_data from a sender verification is now available in
4267 $sender_address_data in subsequent conditions in the ACL statement.
4269 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4271 15. Added a new option "connect=<time>" to callout options, to set a different
4274 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4275 was its contents. (It was OK if the option was not defined at all.)
4277 17. A "Completed" log line is now written for messages that are removed from
4278 the spool by the -Mrm option.
4280 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4281 information about exactly what failed.
4283 19. Added -dd to debug only the daemon process.
4285 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4286 handles timeouts, both on the server side and network timeouts. Renamed the
4287 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4289 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4290 It is now set to "smtps".
4292 22. $host_address is now set to the target address during the checking of
4293 ignore_target_hosts.
4295 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4296 being passed; this would have caused $sender_host_name to have been used if
4297 matching the list had actually called for a host name (not very likely,
4298 since this list is usually IP addresses). A host name is now passed as
4301 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4302 code that specifies a non-restarting handler (typically sigaction() in
4303 modern systems) in an attempt to fix a rare and obscure crash bug.
4305 25. Narrowed the window for a race in the daemon that could cause it to ignore
4306 SIGCHLD signals. This is not a major problem, because they are used only to
4307 wake it up if nothing else does.
4309 26. A malformed maildirsize file could cause Exim to calculate negative values
4310 for the mailbox size or file count. Odd effects could occur as a result.
4311 The maildirsize information is now recalculated if the size or filecount
4314 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4315 support for a long time. Removed HAVE_SYS_VFS_H.
4317 28. Installed the latest version of exipick from John Jetmore.
4319 29. In an address list, if the pattern was not a regular expression, an empty
4320 subject address (from a bounce message) matched only if the pattern was an
4321 empty string. Non-empty patterns were not even tested. This was the wrong
4322 because it is perfectly reasonable to use an empty address as part of a
4323 database query. An empty address is now tested by patterns that are
4324 lookups. However, all the other forms of pattern expect the subject to
4325 contain a local part and a domain, and therefore, for them, an empty
4326 address still always fails if the pattern is not itself empty.
4328 30. Exim went into a mad DNS loop when attempting to do a callout where the
4329 host was specified on an smtp transport, and looking it up yielded more
4330 than one IP address.
4332 31. Re-factored the code for checking spool and log partition space into a
4333 function that finds that data and another that does the check. The former
4334 is then used to implement four new variables: $spool_space, $log_space,
4335 $spool_inodes, and $log_inodes.
4337 32. The RFC2047 encoding function was originally intended for short strings
4338 such as real names; it was not keeping to the 75-character limit for
4339 encoded words that the RFC imposes. It now respects the limit, and
4340 generates multiple encoded words if necessary. To be on the safe side, I
4341 have increased the buffer size for the ${rfc2047: expansion operator from
4344 33. It is now permitted to omit both strings after an "if" condition; if the
4345 condition is true, the result is "true". As before, when the second string
4346 is omitted, a false condition yields an empty string. This makes it less
4347 cumbersome to write custom ACL and router conditions.
4349 34. Failure to deliver a bounce message always caused it to be frozen, even if
4350 there was an errors_to setting on the router. The errors_to setting is now
4353 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4354 canonical form (fully expanded) before being placed in
4355 $sender_host_address.
4357 36. The table in the code that translates DNS record types into text (T_A to
4358 "A" for instance) was missing entries for NS and CNAME. It is just possible
4359 that this could have caused confusion if both these types were looked up
4360 for the same domain, because the text type is used as part of Exim's
4361 per-process caching. But the chance of anyone hitting this buglet seems
4364 37. The dnsdb lookup has been extended in a number of ways.
4366 (1) There is a new type, "zns", which walks up the domain tree until it
4367 finds some nameserver records. It should be used with care.
4369 (2) There is a new type, "mxh", which is like "mx" except that it returns
4370 just the host names, not the priorities.
4372 (3) It is now possible to give a list of domains (or IP addresses) to be
4373 looked up. The behaviour when one of the lookups defers can be
4374 controlled by a keyword.
4376 (4) It is now possible to specify the separator character for use when
4377 multiple records are returned.
4379 38. The dnslists ACL condition has been extended: it is now possible to supply
4380 a list of IP addresses and/or domains to be looked up in a particular DNS
4383 39. Added log_selector=+queue_time_overall.
4385 40. When running the queue in the test harness, wait just a tad after forking a
4386 delivery process, to get repeatability of debugging output.
4388 41. Include certificate and key file names in error message when GnuTLS fails
4389 to set them up, because the GnuTLS error message doesn't include the name
4390 of the failing file when there is a problem reading it.
4392 42. Allow both -bf and -bF in the same test run.
4394 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4396 44. The "Exiscan patch" is now merged into the mainline Exim source.
4398 45. Sometimes the final signoff response after QUIT could fail to get
4399 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4400 < 0 before doing a fflush(). This bug looks as though it goes back to the
4401 introduction of TLS in release 3.20, but "sometimes" must have been rare
4402 because the tests only now provoked it.
4404 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4405 (this can affect the format of dates).
4407 47. exim_tidydb, when checking for the continued existence of a message for
4408 which it has found a message-specific retry record, was not finding
4409 messages that were in split spool directories. Consequently, it was
4410 deleting retry records that should have stayed in existence.
4412 48. Steve fixed some bugs in eximstats.
4414 49. The SPA authentication driver was not abandoning authentication and moving
4415 on to the next authenticator when an expansion was forced to fail,
4416 contradicting the general specification for all authenticators. Instead it
4417 was generating a temporary error. It now behaves as specified.
4419 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4420 (the order specifies the preference for clients). The order is now AES256,
4421 AES128, 3DES, ARCFOUR128.
4423 51. Small patch to Sieve code - explicitly set From: when generating an
4426 52. Exim crashed if a remote delivery caused a very long error message to be
4427 recorded - for instance if somebody sent an entire SpamAssassin report back
4428 as a large number of 550 error lines. This bug was coincidentally fixed by
4429 increasing the size of one of Exim's internal buffers (big_buffer) that
4430 happened as part of the Exiscan merge. However, to be on the safe side, I
4431 have made the code more robust (and fixed the comments that describe what
4434 53. Now that there can be additional text after "Completed" in log lines (if
4435 the queue_time_overall log selector is set), a one-byte patch to exigrep
4436 was needed to allow it to recognize "Completed" as not the last thing in
4439 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4440 patch that reportedly fixes this has been added. I am not expert enough to
4441 create a test for it. This is what the patch creator wrote:
4443 "I found a little strange behaviour of ldap code when working with
4444 Windows 2003 AD Domain, where users was placed in more than one
4445 Organization Units. When I tried to give exim partial DN, the exit code
4446 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4447 But simultaneously result of request was absolutely normal ldap result,
4448 so I produce this patch..."
4450 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4451 REFERENCE, so I have modified the code to exclude the patch when that macro
4454 55. Some experimental protocols are using DNS PTR records for new purposes. The
4455 keys for these records are domain names, not reversed IP addresses. The
4456 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4457 leaves it alone. Component reversal etc. now happens only for IP addresses.
4460 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4462 57. Double the size of the debug message buffer (to 2048) so that more of very
4463 long debug lines gets shown.
4465 58. The exicyclog utility now does better if the number of log files to keep
4466 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4468 59. Two changes related to the smtp_active_hostname option:
4470 (1) $smtp_active_hostname is now available as a variable.
4471 (2) The default for smtp_banner uses $smtp_active_hostname instead
4472 of $primary_hostname.
4474 60. The host_aton() function is supposed to be passed a string that is known
4475 to be a valid IP address. However, in the case of IPv6 addresses, it was
4476 not checking this. This is a hostage to fortune. Exim now panics and dies
4477 if the condition is not met. A case was found where this could be provoked
4478 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4479 components; fortuitously, this particular loophole had already been fixed
4480 by change 4.50/55 above.
4482 If there are any other similar loopholes, the new check in host_aton()
4483 itself should stop them being exploited. The report I received stated that
4484 data on the command line could provoke the exploit when Exim was running as
4485 exim, but did not say which command line option was involved. All I could
4486 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4487 running as the user.
4490 61. There was a buffer overflow vulnerability in the SPA authentication code
4491 (which came originally from the Samba project). I have added a test to the
4492 spa_base64_to_bits() function which I hope fixes it.
4495 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4496 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4498 63. The daemon start-up calls getloadavg() while still root for those OS that
4499 need the first call to be done as root, but it missed one case: when
4500 deliver_queue_load_max is set with deliver_drop_privilege. This is
4501 necessary for the benefit of the queue runner, because there is no re-exec
4502 when deliver_drop_privilege is set.
4504 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4505 This has been fixed.
4507 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4508 was in use, was not putting the data itself into the right store pool;
4509 consequently, it could be overwritten for a subsequent message in the same
4510 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4513 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4515 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4516 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4517 the function were treating the return as a boolean value, which happened to
4518 work because 0=false and not-0=true, but is not correct code.
4520 68. The host_aton() function was not handling scoped IPv6 addresses (those
4521 with, for example, "%eth0" on the end) correctly.
4523 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4524 negated items (that is, ~something) in unsigned ints. Some compilers
4525 apparently mutter when there is no cast.
4527 70. If an address verification called from an ACL failed, and did not produce a
4528 user-specific message (i.e. there was only a "system" message), nothing was
4529 put in $acl_verify_message. In this situation, it now puts the system
4532 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4533 session; change 4.31/43 added the unwanted input to the log line - except
4534 that it did not do this in the start of session case. It now does.
4536 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4537 This is wrong and can cause the other end to generate a synchronization
4538 error if it is another Exim or anything else that does the synchronization
4539 check. A QUIT command is no longer sent after a timeout.
4541 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4542 during host lookups.
4544 74. The defer_ok option of callout verification was not working if it was used
4545 when verifying addresses in header lines, that is, for this case:
4547 verify = header_sender/callout=defer_ok
4549 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4550 those file descriptors could be used for SMTP connections. If anything
4551 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4552 could be sent to the SMTP client, causing chaos. The daemon now opens
4553 stdin, stdout, and stderr to /dev/null when it puts itself into the
4556 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4557 log by default. The user can override this with suitable Perl magic.
4559 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4560 the log message when discard triggers, was not working for the DATA ACL or
4561 for the non-SMTP ACL.
4563 78. Error message wording change in sieve.c.
4565 79. If smtp_accept_max_per_host was set, the number of connections could be
4566 restricted to fewer than expected, because the daemon was trying to set up
4567 a new connection before checking whether the processes handling previous
4568 connections had finished. The check for completed processes is now done
4569 earlier. On busy systems, this bug wouldn't be noticed because something
4570 else would have woken the daemon, and it would have reaped the completed
4573 80. If a message was submitted locally by a user whose login name contained one
4574 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4575 It caused a spool format error. I have fixed the spool reading code. A
4576 related problem was that the "from" clause in the Received: line became
4577 illegal because of the space(s). It is now covered by ${quote_local_part.
4579 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4582 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4584 83. Patch from Sieve maintainer to fix unterminated string problem in
4585 "vacation" handling.
4587 84. Some minor changes to the Linux configuration files to help with other
4588 OS variants using glibc.
4590 85. One more patch for Sieve to update vacation handling to latest spec.
4593 ----------------------------------------------------
4594 See the note above about the 4.44 and 4.50 releases.
4595 ----------------------------------------------------
4601 1. Change 4.43/35 introduced a bug that caused file counts to be
4602 incorrectly computed when quota_filecount was set in an appendfile
4605 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4606 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4609 3. Give more explanation in the error message when the command for a transport
4610 filter fails to execute.
4612 4. There are several places where Exim runs a non-Exim command in a
4613 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4614 was being done only for the command run by the queryprogram router. It is
4615 now done for all such subprocesses. The other cases are: ${run, transport
4616 filters, and the commands run by the lmtp and pipe transports.
4618 5. Some older OS have a limit of 256 on the maximum number of file
4619 descriptors. Exim was using setrlimit() to set 1000 as a large value
4620 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4621 systems. I've change it so that if it can't get 1000, it tries for 256.
4623 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4624 was an oversight, and furthermore, ever since the addition of extra
4625 controls (e.g. 4.43/32), the checks on when to allow different forms of
4626 "control" were broken. There should now be diagnostics for all cases when a
4627 control that does not make sense is encountered.
4629 7. $recipients is now available in the predata ACL (oversight).
4631 8. Tidy the search cache before the fork to do a delivery from a message
4632 received from the command line. Otherwise the child will trigger a lookup
4633 failure and thereby defer the delivery if it tries to use (for example) a
4634 cached ldap connection that the parent has called unbind on.
4636 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4637 of $address_data from the recipient verification was clobbered by the
4638 sender verification.
4640 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4641 was its contents. (It was OK if the option was not defined at all.)
4643 11. A "Completed" log line is now written for messages that are removed from
4644 the spool by the -Mrm option.
4646 12. $host_address is now set to the target address during the checking of
4647 ignore_target_hosts.
4649 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4650 being passed; this would have caused $sender_host_name to have been used if
4651 matching the list had actually called for a host name (not very likely,
4652 since this list is usually IP addresses). A host name is now passed as
4655 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4656 code that specifies a non-restarting handler (typically sigaction() in
4657 modern systems) in an attempt to fix a rare and obscure crash bug.
4659 15. Narrowed the window for a race in the daemon that could cause it to ignore
4660 SIGCHLD signals. This is not a major problem, because they are used only to
4661 wake it up if nothing else does.
4663 16. A malformed maildirsize file could cause Exim to calculate negative values
4664 for the mailbox size or file count. Odd effects could occur as a result.
4665 The maildirsize information is now recalculated if the size or filecount
4668 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4669 support for a long time. Removed HAVE_SYS_VFS_H.
4671 18. Updated exipick to current release from John Jetmore.
4673 19. Allow an empty sender to be matched against a lookup in an address list.
4674 Previously the only cases considered were a regular expression, or an
4677 20. Exim went into a mad DNS lookup loop when doing a callout where the
4678 host was specified on the transport, if the DNS lookup yielded more than
4681 21. The RFC2047 encoding function was originally intended for short strings
4682 such as real names; it was not keeping to the 75-character limit for
4683 encoded words that the RFC imposes. It now respects the limit, and
4684 generates multiple encoded words if necessary. To be on the safe side, I
4685 have increased the buffer size for the ${rfc2047: expansion operator from
4688 22. Failure to deliver a bounce message always caused it to be frozen, even if
4689 there was an errors_to setting on the router. The errors_to setting is now
4692 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4693 canonical form (fully expanded) before being placed in
4694 $sender_host_address.
4696 24. Updated eximstats to version 1.33
4698 25. Include certificate and key file names in error message when GnuTLS fails
4699 to set them up, because the GnuTLS error message doesn't include the name
4700 of the failing file when there is a problem reading it.
4702 26. Expand error message when OpenSSL has problems setting up cert/key files.
4705 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4706 (this can affect the format of dates).
4708 28. exim_tidydb, when checking for the continued existence of a message for
4709 which it has found a message-specific retry record, was not finding
4710 messages that were in split spool directories. Consequently, it was
4711 deleting retry records that should have stayed in existence.
4713 29. eximstats updated to version 1.35
4714 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4715 1.35 - bugfix such that pie charts by volume are generated correctly
4717 30. The SPA authentication driver was not abandoning authentication and moving
4718 on to the next authenticator when an expansion was forced to fail,
4719 contradicting the general specification for all authenticators. Instead it
4720 was generating a temporary error. It now behaves as specified.
4722 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4723 (the order specifies the preference for clients). The order is now AES256,
4724 AES128, 3DES, ARCFOUR128.
4726 31. Small patch to Sieve code - explicitly set From: when generating an
4729 32. Exim crashed if a remote delivery caused a very long error message to be
4730 recorded - for instance if somebody sent an entire SpamAssassin report back
4731 as a large number of 550 error lines. This bug was coincidentally fixed by
4732 increasing the size of one of Exim's internal buffers (big_buffer) that
4733 happened as part of the Exiscan merge. However, to be on the safe side, I
4734 have made the code more robust (and fixed the comments that describe what
4737 33. Some experimental protocols are using DNS PTR records for new purposes. The
4738 keys for these records are domain names, not reversed IP addresses. The
4739 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4740 leaves it alone. Component reversal etc. now happens only for IP addresses.
4743 34. The host_aton() function is supposed to be passed a string that is known
4744 to be a valid IP address. However, in the case of IPv6 addresses, it was
4745 not checking this. This is a hostage to fortune. Exim now panics and dies
4746 if the condition is not met. A case was found where this could be provoked
4747 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4748 components; fortuitously, this particular loophole had already been fixed
4749 by change 4.50/55 or 4.44/33 above.
4751 If there are any other similar loopholes, the new check in host_aton()
4752 itself should stop them being exploited. The report I received stated that
4753 data on the command line could provoke the exploit when Exim was running as
4754 exim, but did not say which command line option was involved. All I could
4755 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4756 running as the user.
4759 35. There was a buffer overflow vulnerability in the SPA authentication code
4760 (which came originally from the Samba project). I have added a test to the
4761 spa_base64_to_bits() function which I hope fixes it.
4764 36. The daemon start-up calls getloadavg() while still root for those OS that
4765 need the first call to be done as root, but it missed one case: when
4766 deliver_queue_load_max is set with deliver_drop_privilege. This is
4767 necessary for the benefit of the queue runner, because there is no re-exec
4768 when deliver_drop_privilege is set.
4770 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4771 was in use, was not putting the data itself into the right store pool;
4772 consequently, it could be overwritten for a subsequent message in the same
4773 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4776 38. Sometimes the final signoff response after QUIT could fail to get
4777 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4778 < 0 before doing a fflush(). This bug looks as though it goes back to the
4779 introduction of TLS in release 3.20, but "sometimes" must have been rare
4780 because the tests only now provoked it.
4786 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4787 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4788 Now it returns an integer. A number of calls were still expecting a T/F
4789 return. Fortuitously, in all cases, the tests worked in OK situations,
4790 which is the norm. However, things would have gone wrong on any write
4791 failures on the smtp file descriptor. This function is used when sending
4792 messages over SMTP and also when doing verify callouts.
4794 2. When Exim is called to do synchronous delivery of a locally submitted
4795 message (the -odf or -odi options), it no longer closes stderr before doing
4798 3. Implemented the mua_wrapper option.
4800 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4802 5. Implemented the functions header_remove(), header_testname(),
4803 header_add_at_position(), and receive_remove_recipient(), and exported them
4806 6. If an ACL "warn" statement specified the addition of headers, Exim already
4807 inserted X-ACL-Warn: at the start if there was no header name. However, it
4808 was not making this test for the second and subsequent header lines if
4809 there were newlines in the string. This meant that an invalid header could
4810 be inserted if Exim was badly configured.
4812 7. Allow an ACL "warn" statement to add header lines at the start or after all
4813 the Received: headers, as well as at the end.
4815 8. Added the rcpt_4xx retry error code.
4817 9. Added postmaster_mailfrom=xxx to callout verification option.
4819 10. Added mailfrom=xxxx to the callout verification option, for verify=
4822 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4823 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4825 12. Inserted some casts to stop certain compilers warning when using pointer
4826 differences as field lengths or precisions in printf-type calls (mostly
4827 affecting debugging statements).
4829 13. Added optional readline() support for -be (dynamically loaded).
4831 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4832 same clock tick as a message's arrival, so that its received time was the
4833 same as the "first fail" time on the retry record, and that message
4834 remained on the queue past the ultimate address timeout, every queue runner
4835 would try a delivery (because it was past the ultimate address timeout) but
4836 after another failure, the ultimate address timeout, which should have then
4837 bounced the address, did not kick in. This was a "< instead of <=" error;
4838 in most cases the first failure would have been in the next clock tick
4839 after the received time, and all would be well.
4841 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4842 being recognized when the domain list was tested by the match_domain
4843 condition in an expansion string.
4845 16. Added the ${str2b64: operator.
4847 17. Exim was always calling setrlimit() to set a large limit for the number of
4848 processes, without checking whether the existing limit was already
4849 adequate. (It did check for the limit on file descriptors.) Furthermore,
4850 errors from getrlimit() and setrlimit() were being ignored. Now they are
4851 logged to the main and panic logs, but Exim does carry on, to try to do its
4852 job under whatever limits there are.
4854 18. Imported PCRE 5.0.
4856 19. Trivial typo in log message " temporarily refused connection" (the leading
4859 20. If the log selector return_path_on_delivery was set and an address was
4860 redirected to /dev/null, the delivery process crashed because it assumed
4861 that a return path would always be set for a "successful" delivery. In this
4862 case, the whole delivery is bypassed as an optimization, and therefore no
4865 21. Internal re-arrangement: the function for sending a challenge and reading
4866 a response while authentication was assuming a zero-terminated challenge
4867 string. It's now changed to take a pointer and a length, to allow for
4868 binary data in such strings.
4870 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4872 23. Exim was not respecting finduser_retries when seeking the login of the
4873 uid under which it was called; it was always trying 10 times. (The default
4874 setting of finduser_retries is zero.) Also, it was sleeping after the final
4875 failure, which is pointless.
4877 24. Implemented tls_on_connect_ports.
4879 25. Implemented acl_smtp_predata.
4881 26. If the domain in control=submission is set empty, Exim assumes that the
4882 authenticated id is a complete email address when it generates From: or
4883 Sender: header lines.
4885 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4886 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4887 chown and chgrp in /bin and hostname in /usr/bin.
4889 28. Exim was keeping the "process log" file open after each use, just as it
4890 does for the main log. This opens the possibility of it remaining open for
4891 long periods when the USR1 signal hits a daemon. Occasional processlog
4892 errors were reported, that could have been caused by this. Anyway, it seems
4893 much more sensible not to leave this file open at all, so that is what now
4896 29. The long-running daemon process does not normally write to the log once it
4897 has entered its main loop, and it closes the log before doing so. This is
4898 so that log files can straightforwardly be renamed and moved. However,
4899 there are a couple of unusual error situations where the daemon does write
4900 log entries, and I had neglected to close the log afterwards.
4902 30. The text of an SMTP error response that was received during a remote
4903 delivery was being truncated at 512 bytes. This is too short for some of
4904 the long messages that one sometimes sees. I've increased the limit to
4907 31. It is now possible to make retry rules that apply only when a message has a
4908 specific sender, in particular, an empty sender.
4910 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4911 it possible to be selective about when SMTP synchronization is enforced.
4913 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4915 32. Implemented hosts_connection_nolog.
4917 33. Added an ACL for QUIT.
4919 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4922 35. Added mailbox_size and mailbox_filecount to appendfile.
4924 36. Added control = no_multiline_responses to ACLs.
4926 37. There was a bug in the logic of the code that waits for the clock to tick
4927 in the case where the clock went backwards by a substantial amount such
4928 that the microsecond fraction of "now" was more than the microsecond
4929 fraction of "then" (but the whole seconds number was less).
4931 38. Added support for the libradius Radius client library this is found on
4932 FreeBSD (previously only the radiusclient library was supported).
4938 1. When certain lookups returned multiple values in the form name=value, the
4939 quoting of the values was not always being done properly. Specifically:
4940 (a) If the value started with a double quote, but contained no whitespace,
4942 (b) If the value contained whitespace other than a space character (i.e.
4943 tabs or newlines or carriage returns) it was not quoted.
4944 This fix has been applied to the mysql and pgsql lookups by writing a
4945 separate quoting function and calling it from the lookup code. The fix
4946 should probably also be applied to nisplus, ibase and oracle lookups, but
4947 since I cannot test any of those, I have not disturbed their existing code.
4949 2. A hit in the callout cache for a specific address caused a log line with no
4950 reason for rejecting RCPT. Now it says "Previous (cached) callout
4951 verification failure".
4953 3. There was an off-by-one bug in the queryprogram router. An over-long
4954 return line was truncated at 256 instead of 255 characters, thereby
4955 overflowing its buffer with the terminating zero. As well as fixing this, I
4956 have increased the buffer size to 1024 (and made a note to document this).
4958 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4959 when Exim is waiting for an SMTP response from a remote server, Exim
4960 restarts its select() call on the socket, thereby resetting its timeout.
4961 This is not a problem when such interrupts are rare. Somebody set up a cron
4962 job to run exiwhat every 2 minutes, which is less than the normal select()
4963 timeout (5 or 10 minutes). This meant that the select() timeout never
4964 kicked in because it was always reset. I have fixed this by comparing the
4965 time when an interrupt arrives with the time at the start of the first call
4966 to select(). If more time than the timeout has elapsed, the interrupt is
4967 treated as a timeout.
4969 5. Some internal re-factoring in preparation for the addition of Sieve
4970 extensions (by MH). In particular, the "personal" test is moved to a
4971 separate function, and given an option for scanning Cc: and Bcc: (which is
4972 not set for Exim filters).
4974 6. When Exim created an email address using the login of the caller as the
4975 local part (e.g. when creating a From: or Sender: header line), it was not
4976 quoting the local part when it contained special characters such as @.
4978 7. Installed new OpenBSD configuration files.
4980 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4981 try to make them clearer.
4983 9. Callout options, other than the timeout value, were being ignored when
4984 verifying sender addresses in header lines. For example, when using
4986 verify = header_sender/callout=no_cache
4988 the cache was (incorrectly) being used.
4990 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4991 only the Cygwin environment.
4993 11. When return_path_on_delivery was set as a log selector, if different remote
4994 addresses in the same message used different return paths and parallel
4995 remote delivery occurred, the wrong values would sometimes be logged.
4996 (Whenever a remote delivery process finished, the return path value from
4997 the most recently started remote delivery process was logged.)
4999 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5000 lines when AUTH and/or TLS are in use. This is the "received protocol"
5001 field. Exim used to use "asmtp" for authenticated SMTP, without any
5002 indication (in the protocol name) for TLS use. Now it follows the RFC and
5003 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5004 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5005 names appear in log lines as well as in Received: header lines.
5007 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5008 extensions, and comparison tests, and to fix some bugs.
5010 14. Changes to the "personal" filter test:
5012 (1) The test was buggy in that it was just doing the equivalent of
5013 "contains" tests on header lines. For example, if a user's address was
5014 anne@some.where, the "personal" test would incorrectly be true for
5016 To: susanne@some.where
5018 This test is now done by extracting each address from the header in turn,
5019 and checking the entire address. Other tests that are part of "personal"
5020 are now done using regular expressions (for example, to check local parts
5021 of addresses in From: header lines).
5023 (2) The list of non-personal local parts in From: addresses has been
5024 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5025 taken from the Sieve specification recommendations.
5027 (3) If the message contains any header line starting with "List-" it is
5028 treated as non-personal.
5030 (4) The test for "circular" in the Subject: header line has been removed
5031 because it now seems ill-conceived.
5033 15. Minor typos in src/EDITME comments corrected.
5035 16. Installed latest exipick from John Jetmore.
5037 17. If headers_add on a router specified a text string that was too long for
5038 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5039 of string_sprintf() is now avoided.
5041 18. $message_body_size was not set (it was always zero) when running the DATA
5042 ACL and the local_scan() function.
5044 19. For the "mail" command in an Exim filter, no default was being set for
5045 the once_repeat time, causing a random time value to be used if "once" was
5046 specified. (If the value happened to be <= 0, no repeat happened.) The
5047 default is now 0s, meaning "never repeat". The "vacation" command was OK
5048 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5049 (I found it when inspecting the code).
5051 20. There is now an overall timeout for performing a callout verification. It
5052 defaults to 4 times the callout timeout, which applies to individual SMTP
5053 commands during the callout. The overall timeout applies when there is more
5054 than one host that can be tried. The timeout is checked before trying the
5055 next host. This prevents very long delays if there are a large number of
5056 hosts and all are timing out (e.g. when the network connections are timing
5057 out). The value of the overall timeout can be changed by specifying an
5058 additional sub-option for "callout", called "maxwait". For example:
5060 verify = sender/callout=5s,maxwait=20s
5062 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5063 to the end before writing, but this should make it even safer).
5065 22. Exim was forgetting that it had advertised PIPELINING for the second and
5066 subsequent messages on an SMTP connection. It was also not resetting its
5067 memory on STARTTLS and an internal HELO.
5069 23. When Exim logs an SMTP synchronization error within a session, it now
5070 records whether PIPELINING has been advertised or not.
5072 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5073 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5074 rather than long int.
5076 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5078 26. Added the never_mail option to autoreply.
5084 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5085 crash if the getsockname() call failed; this can happen if a connection is
5086 closed very soon after it is established. The problem was simply in the
5087 order in which certain operations were done, causing Exim to try to write
5088 to the SMTP stream before it had set up the file descriptor. The bug has
5089 been fixed by making things happen in the correct order.
5095 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5096 before the connection was closed, thus losing the rejection response.
5098 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5099 some early Solaris releases, but causes trouble in current releases where
5100 socklen_t is defined.
5102 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5105 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5108 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5109 that releasing the top of it at the end releases what was used for sub-
5110 expansions (unless the block got too big). However, discard this block if
5111 the first thing is a variable or header, so that we can use its block when
5112 it is dynamic (useful for very large $message_headers, for example).
5114 6. Lookups now cache *every* query, not just the most recent. A new, separate
5115 store pool is used for this. It can be recovered when all lookup caches are
5116 flushed. Lookups now release memory at the end of their result strings.
5117 This has involved some general refactoring of the lookup sources.
5119 7. Some code has been added to the store_xxx() functions to reduce the amount
5120 of flapping under certain conditions.
5122 8. log_incoming_interface used to affect only the <= reception log lines. Now
5123 it causes the local interface and port to be added to several more SMTP log
5124 lines, for example "SMTP connection from", and rejection lines.
5126 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5128 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5130 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5131 could be overwritten at the end of the current message (or the start of a
5132 new message if it was set in a HELO ACL). The value is now preserved for
5133 the duration of the SMTP connection.
5135 12. If a transport had a headers_rewrite setting, and a matching header line
5136 contained an unqualified address, that address was qualified, even if it
5137 did not match any rewriting rules. The underlying bug was that the values
5138 of the flags that permit the existence of unqualified sender and recipient
5139 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5140 non-local messages, and by -bnq for local messages) were not being
5141 preserved with the message after it was received.
5143 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5144 "next input=" as part of the text comprising the host identity instead of
5145 the correct text. The code was using the same buffer for two different
5146 strings. However, depending on which order the printing function evaluated
5147 its arguments, the bug did not always show up. Under Linux, for example, my
5148 test suite worked just fine.
5150 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5151 doesn't work with some older versions of Perl. It has been changed to "my",
5152 which in any case is probably the better facility to use.
5154 15. A really picky compiler found some instances of statements for creating
5155 error messages that either had too many or two few arguments for the format
5158 16. The size of the buffer for calls to the DNS resolver has been increased
5159 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5160 for addresses that have a lot of PTR records. This alleviates a problem; it
5161 does not fully solve it.
5163 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5164 buffer now truncates the list and logs the incident, which is the same
5165 action as happens when Exim is looking up a host name and its aliases.
5166 Previously in this situation something unpredictable would happen;
5167 sometimes it was "internal error: store_reset failed".
5169 18. If a server dropped the connection unexpectedly when an Exim client was
5170 using GnuTLS and trying to read a response, the client delivery process
5171 crashed while trying to generate an error log message.
5173 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5174 string, for example:
5176 warn message = H1: something\nH2: something
5178 the text was added as a single header line from Exim's point of view
5179 though it ended up OK in the delivered message. However, searching for the
5180 second and subsequent header lines using $h_h2: did not work. This has been
5181 fixed. Similarly, if a system filter added multiple headers in this way,
5182 the routers could not see them.
5184 20. Expanded the error message when iplsearch is called with an invalid key to
5185 suggest using net-iplsearch in a host list.
5187 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5188 ACLs are no longer actually imposed (and a message to that effect is
5191 22. If a "gecos" field in a passwd entry contained escaped characters, in
5192 particular, if it contained a \" sequence, Exim got it wrong when building
5193 a From: or a Sender: header from that name. A second bug also caused
5194 incorrect handling when an unquoted " was present following a character
5195 that needed quoting.
5197 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5198 was not being matched caselessly.
5200 24. Arranged for all hyphens in the exim.8 source to be escaped with
5203 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5204 properly. Recipient callout cache records were still being keyed to include
5205 the sender, even when use_sender was set false. This led to far more
5206 callouts that were necessary. The sender is no longer included in the key
5207 when use_sender is false.
5209 26. Added "control = submission" modifier to ACLs.
5211 27. Added the ${base62d: operator to decode base 62 numbers.
5213 28. dnsdb lookups can now access SRV records.
5215 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5216 the configuration file.
5218 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5219 (-v) mode. This makes the output for a verbose queue run more intelligible.
5221 31. Added a use_postmaster feature to recipient callouts.
5223 32. Added the $body_zerocount variable, containing the number of binary zero
5224 bytes in the message body.
5226 33. The time of last modification of the "new" subdirectory is now used as the
5227 "mailbox time last read" when there is a quota error for a maildir
5230 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5232 35. Added +ignore_unknown as a special item in host lists.
5234 36. Code for decoding IPv6 addresses in host lists is now included, even if
5235 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5236 address was recognized as an IP address, but was then not correctly decoded
5237 into binary, causing unexpected and incorrect effects when compared with
5244 1. Very minor rewording of debugging text in manualroute to say "list of
5245 hosts" instead of "hostlist".
5247 2. If verify=header_syntax was set, and a header line with an unqualified
5248 address (no domain) and a large number of spaces between the end of the
5249 name and the colon was received, the reception process suffered a buffer
5250 overflow, and (when I tested it) crashed. This was caused by some obsolete
5251 code that should have been removed. The fix is to remove it!
5253 3. When running in the test harness, delay a bit after writing a bounce
5254 message to get a bit more predictability in the log output.
5256 4. Added a call to search_tidyup() just before forking a reception process. In
5257 theory, someone could use a lookup in the expansion of smtp_accept_max_
5258 per_host which, without the tidyup, could leave open a database connection.
5260 5. Added the variables $recipient_data and $sender_data which get set from a
5261 lookup success in an ACL "recipients" or "senders" condition, or a router
5262 "senders" option, similar to $domain_data and $local_part_data.
5264 6. Moved the writing of debug_print from before to after the "senders" test
5267 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5268 problems for message scanning, either using a data ACL, or using
5269 local_scan() because the Received: header was not generated till after they
5270 were called (in order to set the time as the time of reception completion).
5271 I have revised the way this works. The header is now generated after the
5272 body is received, but before the ACL or local_scan() are called. After they
5273 are run, the timestamp in the header is updated.
5279 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5280 before starting a queue runner without re-exec. This happened only when
5281 deliver_drop_privilege was set or when the Exim user was set to root. The
5282 effect of the bug was that timeouts during subsequent deliveries caused
5283 crashes instead of being properly handled. The handler is now left at its
5284 default (and expected) setting.
5286 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5287 message, again when deliver_drop_privilege is set or Exim is run as root.
5288 The bug described in (1) was not present in this case, but the tidying up
5289 of the other signals was missing. I have made the two cases consistent.
5291 3. The ignore_target_hosts setting on a manualroute router was being ignored
5292 for hosts that were looked up using the /MX notation.
5294 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5297 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5298 operated on the sender address. After changing the $sender_address to <>
5299 for the sender address verify, Exim was re-instated it as the original
5300 (before rewriting) address, but remembering that it had rewritten it, so it
5301 wasn't rewriting it again. This bug also had the effect of breaking the
5302 sender address verification caching when the sender address was rewritten.
5304 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5305 This has been changed so that if the ip literal address matches
5306 ignore_target_hosts, the router declines.
5308 7. Added expansion conditions match_domain, match_address, and match_local_
5309 part (NOT match_host).
5311 8. The placeholder for the Received: header didn't have a length field set.
5313 9. Added code to Exim itself and to exim_lock to test for a specific race
5314 condition that could lead to file corruption when using MBX delivery. The
5315 issue is with the lockfile that is created in /tmp. If this file is removed
5316 after a process has opened it but before that process has acquired a lock,
5317 there is the potential for a second process to recreate the file and also
5318 acquire a lock. This could lead to two Exim processes writing to the file
5319 at the same time. The added code performs the same test as UW imapd; it
5320 checks after acquiring the lock that its file descriptor still refers to
5321 the same named file.
5323 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5324 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5327 11. Added the smtp_active_hostname option. If used, this will typically be made
5328 to depend on the incoming interface address. Because $interface_address is
5329 not set up until the daemon has forked a reception process, error responses
5330 that can happen earlier (such as "too many connections") no longer contain
5333 12. If an expansion in a condition on a "warn" statement fails because a lookup
5334 defers, the "warn" statement is abandoned, and the next ACL statement is
5335 processed. Previously this caused the whole ACL to be aborted.
5337 13. Added the iplsearch lookup type.
5339 14. Added ident_timeout as a log selector.
5341 15. Added tls_certificate_verified as a log selector.
5343 16. Added a global option tls_require_ciphers (compare the smtp transport
5344 option of the same name). This controls incoming TLS connections.
5346 17. I finally figured out how to make tls_require_ciphers do a similar thing
5347 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5348 before starting the TLS session.
5350 18. Tabs are now shown as \t in -bP output.
5352 19. If the log selector return_path_on_delivery was set, Exim crashed when
5353 bouncing a message because it had too many Received: header lines.
5355 20. If two routers both had headers_remove settings, and the first one included
5356 a superfluous trailing colon, the final name in the first list and the
5357 first name in the second list were incorrectly joined into one item (with a
5358 colon in the middle).
5364 1. Added -C and -D options to the exinext utility, mainly to make it easier
5365 to include in the automated testing, but these could be helpful when
5366 multiple configurations are in use.
5368 2. The exinext utility was not formatting the output nicely when there was
5369 an alternate port involved in the retry record key, nor when there was a
5370 message id as well (for retries that were specific to a specific message
5371 and a specific host). It was also confused by IPv6 addresses, because of
5372 the additional colons they contain. I have fixed the IPv4 problem, and
5373 patched it up to do a reasonable job for IPv6.
5375 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5376 delivery, the log line now contains "pipelined" if PIPELINING was used.
5378 4. An SMTP transport process used to panic and die if the bind() call to set
5379 an explicit outgoing interface failed. This has been changed; it is now
5380 treated in the same way as a connect() failure.
5382 5. A reference to $sender_host_name in the part of a conditional expansion
5383 that was being skipped was still causing a DNS lookup. This no longer
5386 6. The def: expansion condition was not recognizing references to header lines
5387 that used bh_ and bheader_.
5389 7. Added the _cache feature to named lists.
5391 8. The code for checking quota_filecount in the appendfile transport was
5392 allowing one more file than it should have been.
5394 9. For compatibility with Sendmail, the command line option
5402 and sets the incoming protocol and host name (for trusted callers). The
5403 host name and its colon can be omitted when only the protocol is to be set.
5404 Note the Exim already has two private options, -pd and -ps, that refer to
5405 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5406 "s", but I don't think that's a major issue.
5408 10. A number of refactoring changes to the code, none of which should affect
5411 (a) The number of logging options was getting close to filling up the
5412 32-bit word that was used as a bit map. I have split them into two classes:
5413 those that are passed in the argument to log_write(), and those that are
5414 only ever tested independently outside of that function. These are now in
5415 separate 32-bit words, so there is plenty of room for expansion again.
5416 There is no change in the user interface or the logging behaviour.
5418 (b) When building, for example, log lines, the code previously used a
5419 macro that called string_cat() twice, in order to add two strings. This is
5420 not really sufficiently general. Furthermore, there was one instance where
5421 it was actually wrong because one of the argument was used twice, and in
5422 one call a function was used. (As it happened, calling the function twice
5423 did not affect the overall behaviour.) The macro has been replaced by a
5424 function that can join an arbitrary number of extra strings onto a growing
5427 (c) The code for expansion conditions now uses a table and a binary chop
5428 instead of a serial search (which was left over from when there were very
5429 few conditions). Also, it now recognizes conditions like "pam" even when
5430 the relevant support is not compiled in: a suitably worded error message is
5431 given if an attempt is made to use such a condition.
5433 11. Added ${time_interval:xxxxx}.
5435 12. A bug was causing one of the ddress fields not to be passed back correctly
5436 from remote delivery subprocesses. The field in question was not being
5437 subsequently used, so this caused to problems in practice.
5439 13. Added new log selectors queue_time and deliver_time.
5441 14. Might have fixed a bug in maildirsizefile handling that threw up
5442 "unexpected character" debug warnings, and recalculated the data
5443 unnecessarily. In any case, I expanded the warning message to give more
5446 15. Added the message "Restricted characters in address" to the statements in
5447 the default ACL that block characters like @ and % in local parts.
5449 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5450 Three changes have been made:
5452 (a) There was a serious bug; a negative response to MAIL caused the whole
5453 recipient domain to be cached as invalid, thereby blocking all messages
5454 to all local parts at the same domain, from all senders. This bug has
5455 been fixed. The domain is no longer cached after a negative response to
5456 MAIL if the sender used is not empty.
5458 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5461 (c) A new callout option, "use_sender" has been added for people who want
5462 the modified behaviour.
5468 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5471 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5472 indeed breaks things for older releases.
5474 3. Added additional logging to the case where there is a problem reading data
5475 from a filter that is running in a subprocess using a pipe, in order to
5476 try to track down a specific problem.
5478 4. Testing facility fudge: when running in the test harness and attempting
5479 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5480 getting "No route to host". Convert this to a timeout.
5482 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5485 6. Some OS don't have socklen_t but use size_t instead. This affects the
5486 fifth argument of getsockopt() amongst other things. This is now
5487 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5488 can be set for individual OS. I have set it for SunOS5, OSF1, and
5489 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5490 some earlier ones do not.
5492 7. Change 4.30/15 was not doing the test caselessly.
5494 8. The standard form for an IPv6 address literal was being rejected by address
5495 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5496 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5497 this, as well as the form without the "IPv6" on the front (but only when
5498 address literals are enabled, of course).
5500 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5502 10. Exim crashed if a message with an empty sender address specified by -f
5503 encountered a router with an errors_to setting. This could be provoked only
5504 by a command such as
5508 where an empty string was supplied; "<>" did not hit this bug.
5510 11. Installed PCRE release 4.5.
5512 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5513 remained set. It is now erased.
5515 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5516 times from message ids (which are base 36 rather than the normal 62).
5518 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5519 were being counted as actual protocol errors, and logged if the log
5520 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5521 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5522 and DATA following a set of rejected RCPTs do not count as protocol errors.
5523 In other words, Exim assumes they were pipelined, though this may not
5524 actually be the case. Of course, in all cases the client gets an
5525 appropriate error code.
5527 15. If a lookup fails in an ACL condition, a message about the failure may
5528 be available; it is used if testing the ACL cannot continue, because most
5529 such messages specify what the cause of the deferral is. However, some
5530 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5531 that caused an old message to be retained and used if a later statement
5532 caused a defer, replacing the real cause of the deferral.
5534 16. If an IP address had so many PTR records that the DNS lookup buffer
5535 was not large enough to hold them, Exim could crash while trying to process
5536 the truncated data. It now detects and logs this case.
5538 17. Further to 4.21/58, another change has been made: if (and only if) the
5539 first line of a message (the first header line) ends with CRLF, a bare LF
5540 in a subsequent header line has a space inserted after it, so as not to
5541 terminate the header.
5543 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5544 unnecessarily, used atoi() instead of strtol(), and didn't check the
5545 termination when getting file sizes from file names by regex.
5547 19. Completely re-implemented the support for maildirsize files, in the light
5548 of a number of problems with the previous contributed implementation
5549 (4.30/29). In particular:
5551 . If the quota is zero, the maildirsize file is maintained, but no quota is
5554 . If the maildir directory does not exist, it is created before any attempt
5555 to write a maildirsize file.
5557 . The quota value in the file is just a cache; if the quota is changed in
5558 the transport, the new value overrides.
5560 . A regular expression is available for excluding directories from the
5563 20. The autoreply transport checks the characters in options that define the
5564 message's headers; it allows continued headers, but it was checking with
5565 isspace() after an embedded newline instead of explicitly looking for a
5568 21. If all the "regular" hosts to which an address was routed had passed their
5569 expiry times, and had not reached their retry times, the address was
5570 bounced, even if fallback hosts were defined. Now Exim should go on to try
5573 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5574 equivalent code in the SMTP transport. Some hosts send humungous responses
5575 to HELO/EHLO, more than 1024 it seems.
5577 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5578 gives compiler warnings in some environments. I've now done it "properly",
5581 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5582 (because of problems with the built-in one) was declared to return uschar *
5583 instead of char *, causing compiler failure.
5585 25. Fixed a file descriptor leak when processing alias/forward files.
5587 26. Fixed a minor format string issue in dbfn.c.
5589 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5591 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5592 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5595 29. When a sender address is verified, it is cached, to save repeating the test
5596 when there is more than one recipient in a message. However, when the
5597 verification involves a callout, it is possible for different callout
5598 options to be set for different recipients. It is too complicated to keep
5599 track of this in the cache, so now Exim always runs a verification when a
5600 callout is required, relying on the callout cache for the optimization.
5601 The overhead is duplication of the address routing, but this should not be
5604 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5605 to be verified with callout=postmaster, and the main callout worked but the
5606 postmaster check failed, the verification correctly failed. However, if a
5607 subsequent RCPT command asked for sender verification *without* the
5608 postmaster check, incorrect caching caused this verification also to fail,
5611 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5612 it was not caching the DNS options (qualify_single, search_parents) that
5613 were used when the lookup failed. A subsequent lookup with different
5614 options therefore always gave the same answer, though there were cases
5615 where it should not have. (Example: a "domains = !$mx_any" option on a
5616 dnslookup router: the "domains" option is always processed without any
5617 widening, but the router might have qualify_single set.) Now Exim uses the
5618 cached value only when the same options are set.
5620 32. Added John Jetmore's "exipick" utility to the distribution.
5622 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5623 than a timeout (e.g. a certificate is required, and is not provided), an
5624 Exim server now closes the connection immediately. Previously it waited for
5625 the client to close - but if the client is SSL, it seems that they each
5626 wait for each other, leading to a delay before one of them times out.
5628 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5629 maintained 0.8.x compatibility because I don't think many are using it, and
5630 it is clearly obsolete.
5632 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5635 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5636 client certificate was expired. A simple patch fixes this, though I don't
5637 understand the full logic of why the verify callback is called multiple
5640 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5641 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5642 which causes problems with some clients (such as the Certicom SSL Plus
5643 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5644 disables the coutermeasure allowing Eudora to connect."
5646 38. Exim was not checking that a write() to a log file succeeded. This could
5647 lead to Bad Things if a log got too big, in particular if it hit a file
5648 size limit. Exim now panics and dies if it cannot write to a log file, just
5649 as it does if it cannot open a log file.
5651 39. Modified OS/Makefile-Linux so that it now contains
5653 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5655 The two -D definitions ensure that Exim is compiled with large file
5656 support, which makes it possible to handle log files that are bigger than
5659 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5660 instance) a domain was checked against a named list that involved a lookup,
5661 causing $domain_data to be set, then another domain was checked against the
5662 same list, then the first domain was re-checked, the value of $domain_data
5663 after the final check could be wrong. In particular, if the second check
5664 failed, it could be set empty. This bug probably also applied to
5667 41. The strip_trailing_dot option was not being applied to the address given
5668 with the -f command-line option.
5670 42. The code for reading a message's header from the spool was incrementing
5671 $received_count, but never initializing it. This meant that the value was
5672 incorrect (doubled) while delivering a message in the same process in which
5673 it was received. In the most common configuration of Exim, this never
5674 happens - a fresh exec is done - but it can happen when
5675 deliver_drop_privilege is set.
5677 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5678 it now includes up to 150 characters of the unexpected data in the log
5681 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5682 and building data strings. The size of both of these buffers was 10 000
5683 bytes - far larger than anybody would *ever* want, thought I. Needless to
5684 say, somebody hit the limit. I have increased the maximum line length to
5685 20 000 and the maximum data length of concatenated lines to 100 000. I have
5686 also fixed two bugs, because there was no checking on these buffers. Tsk,
5687 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5688 buffer is too small.
5690 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5691 lsearch lookups. Now it does.
5693 46. When parsing a route_list item in a manualroute router, a fixed-length
5694 buffer was used for the list of hosts. I made this 1024 bytes long,
5695 thinking that nobody would ever have a list of hosts that long. Wrong.
5696 Somebody had a whole pile of complicated expansion conditions, and the
5697 string was silently truncated, leading to an expansion error. It turns out
5698 that it is easier to change to an unlimited length (owing to other changes
5699 that have happened since this code was originally written) than to build
5700 structure for giving a limitation error. The length of the item that
5701 expands into the list of hosts is now unlimited.
5703 47. The lsearch lookup could not handle data where the length of text line was
5704 more than 4095 characters. Such lines were truncated, leading to shortened
5705 data being returned. It should now handle lines of any length.
5707 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5708 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5711 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5712 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5713 no longer happens in comment lines. A list of replacements is now placed
5714 at the head of all of the source files, except those whose only change is
5715 to replace PERL_COMMAND in the very first #! line.
5717 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5718 messages on the queue, with a bottom-up merge sort, using code contributed
5719 by Michael Haardt. This should make operations like -bp somewhat faster on
5720 large queues. It won't affect queue runners, except when queue_run_in_order
5723 51. Installed eximstats 1.31 in the distribution.
5725 52. Added support for SRV lookups to the dnslookup router.
5727 53. If an ACL referred to $message_body or $message_body_end, the value was not
5728 reset for any messages that followed in the same SMTP session.
5730 54. The store-handling optimization for building very long strings was not
5731 differentiating between the different store pools. I don't think this
5732 actually made any difference in practice, but I've tidied it.
5734 55. While running the routers to verify a sender address, $sender_address
5735 was still set to the sender address. This is wrong, because when routing to
5736 send a bounce to the sender, it would be empty. Therefore, I have changed
5737 it so that, while verifying a sender address, $sender_address is set to <>.
5738 (There is no change to what happens when verifying a recipient address.)
5740 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5741 target A or AAAA records (if not already returned) without resetting the
5742 qualify_single or search_parents options of the DNS resolver. These are
5743 inappropriate in this case because the targets of MX and SRV records must
5744 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5745 target that, when qualified, matched something in the local domain. These
5746 two options are now turned off when doing these lookups.
5748 57. It seems that at least some releases of Reiserfs (which does not have the
5749 concept of a fixed number of inodes) returns zero and not -1 for the
5750 number of available inodes. This interacted badly with check_spool_inodes,
5751 which assumed that -1 was the "no such thing" setting. What I have done is
5752 to check that the total number of inodes is greater than zero before doing
5753 the test of how many are available.
5755 58. When a "warn" ACL statement has a log_message modifier, the message is
5756 remembered, and not repeated. This is to avoid a lot of repetition when a
5757 message has many recipients that cause the same warning to be written.
5758 However, Exim was preserving the list of already written lines for an
5759 entire SMTP session, which doesn't seem right. The memory is now reset if a
5760 new message is started.
5762 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5763 debugging output unless log_rewrite was also set.
5765 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5766 of (int)(handle) when we know that handle contains (void *)(-1).
5768 61. The Exim daemon panic-logs an error return when it closes the incoming
5769 connection. However "connection reset by peer" seems to be common, and
5770 isn't really an error worthy of noting specially, so that particular error
5773 62. When Exim is trying to find all the local interfaces, it used to panic and
5774 die if the ioctl to get the interface flags failed. However, it seems that
5775 on at least one OS (Solaris 9) it is possible to have an interface that is
5776 included in the list of interfaces, but for which you get a failure error
5777 for this call. This happens when the interface is not "plumbed" into a
5778 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5779 failure of the "get flags" call assumes that the interface is down.
5781 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5782 makes life easier for people who are doing arithmetic on fields extracted
5783 from dates, where you often get leading zeros that should not be
5784 interpreted as octal.
5786 64. Added qualify_domain to the redirect router, to override the global
5789 65. If a pathologically long header line contained very many addresses (the
5790 report of this problem mentioned 10 000) and each of them was rewritten,
5791 Exim could use up a very large amount of memory. (It kept on making new
5792 copies of the header line as it rewrote, and never released the old ones.)
5793 At the expense of a bit more processing, the header rewriting function has
5794 been changed so that it no longer eats memory in this way.
5796 66. The generation of the Received: header has been moved from the time that a
5797 message starts to be received, to the time that it finishes. The timestamp
5798 in the Received: header should now be very close to that of the <= log
5799 line. There are two side-effects of this change:
5801 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5802 logged header lines no longer include the local Received: line, because
5803 it has not yet been created. The same applies to a copy of the message
5804 that is returned to a non-SMTP sender when a message is rejected.
5806 (b) When a filter file is tested using -bf, no additional Received: header
5807 is added to the test message. After some thought, I decided that this
5810 This change does not affect the value of $received_for. It is still set
5811 after address rewriting, but before local_scan() is called.
5813 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5815 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5816 gave an unhelpful panic error message, and a defer error. I have managed to
5817 change this behaviour so that it now rejects any supplied certificate,
5818 which seems right, as the list of acceptable certificates is empty.
5820 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5821 gave an unhelpful defer error. I have not managed to make this reject any
5822 supplied certificates, but the error message it gives is "no certificate
5823 supplied", which is not helpful.
5825 70. exigrep's output now also includes lines that are not associated with any
5826 message, but which match the given pattern. Implemented by a patch from
5827 Martin Sluka, which also tidied up the Perl a bit.
5829 71. Recipient callout verification, like sender verification, was using <> in
5830 the MAIL FROM command. This isn't really the right thing, since the actual
5831 sender may affect whether the remote host accepts the recipient or not. I
5832 have changed it to use the actual sender in the callout; this means that
5833 the cache record is now keyed on a recipient/sender pair, not just the
5834 recipient address. There doesn't seem to be a real danger of callout loops,
5835 since a callout by the remote host to check the sender would use <>.
5836 [SEE ABOVE: changed after hitting problems.]
5838 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5839 temporary errors. However, in the case of such a code being given after
5840 the end of a data transmission (i.e. after ".") Exim was failing to write
5841 a retry record for the message. (Yes, there was some broken host that was
5842 actually sending 8xx at this point.)
5844 73. An unknown lookup type in a host list could cause Exim to panic-die when
5845 the list was checked. (An example that provoked this was putting <; in the
5846 middle of a list instead of at the start.) If this happened during a DATA
5847 ACL check, a -D file could be left lying around. This kind of configuration
5848 error no longer causes Exim to die; instead it causes a defer error. The
5849 incident is still logged to the main and panic logs.
5851 74. Buglet left over from Exim 3 conversion. The message "too many messages
5852 in one connection" was written to the rejectlog but not the mainlog, except
5853 when address rewriting (yes!) was being logged.
5855 75. Added write_rejectlog option.
5857 76. When a system filter was run not as root (that is, when system_filter_user
5858 was set), the values of the $n variables were not being returned to the
5859 main process; thus, they were not subsequently available in the $sn
5862 77. Added +return_path_on_delivery log selector.
5864 78. A connection timeout was being treated differently from recipients deferred
5865 when testing hosts_max_try with a message that was older than the host's
5866 retry timeout. (The host should not be counted, thus allowing all hosts to
5867 be tried at least once before bouncing.) This may have been the cause of an
5868 occasionally reported bug whereby a message would remain on the queue
5869 longer than the retry timeout, but would be bounced if a delivery was
5870 forced. I say "may" because I never totally pinned down the problem;
5871 setting up timeout/retry tests is difficult. See also the next item.
5873 79. The ultimate address timeout was not being applied to errors that involved
5874 a combination of host plus message (for example, a timeout on a MAIL
5875 command). When an address resolved to a number of possible hosts, and they
5876 were not all tried for each delivery (e.g. because of hosts_max_try), a
5877 message could remain on the queue longer than the retry timeout.
5879 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5882 81. Fixed an obscure SMTP outgoing bug which required at least the following
5883 conditions: (a) there was another message waiting for the same server;
5884 (b) the server returned 5xx to all RCPT commands in the first message so
5885 that the message was not completed; (c) the server dropped the connection
5886 or gave a negative response to the RSET that Exim sends to abort the
5887 transaction. The observed case was a dropped connection after DATA that had
5888 been sent in pipelining mode. That is, the server had advertised PIPELINING
5889 but was not implementing it correctly. The effect of the bug was incorrect
5890 behaviour, such as trying another host, and this could lead to a crash.
5896 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5897 and daemon.c were passed as pointers to ints; they should have been
5898 pointers to socklen_t variables (which are typically unsigned ints).
5900 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5903 3. Fixed a really odd bug that affected only the testing scheme; patching a
5904 certain fixed string in the binary changed the value of another string that
5905 happened to be identical to the end of the original first string.
5907 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5908 name", it returns that address as the IP address. On some operating
5909 systems (e.g. Solaris), it also passes back the IP address string as the
5910 "host name". However, on others (e.g. Linux), it passes back an empty
5911 string. Exim wasn't checking for this, and was changing the host name to an
5912 empty string, assuming it had been canonicalized.
5914 5. Although rare, it is permitted to have more than one PTR record for a given
5915 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5916 all the names associated with an address, because they do in Solaris.
5917 However, it seems that they do not in Linux for data that comes from the
5918 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5919 I found this out when I moved to a new Linux workstation and tried to run
5920 the Exim test suite.
5922 To get round this problem I have changed the code so that it now does its
5923 own call to the DNS to look up PTR records when searching for a host name.
5924 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5925 addresses that are only in /etc/hosts are still found.
5927 This behaviour is, however, controlled by an option called host_lookup_
5928 order, which defaults to "bydns:byaddr". If people want to use the other
5929 order, or indeed, just use one or the other means of lookup, they can
5930 specify it in this variable.
5932 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5933 some operating systems, this comes back from gethostbyaddr() as an empty
5934 string, and this is what Exim used to test for. However, it seems that in
5935 other systems, "." is yielded. Exim now tests for this case too.
5937 7. The values of check_spool_space and check_log_space are now held internally
5938 as a number of kilobytes instead of an absolute number of bytes. If a
5939 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5940 kilobyte. This means that much larger values can be stored.
5942 8. Exim monitor: an attempt to get the action menu when not actually pointing
5943 at a message produces an empty menu entitled "No message selected". This
5944 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5945 no entries in it ("Shell widget menu has zero width and/or height"). So I
5946 have added a single, blank menu entry in this case.
5948 9. Added ${quote_local_part.
5950 10. MIME decoding is now applied to the contents of Subject: header lines when
5953 11. Now that a reference to $sender_host_address automatically causes a reverse
5954 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5955 host lookup before query-style lookups in lists that might use this
5956 variable. This has therefore been abolished, and the "net-" prefix is no
5957 longer necessary for query-style lookups.
5959 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5960 have been a typo for LFLAGS, so it has been changed.
5962 13. The install script calls Exim with "-C /dev/null" in order to find the
5963 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5964 to be output. However, since Exim outputs its version number before the
5965 error, it didn't break the script. It just looked ugly. I fixed this by
5966 always allowing "-C /dev/null" if the caller is root.
5968 14. Ignore overlarge ACL variable number when reading spool file - insurance
5969 against a later release with more variables having written the file.
5971 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5972 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5973 this, as well as the form without the "IPv6" on the front.
5975 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5976 OS/Makefile-Darwin file.
5978 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5979 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5981 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5982 variable. However, some OS use ints for inodes. I've added cast to long int
5983 to get rid of the compiler warning.
5985 19. I had forgotten to lock out "/../" in configuration file names when
5986 ALT_CONFIG_PREFIX was set.
5988 20. Routers used for verification do not need to specify transports. However,
5989 if such a router generated a host list, and callout was configured, Exim
5990 crashed, because it could not find a port number from the (non-existent)
5991 transport. It now assumes port 25 in this circumstance.
5993 21. Added the -t option to exigrep.
5995 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5996 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5997 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5998 them is set, LOOKUP_LSEARCH is forced.
6000 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6002 24. Added sender and host information to the "rejected by local_scan()" log
6003 line; previously there was no indication of these.
6005 25. Added .include_if_exists.
6007 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6008 Linux. It turns out that not all file systems support this. Apparently some
6009 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6010 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6011 not supported on the file descriptor, is now ignored when Exim is trying to
6012 sync a directory. This applies only to Linux.
6014 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6016 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6017 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6020 29. Added support for maildirsize files from supplied patch (modified a bit).
6022 30. The use of :fail: followed by an empty string could lead Exim to respond to
6023 sender verification failures with (e.g.):
6025 550 Verification failed for <xxx>
6026 550 Sender verify failed
6028 where the first response line was missing the '-' that indicates it is not
6029 the final line of the response.
6031 31. The loop for finding the name of the user that called Exim had a hardwired
6032 limit of 10; it now uses the value of finduser_retries, which is used for
6033 all other user lookups.
6035 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6038 33. Exim was neglecting to zero errno before one call of strtol() when
6039 expanding a string and expecting an integer value. On some systems this
6040 resulted in spurious "integer overflow" errors. Also, it was casting the
6041 result into an int without checking.
6043 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6044 did not work. The code looks as if it has *never* worked, though it appears
6045 to have been documented since at least release 1.62. I have made it work.
6047 35. The "timeout_DNS" error in retry rules, also documented since at least
6048 1.62, also never worked. As it isn't clear exactly what this means, and
6049 clearly it isn't a major issue, I have abolished the feature by treating it
6050 as "timeout", and writing a warning to the main and panic logs.
6052 36. The display of retry rules for -brt wasn't always showing the error code
6055 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6056 timeout_connect_A, timeout_connect_MX.
6058 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6059 to the empty sender.
6061 39. The daemon was not analysing the content of -oX till after it had closed
6062 stderr and disconnected from the controlling terminal. This meant that any
6063 syntax errors were only noted on the panic log, and the return code from
6064 the command was 0. By re-arranging the code a little, I've made the
6065 decoding happen first, so such errors now appear on stderr, and the return
6066 code is 1. However, the actual setting up of the sockets still happens in
6067 the disconnected process, so errors there are still only recorded on the
6070 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6071 connections (as happens on some IP stacks) was logged at start up time as
6072 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6073 it from "IPv6 and IPv4", which means that two separate sockets are being
6076 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6077 says whether AF_INET or AF_INET6 was passed as an argument.
6079 42. Exiwhat output was messed up when time zones were included in log
6082 43. Exiwhat now gives more information about the daemon's listening ports,
6083 and whether -tls-on-connect was used.
6085 44. The "port" option of the smtp transport is now expanded.
6087 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6088 silently ignored. Now an error message is written to the main and panic
6091 46. There's a new ACL modifier called "logwrite" which writes to a log file
6092 as soon as it is encountered.
6094 47. Added $local_user_uid and $local_user_gid at routing time.
6096 48. Exim crashed when trying to verify a sender address that was being
6099 49. Exim was recognizing only a space character after ".include". It now also
6100 recognizes a tab character.
6102 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6103 extracting the relevant information from the specification. The man page no
6104 longer contains scrambled data for the -d option, and I've added a section
6105 at the front about calling Exim under different names.
6107 51. Added "extra_headers" argument to the "mail" command in filter files.
6109 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6112 53. Installed eximstats 1.29.
6114 54. Added transport_filter_timeout as a generic transport option.
6116 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6117 Cc: header lines. This was required by RFC 822, but it not required by RFC
6120 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6121 incoming messages that did not have them. Now it does so only if the
6122 message originates locally, that is, if there is no associated remote host
6123 address. When Resent- header lines are present, this applies to the Resent-
6124 lines rather than the non-Resent- lines.
6126 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6127 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6129 58. Messages for configuration errors now include the name of the main
6130 configuration file - useful now that there may be more than one file in a
6131 list (.included file names were always shown).
6133 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6134 for those rare installations that do not start the daemon as root or run it
6135 setuid root. I've cut out the call to initgroups() if the daemon is not
6138 60. The Exim user and group can now be bound into the binary as text strings
6139 that are looked up at the start of Exim's processing.
6141 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6143 62. Added $mailstore_basename variable.
6145 63. Installed patch to sieve.c from Michael Haardt.
6147 64. When Exim failed to open the panic log after failing to open the main log,
6148 the original message it was trying to log was written to stderr and debug
6149 output, but if they were not available (the usual case in production), it
6150 was lost. Now it is written to syslog before the two lines that record the
6151 failures to open the logs.
6153 65. Users' Exim filters run in subprocesses under the user's uid. It is
6154 possible for a "deliver" command or an alias in a "personal" command to
6155 provoke an address rewrite. If logging of address rewriting is configured,
6156 this fails because the process is not running as root or exim. There may be
6157 a better way of dealing with this, but for the moment (because 4.30 needs
6158 to be released), I have disabled address rewrite logging when running a
6159 filter in a non-root, non-exim process.
6165 1. The buildconfig auxiliary program wasn't quoting the value set for
6166 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6167 not defined. This bug was masked in 4.22 by the effect that was fixed in
6170 2. Some messages that were rejected after a message id was allocated were
6171 shown as "incomplete" by exigrep. It no longer does this for messages that
6172 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6174 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6175 have allow_domain_literals set, the ID did not get logged in the <= line.
6176 Domain literals are now always recognized in Message-ID: header lines.
6178 4. The first argument for a ${extract expansion item is the key name or field
6179 number. Leading and trailing spaces in this item were not being ignored,
6180 causing some misleading effects.
6182 5. When deliver_drop_privilege was set, single queue runner processes started
6183 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6184 same command in the process it spins off) were not dropping privilege.
6186 6. When the daemon running as "exim" started a queue runner, it always
6187 re-executed Exim in the spun-off process. This is a waste of effort when
6188 deliver_drop_privilege is set. The new process now just calls the
6189 queue-runner function directly.
6195 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6198 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6199 ignored. Though the use of -oP was forcing the writing of a pid file, it
6200 was always written to the default place.
6202 3. If the message "no IP address found for host xxxx" is generated during
6203 incoming verification, it is now followed by identification of the incoming
6204 connection (so you can more easily find what provoked it).
6206 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6208 5. Added some features to "harden" Exim a bit more against certain attacks:
6210 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6211 be put in Local/Makefile. This is like the never_users runtime option,
6212 but it cannot be overridden. The default setting is "root".
6214 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6215 prefix string with which any file named in a -C command line option
6218 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6219 is retained for -C and -D only if the caller of Exim is root. Without
6220 it, the exim user may also use -C and -D and retain privilege.
6222 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6223 command line option is disabled.
6225 6. Macro names set by the -D option must start with an upper case letter, just
6226 like macro names defined in the configuration file.
6228 7. Added "dereference=" facility to LDAP.
6230 8. Two instances of the typo "uknown" in the source files are fixed.
6232 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6233 the Configure-Makefile script screwed up while processing it.
6235 10. Incorporated PCRE 4.4.
6237 11. The SMTP synchronization check was not operating right at the start of an
6238 SMTP session. For example, it could not catch a HELO sent before the client
6239 waited for the greeting. There is now a check for outstanding input at the
6240 point when the greeting is written. Because of the duplex, asynchronous
6241 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6242 way, but not yet received, when the check is performed.
6244 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6245 on TCP/IP sockets, because this apparently causes some broken clients to
6248 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6249 unchanged) from the Cygwin maintainer.
6251 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6252 was supported instead of testing for mbx. Effectively a typo.
6254 15. The spa authenticator server code was not checking that the input it
6255 received was valid base64.
6257 16. The debug output line for the "set" modifier in ACLs was not showing the
6258 name of the variable that was being set.
6260 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6262 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6263 lookup on its own. Something else was needed to trigger the lookup. For
6264 example, a match in host_lookup or the need for a host name in a host list.
6265 Now, if $sender_host_name is referenced and the host name has not yet been
6266 looked up, a lookup is performed. If the lookup fails, the variable remains
6267 empty, and $host_lookup_failed is set to "1".
6269 19. Added "eqi" as a case-independent comparison operator.
6271 20. The saslauthd authentication condition could segfault if neither service
6272 nor realm was specified.
6274 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6275 error message that was logged was misleading, and incoming SMTP
6276 connections were dropped. The message is now more accurate, and temporary
6277 errors are given to SMTP connections.
6279 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6280 (except RSET) with a 421 error, until QUIT is received. However, it was
6281 failing to send a response to QUIT.
6283 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6284 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6285 ACL is now run after the helo_try_verify_hosts code.
6287 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6288 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6289 case-independent, so other case variants are also recognized). Apparently
6290 some systems use these upper case variants.
6292 25. If more than two messages were waiting for the same host, and a transport
6293 filter was specified for the transport, Exim sent two messages over the
6294 same TCP/IP connection, and then failed with "socket operation on non-
6295 socket" when it tried to send the third.
6297 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6299 27. The extern definition of crypt16() in expand.c was not being excluded when
6300 the OS had its own crypt16() function.
6302 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6303 as a preferred synonym for return_size_limit, both as an option and as an
6306 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6308 30. Changed the default configuration ACL to relax the local part checking rule
6309 for addresses that are not in any local domains. For these addresses,
6310 slashes and pipe symbols are allowed within local parts, but the sequence
6311 /../ is explicitly forbidden.
6313 31. SPA server authentication was not clearing the challenge buffer before
6316 32. log_message in a "warn" ACL statement was writing to the reject log as
6317 well as to the main log, which contradicts the documentation and doesn't
6318 seem right (because no rejection is happening). So I have stopped it.
6320 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6321 However, I am unable to do any testing of this.
6323 34. Fixed an infelicity in the appendfile transport. When checking directories
6324 for a mailbox, to see if any needed to be created, it was accidentally
6325 using path names with one or more superfluous leading slashes; tracing
6326 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6328 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6329 contents are added to the log line that is written for every discarded
6330 recipient. (Previously a log_message setting was ignored.)
6332 36. The ${quote: operator now quotes the string if it is empty.
6334 37. The install script runs exim in order to find its version number. If for
6335 some reason other than non-existence or emptiness, which it checks, it
6336 could not run './exim', it was installing it with an empty version number,
6337 i.e. as "exim-". This error state is now caught, and the installation is
6340 38. An argument was missing from the function that creates an error message
6341 when Exim fails to connect to the socket for saslauthd authentication.
6342 This could cause Exim to crash, or give a corrupted message.
6344 39. Added isip, isip4, and isip6 to ${if conditions.
6346 40. The ACL variables $acl_xx are now saved with the message, and can be
6347 accessed later in routers, transports, and filters.
6349 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6350 strings in the file are not string-expanded.
6352 42. If a MAIL command specified a SIZE value that was too large to fit into an
6353 int variable, the check against message_size_limit failed. Such values are
6354 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6355 day this will have to be increased, but I don't think I want to be around
6356 when emails are that large.
6363 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6364 iconv() is not standard in FreeBSD.
6366 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6367 IPv6 enabled. The observed symptom was a segmentation fault on return from
6368 the function os_common_find_running_interfaces() in src/os.c.
6370 3. In the check_special_case() function in daemon.c I had used "errno" as an
6371 argument name, which causes warnings on some systems. This was basically a
6372 typo, since it was named "eno" in the comments!
6374 4. The code that waits for the clock to tick (at a resolution of some fraction
6375 of a second) so as to ensure message-id uniqueness was always waiting for
6376 at least one whole tick, when it could have waited for less. [This is
6377 almost certainly not relevant at current processor speeds, where it is
6378 unlikely to ever wait at all. But we try to future-proof.]
6380 5. The function that sleeps for a time interval that includes fractions of a
6381 second contained a race. It did not block SIGALRM between setting the
6382 timer, and suspending (a couple of lines later). If the interval was short
6383 and the sigsuspend() was delayed until after it had expired, the suspension
6384 never ended. On busy systems this could lead to processes getting stuck for
6387 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6388 process, before it forks any delivery processes. The open lookup caching
6389 mechanism meant that the open file or database connection was passed into
6390 the delivery process. The problem was that delivery processes always tidy
6391 up cached lookup data. This could cause a problem for the next delivery
6392 process started by the queue runner, because the external queue runner
6393 process does not know about the closure. So the next delivery process
6394 still has data in the lookup cache. In the case of a file lookup, there was
6395 no problem because closing a file descriptor in a subprocess doesn't affect
6396 the parent. However, if the lookup was caching a connection to a database,
6397 the connection was closed, and the second delivery process was likely to
6398 see errors such as "PGSQL: query failed: server closed the connection
6399 unexpectedly". The problem has been fixed by closing all cached lookups
6400 in a queue runner before running a delivery process.
6402 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6403 seem to have the "const" qualifier which it has on other OS. I've
6406 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6407 *of the same type* (client or server) with the same public name that an
6408 error should be diagnosed.
6410 9. When Exim looked up a host name for an IP address, but failed to find the
6411 original IP address when looking up the host name (a safety check), it
6412 output the message "<ip address> does not match any IP for NULL", which was
6413 confusing, to say the least. The bug was that the host name should have
6414 appeared instead of "NULL".
6416 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6417 user that is built into the binary, and the -C or -D options is used, root
6418 privilege is dropped before the configuration file is read. In addition,
6419 logging is switched to stderr instead of the normal log files. If the
6420 configuration then re-defines the Exim user, the unprivileged environment
6421 is probably not what is expected, so Exim logs a panic warning message (but
6424 However, if deliver_drop_privilege is set, the unprivileged state may well
6425 be exactly what is intended, so the warning has been cut out in that case,
6426 and Exim is allowed to try to write to its normal log files.
6432 1. smtp_return_error_details was not giving details for temporary sender
6433 or receiver verification errors.
6435 2. Diagnose a configuration error if two authenticators have the same public
6438 3. Exim used not to create the message log file for a message until the first
6439 delivery attempt. This could be confusing when incoming messages were held
6440 for policy or load reasons. The message log file is now created at the time
6441 the message is received, and an initial "Received" line is written to it.
6443 4. The automatically generated man page for command line options had a minor
6444 bug that caused no ill effects; however, a more serious problem was that
6445 the procedure for building the man page automatically didn't always
6446 operate. Consequently, release 4.20 contains an out-of-date version. This
6447 shouldn't happen again.
6449 5. When building Exim with embedded Perl support, the script that builds the
6450 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6451 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6453 6. The freeze_tell option was not being used for messages that were frozen on
6454 arrival, either by an ACL or by local_scan().
6456 7. Added the smtp_incomplete_transaction log selector.
6458 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6459 was accepting AUTH without a new EHLO.
6461 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6462 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6465 10. Logging of TCP/IP connections (when configured) now happens in the main
6466 daemon process instead of the child process, so that the TCP/IP connection
6467 count is more accurate (but it can never be perfect).
6469 11. The use of "drop" in a nested ACL was not being handled correctly in the
6470 outer ACL. Now, if condition failure induced by the nested "drop" causes
6471 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6472 or "require"), the connection is dropped.
6474 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6475 that yield "discard" can now be used with an "accept" or a "discard" verb,
6476 but an error is generated for any others (because I can't see a useful way
6477 to define what should happen).
6479 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6480 are now processed in the same way as lines in the Exim configuration file.
6481 In particular, continuation lines are supported.
6483 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6485 15. Added -ti meaning -t -i.
6487 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6488 domains, and warn by logging if others are found.
6490 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6491 returned from GIFCONF when Exim is trying to find the list of interfaces on
6492 a host. The code in os.c has been modified to copy each ifreq to an aligned
6493 structure in all cases.
6495 Also, in some cases, the returned ifreq's were being copied to a 'struct
6496 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6497 means the last couple of bytes of an IPv6 address could be chopped if the
6498 ifreq contained only a normal sockaddr (14 bytes storage).
6500 18. Named domain lists were not supported in the hosts_treat_as_local option.
6501 An entry such as +xxxx was not recognized, and was treated as a literal
6504 19. Ensure that header lines added by a DATA ACL are included in the reject log
6505 if the ACL subsequently rejects the message.
6507 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6508 MD5 (which is deprecated).
6510 21. When testing a filter file using -bf, Exim was writing a message when it
6511 took the sender from a "From " line in the message, but it was not doing so
6512 when it took $return_path from a Return-Path: header line. It now does.
6514 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6515 with a valid header line field name (a series of printing characters
6516 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6518 23. Changed "disc" in the source to "disk" to conform to the documentation and
6519 the book and for uniformity.
6521 24. Ignore Sendmail's -Ooption=value command line item.
6523 25. When execve() failed while trying to run a command in a pipe transport,
6524 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6525 could be confused with a return value of 69 from the command itself. This
6526 has been changed to 127, the value the shell returns if it is asked to run
6527 a non-existent command. The wording for the related log line suggests a
6528 non-existent command as the problem.
6530 26. If received_header_text expands to an empty string, do not add a Received:
6531 header line to the message. (Well, it adds a token one on the spool, but
6532 marks it "old" so that it doesn't get used or transmitted.)
6534 27. Installed eximstats 1.28 (addition of -nt option).
6536 28. There was no check for failure on the call to getsockname() in the daemon
6537 code. This can fail if there is a shortage of resources on the system, with
6538 ENOMEM, for example. A temporary error is now given on failure.
6540 29. Contrary to the C standard, it seems that in some environments, the
6541 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6542 program. Exim now does this explicitly; it affects the formatting of
6543 timestamps using strftime().
6545 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6546 complaints. I've now initialized all the variables, to avoid this.
6548 32. Header lines added by a system filter were not being "seen" during
6549 transport-time rewrites.
6551 33. The info_callback() function passed to OpenSSL is set up with type void
6552 (*)(SSL *, int, int), as described somewhere. However, when calling the
6553 function (actually a macro) that sets it up, the type void(*)() is
6554 expected. I've put in a cast to prevent warnings from picky compilers.
6556 34. If a DNS black list lookup found a CNAME record, but there were no A
6557 records associated with the domain it pointed at, Exim crashed.
6559 35. If a DNS black list lookup returned more than one A record, Exim ignored
6560 all but the first. It now scans all returned addresses if a particular IP
6561 value is being sought. In this situation, the contents of the
6562 $dnslist_value variable are a list of all the addresses, separated by a
6565 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6566 to accept a host name and all its aliases if the forward lookup for any of
6567 them yielded the IP address of the incoming connection. Now it accepts only
6568 those names whose forward lookup yields the correct IP address. Any other
6569 names are discarded. This closes a loophole whereby a rogue DNS
6570 administrator could create reverse DNS records to break through a
6571 wildcarded host restriction in an ACL.
6573 37. If a user filter or a system filter that ran in a subprocess used any of
6574 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6575 the wrong values were passed to the pipe command ($thisaddress had the
6576 value of $0, $0 had the value of $1, etc). This bug was introduced by
6577 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6579 38. Improved the line breaking for long SMTP error messages from ACLs.
6580 Previously, if there was no break point between 40 and 75 characters, Exim
6581 left the rest of the message alone. Two changes have been made: (a) I've
6582 reduced the minimum length to 35 characters; (b) if it can't find a break
6583 point between 35 and 75 characters, it looks ahead and uses the first one
6584 that it finds. This may give the occasional overlong line, but at least the
6585 remaining text gets split now.
6587 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6588 file descriptors might be low, and that setting 1000 would always raise it.
6589 It turns out that in some environments, the limit is already over 1000 and
6590 that lowering it causes trouble. So now Exim takes care not to decrease it.
6592 40. When delivering a message, the value of $return_path is set to $sender_
6593 address at the start of routing (routers may change the value). By an
6594 oversight, this default was not being set up when an address was tested by
6595 -bt or -bv, which affected the outcome if any router or filter referred to
6598 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6599 log only when "message" or "log_message" are set. However, if one of the
6600 conditions was an address verification, or a call to a nested ACL, the
6601 messages generated by the underlying test were being passed through. This
6602 no longer happens. The underlying message is available in $acl_verify_
6603 message for both "message" and "log_message" expansions, so it can be
6604 passed through if needed.
6606 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6607 new expansion $bh_ to give the encoded byte string without charset
6608 translation. Translation happens only if iconv() is available; HAVE_ICONV
6609 indicates this at build time. HEADERS_CHARSET gives the charset to
6610 translate to; headers_charset can change it in the configuration, and
6611 "headers charset" can change it in an individual filter file.
6613 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6614 that creates RFC 2047 encoded "words" labels them as that charset instead
6615 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6616 expansion operator; (ii) when Exim creates a From: line for a local
6617 message; (iii) when a header line is rewritten to include a "phrase" part.
6619 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6620 buggy, causing it to skip the first lines of messages whose message ID
6621 ended in 'D'. This would not have bitten before Exim release 4.14, because
6622 message IDs were unlikely to end in 'D' before then. The effect was to have
6623 incorrect size information for certain domains.
6625 45. #include "config.h" was missing at the start of the crypt16.c module. This
6626 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6629 46. If there was a timeout during a "random" callout check, Exim treated it as
6630 a failure of the random address, and carried on sending RSET and the real
6631 address. If the delay was just some slowness somewhere, the response to the
6632 original RCPT would be taken as a response to RSET and so on, causing
6633 mayhem of various kinds.
6635 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6636 when I implemented it. It didn't allow for the fact that some option values
6637 may legitimately be negative (e.g. size_addition), and it didn't even do
6638 the right test for positive values.
6640 48. Domain names in DNS records are case-independent. Exim always looks them up
6641 in lower case. Some resolvers return domain names in exactly the case they
6642 appear in the zone file, that is, they may contain uppercase letters. Not
6643 all resolvers do this - some return always lower case. Exim was treating a
6644 change of case by a resolver as a change of domain, similar to a widening
6645 of a domain abbreviation. This triggered its re-routing code and so it was
6646 trying to route what was effectively the same domain again. This normally
6647 caused routing to fail (because the router wouldn't handle the domain
6648 twice). Now Exim checks for this case specially, and just changes the
6649 casing of the domain that it ultimately uses when it transmits the message
6652 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6655 50. If a filter generated a file delivery with a non-absolute name (possible if
6656 no home directory exists for the router), the forbid_file option was not
6659 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6660 the existing equality matching.
6662 52. Exim was using ints instead of ino_t variables in some places where it was
6663 dealing with inode numbers.
6665 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6666 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6667 called TMPDIR, and if it finds it is different, it changes its value.
6669 54. The smtp_printf() function is now made available to local_scan() so
6670 additional output lines can be written before returning. There is also an
6671 smtp_fflush() function to enable the detection of a dropped connection.
6672 The variables smtp_input and smtp_batched_input are exported to
6675 55. Changed the default runtime configuration: the message "Unknown user"
6676 has been removed from the ACL, and instead placed on the localuser router,
6677 using the cannot_route_message feature. This means that any verification
6678 failures that generate their own messages won't get overridden. Similarly,
6679 the "Unrouteable address" message that was in the ACL for unverifiable
6680 relay addresses has also been removed.
6682 56. Added hosts_avoid_esmtp to the smtp transport.
6684 57. The exicyclog script was not checking for the esoteric option
6685 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6686 will work only if exicyclog is run under the appropriate euid.
6688 58. Following a discussion on the list, the rules by which Exim recognises line
6689 endings on incoming messages have been changed. The -dropcr and drop_cr
6690 options are now no-ops, retained only for backwards compatibility. The
6691 following line terminators are recognized: LF CRLF CR. However, special
6692 processing applies to CR:
6694 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6695 nor a local message in the state where . is a terminator.
6697 (ii) If a bare CR is encountered in a header line, an extra space is added
6698 after the line terminator so as not to end the header. The reasoning
6699 behind this is that bare CRs in header lines are most likely either
6700 to be mistakes, or people trying to play silly games.
6702 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6703 was being incorrectly given as 18 bytes larger than it should have been.
6704 This is a VOB (very old bug).
6706 60. This may never have affected anything current, but just in case it has:
6707 When the local host is found other than at the start of a list of hosts,
6708 the local host, those with the same MX, and any that follow, are discarded.
6709 When the list in question was part of a longer list of hosts, the following
6710 hosts (not currently being processed) were also being discarded. This no
6711 longer happens. I'm not sure if this situation could ever has previously
6714 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6717 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6718 header. This is something that is recommended in a new Internet Draft, and
6719 is something that is documented as being done by Sendmail. There are two
6720 possible values. For messages generated by the autoreply transport, Exim
6723 Auto-Submitted: auto-replied
6725 whereas for all other generated messages (e.g. bounces) it adds
6727 Auto-Submitted: auto-generated
6729 63. The "personal" condition in filters now includes a test for the
6730 Auto-Submitted: header. If it contains the string "auto-" the message it
6731 not considered personal.
6733 64. Added rcpt_include_affixes as a generic transport option.
6735 65. Added queue_only_override (default true).
6737 66. Added the syslog_duplication option.
6739 67. If what should have been the first header line of a message consisted of
6740 a space followed by a colon, Exim was mis-interpreting it as a header line.
6741 It isn't of course - it is syntactically invalid and should therefore be
6742 treated as the start of the message body. The misbehaviour could have
6743 caused a number of strange effects, including loss of data in subsequent
6744 header lines, and spool format errors.
6746 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6747 client host had authenticated. This control can now be exercised by an ACL
6748 for more flexibility.
6750 69. By default, callouts do not happen when testing with -bh. There is now a
6751 variant, -bhc, which does actually run the callout code, including
6752 consulting and updating the callout cache.
6754 70. Added support for saslauthd authentication, courtesy of Alexander
6757 71. If statvfs() failed on the spool or log directories while checking their
6758 size for availability, Exim confusingly gave the error "space shortage".
6759 Furthermore, in debugging mode it crashed with a floating point exception.
6760 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6761 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6762 serious problem, Exim now writes to the main and panic logs when this
6763 happens, with details of the failure. It then refuses to accept the
6764 incoming message, giving the message "spool directory problem" or "log
6765 directory problem" with a 421 code for SMTP messages.
6767 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6768 0, 1, and 2 exist, because some OS complain for execs without them (see
6769 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6770 descriptors. However, the code omitted to check that the open succeeded,
6771 causing mysterious errors if for some reason the permissions on /dev/null
6772 got screwed. Now Exim writes a message to the main and panic logs, and
6773 bombs out if it can't open /dev/null.
6775 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6776 interact so that it is all more flexible. It is supposed to remain
6777 backwards compatible. Also added extra_local_interfaces.
6779 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6780 to bomb out with an assertion failure - to the client this appears as a
6781 connection drop. This problem occurs in the part of the code that was taken
6782 from the Samba project. Fortunately, the assertion is in a very simple
6783 function, so I have fixed this by reproducing the function inline in the
6784 one place where it is called, and arranging for authentication to fail
6785 instead of killing the process with assert().
6787 75. The SPA client code was not working when the server requested OEM rather
6788 than Unicode encoding.
6790 76. Added code to make require_files with a specific uid setting more usable in
6791 the case where statting the file as root fails - usually a non-root-mounted
6792 NFS file system. When this happens and the failure is EACCES, Exim now
6793 forks a subprocess and does the per-uid checking as the relevant uid.
6795 77. Added process_log_path.
6797 78. If log_file_path was not explicitly set, a setting of check_log_space or
6798 check_log_inodes was ignored.
6800 79. If a space check for the spool or log partitions fails, the incident is now
6801 logged. Of course, in the latter case the data may get lost...
6803 80. Added the %p formatting code to string_format() so that it can be used to
6804 print addresses in debug_print(). Adjusted all the address printing in the
6805 debugging in store.c to use %p rather than %d.
6807 81. There was a concern that a line of code in smtp_in.c could overflow a
6808 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6809 initially expressed, the concern was not well-founded, because trailing
6810 spaces are removed early. However, if the trailing spaces were followed by
6811 a NULL, they did not get removed, so the overflow was possible. Two fixes
6814 (a) I re-wrote the offending code in a cleaner fashion.
6815 (b) If an incoming SMTP command contains a NULL character, it is rejected
6818 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6819 runs initgroups(), so that if the Exim user is in any additional groups,
6820 they will be used during message reception.
6826 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob