1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
64 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
65 4.77 as the kernel support then wasn't solid, having issues in 64bit
66 mode. Now, it's been long enough.
68 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
69 case where the routing stage had gathered several addresses to send to
70 a host before calling the transport for the first, we previously failed
71 to close down TLS in the old transport process before passing the TCP
72 connection to the new process. The new one sent a STARTTLS command
73 which naturally failed, giving a failed delivery and bloating the retry
74 database. Investigation and fix prototype from Wolfgang Breyha.
80 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
81 than -2003 did; needs libidn2 in addition to libidn.
83 JH/02 The path option on a pipe transport is now expanded before use.
85 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
86 Patch provided by "Björn", documentation fix added too.
88 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
89 missing a wire-to-host endian conversion.
91 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
92 close after a BDAT command line could be taken as a following command,
93 giving a synch failure. Fix by only checking for synch immediately
94 before acknowledging the chunk.
96 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
97 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
98 macro. Patches provided by Josh Soref.
100 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
101 Previously we did not; the RFC seems ambiguous and VRFY is not listed
102 by IANA as a service extension. However, John Klensin suggests that we
105 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
106 the dkim code may be unix-mode line endings rather than smtp wire-format
107 CRLF, so prepend a CR to any bare LF.
109 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
110 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
112 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
113 the first were themselves being wrongly included in the feed into dkim
114 processing; with most chunk sizes in use this resulted in an incorrect
115 body hash calculated value.
117 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
118 DKIM signature block, for verification. Although advised against by
119 standards it is specifically not ruled illegal.
121 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
123 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
124 missing a body hash (the bh= tag).
126 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
127 It seems that HAProxy sends the Proxy Protocol information in clear and
128 only then does a TLS startup, so do the same.
130 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
131 TCP connections (such as for Spamd) unless the daemon successfully set
132 Fast Open mode on its listening sockets. This fixes breakage seen on
133 too-old kernels or those not configured for Fast Open, at the cost of
134 requiring both directions being enabled for TFO, and TFO never being used
135 by non-daemon-related Exim processes.
137 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
138 endings, at least on the first header line. Try to canonify any that get
139 past that check, despite the cost.
141 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
142 now limited to an arbitrary five deep, while parsing addresses with the
143 strip_excess_angle_brackets option enabled.
145 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
146 instead leave the unprompted TLS handshake in socket buffer for the
147 TLS library to consume.
149 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
151 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
153 JH/16 Drop variables when they go out of scope. Memory management drops a whole
154 region in one operation, for speed, and this leaves assigned pointers
155 dangling. Add checks run only under the testsuite which checks all
156 variables at a store-reset and panics on a dangling pointer; add code
157 explicitly nulling out all the variables discovered. Fixes one known
158 bug: a transport crash, where a dangling pointer for $sending_ip_address
159 originally assigned in a verify callout, is re-used.
161 PP/06 Drop '.' from @INC in various Perl scripts.
163 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
165 PP/08 Reduce a number of compilation warnings under clang; building with
166 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
167 should be warning-free.
169 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
171 HS/01 Fix portability problems introduced by PP/08 for platforms where
172 realloc(NULL) is not equivalent to malloc() [SunOS et al].
174 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
175 chunk. This allows us to accept broken chunked messages. We need a more
176 general solution here.
178 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
179 already-broken messages in the queue.
181 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
183 JH/19 Fix reference counting bug in routing-generated-address tracking.
189 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
190 supports it and a size is available (ie. the sending peer gave us one).
192 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
193 years of being deprecated). The replacements are the ACLs
194 acl_smtp_mime and acl_not_smtp_mime.
196 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
197 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
198 or even an in-clear connection were permitted. Now, if the host lookup
199 was dnssec and dane was requested then the host is only used if the
200 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
201 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
202 if one fails this test.
203 This means that a poorly-configured remote DNS will make it incommunicado;
204 but it protects against a DNS-interception attack on it.
206 JH/04 Bug 1810: make continued-use of an open smtp transport connection
207 non-noisy when a race steals the message being considered.
209 JH/05 If main configuration option tls_certificate is unset, generate a
210 self-signed certificate for inbound TLS connections.
212 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
213 in rewrites and routers.
215 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
216 and logged a warning sing 4.83; now they are a configuration file error.
218 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
219 (lacking @domain). Apply the same qualification processing as RCPT.
221 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
223 JH/10 Support ${sha256:} applied to a string (as well as the previous
226 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
227 a cutthrough deliver is pending, as we always want to make a connection.
228 This also avoids re-routing the message when later placing the cutthrough
229 connection after a verify cache hit.
230 Do not update it with the verify result either.
232 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
233 when routing results in more than one destination address.
235 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
236 signing (which inhibits the cutthrough capability). Previously only
237 the presence of an option was tested; now an expansion evaluating as
238 empty is permissible (obviously it should depend only on data available
239 when the cutthrough connection is made).
241 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
242 the relevant preceding SMTP command did not note the pipelining mode.
244 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
245 Previously they were not counted.
247 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
248 as one having no matching records. Previously we deferred the message
249 that needed the lookup.
251 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
252 distinguished as "(=".
254 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
255 for missing MX records. Previously it only worked for missing A records.
257 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
259 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
260 after the data-go-ahead and data-ack. Patch from Jason Betts.
262 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
263 even for a "none" policy. Patch from Tony Meyer.
265 JH/22 Fix continued use of a connection for further deliveries. If a port was
266 specified by a router, it must also match for the delivery to be
269 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
270 When one of the recipients of a message was unsuitable for the connection
271 (has no matching addresses), we lost track of needing to mark it
272 deferred. As a result mail would be lost.
274 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
276 JH/25 Decoding ACL controls is now done using a binary search; the source code
277 takes up less space and should be simpler to maintain. Merge the ACL
278 condition decode tables also, with similar effect.
280 JH/26 Fix problem with one_time used on a redirect router which returned the
281 parent address unchanged. A retry would see the parent address marked as
282 delivered, so not attempt the (identical) child. As a result mail would
285 JH/27 Fix a possible security hole, wherein a process operating with the Exim
286 UID can gain a root shell. Credit to http://www.halfdog.net/ for
287 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
290 JH/28 Enable {spool,log} filesystem space and inode checks as default.
291 Main config options check_{log,spool}_{inodes,space} are now
292 100 inodes, 10MB unless set otherwise in the configuration.
294 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
295 Previously it only applied to the main-section connection policy
298 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
300 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
301 by me. Added RFC7919 DH primes as an alternative.
303 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
304 are not in the system include path.
306 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
307 GnuTLS, when a session startup failed (eg because the client disconnected)
308 Exim did stdio operations after fclose. This was exposed by a recent
309 change which nulled out the file handle after the fclose.
311 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
312 signed directly by the cert-signing cert, rather than an intermediate
313 OCSP-signing cert. This is the model used by LetsEncrypt.
315 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
317 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
318 an incoming connection.
320 HS/02 Bug 1802: Do not half-close the connection after sending a request
323 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
324 fallback to "prime256v1".
326 JH/34 SECURITY: Use proper copy of DATA command in error message.
327 Could leak key material. Remotely exploitable. CVE-2016-9963.
333 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
334 and 3.4.4 - once the server is enabled to respond to an OCSP request
335 it does even when not requested, resulting in a stapling non-aware
336 client dropping the TLS connection.
338 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
339 support variable-length bit vectors. No functional change.
341 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
342 The I= interface field on outgoing lines is now after the H= remote
343 host field, same as incoming lines. There is a separate
344 outgoing_interface log selector which allows you to disable the
347 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
348 If not running log_selector +smtp_connection the mainlog would be held
349 open indefinitely after a "too many connections" event, including to a
350 deleted file after a log rotate. Leave the per net connection logging
351 leaving it open for efficiency as that will be quickly detected by the
352 check on the next write.
354 HS/01 Bug 1671: Fix post transport crash.
355 Processing the wait-<transport> messages could crash the delivery
356 process if the message IDs didn't exist for some reason. When
357 using 'split_spool_directory=yes' the construction of the spool
358 file name failed already, exposing the same netto behaviour.
360 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
361 mime_regex ACL conditions.
363 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
364 to DSN fail messages (bounces): remote IP, remote greeting, remote response
365 to HELO, local diagnostic string.
367 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
368 log line to debug. Even when configured with a tls authenticator many
369 client connections are expected to not authenticate in this way, so
370 an authenticate fail is not an error.
372 HS/02 Add the Exim version string to the process info. This way exiwhat
373 gives some more detail about the running daemon.
375 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
376 matter for fast-change records such as DNSBLs.
378 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
379 retry record, even if constant. There may be multiple transports with
380 different interface settings and the retry behaviour needs to be kept
383 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
385 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
387 JH/10 Bug 840: fix log_defer_output option of pipe transport
389 JH/11 Bug 830: use same host for all RCPTS of a message, even under
390 hosts_randomize. This matters a lot when combined with mua_wrapper.
392 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
393 ${quote_pgsql:<string>} operator.
395 JH/13 Bug 1708: avoid misaligned access in cached lookup.
397 JH/14 Change header file name for freeradius-client. Relevant if compiling
398 with Radius support; from the Gentoo tree and checked under Fedora.
400 JH/15 Bug 1712: Introduce $prdr_requested flag variable
402 JH/16 Bug 1714: Permit an empty string as expansion result for transport
403 option transport_filter, meaning no filtering.
405 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
407 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
408 defaults to "*" (all hosts). The variable is now available when not built
409 with TLS, default unset, mainly to enable keeping the testsuite sane.
410 If a server certificate is not supplied (via tls_certificate) an error is
411 logged, and clients will find TLS connections fail on startup. Presumably
412 they will retry in-clear.
413 Packagers of Exim are strongly encouraged to create a server certificate
414 at installation time.
416 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
417 with the $config_file variable.
419 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
420 in transport context, after the attempt, and per-recipient. The latter type
421 is per host attempted. The event data is the error message, and the errno
422 information encodes the lookup type (A vs. MX) used for the (first) host,
423 and the trailing two digits of the smtp 4xx response.
425 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
426 to write to mainlog (or rejectlog, paniclog) in the window between file
427 creation and permissions/ownership being changed. Particularly affects
428 installations where exicyclog is run as root, rather than exim user;
429 result is that the running daemon panics and dies.
431 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
433 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
434 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
435 "pri" and "weight". Note that the previous implicit priority given by the
436 list order is no longer honoured.
438 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
441 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
442 by defining SUPPORT_SOCKS.
444 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
445 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
446 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
447 variables are renamed to proxy_{local,external}_{address,port}.
449 JH/27 Move Internationalisation support from Experimental to mainline, enabled
450 for a build by defining SUPPORT_I18N
452 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
453 of the query string, and make ${quote_redis:} do that quoting.
455 JH/29 Move Events support from Experimental to mainline, enabled by default
456 and removable for a build by defining DISABLE_EVENT.
458 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
460 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
461 cached by the daemon.
463 JH/32 Move Redis support from Experimental to mainline, enabled for a build
464 by defining LOOKUP_REDIS. The libhiredis library is required.
466 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
467 keys are given for lookup.
469 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
470 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
471 only supported when built with TLS support. The PolarSSL SHA routines
472 are still used when the TLS library is too old for convenient support.
474 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
475 openssl_options), for security. OpenSSL forces this from version 1.1.0
476 server-side so match that on older versions.
478 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
479 allocation for $value could be released as the expansion processing
480 concluded, but leaving the global pointer active for it.
482 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
483 and to use the domains and local_parts ACL conditions.
485 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
486 incorrectly not doubled on cutthrough transmission, hence seen as a
487 body-termination at the receiving system - resulting in truncated mails.
488 Commonly the sender saw a TCP-level error, and retransmitted the message
489 via the normal store-and-forward channel. This could result in duplicates
490 received - but deduplicating mailstores were liable to retain only the
491 initial truncated version.
493 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
495 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
497 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
498 we're in there, support oversigning also; bug 1309.
500 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
502 HS/04 Add support for keep_environment and add_environment options.
504 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
505 either intentional arithmetic overflow during PRNG, or testing config-
508 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
509 delivery resulted in actual delivery. Cancel cutthrough before DATA
512 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
513 rejects a recipient: pass the reject to the originator.
515 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
516 Many were false-positives and ignorable, but it's worth fixing the
519 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
520 for the new environment-manipulation done at startup. Move the routines
521 from being local to tls.c to being global via the os.c file.
523 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
524 an extract embedded as result-arg for a map, the first arg for extract
525 is unavailable so we cannot tell if this is a numbered or keyed
526 extraction. Accept either.
532 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
535 JH/02 The smtp transport option "multi_domain" is now expanded.
537 JH/03 The smtp transport now requests PRDR by default, if the server offers
540 JH/04 Certificate name checking on server certificates, when exim is a client,
541 is now done by default. The transport option tls_verify_cert_hostnames
542 can be used to disable this per-host. The build option
543 EXPERIMENTAL_CERTNAMES is withdrawn.
545 JH/05 The value of the tls_verify_certificates smtp transport and main options
546 default to the word "system" to access the system default CA bundle.
547 For GnuTLS, only version 3.0.20 or later.
549 JH/06 Verification of the server certificate for a TLS connection is now tried
550 (but not required) by default. The verification status is now logged by
551 default, for both outbound TLS and client-certificate supplying inbound
554 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
557 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
558 Status Notification (bounce) messages are now MIME format per RFC 3464.
559 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
560 under the control of the dsn_advertise_hosts option, and routers may
561 have a dsn_lasthop option.
563 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
564 default, modifiable by a malware= option. The list separator for
565 the options can now be changed in the usual way. Bug 68.
567 JH/10 The smtp_receive_timeout main option is now expanded before use.
569 JH/11 The incoming_interface log option now also enables logging of the
570 local interface on delivery outgoing connections.
572 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
573 if the interface and destination host and port all match.
575 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
578 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
579 Patch from Andrew Lewis.
581 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
582 now supports optional time-restrictions, weighting, and priority
583 modifiers per server. Patch originally by <rommer@active.by>.
585 JH/16 The spamd_address main option now supports a mixed list of local
586 and remote servers. Remote servers can be IPv6 addresses, and
587 specify a port-range.
589 JH/17 Bug 68: The spamd_address main option now supports an optional
590 timeout value per server.
592 JH/18 Bug 1581: Router and transport options headers_add/remove can
593 now have the list separator specified.
595 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
598 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
601 JH/21 Support for the A6 type of dns record is withdrawn.
603 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
604 rather than the verbs used.
606 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
607 from 255 to 1024 chars.
609 JH/24 Verification callouts now attempt to use TLS by default.
611 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
612 are generic router options now. The defaults didn't change.
614 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
615 Original patch from Alexander Shikoff, worked over by JH.
617 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
618 files if ZCAT_COMMAND is not executable.
620 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
622 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
624 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
625 Normally benign, it bites when the pair was led to by a CNAME;
626 modern usage is to not canonicalize the domain to a CNAME target
627 (and we were inconsistent anyway for A-only vs AAAA+A).
629 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
631 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
632 when evaluating $sender_host_dnssec.
634 JH/31 Check the HELO verification lookup for DNSSEC, adding new
635 $sender_helo_dnssec variable.
637 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
639 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
641 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
643 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
644 documented as working, but never had. Support all but $spam_report.
646 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
647 added for tls authenticator.
649 HS/03 Add perl_taintmode main config option
655 TL/01 When running the test suite, the README says that variables such as
656 no_msglog_check are global and can be placed anywhere in a specific
657 test's script, however it was observed that placement needed to be near
658 the beginning for it to behave that way. Changed the runtest perl
659 script to read through the entire script once to detect and set these
660 variables, reset to the beginning of the script, and then run through
661 the script parsing/test process like normal.
663 TL/02 The BSD's have an arc4random API. One of the functions to induce
664 adding randomness was arc4random_stir(), but it has been removed in
665 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
666 function when detected.
668 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
669 cause callback expansion.
671 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
672 syntax errors in an expansion can be treated as a string instead of
673 logging or causing an error, due to the internal use of bool_lax
674 instead of bool when processing it.
676 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
677 server certificates when making smtp deliveries.
679 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
681 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
683 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
685 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
686 Merged patch from Sebastian Wiedenroth.
688 JH/05 Fix results-pipe from transport process. Several recipients, combined
689 with certificate use, exposed issues where response data items split
690 over buffer boundaries were not parsed properly. This eventually
691 resulted in duplicates being sent. This issue only became common enough
692 to notice due to the introduction of connection certificate information,
693 the item size being so much larger. Found and fixed by Wolfgang Breyha.
695 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
696 size buffer was used, resulting in syntax errors when an expansion
699 JH/07 Add support for directories of certificates when compiled with a GnuTLS
700 version 3.3.6 or later.
702 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
703 is EXPERIMENTAL_EVENT, the main-configuration and transport options
704 both become "event_action", the variables become $event_name, $event_data
705 and $event_defer_errno. There is a new variable $verify_mode, usable in
706 routers, transports and related events. The tls:cert event is now also
707 raised for inbound connections, if the main configuration event_action
710 TL/06 In test suite, disable OCSP for old versions of openssl which contained
711 early OCSP support, but no stapling (appears to be less than 1.0.0).
713 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
714 server certificate names available under the smtp transport option
715 "tls_verify_cert_hostname" now do not permit multi-component wildcard
718 JH/10 Time-related extraction expansions from certificates now use the main
719 option "timezone" setting for output formatting, and are consistent
720 between OpenSSL and GnuTLS compilations. Bug 1541.
722 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
723 encoded parameter in the incoming message. Bug 1558.
725 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
726 include certificate info, eximon was claiming there were spoolfile
729 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
731 JH/14 Log delivery-related information more consistently, using the sequence
732 "H=<name> [<ip>]" wherever possible.
734 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
735 are problematic for Debian distribution, omit them from the release
738 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
740 JH/16 Fix string representation of time values on 64bit time_t architectures.
743 JH/17 Fix a null-indirection in certextract expansions when a nondefault
744 output list separator was used.
749 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
750 checkers that were complaining about end of non-void function with no
753 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
754 This was a regression introduced in 4.83 by another bugfix.
756 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
758 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
759 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
765 TF/01 Correctly close the server side of TLS when forking for delivery.
767 When a message was received over SMTP with TLS, Exim failed to clear up
768 the incoming connection properly after forking off the child process to
769 deliver the message. In some situations the subsequent outgoing
770 delivery connection happened to have the same fd number as the incoming
771 connection previously had. Exim would try to use TLS and fail, logging
772 a "Bad file descriptor" error.
774 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
775 utilities have not been installed.
777 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
778 temporary space as the ACL may create new global variables.
780 TL/01 LDAP support uses per connection or global context settings, depending
781 upon the detected version of the libraries at build time.
783 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
784 to extract and use the src ip:port in logging and expansions as if it
785 were a direct connection from the outside internet. PPv2 support was
786 updated based on HAProxy spec change in May 2014.
788 JH/02 Add ${listextract {number}{list}{success}{fail}}.
790 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
791 Properly escape header and check for NULL return.
793 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
796 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
798 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
799 characters in header names, implemented as a verify condition.
800 Contributed by Michael Fischer v. Mollard.
802 TL/05 Rename SPF condition results err_perm and err_temp to standardized
803 results permerror and temperror. Previous values are deprecated but
804 still accepted. In a future release, err_perm and err_temp will be
805 completely removed, which will be a backward incompatibility if the
806 ACL tests for either of these two old results. Patch contributed by
807 user bes-internal on the mailing list.
809 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
811 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
812 selectors, in both main and reject logs.
814 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
817 JH/07 Add malware type "sock" for talking to simple daemon.
819 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
821 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
822 routers/transports under cutthrough routing.
824 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
825 numbers. Touch up "bool" conditional to keep the same definition.
827 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
829 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
831 JH/12 Expand items in router/transport headers_add or headers_remove lists
832 individually rather than the list as a whole. Bug 1452.
834 Required for reasonable handling of multiple headers_ options when
835 they may be empty; requires that headers_remove items with embedded
836 colons must have them doubled (or the list-separator changed).
838 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
839 view the policy declared in the DMARC record. Currently, $dmarc_status
840 is a combined value of both the record presence and the result of the
843 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
845 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
846 dnslookup router and the smtp transport (applying to the forward
849 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
850 of ldap servers used for a specific lookup. Patch provided by Heiko
853 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
854 New variable $lookup_dnssec_authenticated for observability.
856 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
857 Patch submitted by Lars Timman.
859 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
861 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
862 Requires trusted mode and valid format message id, aborts otherwise.
863 Patch contributed by Heiko Schlichting.
865 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
866 certextract with support for various fields. Bug 1358.
868 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
869 is requested by default, modifiable by smtp transport option
872 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
873 operate on certificate variables to give certificate fingerprints
874 Also new ${sha256:cert_variable}.
876 JH/23 The PRDR feature is moved from being Experimental into the mainline.
878 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
879 Christian Aistleitner.
881 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
883 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
884 file. Patch from Wolfgang Breyha.
886 JH/25 Expand the coverage of the delivery $host and $host_address to
887 client authenticators run in verify callout. Bug 1476.
889 JH/26 Port service names are now accepted for tls_on_connect_ports, to
890 align with daemon_smtp_ports. Bug 72.
892 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
893 support and error reporting did not work properly.
895 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
896 and is readable. Patch from Andrew Colin Kissa.
898 TL/14 Enhance documentation of ${run expansion and how it parses the
899 commandline after expansion, particularly in the case when an
900 unquoted variable expansion results in an empty value.
902 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
904 PP/02 Fix internal collision of T_APL on systems which support RFC3123
905 by renaming away from it. Addresses GH issue 15, reported by
908 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
910 TL/15 SECURITY: prevent double expansion in math comparison functions
911 (can expand unsanitized data). Not remotely exploitable.
918 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
920 PP/02 Make -n do something, by making it not do something.
921 When combined with -bP, the name of an option is not output.
923 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
926 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
927 $sender_host_name and config options to manage this, and basic check
930 PP/05 DSCP support for outbound connections and control modifier for inbound.
932 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
933 (Only plugin which currently uses this is kerberos4, which nobody should
934 be using, but we should make it available and other future plugins might
935 conceivably use it, even though it would break NAT; stuff *should* be
936 using channel bindings instead).
938 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
939 name; added for Sendmail compatibility; requires admin caller.
940 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
941 just ignore it); requires trusted caller.
942 Also parse but ignore: -Ac -Am -X<logfile>
945 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
947 TL/02 Add +smtp_confirmation as a default logging option.
949 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
950 Patch by Magnus Holmgren from 2007-02-20.
952 TL/04 Bugzilla 1281 - Spec typo.
953 Bugzilla 1283 - Spec typo.
954 Bugzilla 1290 - Spec grammar fixes.
956 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
958 TL/06 Add Experimental DMARC support using libopendmarc libraries.
960 TL/07 Fix an out of order global option causing a segfault. Reported to dev
961 mailing list by by Dmitry Isaikin.
963 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
965 JH/02 Support "G" suffix to numbers in ${if comparisons.
967 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
969 NM/01 Bugzilla 1197 - Spec typo
970 Bugzilla 1196 - Spec examples corrections
972 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
974 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
975 gnutls_enable_pkcs11, but renamed to more accurately indicate its
978 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
979 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
981 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
982 "acl {{name}{arg}...}", and optional args on acl condition
985 JH/05 Permit multiple router/transport headers_add/remove lines.
987 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
989 JH/07 Avoid using a waiting database for a single-message-only transport.
990 Performance patch from Paul Fisher. Bugzilla 1262.
992 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
995 JH/09 Add $headers_added variable, with content from use of ACL modifier
996 add_header (but not yet added to the message). Bugzilla 199.
998 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
999 Pulled from Bugzilla 817 by Wolfgang Breyha.
1001 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1003 (nb: this is the same fix as in Exim 4.80.1)
1005 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1008 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1011 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1013 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1014 advertise SMTP AUTH mechanism to us, instead of a generic
1015 protocol violation error. Also, make Exim more robust to bad
1016 data from the Dovecot auth socket.
1018 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1020 When a queue runner is handling a message, Exim first routes the
1021 recipient addresses, during which it prunes them based on the retry
1022 hints database. After that it attempts to deliver the message to
1023 any remaining recipients. It then updates the hints database using
1026 So if a recipient address works intermittently, it can get repeatedly
1027 deferred at routing time. The retry hints record remains fresh so the
1028 address never reaches the final cutoff time.
1030 This is a fairly common occurrence when a user is bumping up against
1031 their storage quota. Exim had some logic in its local delivery code
1032 to deal with this. However it did not apply to per-recipient defers
1033 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1035 This change adds a proper retry rule check during routing so that the
1036 final cutoff time is checked against the message's age. We only do
1037 this check if there is an address retry record and there is not a
1038 domain retry record; this implies that previous attempts to handle
1039 the address had the retry_use_local_parts option turned on. We use
1040 this as an approximation for the destination being like a local
1041 delivery, as in LMTP.
1043 I suspect this new check makes the old local delivery cutoff check
1044 redundant, but I have not verified this so I left the code in place.
1046 TF/02 Correct gecos expansion when From: is a prefix of the username.
1048 Test 0254 submits a message to Exim with the header
1052 When I ran the test suite under the user fanf2, Exim expanded
1053 the header to contain my full name, whereas it should have added
1054 a Resent-Sender: header. It erroneously treats any prefix of the
1055 username as equal to the username.
1057 This change corrects that bug.
1059 GF/01 DCC debug and logging tidyup
1060 Error conditions log to paniclog rather than rejectlog.
1061 Debug lines prefixed by "DCC: " to remove any ambiguity.
1063 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1065 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1066 Bug spotted by Jeremy Harris; was flawed since initial commit.
1067 Would have resulted in OCSP responses post-SNI triggering an Exim
1068 NULL dereference and crash.
1070 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1072 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1073 Bug detection, analysis and fix by Samuel Thibault.
1074 Bugzilla 1331, Debian bug #698092.
1076 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1078 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1079 Server implementation by Todd Lyons, client by JH.
1080 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1081 config variable "prdr_enable" controls whether the server
1082 advertises the facility. If the client requests PRDR a new
1083 acl_data_smtp_prdr ACL is called once for each recipient, after
1084 the body content is received and before the acl_smtp_data ACL.
1085 The client is controlled by both of: a hosts_try_prdr option
1086 on the smtp transport, and the server advertisement.
1087 Default client logging of deliveries and rejections involving
1088 PRDR are flagged with the string "PRDR".
1090 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1091 fclose(). Diagnosis by Todd Lyons.
1093 PP/17 Update configure.default to handle IPv6 localhost better.
1094 Patch by Alain Williams (plus minor tweaks).
1097 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1098 This is now consistent with GnuTLS, and is now documented: the
1099 previous undocumented portable approach to treating the option as
1100 unset was to force an expansion failure. That still works, and
1101 an empty string is now equivalent.
1103 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1104 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1105 not performing validation itself.
1107 PP/20 Added force_command boolean option to pipe transport.
1108 Patch from Nick Koston, of cPanel Inc.
1110 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1113 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1115 PP/21 Fix eximon continuous updating with timestamped log-files.
1116 Broken in a format-string cleanup in 4.80, missed when I repaired the
1117 other false fix of the same issue.
1118 Report and fix from Heiko Schlichting.
1121 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1122 Report from Prashanth Katuri.
1124 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1125 It's SecureTransport, so affects any MacOS clients which use the
1126 system-integrated TLS libraries, including email clients.
1128 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1129 using a MIME ACL for non-SMTP local injection.
1130 Report and assistance in diagnosis by Warren Baker.
1132 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1134 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1136 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1137 last id that failed so it may be referenced in subsequent ACL's.
1139 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1142 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1143 ldap library initialization, allowing self-signed CA's to be
1144 used. Also properly sets require_cert option later in code by
1145 using NULL (global ldap config) instead of ldap handle (per
1146 session). Bug diagnosis and testing by alxgomz.
1148 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1149 the src/util/ subdirectory.
1151 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1152 renamed to Transport Post Delivery Action by Jeremy Harris, as
1155 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1156 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1157 redis_servers = needs to be configured which will be used by the redis
1158 lookup. Patch from Warren Baker, of The Packet Hub.
1160 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1162 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1163 hostname or reverse DNS when processing a host list. Used suggestions
1164 from multiple comments on this bug.
1166 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1168 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1169 Missed a few lines, added it to make the runtest require no keyboard
1172 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1173 contains upper case chars. Make router use caseful_local_part.
1175 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1176 support when GnuTLS has been built with p11-kit.
1182 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1184 This, or similar/improved, will also be change PP/11 of 4.82.
1190 PP/01 Handle short writes when writing local log-files.
1191 In practice, only affects FreeBSD (8 onwards).
1192 Bugzilla 1053, with thanks to Dmitry Isaikin.
1194 NM/01 Bugzilla 949 - Documentation tweak
1196 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1199 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1201 PP/02 Implemented gsasl authenticator.
1203 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1205 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1206 `pkg-config foo` for cflags/libs.
1208 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1209 with rest of GSASL and with heimdal_gssapi.
1211 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1212 `pkg-config foo` for cflags/libs for the TLS implementation.
1214 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1215 properties get this fed in as external SSF. A number of robustness
1216 and debugging improvements to the cyrus_sasl authenticator.
1218 PP/08 cyrus_sasl server now expands the server_realm option.
1220 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1221 Patch by Jeremy Harris.
1223 PP/10 Added dbmjz lookup type.
1225 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1227 PP/12 MAIL args handles TAB as well as SP, for better interop with
1228 non-compliant senders.
1229 Analysis and variant patch by Todd Lyons.
1231 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1232 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1233 Patch from Dirk Mueller <dmueller@suse.com>
1235 PP/13 tls_peerdn now print-escaped for spool files.
1236 Observed some $tls_peerdn in wild which contained \n, which resulted
1237 in spool file corruption.
1239 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1240 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1241 or write after TLS renegotiation, which otherwise led to messages
1244 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1245 as a tracking header (ie: a signed header comes before the signature).
1246 Patch from Wolfgang Breyha.
1248 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1249 comma-sep list; embedded commas doubled.
1251 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1253 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1255 Report and patch from Dmitry Banschikov.
1257 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1258 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1259 failures. We appear to now support TLS1.1+ with Exim.
1261 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1262 lets Exim select keys and certificates based upon TLS SNI from client.
1263 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1264 before an outbound SMTP session. New log_selector, +tls_sni.
1266 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1267 NULL dereference. Report and patch from Alun Jones.
1269 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1270 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1271 Not seeing resolver debug output on NetBSD, but suspect this is a
1272 resolver implementation change.
1274 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1275 Left warnings. Added "eximon gdb" invocation mode.
1277 PP/21 Defaulting "accept_8bitmime" to true, not false.
1279 PP/22 Added -bw for inetd wait mode support.
1281 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1282 locate the relevant includes and libraries. Made this the default.
1284 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1285 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1287 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1288 This may cause build issues on older platforms.
1290 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1291 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1292 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1293 Added SNI support via GnuTLS too.
1294 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1296 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1298 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1299 Applied second patch from Janne, implementing suggestion to default
1300 multiple-strings-in-record handling to match SPF spec.
1302 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1304 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1305 read-only, out of scope).
1306 Patch from Wolfgang Breyha, report from Stuart Northfield.
1308 PP/29 Fix three issues highlighted by clang analyser static analysis.
1309 Only crash-plausible issue would require the Cambridge-specific
1310 iplookup router and a misconfiguration.
1311 Report from Marcin Mirosław.
1313 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1315 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1316 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1317 As part of this, removing so much warning spew let me fix some minor
1318 real issues in debug logging.
1320 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1321 assignment on my part. Fixed.
1323 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1324 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1325 Janne Snabb (who went above and beyond: thank you).
1327 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1328 string otherwise requires a connection and a bunch more work and it's
1329 relatively easy to get wrong. Should also expose TLS library linkage
1332 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1333 64-bit ${eval} (JH/03).
1335 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1336 GNU libc to support some of the 64-bit stuff, should not lead to
1337 conflicts. Defined before os.h is pulled in, so if a given platform
1338 needs to override this, it can.
1340 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1341 protection layer was required, which is not implemented.
1342 Bugzilla 1254, patch from Wolfgang Breyha.
1344 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1345 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1346 tls_dhparam take prime identifiers. Also unbreak combination of
1347 OpenSSL+DH_params+TLSSNI.
1349 PP/39 Disable SSLv2 by default in OpenSSL support.
1355 PP/01 Solaris build fix for Oracle's LDAP libraries.
1356 Bugzilla 1109, patch from Stephen Usher.
1358 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1360 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1363 TF/02 Fix a couple more cases where we did not log the error message
1364 when unlink() failed. See also change 4.74-TF/03.
1366 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1367 lock up or crash if it happened to be inside a call to libc when it
1368 got a SIGUSR1 from exiwhat.
1370 The SIGUSR1 handler appends the current process status to the process
1371 log which is later printed by exiwhat. It used to use the general
1372 purpose logging code to do this, but several functions it calls are
1373 not safe for signals.
1375 The new output code in the SIGUSR1 handler is specific to the process
1376 log, and simple enough that it's easy to inspect for signal safety.
1377 Removing some special cases also simplifies the general logging code.
1378 Removing the spurious timestamps from the process log simplifies
1381 TF/04 Improved ratelimit ACL condition.
1383 The /noupdate option has been deprecated in favour of /readonly which
1384 has clearer semantics. The /leaky, /strict, and /readonly update modes
1385 are mutually exclusive. The update mode is no longer included in the
1386 database key; it just determines when the database is updated. (This
1387 means that when you upgrade Exim will forget old rate measurements.)
1389 Exim now checks that the per_* options are used with an update mode that
1390 makes sense for the current ACL. For example, when Exim is processing a
1391 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1392 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1393 must specify per_mail/readonly. If you omit the update mode it defaults to
1394 /leaky where that makes sense (as before) or /readonly where required.
1396 The /noupdate option is now undocumented but still supported for
1397 backwards compatibility. It is equivalent to /readonly except that in
1398 ACLs where /readonly is required you may specify /leaky/noupdate or
1399 /strict/noupdate which are treated the same as /readonly.
1401 A useful new feature is the /count= option. This is a generalization
1402 of the per_byte option, so that you can measure the throughput of other
1403 aggregate values. For example, the per_byte option is now equivalent
1404 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1406 The per_rcpt option has been generalized using the /count= mechanism
1407 (though it's more complicated than the per_byte equivalence). When it is
1408 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1409 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1410 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1411 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1412 Note that using per_rcpt with a non-readonly update mode in more than
1413 one ACL will cause the recipients to be double-counted. (The per_mail
1414 and per_byte options don't have this problem.)
1416 The handling of very low rates has changed slightly. If the computed rate
1417 is less than the event's count (usually one) then this event is the first
1418 after a long gap. In this case the rate is set to the same as this event's
1419 count, so that the first message of a spam run is counted properly.
1421 The major new feature is a mechanism for counting the rate of unique
1422 events. The new per_addr option counts the number of different
1423 recipients that someone has sent messages to in the last time period. It
1424 behaves like per_rcpt if all the recipient addresses are different, but
1425 duplicate recipient addresses do not increase the measured rate. Like
1426 the /count= option this is a general mechanism, so the per_addr option
1427 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1428 example, measure the rate that a client uses different sender addresses
1429 with the options per_mail/unique=$sender_address. There are further
1430 details in the main documentation.
1432 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1434 TF/06 Removed a few PCRE remnants.
1436 TF/07 Automatically extract Exim's version number from tags in the git
1437 repository when doing development or release builds.
1439 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1440 Bugzilla 879. Patch from Paul Fisher.
1442 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1443 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1446 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1448 PP/05 Variable $av_failed, true if the AV scanner deferred.
1449 Bugzilla 1078. Patch from John Horne.
1451 PP/06 Stop make process more reliably on build failure.
1452 Bugzilla 1087. Patch from Heiko Schlittermann.
1454 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1455 Bugzilla 1089. Patch from Heiko Schlittermann.
1457 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1458 Bugzilla 1131. Patch from Holger Weiß.
1460 PP/09 Handle IPv6 addresses with SPF.
1461 Bugzilla 860. Patch from Wolfgang Breyha.
1463 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1465 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1468 PP/11 match_* no longer expand right-hand-side by default.
1469 New compile-time build option, EXPAND_LISTMATCH_RHS.
1470 New expansion conditions, "inlist", "inlisti".
1472 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1474 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1476 PP/14 fix log_write() format string regression from TF/03.
1477 Bugzilla 1152. Patch from Dmitry Isaikin.
1483 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1485 PP/02 Harmonised TLS library version reporting; only show if debugging.
1486 Layout now matches that introduced for other libraries in 4.74 PP/03.
1488 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1490 PP/04 New "dns_use_edns0" global option.
1492 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1495 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1496 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1498 TK/01 Updated PolarSSL code to 0.14.2.
1499 Bugzilla 1097. Patch from Andreas Metzler.
1501 PP/07 Catch divide-by-zero in ${eval:...}.
1502 Fixes bugzilla 1102.
1504 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1507 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1508 format-string attack -- SECURITY: remote arbitrary code execution.
1510 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1511 time unintentionally subject to list matching rules, letting the header
1512 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1513 arbitrary string expansion). This allowed for information disclosure.
1515 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1516 INT_MIN/-1 -- value coerced to INT_MAX.
1522 NM/01 Workaround for PCRE version dependency in version reporting
1525 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1526 This fixes portability to compilers other than gcc, notably
1527 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1529 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1530 makefiles for portability to HP-UX and POSIX correctness.
1532 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1533 Also via indented variable definition in the Makefile.
1534 (Debugging by Oliver Heesakkers).
1536 PP/02 Restore caching of spamd results with expanded spamd_address.
1537 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1539 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1540 Improves build reliability. Fix from: Frank Elsner
1542 NM/02 Fix wide character breakage in the rfc2047 coding
1543 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1545 NM/03 Allow underscore in dnslist lookups
1546 Fixes bug 1026. Patch from Graeme Fowler
1548 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1549 Code patches from Adam Ciarcinski of NetBSD.
1551 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1554 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1555 is logged, to avoid truncation. Patch from John Horne.
1557 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1558 Patch from Jakob Hirsch.
1560 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1561 SQL string expansion failure details.
1562 Patch from Andrey Oktyabrski.
1564 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1565 Patch from Simon Arlott.
1567 PP/09 New lookups functionality failed to compile on old gcc which rejects
1568 extern declarations in function scope.
1569 Patch from Oliver Fleischmann
1571 PP/10 Use sig_atomic_t for flags set from signal handlers.
1572 Check getgroups() return and improve debugging.
1573 Fixed developed for diagnosis in bug 927 (which turned out to be
1576 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1577 Patch from Mark Zealey.
1579 PP/12 Bugzilla 1056: Improved spamd server selection.
1580 Patch from Mark Zealey.
1582 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1583 Based on patch from Heiko Schlittermann.
1585 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1586 Patch from Uwe Doering, sign-off by Michael Haardt.
1588 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1589 variable declaration deep within a block. Bug and patch from
1592 PP/15 lookups-Makefile IRIX compatibility coercion.
1594 PP/16 Make DISABLE_DKIM build knob functional.
1596 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1597 Patch by Simon Arlott
1599 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1600 variable argument macros. Our copy now differs from upstream.
1606 TF/01 Failure to get a lock on a hints database can have serious
1607 consequences so log it to the panic log.
1609 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1610 controlled using the smtp_confirmation log selector.
1612 TF/03 Include the error message when we fail to unlink a spool file.
1614 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1615 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1616 for maintaining out-of-tree patches for some time.
1618 PP/01 Bugzilla 139: Documentation and portability issues.
1619 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1620 Handle per-OS dynamic-module compilation flags.
1622 PP/02 Let /dev/null have normal permissions.
1623 The 4.73 fixes were a little too stringent and complained about the
1624 permissions on /dev/null. Exempt it from some checks.
1625 Reported by Andreas M. Kirchwitz.
1627 PP/03 Report version information for many libraries, including
1628 Exim version information for dynamically loaded libraries. Created
1629 version.h, now support a version extension string for distributors
1630 who patch heavily. Dynamic module ABI change.
1632 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1633 privilege escalation vulnerability whereby the Exim run-time user
1634 can cause root to append content of the attacker's choosing to
1637 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1640 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1641 If dropping privileges for untrusted macros, we disabled normal logging
1642 on the basis that it would fail; for the Exim run-time user, this is not
1643 the case, and it resulted in successful deliveries going unlogged.
1644 Fixed. Reported by Andreas Metzler.
1650 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1651 only prepend for the Resent-* case. Fixes regression introduced in
1652 Exim 4.70 by NM/22 for Bugzilla 607.
1654 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1655 increasing numbers of administrators be bitten by this.
1657 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1659 PP/03 Bugzilla 994: added openssl_options main configuration option.
1661 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1663 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1665 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1667 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1668 without a peer certificate, leading to a segfault because of an
1669 assumption that peers always have certificates. Be a little more
1670 paranoid. Problem reported by Martin Tscholak.
1672 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1673 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1674 NB: ClamAV planning to remove STREAM in "middle of 2010".
1675 CL also introduces -bmalware, various -d+acl logging additions and
1676 more caution in buffer sizes.
1678 PP/09 Implemented reverse_ip expansion operator.
1680 PP/10 Bugzilla 937: provide a "debug" ACL control.
1682 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1684 PP/12 Bugzilla 973: Implement --version.
1686 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1688 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1690 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1692 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1693 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1694 ignore trailing whitespace.
1696 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1698 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1701 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1702 Notification from Dr Andrew Aitchison.
1704 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1705 ExtendedDetectionInfo response format.
1706 Notification from John Horne.
1708 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1711 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1712 XSL and documented dependency on system catalogs, with examples of how
1715 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1718 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1719 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1720 configuration file which is writeable by the Exim user or group.
1722 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1723 of configuration files to cover files specified with the -C option if
1724 they are going to be used with root privileges, not just the default
1727 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1728 option (effectively making it always true).
1730 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1731 files to be used while preserving root privileges.
1733 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1734 that rogue child processes cannot use them.
1736 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1737 run-time user, instead of root.
1739 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1740 Exim run-time user without dropping privileges.
1742 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1743 result string, instead of calling string_vformat() twice with the same
1746 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1747 for other users. Others should always drop root privileges if they use
1748 -C on the command line, even for a whitelisted configure file.
1750 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1752 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1758 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1759 $data_path, and $header_path variables; fixed documentation bugs and
1762 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1763 exipick to access non-standard spools, including the "frozen" queue
1766 NM/01 Bugzilla 965: Support mysql stored procedures.
1767 Patch from Alain Williams
1769 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1771 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1772 Patch from Andreas Metzler
1774 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1775 Patch from Kirill Miazine
1777 NM/05 Bugzilla 671: Added umask to procmail example.
1779 JJ/03 installed exipick 20100323.0, fixing doc bug
1781 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1782 directory. Notification and patch from Dan Rosenberg.
1784 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1786 TK/02 Improve log output when DKIM signing operation fails.
1788 MH/01 Treat the transport option dkim_domain as a colon separated
1789 list, not as a single string, and sign the message with each element,
1790 omitting multiple occurences of the same signer.
1792 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1793 Bugzilla 985, 986. Patch by Simon Arlott
1795 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1796 Patch by Simon Arlott
1798 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1799 MBX locking. Notification from Dan Rosenberg.
1805 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1807 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1809 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1811 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1813 NM/04 Bugzilla 915: Flag broken perl installation during build.
1819 TK/01 Added patch by Johannes Berg that expands the main option
1820 "spamd_address" if it starts with a dollar sign.
1822 TK/02 Write list of recipients to X-Envelope-Sender header when building
1823 the mbox-format spool file for content scanning (suggested by Jakob
1826 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1827 (http://www.dcc-servers.net/) support via dccifd. Activated by
1828 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1830 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1831 by Mark Daniel Reidel <mr@df.eu>.
1833 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1834 When building exim an external PCRE library is now needed -
1835 PCRE is a system library on the majority of modern systems.
1836 See entry on PCRE_LIBS in EDITME file.
1838 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1839 conversation. Added nologin parameter to request.
1840 Patch contributed by Kirill Miazine.
1842 TF/01 Do not log submission mode rewrites if they do not change the address.
1844 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1846 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1847 log files in place. Contributed by Roberto Lima.
1849 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1851 TF/03 Bugzilla 615: When checking the local_parts router precondition
1852 after a local_part_suffix or local_part_prefix option, Exim now
1853 does not use the address's named list lookup cache, since this
1854 contains cached lookups for the whole local part.
1856 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1857 Robert Millan. Documentation is in experimental-spec.txt.
1859 TF/04 Bugzilla 668: Fix parallel build (make -j).
1861 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1863 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1864 Patch provided by Jan Srzednicki.
1866 TF/05 Leading white space used to be stripped from $spam_report which
1867 wrecked the formatting. Now it is preserved.
1869 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1870 that they are available at delivery time.
1872 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1874 TF/08 TLS error reporting now respects the incoming_interface and
1875 incoming_port log selectors.
1877 TF/09 Produce a more useful error message if an SMTP transport's hosts
1878 setting expands to an empty string.
1880 NM/06 Bugzilla 744: EXPN did not work under TLS.
1881 Patch provided by Phil Pennock.
1883 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1884 Patch provided by Richard Godbee.
1886 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1887 acl_smtp_notquit, added index entry.
1889 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1890 Patch provided by Eugene Bujak.
1892 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1893 accept(). Patch provided by Maxim Dounin.
1895 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1896 Patch provided by Phil Pennock.
1898 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1900 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1901 Patch provided by Brad "anomie" Jorsch.
1903 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1904 Patch provided by Dean Brooks.
1906 TK/05 Add native DKIM support (does not depend on external libraries).
1908 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1909 Patch provided by Graeme Fowler.
1911 NM/16 Bugzilla 851: Documentation example syntax fix.
1913 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1915 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1918 NM/19 Bugzilla 745: TLS version reporting.
1919 Patch provided by Phil Pennock.
1921 NM/20 Bugzilla 167: bool: condition support.
1922 Patch provided by Phil Pennock.
1924 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1925 clients. Patch provided by Phil Pennock.
1927 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1928 Patch provided by Brad "anomie" Jorsch.
1930 NM/23 Bugzilla 687: Fix misparses in eximstats.
1931 Patch provided by Heiko Schlittermann.
1933 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1934 Patch provided by Heiko Schlittermann.
1936 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1937 plus update to original patch.
1939 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1941 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1942 Patch provided by David Brownlee.
1944 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1946 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1948 NM/30 Bugzilla 888: TLS documentation bugfixes.
1950 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1952 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1953 Unlike the original bugzilla I have changed all shell scripts in src tree.
1955 NM/33 Bugzilla 898: Transport filter timeout fix.
1956 Patch by Todd Rinaldo.
1958 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1959 Patch by Serge Demonchaux.
1961 NM/35 Bugzilla 39: Base64 decode bug fixes.
1962 Patch by Jakob Hirsch.
1964 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1966 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1968 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1970 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1976 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1977 ALT-N's libdkim that I have put here:
1978 http://duncanthrax.net/exim-experimental/
1980 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1981 were called 'true' and it seems that C99 defines that as a reserved
1982 keyword to be used with 'bool' variable types. That means you could
1983 not include C99-style headers which use bools without triggering
1984 build errors in sieve.c.
1986 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1987 as mailq or other aliases. Changed the --help handling significantly
1988 to do whats expected. exim_usage() emits usage/help information.
1990 SC/01 Added the -bylocaldomain option to eximstats.
1992 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1994 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1996 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2002 PH/01 Another patch from the Sieve maintainer.
2004 PH/02 When an IPv6 address is converted to a string for single-key lookup
2005 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2006 dots are used instead of colons so that keys in lsearch files need not
2007 contain colons. This was done some time before quoting was made available
2008 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2009 (notated using the quote facility) so as to distinguish them from IPv4
2010 keys. This meant that lookups for IP addresses in host lists did not work
2011 for iplsearch lookups.
2013 This has been fixed by arranging for IPv6 addresses to be expressed with
2014 colons if the lookup type is iplsearch. This is not incompatible, because
2015 previously such lookups could never work.
2017 The situation is now rather anomalous, since one *can* have colons in
2018 ordinary lsearch keys. However, making the change in all cases is
2019 incompatible and would probably break a number of configurations.
2021 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2024 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2025 conversion specification without a maximum field width, thereby enabling
2026 a rogue spamd server to cause a buffer overflow. While nobody in their
2027 right mind would setup Exim to query an untrusted spamd server, an
2028 attacker that gains access to a server running spamd could potentially
2029 exploit this vulnerability to run arbitrary code as the Exim user.
2031 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2032 $primary_hostname instead of what libspf2 thinks the hosts name is.
2034 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2035 a directory entry by the name of the lookup key. Previously, if a
2036 symlink pointed to a non-existing file or a file in a directory that
2037 Exim lacked permissions to read, a lookup for a key matching that
2038 symlink would fail. Now it is enough that a matching directory entry
2039 exists, symlink or not. (Bugzilla 503.)
2041 PH/03 The body_linecount and body_zerocount variables are now exported in the
2044 PH/04 Added the $dnslist_matched variable.
2046 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2047 This means they are set thereafter only if the connection becomes
2050 PH/06 Added the client_condition to authenticators so that some can be skipped
2051 by clients under certain conditions.
2053 PH/07 The error message for a badly-placed control=no_multiline_responses left
2054 "_responses" off the end of the name.
2056 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2058 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2059 (without spaces) instead of just copying the configuration text.
2061 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2063 PH/11 Added $max_received_linelength.
2065 PH/12 Added +ignore_defer and +include_defer to host lists.
2067 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2068 way in which PCRE > 7.0 is built.
2070 PH/14 Implemented queue_only_load_latch.
2072 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2073 MAIL command. The effect was to mangle the value on 64-bit systems.
2075 PH/16 Another patch from the Sieve maintainer.
2077 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2079 PH/18 If a system quota error occurred while trying to create the file for
2080 a maildir delivery, the message "Mailbox is full" was not appended to the
2081 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2082 only to a quota excession during the actual writing of the file.
2084 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2085 characters?) which causes problems in log lines. The DN values are now
2086 passed through string_printing() before being added to log lines.
2088 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2089 and InterBase are left for another time.)
2091 PH/21 Added message_body_newlines option.
2093 PH/22 Guard against possible overflow in moan_check_errorcopy().
2095 PH/23 POSIX allows open() to be a macro; guard against that.
2097 PH/24 If the recipient of an error message contained an @ in the local part
2098 (suitably quoted, of course), incorrect values were put in $domain and
2099 $local_part during the evaluation of errors_copy.
2105 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2106 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2109 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2110 issue a MAIL command.
2112 PH/02 In an ACL statement such as
2114 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2116 if a client was not listed at all, or was listed with a value other than
2117 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2118 the condition was not true (as it should be), so access was not denied.
2119 The bug was that the ! inversion was incorrectly passed on to the second
2120 item. This has been fixed.
2122 PH/03 Added additional dnslists conditions == and =& which are different from
2123 = and & when the dns lookup returns more than one IP address.
2125 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2126 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2128 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2129 FSYNC, which compiles an option called disable_fsync that allows for
2130 bypassing fsync(). The documentation is heavily laced with warnings.
2132 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2134 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2135 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2136 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2137 including adding "make clean"; (3) Added -fPIC when compiling the test
2138 dynamically loaded module, to get rid of a warning.
2140 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2141 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2142 = 0s. The bug is otherwise harmless.
2144 PH/07 There was a bug in the dovecot authenticator such that the value of
2145 $auth1 could be overwritten, and so not correctly preserved, after a
2146 successful authentication. This usually meant that the value preserved by
2147 the server_setid option was incorrect.
2149 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2151 PH/09 Installed PCRE release 7.0.
2153 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2154 run for batched SMTP input. It is now run at the start of every message
2155 in the batch. While fixing this I discovered that the process information
2156 (output by running exiwhat) was not always getting set for -bs and -bS
2157 input. This is fixed, and it now also says "batched" for BSMTP.
2159 PH/11 Added control=no_pipelining.
2161 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2162 patch, slightly modified), and move the expansion of helo_data till after
2163 the connection is made in the smtp transport (so it can use these
2166 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2168 PH/14 Added log_selector = +pid.
2170 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2172 PH/16 Add ${if forany and ${if forall.
2174 PH/17 Added dsn_from option to vary the From: line in DSNs.
2176 PH/18 Flush SMTP output before performing a callout, unless control =
2177 no_callout_flush is set.
2179 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2180 was true (the default) a successful delivery failed to delete the retry
2181 item, thus causing premature timeout of the address. The bug is now
2184 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2186 PH/21 Long custom messages for fakedefer and fakereject are now split up
2187 into multiline responses in the same way that messages for "deny" and
2188 other ACL rejections are.
2190 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2191 with slight modification.
2193 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2194 draft, changing the syntax and factoring some duplicate code".
2196 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2197 for deliveries of the second and subsequent messages over the same SMTP
2200 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2201 ${reduce, with only minor "tidies".
2203 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2205 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2206 expansion side effects.
2208 PH/27 When a message times out after an over-quota error from an Exim-imposed
2209 quota, the bounce message says "mailbox is full". This message was not
2210 being given when it was a system quota that was exceeded. It now should
2213 MH/03 Made $recipients available in local_scan(). local_scan() already has
2214 better access to the recipient list through recipients_list[], but
2215 $recipients can be useful in postmaster-provided expansion strings.
2217 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2218 in the case of a MAIL command with additional options following the
2219 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2220 were accidentally chopped off.
2222 PH/29 SMTP synchronization checks are implemented when a command is read -
2223 there is a check that no more input is waiting when there shouldn't be
2224 any. However, for some commands, a delay in an ACL can mean that it is
2225 some time before the response is written. In this time, more input might
2226 arrive, invalidly. So now there are extra checks after an ACL has run for
2227 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2228 pipelining has not been advertised.
2230 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2232 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2233 returned in the SMTP response when smtp_return_error_details was false.
2234 This has been fixed.
2236 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2237 instead of the C I/O that was originally supplied, because problems were
2238 reported on Solaris.
2240 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2241 Exim which did not show up earlier: it was assuming that a call to
2242 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2243 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2244 was a macro that became an assignment, so it seemed to work. This has
2245 changed to a proper function call with a void return, hence the compile
2246 error. Exim's code has been fixed.
2248 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2251 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2253 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2255 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2256 because it turns out to be too big or there is a timeout) there is no
2257 "Completed" line in the log. When some messages of this type were
2258 selected by exigrep, they were listed as "not completed". Others were
2259 picked up by some special patterns. I have improved the selection
2260 criteria to be more general.
2262 PH/38 The host_find_failed option in the manualroute router can now be set
2263 to "ignore", to completely ignore a host whose IP address cannot be
2264 found. If all hosts are ignored, the behaviour is controlled by the new
2265 host_all_ignored option.
2267 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2268 homing or because of multiple MX records with /mx) generated more than
2269 one IP address, and the following item turned out to be the local host,
2270 all the secondary addresses of the first item were incorrectly removed
2271 from the list, along with the local host and any following hosts (which
2272 is what is supposed to happen).
2274 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2275 whoever called Exim into the -H file. In the case of the daemon it was
2276 behaving confusingly. When first started, it used values for whoever
2277 started the daemon, but after a SIGHUP it used the Exim user (because it
2278 calls itself on a restart). I have changed the code so that it now always
2281 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2282 message are rejected with the same error (e.g. no authentication or bad
2283 sender address), and a DATA command is nevertheless sent (as can happen
2284 with PIPELINING or a stupid MUA), the error message that was given to the
2285 RCPT commands is included in the rejection of the DATA command. This is
2286 intended to be helpful for MUAs that show only the final error to their
2289 PH/42 Another patch from the Sieve maintainer.
2291 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2292 Eximstats - Fixed some broken HTML links and added missing column headers
2294 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2295 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2297 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2299 PH/43 Yet another patch from the Sieve maintainer.
2301 PH/44 I found a way to check for a TCP/IP connection going away before sending
2302 the response to the final '.' that terminates a message, but only in the
2303 case where the client has not sent further data following the '.'
2304 (unfortunately, this is allowed). However, in many cases there won't be
2305 any further data because there won't be any more messages to send. A call
2306 to select() can be used: if it shows that the input is "ready", there is
2307 either input waiting, or the socket has been closed. An attempt to read
2308 the next input character can distinguish the two cases. Previously, Exim
2309 would have sent an OK response which the client would never have see.
2310 This could lead to message repetition. This fix should cure that, at
2311 least in a lot of common cases.
2313 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2314 advertised in response to EHLO.
2320 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2321 fixed by 4.65/MH/01 (is this a record?) are fixed:
2323 (i) An empty string was always treated as zero by the numeric comparison
2324 operators. This behaviour has been restored.
2326 (ii) It is documented that the numeric comparison operators always treat
2327 their arguments as decimal numbers. This was broken in that numbers
2328 starting with 0 were being interpreted as octal.
2330 While fixing these problems I realized that there was another issue that
2331 hadn't been noticed. Values of message_size_limit (both the global option
2332 and the transport option) were treated as octal if they started with 0.
2333 The documentation was vague. These values are now always treated as
2334 decimal, and I will make that clear in the documentation.
2340 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2341 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2344 MH/01 Don't check that the operands of numeric comparison operators are
2345 integers when their expansion is in "skipping" mode (fixes bug
2346 introduced by 4.64-PH/07).
2348 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2349 child addresses, Exim now panics and dies. Previously, because the count
2350 is held in a short int, deliveries were likely to be lost. As such a
2351 large number of recipients for a single message is ridiculous
2352 (performance will be very, very poor), I have chosen to impose a limit
2353 rather than extend the field.
2359 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2360 leftover -K file (the existence of which was triggered by #402).
2361 While we were at it, introduced process PID as part of the -K
2362 filename. This should rule out race conditions when creating
2365 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2366 processing considerably. Previous code took too long for large mails,
2367 triggering a timeout which in turn triggers #401.
2369 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2370 in the DK code in transports.c. sendfile() is not really portable,
2371 hence the _LINUX specificness.
2373 TF/01 In the add_headers option to the mail command in an Exim filter,
2374 there was a bug that Exim would claim a syntax error in any
2375 header after the first one which had an odd number of characters
2378 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2379 callout verification, Exim cached a "reject" for the entire domain. This
2380 is correct for most verifications, but it is not correct for a recipient
2381 verification with use_sender or use_postmaster set, because in that case
2382 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2383 case of MAIL FROM:<> rejection from other early rejections (e.g.
2384 rejection of HELO). When verifying a recipient using a non-null MAIL
2385 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2386 Whatever the result of the callout, the value of the domain cache is
2387 left unchanged (for any other kind of callout, getting as far as trying
2388 RCPT means that the domain itself is ok).
2390 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2393 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2394 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2395 session, the daemon ignores ECONNECT errors and logs others; it now
2396 ignores EPIPE as well.
2398 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2399 (quoted-printable decoding).
2401 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2402 later the small subsequent patch to fix an introduced bug.
2404 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2406 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2408 PH/08 An error is now given if message_size_limit is specified negative.
2410 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2411 to be given (somewhat) arbitrary names.
2413 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2416 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2417 miscellaneous code fixes
2419 PH/10 Added the log_reject_target ACL modifier to specify where to log
2422 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2423 hostname. This is wrong, because it relates to the incoming message (and
2424 probably the interface on which it is arriving) and not to the outgoing
2425 callout (which could be using a different interface). This has been
2426 changed to use the value of the helo_data option from the smtp transport
2427 instead - this is what is used when a message is actually being sent. If
2428 there is no remote transport (possible with a router that sets up host
2429 addresses), $smtp_active_hostname is used.
2431 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2432 tweaks were necessary in order to get it to work (see also 21 below):
2433 (a) The code assumed that strncpy() returns a negative number on buffer
2434 overflow, which isn't the case. Replaced with Exim's string_format()
2436 (b) There were several signed/unsigned issues. I just did the minimum
2437 hacking in of casts. There is scope for a larger refactoring.
2438 (c) The code used strcasecmp() which is not a standard C function.
2439 Replaced with Exim's strcmpic() function.
2440 (d) The code set only $1; it now sets $auth1 as well.
2441 (e) A simple test gave the error "authentication client didn't specify
2442 service in request". It would seem that Dovecot has changed its
2443 interface. Fortunately there's a specification; I followed it and
2444 changed what the client sends and it appears to be working now.
2446 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2449 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2450 address A is aliased to B and C, where B exists and C does not. Without
2451 -v the output is "A verified" because verification stops after a
2452 successful redirection if more than one address is generated. However,
2453 with -v the child addresses are also verified. Exim was outputting "A
2454 failed to verify" and then showing the successful verification for C,
2455 with its parentage. It now outputs "B failed to verify", showing B's
2456 parentage before showing the successful verification of C.
2458 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2459 look up a TXT record in a specific list after matching in a combined
2462 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2463 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2464 they consult the DNS. I had assumed they would set it the way they
2465 wanted; and indeed my experiments on Linux seem to show that in some
2466 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2467 To be on the safe side, however, I have now made the interface to
2468 host_find_byname() similar to host_find_bydns(), with an argument
2469 containing the DNS resolver options. The host_find_byname() function now
2470 sets these options at its start, just as host_find_bydns() does. The smtp
2471 transport options dns_qualify_single and dns_search_parents are passed to
2472 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2473 of host_find_byname() use the default settings of RES_DEFNAMES
2474 (qualify_single) but not RES_DNSRCH (search_parents).
2476 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2477 spool_read_header() do less string testing, by means of a preliminary
2478 switch on the second character of optional "-foo" lines. (This is
2479 overdue, caused by the large number of possibilities that now exist.
2480 Originally there were few.) While I was there, I also converted the
2481 str(n)cmp tests so they don't re-test the leading "-" and the first
2482 character, in the hope this might squeeze out yet more improvement.
2484 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2485 flag allowing group syntax was set by the header_syntax check but not
2486 turned off, possible causing trouble later; (2) The flag was not being
2487 set at all for the header_verify test, causing "group"-style headers to
2488 be rejected. I have now set it in this case, and also caused header_
2489 verify to ignore an empty address taken from a group. While doing this, I
2490 came across some other cases where the code for allowing group syntax
2491 while scanning a header line wasn't quite right (mostly, not resetting
2492 the flag correctly in the right place). These bugs could have caused
2493 trouble for malformed header lines. I hope it is now all correct.
2495 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2496 with the "reply" argument non-NULL. The code, however (which originally
2497 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2498 but it didn't always do it. This confused somebody who was copying the
2499 code for some other use. I have removed all the tests.
2501 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2502 feature that was used to support insecure browsers during the U.S. crypto
2503 embargo. It requires special client support, and Exim is probably the
2504 only MTA that supported it -- and would never use it because real RSA is
2505 always available. This code has been removed, because it had the bad
2506 effect of slowing Exim down by computing (never used) parameters for the
2507 RSA_EXPORT functionality.
2509 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2510 authenticator to fail if there's a tab character in the incoming data
2511 (there should never be unless someone is messing about, as it's supposed
2512 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2513 if the connection is using TLS or if the remote IP is the same as the
2514 local IP, and the "valid-client-cert option" if a client certificate has
2517 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2518 authenticators. This can be used for authorization after authentication
2519 succeeds. (In the case of plaintext, it servers for both authentication
2522 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2523 if any retry times were supplied.
2525 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2526 connection, where there is no client IP address to check. In this
2527 situation, the verify now always succeeds.
2529 PH/25 Applied John Jetmore's -Mset patch.
2531 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2533 PH/27 In a string expansion for a processed (not raw) header when multiple
2534 headers of the same name were present, leading whitespace was being
2535 removed from all of them, but trailing whitespace was being removed only
2536 from the last one. Now trailing whitespace is removed from each header
2537 before concatenation. Completely empty headers in a concatenation (as
2538 before) are ignored.
2540 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2541 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2543 PH/29 [Removed. This was a change that I later backed out, and forgot to
2544 correct the ChangeLog entry (that I had efficiently created) before
2545 committing the later change.]
2547 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2548 address errors (4xx response to RCPT) over the same connection as other
2549 messages routed to the same hosts. Such deliveries are always "forced",
2550 so retry times are not inspected. This resulted in far too many retries
2551 for the affected addresses. The effect occurred only when there were more
2552 hosts than the hosts_max_try setting in the smtp transport when it had
2553 the 4xx errors. Those hosts that it had tried were not added to the list
2554 of hosts for which the message was waiting, so if all were tried, there
2555 was no problem. Two fixes have been applied:
2557 (i) If there are any address or message errors in an SMTP delivery, none
2558 of the hosts (tried or untried) are now added to the list of hosts
2559 for which the message is waiting, so the message should not be a
2560 candidate for sending over the same connection that was used for a
2561 successful delivery of some other message. This seems entirely
2562 reasonable: after all the message is NOT "waiting for some host".
2563 This is so "obvious" that I'm not sure why it wasn't done
2564 previously. Hope I haven't missed anything, but it can't do any
2565 harm, as the worst effect is to miss an optimization.
2567 (ii) If, despite (i), such a delivery is accidentally attempted, the
2568 routing retry time is respected, so at least it doesn't keep
2569 hammering the server.
2571 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2572 in ${readsocket because some servers need this prod.
2574 PH/32 Added some extra debug output when updating a wait-xxx database.
2576 PH/33 The hint "could be header name not terminated by colon", which has been
2577 given for certain expansion errors for a long time, was not being given
2578 for the ${if def:h_colon_omitted{... case.
2580 PH/34 The spec says: "With one important exception, whenever a domain list is
2581 being scanned, $domain contains the subject domain." There was at least
2582 one case where this was not true.
2584 PH/35 The error "getsockname() failed: connection reset by peer" was being
2585 written to the panic log as well as the main log, but it isn't really
2586 panic-worthy as it just means the connection died rather early on. I have
2587 removed the panic log writing for the ECONNRESET error when getsockname()
2590 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2591 runs only) independently of the message's sender address. This meant
2592 that, if the 4xx error was in fact related to the sender, a different
2593 message to the same recipient with a different sender could confuse
2594 things. In particular, this can happen when sending to a greylisting
2595 server, but other circumstances could also provoke similar problems.
2596 I have changed the default so that the retry time for these errors is now
2597 based a combination of the sender and recipient addresses. This change
2598 can be overridden by setting address_retry_include_sender=false in the
2601 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2602 remote server are returned as part of bounce messages. This was not
2603 happening for LMTP over a pipe (the lmtp transport), but now it is the
2604 same for both kinds of LMTP.
2606 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2607 in header lines that were in fact CNAMEs. This is no longer the case.
2609 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2610 and queue runs started by the daemon processed all messages. This has
2611 been fixed so that -R and -S can now usefully be given with -q<time>.
2613 PH/40 Import PCRE release 6.7 (fixes some bugs).
2615 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2617 PH/42 Give an error if -q is specified more than once.
2619 PH/43 Renamed the variables $interface_address and $interface_port as
2620 $received_ip_address and $received_port, to make it clear that these
2621 values apply to message reception, and not to the outgoing interface when
2622 a message is delivered. (The old names remain recognized, of course.)
2624 PH/44 There was no timeout on the connect() call when using a Unix domain
2625 socket in the ${readsocket expansion. There now is.
2627 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2628 be meaningful with "accept".
2630 SC/01 Eximstats V1.43
2631 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2633 SC/02 Eximstats V1.44
2634 Use a glob alias rather than an array ref in the generated
2635 parser. This improves both readability and performance.
2637 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2638 Collect SpamAssassin and rejection statistics.
2639 Don't display local sender or destination tables unless
2640 there is data to show.
2641 Added average volumes into the top table text output.
2643 SC/04 Eximstats V1.46
2644 Collect data on the number of addresses (recipients)
2645 as well as the number of messages.
2647 SC/05 Eximstats V1.47
2648 Added 'Message too big' to the list of mail rejection
2649 reasons (thanks to Marco Gaiarin).
2651 SC/06 Eximstats V1.48
2652 Mainlog lines which have GMT offsets and are too short to
2653 have a flag are now skipped.
2655 SC/07 Eximstats V1.49 (Alain Williams)
2656 Added the -emptyok flag.
2658 SC/08 Eximstats V1.50
2659 Fixes for obtaining the IP address from reject messages.
2661 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2662 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2663 whitespace changes from 4.64-PH/27
2665 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2668 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2669 are found, allow negative numbers in numeric criteria)
2671 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2673 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2674 to match changes made in 4.64-PH/43
2676 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2678 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2679 "domain" is actually an IP address, and force a failure. This locks out
2680 those revolvers/nameservers that support "A-for-A" lookups, in
2681 contravention of the specifications.
2683 PH/48 When a host name was looked up from an IP address, and the subsequent
2684 forward lookup of the name timed out, the host name was left in
2685 $sender_host_name, contrary to the specification.
2687 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2688 restricted to single-key lookups, Exim was not diagnosing an error if
2689 * or *@ was used with a query-style lookup.
2691 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2693 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2694 long ago, but noone interested enough thought of it. Let's just say that
2695 the "1.1" means that there are some new functions that weren't there at
2696 some point in the past.
2698 PH/51 Error processing for expansion failure of helo_data from an smtp
2699 transport during callout processing was broken.
2701 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2702 tested/used via the -bh/-bhc/-bs options.
2704 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2705 bug, fixed in subsequent PCRE releases).
2707 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2708 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2710 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2716 SC/01 Use a glob alias rather than an array ref in eximstats generated
2717 parser. This improves both readability and performance.
2719 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2720 Don't display local sender or destination tables in eximstats unless
2721 there is data to show.
2722 Added average volumes into the eximstats top table text output.
2724 SC/03 Collect data on the number of addresses (recipients) as well
2725 as the number of messages in eximstats.
2727 TF/01 Correct an error in the documentation for the redirect router. Exim
2728 does (usually) call initgroups() when daemonizing.
2730 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2731 with consistent privilege compared to when running as a daemon.
2733 TF/03 Note in the spec that $authenticated_id is not set for local
2734 submissions from trusted users.
2736 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2737 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2739 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2740 by adding some example configuration directives to the default
2741 configuration file. A little bit of work is required to uncomment the
2742 directives and define how usernames and passwords are checked, but
2743 there is now a framework to start from.
2745 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2746 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2747 without this. I don't know how relevant this is to other LDAP libraries.
2749 PH/02 Add the verb name to the "unknown ACL verb" error.
2751 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2753 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2755 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2756 directory not expanded when it should be if an expanded home directory
2757 was set for the address (which is overridden by the transport).
2759 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2762 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2763 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2764 because it is too late at that time, and has no effect.
2766 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2767 security issue with \' (bugzilla #107). I could not use the
2768 PQescapeStringConn() function, because it needs a PGconn value as one of
2771 PH/08 When testing addresses using -bt, indicate those final addresses that
2772 are duplicates that would not cause an additional delivery. At least one
2773 person was confused, thinking that -bt output corresponded to deliveries.
2774 (Suppressing duplicates isn't a good idea as you lose the information
2775 about possibly different redirections that led to the duplicates.)
2777 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2778 systems where poll() doesn't work, in particular OS X.
2780 PH/10 Added more information to debugging output for retry time not reached.
2782 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2783 operations in malware.c.
2785 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2788 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2789 syslog_duplication set false, log lines that would normally be written
2790 both the the main log and to the reject log were not written to syslog at
2793 PH/14 In the default configuration, change the use of "message" in ACL warn
2794 statements to "add_header".
2796 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2797 not followed by a command (e.g. "seen endif").
2799 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2800 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2803 PH/17 Added extra conditions to the default value of delay_warning_condition
2807 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2808 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2809 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2812 The Auto-Submitted: and various List- headers are standardised, whereas I
2813 don't think Precedence: ever was.
2815 PH/18 Refactored debugging code in route_finduser() to show more information,
2816 in particular, the error code if getpwnam() issues one.
2818 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2819 This is apparently needed in addition to the PH/07 change above to avoid
2820 any possible encoding problems.
2822 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2823 but not after initializing Perl.
2825 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2826 output them only if debugging. By default they are written stderr,
2827 apparently, which is not desirable.
2829 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2832 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2835 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2837 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2838 authenticated or an ident call has been made. Suppress the default
2839 values for $authenticated_id and $authenticated_sender (but permit -oMai
2840 and -oMas) when testing with -bh.
2842 PH/24 Re-jigged the order of the tests in the default configuration so that the
2843 tests for valid domains and recipients precede the DNS black list and CSA
2844 tests, on the grounds that those ones are more expensive.
2846 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2847 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2848 This bug exists in every version of Exim that I still have, right back to
2851 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2852 However, an attempt to turn on case-sensitivity in a regex key by
2853 including (?-i) didn't work because the subject string was already
2854 lowercased, and the effects were non-intuitive. It turns out that a
2855 one-line patch can be used to allow (?-i) to work as expected.
2861 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2862 other effects) broke the use of negated acl sub-conditions.
2864 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2867 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2868 "Deny" causes Exim to reject the incoming connection with a 554 error.
2869 Unfortunately, if there is a major crisis, such as a disk failure,
2870 tcp-wrappers gives "deny", whereas what one would like would be some
2871 kind of temporary error. A kludge has been added to help with this.
2872 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2873 554 error is used if errno is still zero or contains ENOENT (which occurs
2874 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2877 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2879 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2880 errors. Otherwise a message that provokes a temporary error (when other
2881 messages do not) can cause a whole host to time out.
2883 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2884 addresses were routed directly to files or pipes from a redirect router.
2885 File deliveries just didn't batch; pipe deliveries might have suffered
2888 PH/06 A failure to get a lock for a hints database would erroneously always say
2889 "Failed to get write lock", even when it was really a read lock.
2891 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2892 of 0600. This has been changed to use the value of the lockfile_mode
2893 option (which defaults to 0600).
2895 PH/08 Applied small patch from the Sieve maintainer.
2897 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2898 folder from quota calculations, a direct delivery into this folder messed
2899 up the contents of the maildirsize file. This was because the regex was
2900 used only to exclude .Trash (or whatever) when the size of the mailbox
2901 was calculated. There was no check that a delivery was happening into an
2902 excluded directory. This bug has been fixed by ignoring all quota
2903 processing for deliveries into excluded directories.
2905 PH/10 Added the maildirfolder_create_regex option to appendfile.
2911 PH/01 The code for finding all the local interface addresses on a FreeBSD
2912 system running IPv6 was broken. This may well have applied to all BSD
2913 systems, as well as to others that have similar system calls. The broken
2914 code found IPv4 interfaces correctly, but gave incorrect values for the
2915 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2916 that it would not match correctly against @[] and not recognize the IPv6
2919 PH/02 The ipliteral router was not recognizing addresses of the form user@
2920 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2922 PH/03 Added disable_ipv6.
2924 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2925 decoded form, because it is most often used to construct To: headers
2926 lines in autoreplies, and the decoded form may well be syntactically
2927 invalid. However, $reply_address has leading white space removed, and all
2928 newlines turned into spaces so that the autoreply transport does not
2931 PH/05 If group was specified without a user on a router, and no group or user
2932 was specified on a transport, the group from the router was ignored.
2934 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2935 for visible compile-time settings that can be used to change these
2936 numbers, for those that want even more. Backwards compatibility with old
2937 spool files has been maintained. However, going back to a previous Exim
2938 release will lost any variables that are in spool files.
2940 PH/07 Two small changes when running in the test harness: increase delay when
2941 passing a TCP/IP connection to a new process, in case the original
2942 process has to generate a bounce, and remove special handling of
2943 127.0.0.2 (sic), which is no longer necessary.
2945 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2946 be the same on different OS.
2948 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2951 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2952 whether --show-vars was specified or not
2954 JJ/02 exipick: Added support for new ACL variable spool format introduced
2957 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2958 syntactically invalid From: or Reply-to: line, and a filter used this to
2959 generate an autoreply, and therefore failed to obtain an address for the
2960 autoreply, Exim could try to deliver to a non-existent relative file
2961 name, causing unrelated and misleading errors. What now happens is that
2962 it logs this as a hard delivery error, but does not attempt to create a
2965 PH/11 The exinext utility has a -C option for testing purposes, but although
2966 the given file was scanned by exinext itself; it wasn't being passed on
2967 when Exim was called.
2969 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2970 an end-of-file indication when reading a command response.
2972 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2973 compiled. In many other places in Exim, IPv6 addresses are always
2974 recognized, so I have changed this. It also means that IPv4 domain
2975 literals of the form [IPV4:n.n.n.n] are now always recognized.
2977 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2978 used if the router is not running as root, for example, when verifying at
2979 ACL time, or when using -bh. The debugging output from this situation was
2980 non-existent - all you got was a failure to exec. I have made two
2983 (a) Failures to set uid/gid, the current directory, or a process leader
2984 in a subprocess such as that created by queryprogram now generate
2985 suitable debugging output when -d is set.
2987 (b) The queryprogram router detects when it is not running as root,
2988 outputs suitable debugging information if -d is set, and then runs
2989 the subprocess without attempting to change uid/gid.
2991 PH/15 Minor change to Makefile for building test_host (undocumented testing
2994 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2995 additional section of a DNS packet that returns MX or SRV records.
2996 Instead, it always explicitly searches for A/AAAA records. This avoids
2997 major problems that occur when a DNS server includes only records of one
2998 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2999 fixed another bug: if SRV records were looked up and the corresponding
3000 address records were *not* found in the additional section, the port
3001 values from the SRV records were lost.
3003 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3004 using the correct key (the original address) when searching the retry
3005 rules in order to find which one to use for generating the retry hint.
3007 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3008 adding the default one. Similarly, if it contains a Reply-To: header, the
3009 errors_reply_to option, if set, is not used.
3011 PH/19 When calculating a retry time, Exim used to measure the "time since
3012 failure" by looking at the "first failed" field in the retry record. Now
3013 it does not use this if it is later than than the arrival time of the
3014 message. Instead it uses the arrival time. This makes for better
3015 behaviour in cases where some deliveries succeed, thus re-setting the
3016 "first failed" field. An example is a quota failure for a huge message
3017 when small messages continue to be delivered. Without this change, the
3018 "time since failure" will always be short, possible causing more frequent
3019 delivery attempts for the huge message than are intended.
3020 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3022 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3023 $1, $2, $3) because the numerical variables can be reset during some
3024 expansion items (e.g. "match"), thereby losing the authentication data.
3026 PH/21 Make -bV show the size of off_t variables so that the test suite can
3027 decide whether to run tests for quotas > 2G.
3029 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3030 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3031 filecount value is greater than 2G or if a quota value is greater than 2G
3032 on a system where the size of off_t is not greater than 4, a panic error
3035 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3036 never match. The debug and -bh output now contains an explicit error
3037 message indicating a malformed IPv4 address or mask.
3039 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3040 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3041 PH/23 above applies.
3043 PH/25 Do not write to syslog when running in the test harness. The only
3044 occasion when this arises is a failure to open the main or panic logs
3045 (for which there is an explicit test).
3047 PH/26 Added the /no_tell option to "control=freeze".
3049 PH/27 If a host name lookup failed very early in a connection, for example, if
3050 the IP address matched host_lookup and the reverse lookup yielded a name
3051 that did not have a forward lookup, an error message of the form "no IP
3052 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3053 could be logged. Now it outputs the IP address instead of "NULL".
3055 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3056 allows the sender and the authenticated sender to be set when
3057 submitting a message from within Exim. Since child_open_exim() is
3058 documented for local_scan(), the new function should be too.
3060 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3061 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3062 results in an empty string is now treated as unset.
3064 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3066 PH/31 Added sender_verify_fail logging option.
3068 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3069 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3070 tidied the source and removed it altogether.
3072 PH/33 When a queue run was abandoned because the load average was too high, a
3073 log line was always written; now it is written only if the queue_run log
3074 selector is set. In addition, the log line for abandonment now contains
3075 information about the queue run such as the pid. This is always present
3076 in "start" and "stop" lines but was omitted from the "abandon" line.
3078 PH/34 Omit spaces between a header name and the colon in the error message that
3079 is given when verify = headers_syntax fails (if there are lots of them,
3080 the message gets confusing).
3082 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3083 names, as there are now some PTR records that contain slashes. This check
3084 is only to protect against broken name servers that fall over on strange
3085 characters, so the fact that it applies to all lookups doesn't matter.
3087 PH/36 Now that the new test suite is complete, we can remove some of the
3088 special code in Exim that was needed for the old test suite. For example,
3089 sorting DNS records because real resolvers return them in an arbitrary
3090 order. The new test suite's fake resolver always returns records in the
3093 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3094 bounces) except when queue_only is set, to avoid logging races between
3095 the different processes.
3097 PH/38 Panic-die if .include specifies a non-absolute path.
3099 PH/39 A tweak to the "H" retry rule from its user.
3101 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3102 a label. They prevented compilation on older perls.
3104 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3105 a warning to be raised on newish perls.
3107 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3108 on queue. Changes to match documented behaviour of showing count of
3109 messages matching specified criteria.
3111 PH/40 Changed the default ident timeout from 30s to 5s.
3113 PH/41 Added support for the use of login_cap features, on those BSD systems
3114 that have them, for controlling the resources used by pipe deliveries.
3116 PH/42 The content-scanning code uses fopen() to create files in which to put
3117 message data. Previously it was not paying any attention to the mode of
3118 the files. Exim runs with umask(0) because the rest of the code creates
3119 files with open(), and sets the required mode explicitly. Thus, these
3120 files were ending up world-writeable. This was not a big issue, because,
3121 being within the spool directory, they were not world-accessible. I have
3122 created a function called modefopen, which takes an additional mode
3123 argument. It sets umask(777), creates the file, chmods it to the required
3124 mode, then resets the umask. All the relevant calls to fopen() in the
3125 content scanning code have been changed to use this function.
3127 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3128 to 24 hours. This avoids potential overflow problems when processing G
3129 and H retry rules. I suspect nobody ever tinkers with this value.
3131 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3133 PH/45 When the plaintext authenticator is running as a client, the server's
3134 challenges are checked to ensure they are valid base64 strings. By
3135 default, the authentication attempt is cancelled if an invalid string is
3136 received. Setting client_ignore_invalid_base64 true ignores these errors.
3137 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3138 they are received. Thus, the responses can be made to depend on the
3139 challenges. If an invalid string is ignored, an empty string is placed in
3142 PH/46 Messages that are created by the autoreply transport now contains a
3143 References: header, in accordance with RFCs 2822 and 3834.
3145 PH/47 Added authenticated_sender_force to the smtp transport.
3147 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3149 PH/49 Installed latest patch from the Sieve maintainer.
3151 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3152 was also set, the appendfile transport was unnecessarily scanning a
3153 directory of message files (e.g. for maildir delivery) to find the count
3154 of files (along with the size), even though it did not need this
3155 information. It now does the scan only if it needs to find either the
3156 size of the count of files.
3158 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3160 PH/52 Two bugs concerned with error handling when the smtp transport is
3163 (i) Exim was not creating retry information for temporary errors given
3164 for individual recipients after the DATA command when the smtp transport
3165 was used in LMTP mode. This meant that they could be retried too
3166 frequently, and not timed out correctly.
3168 (ii) Exim was setting the flag that allows error details to be returned
3169 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3170 recipients that were returned after the DATA command.
3172 PH/53 This is related to PH/52, but is more general: for any failing address,
3173 when detailed error information was permitted to be returned to the
3174 sender, but the error was temporary, then after the final timeout, only
3175 "retry timeout exceeded" was returned. Now it returns the full error as
3176 well as "retry timeout exceeded".
3178 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3179 do this, and (what is worse) MTAs that accept it.
3181 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3182 will now be deprecated.
3184 PH/56 New os.c-cygwin from the Cygwin maintainer.
3186 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3187 formats (previously only available in exim formats via -bpr, -bpru,
3188 and -bpra. Now also available in native and exiqgrep formats)
3190 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3191 with very large, slow to parse queues
3193 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3195 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3197 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3198 responses to authentication challenges, though it was showing the
3199 challenges; (ii) I've removed the CR characters from the debug output for
3202 PH/58 Allow for the insertion of a newline as well as a space when a string
3203 is turned into more than one encoded-word during RFC 2047 encoding. The
3204 Sieve code now uses this.
3206 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3207 data_4xx, lost_connection, tls_required.
3209 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3210 message was being sent as an SMTP response.
3212 PH/61 Add -l and -k options to exicyclog.
3214 PH/62 When verifying, if an address was redirected to one new address, so that
3215 verification continued, and the new address failed or deferred after
3216 having set something in $address_data, the value of $address_data was not
3217 passed back to the ACL. This was different to the case when no
3218 redirection occurred. The value is now passed back in both cases.
3220 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3221 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3222 use login_cap.h, so on its own it isn't the distinguishing feature. The
3223 new name refers directly to the setclassresources() function.
3225 PH/65 Added configuration files for NetBSD3.
3227 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3229 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3230 is preferred over IPv4.
3232 PH/68 The bounce_return_message and bounce_return_body options were not being
3233 honoured for bounces generated during the reception of non-SMTP messages.
3234 In particular, this applied to messages rejected by the ACL. This bug has
3235 been fixed. However, if bounce_return_message is true and bounce_return_
3236 body is false, the headers that are returned for a non-SMTP message
3237 include only those that have been read before the error was detected.
3238 (In the case of an ACL rejection, they have all been read.)
3240 PH/69 The HTML version of the specification is now built in a directory called
3241 spec_html instead of spec.html, because the latter looks like a path with
3242 a MIME-type, and this confuses some software.
3244 PH/70 Catch two compiler warnings in sieve.c.
3246 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3247 function verify_get_ident() calls ip_connect() to connect a socket, but
3248 if the "connect()" function timed out, ip_connect() used to close the
3249 socket. However, verify_get_ident() also closes the socket later, and in
3250 between Exim writes to the log, which may get opened at this point. When
3251 the socket was closed in ip_connect(), the log could get the same file
3252 descriptor number as the socket. This naturally causes chaos. The fix is
3253 not to close the socket in ip_connect(); the socket should be closed by
3254 the function that creates it. There was only one place in the code where
3255 this was missing, in the iplookup router, which I don't think anybody now
3256 uses, but I've fixed it anyway.
3258 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3259 well as to direct DNS lookups. Otherwise the handling of names in host
3260 lists is inconsistent and therefore confusing.
3266 PH/01 Two changes to the default runtime configuration:
3268 (1) Move the checks for relay_from_hosts and authenticated clients from
3269 after to before the (commented out) DNS black list checks.
3271 (2) Add control=submission to the relay_from_hosts and authenticated
3272 clients checks, on the grounds that messages accepted by these
3273 statements are most likely to be submissions.
3275 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3277 (1) Generate an error if the third argument for the ${prvs expansion is
3280 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3283 (3) Reset the variables that are obtained from the first argument of
3284 ${prvscheck and used in the second argument before leaving the code,
3285 because their memory is reclaimed, so using them afterwards may do
3288 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3289 one (it's much easier than Tom thought :-).
3291 (5) Because of (4), we can now allow for the use of $prvscheck_result
3292 inside the third argument.
3294 PH/03 For some reason, the default setting of PATH when running a command from
3295 a pipe transport was just "/usr/bin". I have changed it to
3298 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3299 anything to be listed in the output from -bV.
3301 PH/05 When a filter generated an autoreply, the entire To: header line was
3302 quoted in the delivery log line, like this:
3304 => >A.N.Other <ano@some.domain> <original@ddress> ...
3306 This has been changed so that it extracts the operative address. There
3307 may be more than one such address. If so, they are comma-separated, like
3310 => >ano@some.domain,ona@other.domain <original@ddress> ...
3312 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3313 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3314 not being looked up in the rDNS to get a host name, Exim was showing the
3315 IP address twice in Received: lines, even though the IP addresses were
3316 identical. For example:
3318 Received: from [1.2.3.4] (helo=[1.2.3.4])
3320 However, if the real host name was known, it was omitting the HELO data
3321 if it matched the actual IP address. This has been tidied up so that it
3322 doesn't show the same IP address twice.
3324 PH/07 When both +timestamp and +memory debugging was on, the value given by
3325 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3326 called by the debug printing, thereby overwriting the timestamp buffer.
3327 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3329 PH/08 When the original message was included in an autoreply transport, it
3330 always said "this is a copy of the message, including all the headers",
3331 even if body_only or headers_only was set. It now gives an appropriate
3334 PH/09 Applied a patch from the Sieve maintainer which:
3336 o fixes some comments
3337 o adds the (disabled) notify extension core
3338 o adds some debug output for the result of if/elsif tests
3339 o points to the current vacation draft in the documentation
3340 and documents the missing references header update
3344 o fixes a bug in processing the envelope test (when testing
3345 multiple envelope elements, the last element determined the
3348 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3349 Electronic Mail") by including:
3351 Auto-submitted: auto-generated
3353 in the messages that it generates (bounce messages and others, such as
3354 warnings). In the case of bounce messages for non-SMTP messages, there was
3355 also a typo: it was using "Auto_submitted" (underscore instead of
3356 hyphen). Since every message generated by Exim is necessarily in response
3357 to another message, thes have all been changed to:
3359 Auto-Submitted: auto-replied
3361 in accordance with these statements in the RFC:
3363 The auto-replied keyword:
3365 - SHOULD be used on messages sent in direct response to another
3366 message by an automatic process,
3368 - MUST NOT be used on manually-generated messages,
3370 - MAY be used on Delivery Status Notifications (DSNs) and Message
3371 Disposition Notifications (MDNs),
3373 - MUST NOT be used on messages generated by automatic or periodic
3374 processes, except for messages which are automatic responses to
3377 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3378 to the default Received: header definition.
3380 PH/12 Added log selector acl_warn_skipped (default on).
3382 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3383 variables because (a) they are in the wrong storage pool and (b) even if
3384 they were copied, it wouldn't work properly because of the caching.
3386 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3387 checking when decoding. Apparently there are clients that generate
3388 overlong encoded strings. Why am I not surprised?
3390 PH/15 If the first argument of "${if match_address" was not empty, but did not
3391 contain an "@" character, Exim crashed. Now it writes a panic log message
3392 and treats the condition as false.
3394 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3396 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3397 extension "envelope-auth". The code is finished and in agreement with
3398 other implementations, but there is no documentation so far and in fact,
3399 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3400 not changing the active code.
3402 Print executed "if" and "elsif" statements when debugging is used. This
3403 helps a great deal to understand what a filter does.
3405 Document more things not specified clearly in RFC3028. I had all this
3406 sorted out, when out of a sudden new issues came to my mind. Oops."
3408 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3411 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3412 canonical form (as documented). However, after a host name lookup from
3413 the IP address, check_host() was doing a simple string comparison with
3414 addresses acquired from the DNS when checking that the found name did
3415 have the original IP as one of its addresses. Since any found IPv6
3416 addresses are likely to be in abbreviated form, the comparison could
3417 fail. Luckily, there already exists a function for doing the comparison
3418 by converting both addresses to binary, so now that is used instead of
3419 the text comparison.
3421 PH/20 There was another similar case to PH/19, when a complete host name was
3422 given in a host list; looking up its IP address could give an abbreviated
3423 form, whereas the current host's name might or might not be abbreviated.
3424 The same fix has been applied.
3430 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3431 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3434 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3435 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3437 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3439 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3440 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3441 Also, for "Auto-Submitted", treat anything other than "no" as
3442 non-personal, in accordance with RFC 3834. (Previously it treated
3443 anything starting "auto-" as non-personal.)
3445 TF/01 The control=submission/name=... option had a problem with syntax
3446 errors if the name included a slash character. The /name= option
3447 now slurps the rest of the string, so it can include any characters
3448 but it must come last in the list of options (after /sender_retain
3451 PH/05 Some modifications to the interface to the fake nameserver for the new
3459 TK/01 Added the "success_on_redirect" address verification option. See
3460 NewStuff for rationale and an example.
3462 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3464 PH/02 Patch to exigrep to allow it to work on syslog lines.
3466 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3467 fread() to skip over the body file's header line, because in Cygwin the
3468 header line is locked and is inaccessible.
3470 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3471 co-exist for some time) to make it clear that it is the Exim ID that is
3472 referenced, not the Message-ID: header line.
3474 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3475 string_format() function, because snprintf() does not exist on all
3478 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3479 ${stat: expansion item.
3481 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3482 protocol synchronization error", to keep the pedants happy.
3484 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3485 well as for IRIX systems, when gcc is being used. See the host.c source
3488 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3490 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3493 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3494 transport and to the smtp transport in LMTP mode.
3496 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3498 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3499 run a filter in a subprocess. This could lead to confusion in subsequent
3500 lookups in the parent process. There should also be a search_tidyup() at
3501 the end of the subprocess.
3503 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3504 only if the host matched helo_try_verify_hosts, which caused the
3505 verification to occur when the EHLO/HELO command was issued. The ACL just
3506 tested the remembered result. Now, if a previous verification attempt has
3507 not happened, "verify = helo" does it there and then.
3509 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3511 TK/03 Fix log output including CR from clamd.
3513 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3514 exist provoked a memory error which could cause a segfault.
3516 PH/15 Installed PCRE 6.2
3518 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3520 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3521 of the problem. Specifically, suggested +O2 rather than +O1 for the
3524 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3526 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3527 record, the port was not correctly logged when the outgoing_port log
3528 selector was set (it logged the transort's default port).
3530 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3531 fallback_hosts, and "hosts" in the smtp transport.
3533 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3534 host errors such as "Connection refused".
3536 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3537 authentication with radiusclient 0.4.9:
3539 - Error returned from rc_read_config was caught wrongly
3540 - Username/password not passed on to radius server due to wrong length.
3542 The presumption is that some radiusclient API changes for 4.51/PH/17
3543 were not taken care of correctly. The code is still untested by me (my
3544 Linux distribution still has 0.3.2 of radiusclient), but it was
3545 contributed by a Radius user.
3547 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3548 expanding the "port" option of the smtp transport.
3550 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3551 while reading a MIME header. Thanks to Tom Hughes for a patch.
3553 PH/24 Include config.h inside local_scan.h so that configuration settings are
3556 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3557 that in an ACL for RCPT (for example), you can examine exactly what was
3560 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3561 commands, but it was not correctly comparing the address with the actual
3562 client host address. Thus, it would show the EHLO address in Received:
3563 header lines when this was not necessary.
3565 PH/27 Added the % operator to ${eval:}.
3567 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3568 it should be ignoring failures (because with -C, for example, it has lost
3569 privilege). It wasn't ignoring creation failures other than "already
3572 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3575 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3576 because an input line was too long, either on its own, or by virtue of
3577 too many continuations, the temporary file was not being removed, and the
3578 return code was incorrect.
3580 PH/31 Missing "BOOL" in function definition in filtertest.c.
3582 PH/32 Applied Sieve patches from the maintainer.
3584 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3586 PH/33 Added "verify = not_blind".
3588 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3589 Local/Makefile (with some defaults set). These are used in built scripts
3590 such as exicyclog, but they have never been used in the exim_install
3591 script (though there are many overriding facilities there). I have
3592 arranged that the exim_install script now takes note of these two
3595 PH/35 Installed configuration files for Dragonfly.
3597 PH/36 When a locally submitted message by a trusted user did not contain a
3598 From: header, and the sender address was obtained from -f or from an SMTP
3599 MAIL command, and the trusted user did not use -F to supply a sender
3600 name, $originator_name was incorrectly used when constructing a From:
3601 header. Furthermore, $originator_name was used for submission mode
3602 messages from external hosts without From: headers in a similar way,
3603 which is clearly wrong.
3605 PH/37 Added control=suppress_local_fixups.
3607 PH/38 When log_selector = +received_sender was set, and the addition of the
3608 sender made the log line's construction buffer exactly full, or one byte
3609 less than full, an overflow happened when the terminating "\n" was
3612 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3613 when the result of a list match is failure because a DNS lookup failed.
3615 PH/40 RM_COMMAND is now used in the building process.
3617 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3618 the "build-* directories that it finds.
3620 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3621 address in a domain literal was a prefix of an interface address.
3623 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3624 when verifying a sender address, unless rewrite_headers is false.
3626 PH/44 Wrote a long comment about why errors_to addresses are verified as
3627 recipients, not senders.
3629 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3630 the ratelimit ACL was added.
3632 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3634 PH/46 Added extra information about PostgreSQL errors to the error string.
3636 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3637 avoiding the need to install special zones in a real server. This is
3638 backwards compatible; if it can't find the fake resolver, it drops back.
3639 Thus, both old and new test suites can be run.
3641 TF/02 Added util/ratelimit.pl
3643 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3644 clock is set back in time.
3646 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3647 Candler <B.Candler@pobox.com>.
3649 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3650 OK for addresses that are the result of redirections.
3652 PH/48 A number of further additions for the benefit of the new test suite,
3653 including a fake gethostbyname() that interfaces to the fake DNS resolver
3656 TF/06 The fix for widen_domains has also been applied to qualify_single and
3657 search_parents which are the other dnslookup options that can cause
3660 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3663 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3665 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3666 service" if it encountered an MX record with an empty target hostname.
3667 The message is now "an MX or SRV record indicated no SMTP service".
3669 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3670 if the DNS of the sending site is misconfigured. This is quite a
3671 common situation. This change restores the behaviour of treating a
3672 helo verification defer as a failure.
3674 PH/51 If self=fail was set on a router, the bounce message did not include the
3675 actual error message.
3681 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3683 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3684 command itself ended in error, the underlying message about the transport
3685 filter timeout was being overwritten with the pipe command error. Now the
3686 underlying error message should be appended to the second error message.
3688 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3690 PH/02 Reduce the amount of output that "make" produces by default. Full output
3691 can still be requested.
3693 PH/03 The warning log line about a condition test deferring for a "warn" verb
3694 was being output only once per connection, rather than after each
3695 occurrence (because it was using the same function as for successful
3696 "warn" verbs). This seems wrong, so I have changed it.
3698 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3699 it should not have, which might have caused a crash in the right
3700 circumstances, but probably never did.
3702 PH/04 Installed a modified version of Tony Finch's patch to make submission
3703 mode fix the return path as well as the Sender: header line, and to
3704 add a /name= option so that you can make the user's friendly name appear
3707 TF/03 Added the control = fakedefer ACL modifier.
3709 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3710 Mark Lowes for thorough testing.
3712 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3714 TK/03 Merged latest SRS patch from Miles Wilton.
3716 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3717 with the definition in sysexits.h (which is #included earlier).
3718 Fortunately, Exim does not actually use EX_OK. The code used to try to
3719 preserve the sysexits.h value, by assuming that macro definitions were
3720 scanned for macro replacements. I have been disabused of this notion,
3721 so now the code just undefines EX_OK before #including unistd.h.
3723 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3724 in the smtp transport. When a block could not be written in a single
3725 write() function, the timeout was being re-applied to each part-write.
3726 This seems wrong - if the receiver was accepting one byte at a time it
3727 would take for ever. The timeout is now adjusted when this happens. It
3728 doesn't have to be particularly precise.
3730 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3731 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3733 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3734 without a domain if the check to <postmaster@domain> fails.
3736 SC/01 Eximstats: added -xls and the ability to specify output files
3737 (patch written by Frank Heydlauf).
3739 SC/02 Eximstats: use FileHandles for outputting results.
3741 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3743 SC/04 Eximstats: fixed display of large numbers with -nvr option
3745 SC/05 Eximstats: fixed merging of reports with empty tables.
3747 SC/06 Eximstats: added the -include_original_destination flag
3749 SC/07 Eximstats: removed tabs and trailing whitespace.
3751 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3753 TK/06 MBOX spool code: Add real "From " MBOX separator line
3754 so the .eml file is really in mbox format (even though
3755 most programs do not really care). Patch from Alex Miller.
3757 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3758 The latter is generated from $received_to and is only set if the
3759 message has one envelope recipient. SA can use these headers,
3760 obviously out-of-the-box. Patch from Alex Miller.
3762 PH/08 The ${def test on a variable was returning false if the variable's
3763 value was "0", contrary to what the specification has always said!
3764 The result should be true unless the variable is empty.
3766 PH/09 The syntax error of a character other than { following "${if
3767 def:variable_name" (after optional whitespace) was not being diagnosed.
3768 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3769 accidental colon was present, for example, could give incorrect results.
3771 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3772 result is used (not including appendfile, where other changes are about
3775 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3776 This involved changing a lot of size variables from int to off_t. It
3777 should work with maildirs and everything.
3779 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3780 spamd dying while we are connected to it.
3782 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3785 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3786 function for BDB 4.3.
3788 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3790 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3791 the sources for PCRE have been re-arranged and more files are now
3794 PH/15 The code I had for printing potentially long long variables in PH/11
3795 above was not the best (it lost precision). The length of off_t variables
3796 is now inspected at build time, and an appropriate printing format (%ld
3797 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3798 to be "long long int" or "long int". This is needed for the internal
3799 formatting function string_vformat().
3801 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3802 the configuration file to be ":syslog", then the script "guesses" where
3803 the logs files are, rather than using the compiled in default. In our
3804 case the guess is not the same as the compiled default, so the script
3805 suddenly stopped working when I started to use syslog. The patch checks
3806 to see if log_file_path is "". If so, it attempts to read it from exim
3807 with no configuration file to get the compiled in version, before it
3808 falls back to the previous guessing code."
3810 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3811 implementing BATV in an Exim configuration. See NewStuff for the gory
3814 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3815 Makefile that are specific to HP-UX.
3817 PH/18 If the "use_postmaster" option was set for a recipient callout together
3818 with the "random" option, the postmaster address was used as the MAIL
3819 FROM address for the random test, but not for the subsequent recipient
3820 test. It is now used for both.
3822 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3823 patch removes a few documentation additions to RFC 3028, because the
3824 latest draft now contains them. It adds the new en;ascii-case comparator
3825 and a new error check for 8bit text in MIME parts. Comparator and
3826 require names are now matched exactly. I enabled the subaddress
3827 extension, but it is not well tested yet (read: it works for me)."
3829 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3830 rework some of the code of TK/09 above to avoid the hardwired use of
3831 "%lld" and "long long". Replaced the call to snprintf() with a call to
3834 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3835 records point to non-existent hosts", "retry timeout exceeded", and
3836 "retry time not reached for any host after a long failure period".
3838 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3839 experimental DomainKeys support:
3841 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3842 (2) On an error such as an illegally used "control", the wrong name for
3843 the control was given.
3845 These problems did NOT occur unless DomainKeys support was compiled.
3847 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3849 PH/24 Added ${if match_ip condition.
3851 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3852 databases so that it will be absolutely obvious if a crash occurs in the
3853 DB library. This is a regular occurrence (often caused by mis-matched
3856 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3857 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3858 picked up on a user's system that detects such things. There doesn't seem
3859 to be a gcc warning option for this - only an attribute that has to be
3860 put on the function's prototype. It seems that in Fedora Core 4 they have
3861 set this on a number of new functions. No doubt there will be more in due
3864 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3865 specify a transport. However, if an address that was verified by such a
3866 router was the subject of a callout, Exim crashed because it tried to
3867 read the rcpt_include_affixes from the non-existent transport. Now it
3868 just assumes that the setting of that option is false. This bug was
3869 introduced by 4.51/PH/31.
3871 PH/28 Changed -d+all to exclude +memory, because that information is very
3872 rarely of interest, but it makes the output a lot bigger. People tend to
3873 do -d+all out of habit.
3875 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3876 code in os-type was giving problems when libc.so lives in lib64, like on
3879 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3880 aren't the modern standard, and it seems that some systems' include files
3881 don't always have them. Exim was already checking for some of the newer
3882 ones like T_AAAA, and defining it itself. I've added checks for all the
3883 record types that Exim uses.
3885 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3886 not automatically generating a new one, as it is supposed to. This
3887 prevented TLS from working. If the file did exist, but contained invalid
3888 data, a new version was generated, as expected. It was only the case of a
3889 non-existent file that was broken.
3891 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3892 with a change in libdomainkeys > 0.64.
3894 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3895 from DNS. If the selector record carries the flag, it now has
3896 precedence over the domain-wide flag.
3898 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3900 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3901 the use of a port name that isn't defined in /etc/services) occurred, the
3902 message was deferred as in a normal delivery, and thus remained on the
3903 spool, instead of being failed because of the mua_wrapper setting. This
3904 is now fixed, and I tidied up some of the mua_wrapper messages at the
3907 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3908 the messages in a hash of arrays rather than using individual hashes.
3909 This is a bit cleaner and results in dramatic memory savings, albeit
3910 at a slight CPU cost.
3912 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3913 as requested by Marc Sherman.
3915 SC/10 Eximstats: added histograms for user specified patterns as requested
3918 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3920 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3921 fopen() in the content-scanning modules that did not already have it.
3927 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3928 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3930 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3932 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3934 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3935 to test Sieve filters that use "vacation".
3937 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3938 that changes the way the GnuTLS parameters are stored in the cache file.
3939 The new format can be generated externally. For backward compatibility,
3940 if the data in the cache doesn't make sense, Exim assumes it has read an
3941 old-format file, and it generates new data and writes a new file. This
3942 means that you can't go back to an older release without removing the
3945 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3946 work if there are any delivery delays because "one_time" forces the
3947 parent to be marked "delivered", so its unseen clone is never tried
3948 again. For this reason, Exim now forbids the simultaneous setting of
3951 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3952 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3953 entry if you want to know the details. The fix, however, neglected to
3954 consider the case when local delivery batching is involved. The test for
3955 "previously delivered" was not happening when checking to see if an
3956 address could be batched with a previous (undelivered) one; under
3957 certain circumstances this could lead to multiple deliveries to the same
3960 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3961 in its include files, and this causes problems building Exim.
3963 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3964 header_syntax) but Exim was just ignoring anything given after a slash.
3965 In particular, this caused confusion with an attempt to use "verify =
3966 reverse_host_lookup/defer_ok". An error is now given when options are
3967 supplied for verify items that do not have them. (Maybe reverse_host_
3968 lookup should have a defer_ok option, but that's a different point.)
3970 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3971 defined by RFC 821) to 2048, because there were problems with some AUTH
3972 commands, and RFC 1869 says the size should be increased for extended
3973 SMTP commands that take arguments.
3975 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3978 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3979 "unknown" error; now it says that the functionality isn't in the binary.
3981 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3982 an address' error message when a string expansion fails (syntax or
3983 whatever). Otherwise the password may appear in the log. Following change
3984 PH/42 below, there is no longer a chance of it appearing in a bounce
3987 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3989 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3990 of its addresses was ever tried. (Bugzilla bug #2.)
3992 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3993 the result incorrectly in the debug output. (It correctly added a newline
3994 to what was transported.)
3996 TF/01 Added $received_time.
3998 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3999 commented out examples of how to interface to a virus scanner and to
4000 SpamAssassin. Also added commented examples of av_scanner and
4001 spamd_address settings.
4003 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4004 and controls are allowed in which ACLs. There were a couple of minor
4005 errors. Some of the entries in the conditions table (which is a table of
4006 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4007 negation of where the condition IS allowed.
4009 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4011 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4012 header file does not have a version number, so I've had to invent a new
4013 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4014 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4015 radiusclient), but it was contributed by a Radius user.
4017 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4018 files or directories, for OpenSSL.
4020 PH/19 When an Exim process that is running as root has to create an Exim log
4021 file, it does so in a subprocess that runs as exim:exim so as to get the
4022 ownership right at creation (otherwise, other Exim processes might see
4023 the file with the wrong ownership). There was no test for failure of this
4024 fork() call, which would lead to the process getting stuck as it waited
4025 for a non-existent subprocess. Forks do occasionally fail when resources
4026 run out. I reviewed all the other calls to fork(); they all seem to check
4029 PH/20 When checking for unexpected SMTP input at connect time (before writing
4030 the banner), Exim was not dealing correctly with a non-positive return
4031 from the read() function. If the client had disconnected by this time,
4032 the result was a log entry for a synchronization error with an empty
4033 string after "input=" when read() returned zero. If read() returned -1
4034 (an event I could not check), uninitialized data bytes were printed.
4035 There were reports of junk text (parts of files, etc) appearing after
4038 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4040 PH/22 Added support for macro redefinition, and (re)definition in between
4041 driver and ACL definitions.
4043 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4044 forgetting to use the resulting value; it was using the unexpanded value.
4046 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4047 hadn't been configured. The fix is from Juergen Kreileder, who
4048 understands it better than I do:
4050 "Here's what I see happening with three configured cyrus_sasl
4051 authenticators configured (plain, login, cram-md5):
4053 On startup auth_cyrus_sasl_init() gets called for each of these.
4054 This means three calls to sasl_listmech() without a specified mech_list.
4055 => SASL tests which mechs of all available mechs actually work
4056 => three warnings about OTP not working
4057 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4059 With the patch, sasl_listmech() also gets called three times. But now
4060 SASL's mech_list option is set to the server_mech specified in the the
4061 authenticator. Or in other words, the answer from sasl_listmech()
4062 gets limited to just the mech you're testing for (which is different
4064 => the return list contains just 'plain' or 'login', 'cram-md5' or
4065 nothing depending on the value of ob->server_mech.
4067 I've just tested the patch: Authentication still works fine,
4068 unavailable mechs specified in the exim configuration are still
4069 caught, and the auth.log warnings about OTP are gone."
4071 PH/25 When debugging is enabled, the contents of the command line are added
4072 to the debugging output, even when log_selector=+arguments is not
4075 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4076 answer is "GNU", and only if the return is "GNU/something" is the answer
4079 PH/27 $acl_verify_message is now set immediately after the failure of a
4080 verification in an ACL, and so is available in subsequent modifiers. In
4081 particular, the message can be preserved by coding like this:
4083 warn !verify = sender
4084 set acl_m0 = $acl_verify_message
4086 Previously, $acl_verify_message was set only while expanding "message"
4087 and "log_message" when a very denied access.
4089 PH/28 Modified OS/os.c-Linux with
4091 -#ifndef OS_LOAD_AVERAGE
4092 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4094 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4095 nomenclature these days.)
4097 PH/29 Installed patch from the Sieve maintainer that adds the options
4098 sieve_useraddress and sieve_subaddress to the redirect router.
4100 PH/30 In these circumstances:
4101 . Two addresses routed to the same list of hosts;
4102 . First host does not offer TLS;
4103 . First host accepts first address;
4104 . First host gives temporary error to second address;
4105 . Second host offers TLS and a TLS session is established;
4106 . Second host accepts second address.
4107 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4108 and peerdn, if requested) that were in fact used only for the second
4111 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4112 attention to any local part prefix or suffix that was matched by the
4113 router that accepted the address. It now behaves in the same way as it
4114 does for delivery: the affixes are removed from the local part unless
4115 rcpt_include_affixes is set on the transport.
4117 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4118 timeout during the DATA phase of an incoming message.
4120 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4121 applied a patch sanctioned by the Sieve maintainer.
4123 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4124 the uid or gid is negative. A case of a negative gid caused this to be
4125 noticed. The fix allows for either to be negative.
4127 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4128 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4131 PH/36 The change PH/12 above was broken. Fixed it.
4133 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4134 the grounds that routing the same address twice would always produce the
4135 same answer. This might have been true once, but it is certainly no
4136 longer true now. Routing a child address may depend on the previous
4137 routing that produced that child. Some complicated redirection strategies
4138 went wrong when messages had multiple recipients, and made Exim's
4139 behaviour dependent on the order in which the addresses were given.
4141 I have moved the duplicate checking until after the routing is complete.
4142 Exim scans the addresses that are assigned to local and remote
4143 transports, and removes any duplicates. This means that more work will be
4144 done, as duplicates will always all be routed, but duplicates are
4145 presumably rare, so I don't expect this is of any significance.
4147 For deliveries to pipes, files, and autoreplies, the duplicate checking
4148 still happens during the routing process, since they are not going to be
4151 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4152 It corrects a timeout issue with spamd. This is Ian's comment: "The
4153 background is that sometimes spamd either never reads data from a
4154 connection it has accepted, or it never writes response data. The exiscan
4155 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4156 blindly assumes that writes won't block so it may never time out."
4158 PH/39 Allow G after quota size as well as K and M.
4160 PH/40 The value set for $authenticated_id in an authenticator may not contain
4161 binary zeroes or newlines because the value is written to log lines and
4162 to spool files. There was no check on this. Now the value is run through
4163 the string_printing() function so that such characters are converted to
4164 printable escape sequences.
4166 PH/41 $message_linecount is a new variable that contains the total number of
4167 lines in the message. Compare $body_linecount, which is the count for the
4170 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4171 bounce and delay warning messages, except in certain special cases, which
4174 (a) An SMTP error message from a remote host;
4175 (b) A message specified in a :fail: redirection;
4176 (c) A message specified in a "fail" command in a system filter;
4177 (d) A message specified in a FAIL return from the queryprogram router;
4178 (e) A message specified by the cannot_route_message router option.
4180 In these cases only, Exim does include the error details in bounce and
4181 warning messages. There are also a few cases where bland messages such
4182 as "unrouteable address" or "local delivery error" are given.
4184 PH/43 $value is now also set for the "else" part of a ${run expansion.
4186 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4187 being worked on, but at least Exim now implements the latest version to
4190 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4191 process to complete was treated as a delivery failure, a timeout while
4192 writing the message to the pipe was logged, but erroneously treated as a
4193 successful delivery. Such timeouts include transport filter timeouts. For
4194 consistency with the overall process timeout, these timeouts are now
4195 treated as errors, giving rise to delivery failures by default. However,
4196 there is now a new Boolean option for the pipe transport called
4197 timeout_defer, which, if set TRUE, converts the failures into defers for
4198 both kinds of timeout. A transport filter timeout is now identified in
4201 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4202 systems where "make" and "gmake" are different, calling "gmake" at top
4203 level broke things. I've arranged for the value of $(MAKE) to be passed
4204 from the Makefile to this script so that it can call the same version of
4208 A note about Exim versions 4.44 and 4.50
4209 ----------------------------------------
4211 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4212 changes of various kinds. As a consequence, a big documentation update was
4213 needed. This delayed the release for rather longer than seemed good, especially
4214 in the light of a couple of (minor) security issues. Therefore, the changes
4215 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4216 release. So 4.44 and 4.50 are in effect two different branches that both start
4219 I have left the 4.50 change log unchanged; it contains all the changes since
4220 4.43. The change log for 4.44 is below; many of its items are identical to
4221 those for 4.50. This seems to be the most sensible way to preserve the
4222 historical information.
4228 1. Minor wording change to the doc/README.SIEVE file.
4230 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4231 computation of the current number of files was incorrect.
4233 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4234 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4237 4. Give more explanation in the error message when the command for a transport
4238 filter fails to execute.
4240 5. There are several places where Exim runs a non-Exim command in a
4241 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4242 was being done only for the command run by the queryprogram router. It is
4243 now done for all such subprocesses. The other cases are: ${run, transport
4244 filters, and the commands run by the lmtp and pipe transports.
4246 6. Added CONFIGURE_GROUP build-time option.
4248 7. Some older OS have a limit of 256 on the maximum number of file
4249 descriptors. Exim was using setrlimit() to set 1000 as a large value
4250 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4251 systems. I've change it so that if it can't get 1000, it tries for 256.
4253 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4254 was an oversight, and furthermore, ever since the addition of extra
4255 controls (e.g. 4.43/32), the checks on when to allow different forms of
4256 "control" were broken. There should now be diagnostics for all cases when a
4257 control that does not make sense is encountered.
4259 9. Added the /retain_sender option to "control=submission".
4261 10. $recipients is now available in the predata ACL (oversight).
4263 11. Tidy the search cache before the fork to do a delivery from a message
4264 received from the command line. Otherwise the child will trigger a lookup
4265 failure and thereby defer the delivery if it tries to use (for example) a
4266 cached ldap connection that the parent has called unbind on.
4268 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4269 of $address_data from the recipient verification was clobbered by the
4270 sender verification.
4272 13. The value of address_data from a sender verification is now available in
4273 $sender_address_data in subsequent conditions in the ACL statement.
4275 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4277 15. Added a new option "connect=<time>" to callout options, to set a different
4280 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4281 was its contents. (It was OK if the option was not defined at all.)
4283 17. A "Completed" log line is now written for messages that are removed from
4284 the spool by the -Mrm option.
4286 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4287 information about exactly what failed.
4289 19. Added -dd to debug only the daemon process.
4291 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4292 handles timeouts, both on the server side and network timeouts. Renamed the
4293 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4295 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4296 It is now set to "smtps".
4298 22. $host_address is now set to the target address during the checking of
4299 ignore_target_hosts.
4301 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4302 being passed; this would have caused $sender_host_name to have been used if
4303 matching the list had actually called for a host name (not very likely,
4304 since this list is usually IP addresses). A host name is now passed as
4307 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4308 code that specifies a non-restarting handler (typically sigaction() in
4309 modern systems) in an attempt to fix a rare and obscure crash bug.
4311 25. Narrowed the window for a race in the daemon that could cause it to ignore
4312 SIGCHLD signals. This is not a major problem, because they are used only to
4313 wake it up if nothing else does.
4315 26. A malformed maildirsize file could cause Exim to calculate negative values
4316 for the mailbox size or file count. Odd effects could occur as a result.
4317 The maildirsize information is now recalculated if the size or filecount
4320 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4321 support for a long time. Removed HAVE_SYS_VFS_H.
4323 28. Installed the latest version of exipick from John Jetmore.
4325 29. In an address list, if the pattern was not a regular expression, an empty
4326 subject address (from a bounce message) matched only if the pattern was an
4327 empty string. Non-empty patterns were not even tested. This was the wrong
4328 because it is perfectly reasonable to use an empty address as part of a
4329 database query. An empty address is now tested by patterns that are
4330 lookups. However, all the other forms of pattern expect the subject to
4331 contain a local part and a domain, and therefore, for them, an empty
4332 address still always fails if the pattern is not itself empty.
4334 30. Exim went into a mad DNS loop when attempting to do a callout where the
4335 host was specified on an smtp transport, and looking it up yielded more
4336 than one IP address.
4338 31. Re-factored the code for checking spool and log partition space into a
4339 function that finds that data and another that does the check. The former
4340 is then used to implement four new variables: $spool_space, $log_space,
4341 $spool_inodes, and $log_inodes.
4343 32. The RFC2047 encoding function was originally intended for short strings
4344 such as real names; it was not keeping to the 75-character limit for
4345 encoded words that the RFC imposes. It now respects the limit, and
4346 generates multiple encoded words if necessary. To be on the safe side, I
4347 have increased the buffer size for the ${rfc2047: expansion operator from
4350 33. It is now permitted to omit both strings after an "if" condition; if the
4351 condition is true, the result is "true". As before, when the second string
4352 is omitted, a false condition yields an empty string. This makes it less
4353 cumbersome to write custom ACL and router conditions.
4355 34. Failure to deliver a bounce message always caused it to be frozen, even if
4356 there was an errors_to setting on the router. The errors_to setting is now
4359 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4360 canonical form (fully expanded) before being placed in
4361 $sender_host_address.
4363 36. The table in the code that translates DNS record types into text (T_A to
4364 "A" for instance) was missing entries for NS and CNAME. It is just possible
4365 that this could have caused confusion if both these types were looked up
4366 for the same domain, because the text type is used as part of Exim's
4367 per-process caching. But the chance of anyone hitting this buglet seems
4370 37. The dnsdb lookup has been extended in a number of ways.
4372 (1) There is a new type, "zns", which walks up the domain tree until it
4373 finds some nameserver records. It should be used with care.
4375 (2) There is a new type, "mxh", which is like "mx" except that it returns
4376 just the host names, not the priorities.
4378 (3) It is now possible to give a list of domains (or IP addresses) to be
4379 looked up. The behaviour when one of the lookups defers can be
4380 controlled by a keyword.
4382 (4) It is now possible to specify the separator character for use when
4383 multiple records are returned.
4385 38. The dnslists ACL condition has been extended: it is now possible to supply
4386 a list of IP addresses and/or domains to be looked up in a particular DNS
4389 39. Added log_selector=+queue_time_overall.
4391 40. When running the queue in the test harness, wait just a tad after forking a
4392 delivery process, to get repeatability of debugging output.
4394 41. Include certificate and key file names in error message when GnuTLS fails
4395 to set them up, because the GnuTLS error message doesn't include the name
4396 of the failing file when there is a problem reading it.
4398 42. Allow both -bf and -bF in the same test run.
4400 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4402 44. The "Exiscan patch" is now merged into the mainline Exim source.
4404 45. Sometimes the final signoff response after QUIT could fail to get
4405 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4406 < 0 before doing a fflush(). This bug looks as though it goes back to the
4407 introduction of TLS in release 3.20, but "sometimes" must have been rare
4408 because the tests only now provoked it.
4410 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4411 (this can affect the format of dates).
4413 47. exim_tidydb, when checking for the continued existence of a message for
4414 which it has found a message-specific retry record, was not finding
4415 messages that were in split spool directories. Consequently, it was
4416 deleting retry records that should have stayed in existence.
4418 48. Steve fixed some bugs in eximstats.
4420 49. The SPA authentication driver was not abandoning authentication and moving
4421 on to the next authenticator when an expansion was forced to fail,
4422 contradicting the general specification for all authenticators. Instead it
4423 was generating a temporary error. It now behaves as specified.
4425 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4426 (the order specifies the preference for clients). The order is now AES256,
4427 AES128, 3DES, ARCFOUR128.
4429 51. Small patch to Sieve code - explicitly set From: when generating an
4432 52. Exim crashed if a remote delivery caused a very long error message to be
4433 recorded - for instance if somebody sent an entire SpamAssassin report back
4434 as a large number of 550 error lines. This bug was coincidentally fixed by
4435 increasing the size of one of Exim's internal buffers (big_buffer) that
4436 happened as part of the Exiscan merge. However, to be on the safe side, I
4437 have made the code more robust (and fixed the comments that describe what
4440 53. Now that there can be additional text after "Completed" in log lines (if
4441 the queue_time_overall log selector is set), a one-byte patch to exigrep
4442 was needed to allow it to recognize "Completed" as not the last thing in
4445 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4446 patch that reportedly fixes this has been added. I am not expert enough to
4447 create a test for it. This is what the patch creator wrote:
4449 "I found a little strange behaviour of ldap code when working with
4450 Windows 2003 AD Domain, where users was placed in more than one
4451 Organization Units. When I tried to give exim partial DN, the exit code
4452 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4453 But simultaneously result of request was absolutely normal ldap result,
4454 so I produce this patch..."
4456 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4457 REFERENCE, so I have modified the code to exclude the patch when that macro
4460 55. Some experimental protocols are using DNS PTR records for new purposes. The
4461 keys for these records are domain names, not reversed IP addresses. The
4462 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4463 leaves it alone. Component reversal etc. now happens only for IP addresses.
4466 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4468 57. Double the size of the debug message buffer (to 2048) so that more of very
4469 long debug lines gets shown.
4471 58. The exicyclog utility now does better if the number of log files to keep
4472 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4474 59. Two changes related to the smtp_active_hostname option:
4476 (1) $smtp_active_hostname is now available as a variable.
4477 (2) The default for smtp_banner uses $smtp_active_hostname instead
4478 of $primary_hostname.
4480 60. The host_aton() function is supposed to be passed a string that is known
4481 to be a valid IP address. However, in the case of IPv6 addresses, it was
4482 not checking this. This is a hostage to fortune. Exim now panics and dies
4483 if the condition is not met. A case was found where this could be provoked
4484 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4485 components; fortuitously, this particular loophole had already been fixed
4486 by change 4.50/55 above.
4488 If there are any other similar loopholes, the new check in host_aton()
4489 itself should stop them being exploited. The report I received stated that
4490 data on the command line could provoke the exploit when Exim was running as
4491 exim, but did not say which command line option was involved. All I could
4492 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4493 running as the user.
4496 61. There was a buffer overflow vulnerability in the SPA authentication code
4497 (which came originally from the Samba project). I have added a test to the
4498 spa_base64_to_bits() function which I hope fixes it.
4501 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4502 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4504 63. The daemon start-up calls getloadavg() while still root for those OS that
4505 need the first call to be done as root, but it missed one case: when
4506 deliver_queue_load_max is set with deliver_drop_privilege. This is
4507 necessary for the benefit of the queue runner, because there is no re-exec
4508 when deliver_drop_privilege is set.
4510 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4511 This has been fixed.
4513 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4514 was in use, was not putting the data itself into the right store pool;
4515 consequently, it could be overwritten for a subsequent message in the same
4516 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4519 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4521 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4522 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4523 the function were treating the return as a boolean value, which happened to
4524 work because 0=false and not-0=true, but is not correct code.
4526 68. The host_aton() function was not handling scoped IPv6 addresses (those
4527 with, for example, "%eth0" on the end) correctly.
4529 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4530 negated items (that is, ~something) in unsigned ints. Some compilers
4531 apparently mutter when there is no cast.
4533 70. If an address verification called from an ACL failed, and did not produce a
4534 user-specific message (i.e. there was only a "system" message), nothing was
4535 put in $acl_verify_message. In this situation, it now puts the system
4538 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4539 session; change 4.31/43 added the unwanted input to the log line - except
4540 that it did not do this in the start of session case. It now does.
4542 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4543 This is wrong and can cause the other end to generate a synchronization
4544 error if it is another Exim or anything else that does the synchronization
4545 check. A QUIT command is no longer sent after a timeout.
4547 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4548 during host lookups.
4550 74. The defer_ok option of callout verification was not working if it was used
4551 when verifying addresses in header lines, that is, for this case:
4553 verify = header_sender/callout=defer_ok
4555 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4556 those file descriptors could be used for SMTP connections. If anything
4557 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4558 could be sent to the SMTP client, causing chaos. The daemon now opens
4559 stdin, stdout, and stderr to /dev/null when it puts itself into the
4562 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4563 log by default. The user can override this with suitable Perl magic.
4565 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4566 the log message when discard triggers, was not working for the DATA ACL or
4567 for the non-SMTP ACL.
4569 78. Error message wording change in sieve.c.
4571 79. If smtp_accept_max_per_host was set, the number of connections could be
4572 restricted to fewer than expected, because the daemon was trying to set up
4573 a new connection before checking whether the processes handling previous
4574 connections had finished. The check for completed processes is now done
4575 earlier. On busy systems, this bug wouldn't be noticed because something
4576 else would have woken the daemon, and it would have reaped the completed
4579 80. If a message was submitted locally by a user whose login name contained one
4580 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4581 It caused a spool format error. I have fixed the spool reading code. A
4582 related problem was that the "from" clause in the Received: line became
4583 illegal because of the space(s). It is now covered by ${quote_local_part.
4585 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4588 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4590 83. Patch from Sieve maintainer to fix unterminated string problem in
4591 "vacation" handling.
4593 84. Some minor changes to the Linux configuration files to help with other
4594 OS variants using glibc.
4596 85. One more patch for Sieve to update vacation handling to latest spec.
4599 ----------------------------------------------------
4600 See the note above about the 4.44 and 4.50 releases.
4601 ----------------------------------------------------
4607 1. Change 4.43/35 introduced a bug that caused file counts to be
4608 incorrectly computed when quota_filecount was set in an appendfile
4611 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4612 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4615 3. Give more explanation in the error message when the command for a transport
4616 filter fails to execute.
4618 4. There are several places where Exim runs a non-Exim command in a
4619 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4620 was being done only for the command run by the queryprogram router. It is
4621 now done for all such subprocesses. The other cases are: ${run, transport
4622 filters, and the commands run by the lmtp and pipe transports.
4624 5. Some older OS have a limit of 256 on the maximum number of file
4625 descriptors. Exim was using setrlimit() to set 1000 as a large value
4626 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4627 systems. I've change it so that if it can't get 1000, it tries for 256.
4629 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4630 was an oversight, and furthermore, ever since the addition of extra
4631 controls (e.g. 4.43/32), the checks on when to allow different forms of
4632 "control" were broken. There should now be diagnostics for all cases when a
4633 control that does not make sense is encountered.
4635 7. $recipients is now available in the predata ACL (oversight).
4637 8. Tidy the search cache before the fork to do a delivery from a message
4638 received from the command line. Otherwise the child will trigger a lookup
4639 failure and thereby defer the delivery if it tries to use (for example) a
4640 cached ldap connection that the parent has called unbind on.
4642 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4643 of $address_data from the recipient verification was clobbered by the
4644 sender verification.
4646 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4647 was its contents. (It was OK if the option was not defined at all.)
4649 11. A "Completed" log line is now written for messages that are removed from
4650 the spool by the -Mrm option.
4652 12. $host_address is now set to the target address during the checking of
4653 ignore_target_hosts.
4655 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4656 being passed; this would have caused $sender_host_name to have been used if
4657 matching the list had actually called for a host name (not very likely,
4658 since this list is usually IP addresses). A host name is now passed as
4661 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4662 code that specifies a non-restarting handler (typically sigaction() in
4663 modern systems) in an attempt to fix a rare and obscure crash bug.
4665 15. Narrowed the window for a race in the daemon that could cause it to ignore
4666 SIGCHLD signals. This is not a major problem, because they are used only to
4667 wake it up if nothing else does.
4669 16. A malformed maildirsize file could cause Exim to calculate negative values
4670 for the mailbox size or file count. Odd effects could occur as a result.
4671 The maildirsize information is now recalculated if the size or filecount
4674 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4675 support for a long time. Removed HAVE_SYS_VFS_H.
4677 18. Updated exipick to current release from John Jetmore.
4679 19. Allow an empty sender to be matched against a lookup in an address list.
4680 Previously the only cases considered were a regular expression, or an
4683 20. Exim went into a mad DNS lookup loop when doing a callout where the
4684 host was specified on the transport, if the DNS lookup yielded more than
4687 21. The RFC2047 encoding function was originally intended for short strings
4688 such as real names; it was not keeping to the 75-character limit for
4689 encoded words that the RFC imposes. It now respects the limit, and
4690 generates multiple encoded words if necessary. To be on the safe side, I
4691 have increased the buffer size for the ${rfc2047: expansion operator from
4694 22. Failure to deliver a bounce message always caused it to be frozen, even if
4695 there was an errors_to setting on the router. The errors_to setting is now
4698 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4699 canonical form (fully expanded) before being placed in
4700 $sender_host_address.
4702 24. Updated eximstats to version 1.33
4704 25. Include certificate and key file names in error message when GnuTLS fails
4705 to set them up, because the GnuTLS error message doesn't include the name
4706 of the failing file when there is a problem reading it.
4708 26. Expand error message when OpenSSL has problems setting up cert/key files.
4711 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4712 (this can affect the format of dates).
4714 28. exim_tidydb, when checking for the continued existence of a message for
4715 which it has found a message-specific retry record, was not finding
4716 messages that were in split spool directories. Consequently, it was
4717 deleting retry records that should have stayed in existence.
4719 29. eximstats updated to version 1.35
4720 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4721 1.35 - bugfix such that pie charts by volume are generated correctly
4723 30. The SPA authentication driver was not abandoning authentication and moving
4724 on to the next authenticator when an expansion was forced to fail,
4725 contradicting the general specification for all authenticators. Instead it
4726 was generating a temporary error. It now behaves as specified.
4728 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4729 (the order specifies the preference for clients). The order is now AES256,
4730 AES128, 3DES, ARCFOUR128.
4732 31. Small patch to Sieve code - explicitly set From: when generating an
4735 32. Exim crashed if a remote delivery caused a very long error message to be
4736 recorded - for instance if somebody sent an entire SpamAssassin report back
4737 as a large number of 550 error lines. This bug was coincidentally fixed by
4738 increasing the size of one of Exim's internal buffers (big_buffer) that
4739 happened as part of the Exiscan merge. However, to be on the safe side, I
4740 have made the code more robust (and fixed the comments that describe what
4743 33. Some experimental protocols are using DNS PTR records for new purposes. The
4744 keys for these records are domain names, not reversed IP addresses. The
4745 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4746 leaves it alone. Component reversal etc. now happens only for IP addresses.
4749 34. The host_aton() function is supposed to be passed a string that is known
4750 to be a valid IP address. However, in the case of IPv6 addresses, it was
4751 not checking this. This is a hostage to fortune. Exim now panics and dies
4752 if the condition is not met. A case was found where this could be provoked
4753 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4754 components; fortuitously, this particular loophole had already been fixed
4755 by change 4.50/55 or 4.44/33 above.
4757 If there are any other similar loopholes, the new check in host_aton()
4758 itself should stop them being exploited. The report I received stated that
4759 data on the command line could provoke the exploit when Exim was running as
4760 exim, but did not say which command line option was involved. All I could
4761 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4762 running as the user.
4765 35. There was a buffer overflow vulnerability in the SPA authentication code
4766 (which came originally from the Samba project). I have added a test to the
4767 spa_base64_to_bits() function which I hope fixes it.
4770 36. The daemon start-up calls getloadavg() while still root for those OS that
4771 need the first call to be done as root, but it missed one case: when
4772 deliver_queue_load_max is set with deliver_drop_privilege. This is
4773 necessary for the benefit of the queue runner, because there is no re-exec
4774 when deliver_drop_privilege is set.
4776 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4777 was in use, was not putting the data itself into the right store pool;
4778 consequently, it could be overwritten for a subsequent message in the same
4779 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4782 38. Sometimes the final signoff response after QUIT could fail to get
4783 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4784 < 0 before doing a fflush(). This bug looks as though it goes back to the
4785 introduction of TLS in release 3.20, but "sometimes" must have been rare
4786 because the tests only now provoked it.
4792 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4793 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4794 Now it returns an integer. A number of calls were still expecting a T/F
4795 return. Fortuitously, in all cases, the tests worked in OK situations,
4796 which is the norm. However, things would have gone wrong on any write
4797 failures on the smtp file descriptor. This function is used when sending
4798 messages over SMTP and also when doing verify callouts.
4800 2. When Exim is called to do synchronous delivery of a locally submitted
4801 message (the -odf or -odi options), it no longer closes stderr before doing
4804 3. Implemented the mua_wrapper option.
4806 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4808 5. Implemented the functions header_remove(), header_testname(),
4809 header_add_at_position(), and receive_remove_recipient(), and exported them
4812 6. If an ACL "warn" statement specified the addition of headers, Exim already
4813 inserted X-ACL-Warn: at the start if there was no header name. However, it
4814 was not making this test for the second and subsequent header lines if
4815 there were newlines in the string. This meant that an invalid header could
4816 be inserted if Exim was badly configured.
4818 7. Allow an ACL "warn" statement to add header lines at the start or after all
4819 the Received: headers, as well as at the end.
4821 8. Added the rcpt_4xx retry error code.
4823 9. Added postmaster_mailfrom=xxx to callout verification option.
4825 10. Added mailfrom=xxxx to the callout verification option, for verify=
4828 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4829 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4831 12. Inserted some casts to stop certain compilers warning when using pointer
4832 differences as field lengths or precisions in printf-type calls (mostly
4833 affecting debugging statements).
4835 13. Added optional readline() support for -be (dynamically loaded).
4837 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4838 same clock tick as a message's arrival, so that its received time was the
4839 same as the "first fail" time on the retry record, and that message
4840 remained on the queue past the ultimate address timeout, every queue runner
4841 would try a delivery (because it was past the ultimate address timeout) but
4842 after another failure, the ultimate address timeout, which should have then
4843 bounced the address, did not kick in. This was a "< instead of <=" error;
4844 in most cases the first failure would have been in the next clock tick
4845 after the received time, and all would be well.
4847 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4848 being recognized when the domain list was tested by the match_domain
4849 condition in an expansion string.
4851 16. Added the ${str2b64: operator.
4853 17. Exim was always calling setrlimit() to set a large limit for the number of
4854 processes, without checking whether the existing limit was already
4855 adequate. (It did check for the limit on file descriptors.) Furthermore,
4856 errors from getrlimit() and setrlimit() were being ignored. Now they are
4857 logged to the main and panic logs, but Exim does carry on, to try to do its
4858 job under whatever limits there are.
4860 18. Imported PCRE 5.0.
4862 19. Trivial typo in log message " temporarily refused connection" (the leading
4865 20. If the log selector return_path_on_delivery was set and an address was
4866 redirected to /dev/null, the delivery process crashed because it assumed
4867 that a return path would always be set for a "successful" delivery. In this
4868 case, the whole delivery is bypassed as an optimization, and therefore no
4871 21. Internal re-arrangement: the function for sending a challenge and reading
4872 a response while authentication was assuming a zero-terminated challenge
4873 string. It's now changed to take a pointer and a length, to allow for
4874 binary data in such strings.
4876 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4878 23. Exim was not respecting finduser_retries when seeking the login of the
4879 uid under which it was called; it was always trying 10 times. (The default
4880 setting of finduser_retries is zero.) Also, it was sleeping after the final
4881 failure, which is pointless.
4883 24. Implemented tls_on_connect_ports.
4885 25. Implemented acl_smtp_predata.
4887 26. If the domain in control=submission is set empty, Exim assumes that the
4888 authenticated id is a complete email address when it generates From: or
4889 Sender: header lines.
4891 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4892 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4893 chown and chgrp in /bin and hostname in /usr/bin.
4895 28. Exim was keeping the "process log" file open after each use, just as it
4896 does for the main log. This opens the possibility of it remaining open for
4897 long periods when the USR1 signal hits a daemon. Occasional processlog
4898 errors were reported, that could have been caused by this. Anyway, it seems
4899 much more sensible not to leave this file open at all, so that is what now
4902 29. The long-running daemon process does not normally write to the log once it
4903 has entered its main loop, and it closes the log before doing so. This is
4904 so that log files can straightforwardly be renamed and moved. However,
4905 there are a couple of unusual error situations where the daemon does write
4906 log entries, and I had neglected to close the log afterwards.
4908 30. The text of an SMTP error response that was received during a remote
4909 delivery was being truncated at 512 bytes. This is too short for some of
4910 the long messages that one sometimes sees. I've increased the limit to
4913 31. It is now possible to make retry rules that apply only when a message has a
4914 specific sender, in particular, an empty sender.
4916 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4917 it possible to be selective about when SMTP synchronization is enforced.
4919 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4921 32. Implemented hosts_connection_nolog.
4923 33. Added an ACL for QUIT.
4925 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4928 35. Added mailbox_size and mailbox_filecount to appendfile.
4930 36. Added control = no_multiline_responses to ACLs.
4932 37. There was a bug in the logic of the code that waits for the clock to tick
4933 in the case where the clock went backwards by a substantial amount such
4934 that the microsecond fraction of "now" was more than the microsecond
4935 fraction of "then" (but the whole seconds number was less).
4937 38. Added support for the libradius Radius client library this is found on
4938 FreeBSD (previously only the radiusclient library was supported).
4944 1. When certain lookups returned multiple values in the form name=value, the
4945 quoting of the values was not always being done properly. Specifically:
4946 (a) If the value started with a double quote, but contained no whitespace,
4948 (b) If the value contained whitespace other than a space character (i.e.
4949 tabs or newlines or carriage returns) it was not quoted.
4950 This fix has been applied to the mysql and pgsql lookups by writing a
4951 separate quoting function and calling it from the lookup code. The fix
4952 should probably also be applied to nisplus, ibase and oracle lookups, but
4953 since I cannot test any of those, I have not disturbed their existing code.
4955 2. A hit in the callout cache for a specific address caused a log line with no
4956 reason for rejecting RCPT. Now it says "Previous (cached) callout
4957 verification failure".
4959 3. There was an off-by-one bug in the queryprogram router. An over-long
4960 return line was truncated at 256 instead of 255 characters, thereby
4961 overflowing its buffer with the terminating zero. As well as fixing this, I
4962 have increased the buffer size to 1024 (and made a note to document this).
4964 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4965 when Exim is waiting for an SMTP response from a remote server, Exim
4966 restarts its select() call on the socket, thereby resetting its timeout.
4967 This is not a problem when such interrupts are rare. Somebody set up a cron
4968 job to run exiwhat every 2 minutes, which is less than the normal select()
4969 timeout (5 or 10 minutes). This meant that the select() timeout never
4970 kicked in because it was always reset. I have fixed this by comparing the
4971 time when an interrupt arrives with the time at the start of the first call
4972 to select(). If more time than the timeout has elapsed, the interrupt is
4973 treated as a timeout.
4975 5. Some internal re-factoring in preparation for the addition of Sieve
4976 extensions (by MH). In particular, the "personal" test is moved to a
4977 separate function, and given an option for scanning Cc: and Bcc: (which is
4978 not set for Exim filters).
4980 6. When Exim created an email address using the login of the caller as the
4981 local part (e.g. when creating a From: or Sender: header line), it was not
4982 quoting the local part when it contained special characters such as @.
4984 7. Installed new OpenBSD configuration files.
4986 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4987 try to make them clearer.
4989 9. Callout options, other than the timeout value, were being ignored when
4990 verifying sender addresses in header lines. For example, when using
4992 verify = header_sender/callout=no_cache
4994 the cache was (incorrectly) being used.
4996 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4997 only the Cygwin environment.
4999 11. When return_path_on_delivery was set as a log selector, if different remote
5000 addresses in the same message used different return paths and parallel
5001 remote delivery occurred, the wrong values would sometimes be logged.
5002 (Whenever a remote delivery process finished, the return path value from
5003 the most recently started remote delivery process was logged.)
5005 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5006 lines when AUTH and/or TLS are in use. This is the "received protocol"
5007 field. Exim used to use "asmtp" for authenticated SMTP, without any
5008 indication (in the protocol name) for TLS use. Now it follows the RFC and
5009 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5010 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5011 names appear in log lines as well as in Received: header lines.
5013 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5014 extensions, and comparison tests, and to fix some bugs.
5016 14. Changes to the "personal" filter test:
5018 (1) The test was buggy in that it was just doing the equivalent of
5019 "contains" tests on header lines. For example, if a user's address was
5020 anne@some.where, the "personal" test would incorrectly be true for
5022 To: susanne@some.where
5024 This test is now done by extracting each address from the header in turn,
5025 and checking the entire address. Other tests that are part of "personal"
5026 are now done using regular expressions (for example, to check local parts
5027 of addresses in From: header lines).
5029 (2) The list of non-personal local parts in From: addresses has been
5030 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5031 taken from the Sieve specification recommendations.
5033 (3) If the message contains any header line starting with "List-" it is
5034 treated as non-personal.
5036 (4) The test for "circular" in the Subject: header line has been removed
5037 because it now seems ill-conceived.
5039 15. Minor typos in src/EDITME comments corrected.
5041 16. Installed latest exipick from John Jetmore.
5043 17. If headers_add on a router specified a text string that was too long for
5044 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5045 of string_sprintf() is now avoided.
5047 18. $message_body_size was not set (it was always zero) when running the DATA
5048 ACL and the local_scan() function.
5050 19. For the "mail" command in an Exim filter, no default was being set for
5051 the once_repeat time, causing a random time value to be used if "once" was
5052 specified. (If the value happened to be <= 0, no repeat happened.) The
5053 default is now 0s, meaning "never repeat". The "vacation" command was OK
5054 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5055 (I found it when inspecting the code).
5057 20. There is now an overall timeout for performing a callout verification. It
5058 defaults to 4 times the callout timeout, which applies to individual SMTP
5059 commands during the callout. The overall timeout applies when there is more
5060 than one host that can be tried. The timeout is checked before trying the
5061 next host. This prevents very long delays if there are a large number of
5062 hosts and all are timing out (e.g. when the network connections are timing
5063 out). The value of the overall timeout can be changed by specifying an
5064 additional sub-option for "callout", called "maxwait". For example:
5066 verify = sender/callout=5s,maxwait=20s
5068 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5069 to the end before writing, but this should make it even safer).
5071 22. Exim was forgetting that it had advertised PIPELINING for the second and
5072 subsequent messages on an SMTP connection. It was also not resetting its
5073 memory on STARTTLS and an internal HELO.
5075 23. When Exim logs an SMTP synchronization error within a session, it now
5076 records whether PIPELINING has been advertised or not.
5078 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5079 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5080 rather than long int.
5082 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5084 26. Added the never_mail option to autoreply.
5090 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5091 crash if the getsockname() call failed; this can happen if a connection is
5092 closed very soon after it is established. The problem was simply in the
5093 order in which certain operations were done, causing Exim to try to write
5094 to the SMTP stream before it had set up the file descriptor. The bug has
5095 been fixed by making things happen in the correct order.
5101 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5102 before the connection was closed, thus losing the rejection response.
5104 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5105 some early Solaris releases, but causes trouble in current releases where
5106 socklen_t is defined.
5108 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5111 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5114 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5115 that releasing the top of it at the end releases what was used for sub-
5116 expansions (unless the block got too big). However, discard this block if
5117 the first thing is a variable or header, so that we can use its block when
5118 it is dynamic (useful for very large $message_headers, for example).
5120 6. Lookups now cache *every* query, not just the most recent. A new, separate
5121 store pool is used for this. It can be recovered when all lookup caches are
5122 flushed. Lookups now release memory at the end of their result strings.
5123 This has involved some general refactoring of the lookup sources.
5125 7. Some code has been added to the store_xxx() functions to reduce the amount
5126 of flapping under certain conditions.
5128 8. log_incoming_interface used to affect only the <= reception log lines. Now
5129 it causes the local interface and port to be added to several more SMTP log
5130 lines, for example "SMTP connection from", and rejection lines.
5132 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5134 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5136 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5137 could be overwritten at the end of the current message (or the start of a
5138 new message if it was set in a HELO ACL). The value is now preserved for
5139 the duration of the SMTP connection.
5141 12. If a transport had a headers_rewrite setting, and a matching header line
5142 contained an unqualified address, that address was qualified, even if it
5143 did not match any rewriting rules. The underlying bug was that the values
5144 of the flags that permit the existence of unqualified sender and recipient
5145 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5146 non-local messages, and by -bnq for local messages) were not being
5147 preserved with the message after it was received.
5149 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5150 "next input=" as part of the text comprising the host identity instead of
5151 the correct text. The code was using the same buffer for two different
5152 strings. However, depending on which order the printing function evaluated
5153 its arguments, the bug did not always show up. Under Linux, for example, my
5154 test suite worked just fine.
5156 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5157 doesn't work with some older versions of Perl. It has been changed to "my",
5158 which in any case is probably the better facility to use.
5160 15. A really picky compiler found some instances of statements for creating
5161 error messages that either had too many or two few arguments for the format
5164 16. The size of the buffer for calls to the DNS resolver has been increased
5165 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5166 for addresses that have a lot of PTR records. This alleviates a problem; it
5167 does not fully solve it.
5169 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5170 buffer now truncates the list and logs the incident, which is the same
5171 action as happens when Exim is looking up a host name and its aliases.
5172 Previously in this situation something unpredictable would happen;
5173 sometimes it was "internal error: store_reset failed".
5175 18. If a server dropped the connection unexpectedly when an Exim client was
5176 using GnuTLS and trying to read a response, the client delivery process
5177 crashed while trying to generate an error log message.
5179 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5180 string, for example:
5182 warn message = H1: something\nH2: something
5184 the text was added as a single header line from Exim's point of view
5185 though it ended up OK in the delivered message. However, searching for the
5186 second and subsequent header lines using $h_h2: did not work. This has been
5187 fixed. Similarly, if a system filter added multiple headers in this way,
5188 the routers could not see them.
5190 20. Expanded the error message when iplsearch is called with an invalid key to
5191 suggest using net-iplsearch in a host list.
5193 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5194 ACLs are no longer actually imposed (and a message to that effect is
5197 22. If a "gecos" field in a passwd entry contained escaped characters, in
5198 particular, if it contained a \" sequence, Exim got it wrong when building
5199 a From: or a Sender: header from that name. A second bug also caused
5200 incorrect handling when an unquoted " was present following a character
5201 that needed quoting.
5203 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5204 was not being matched caselessly.
5206 24. Arranged for all hyphens in the exim.8 source to be escaped with
5209 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5210 properly. Recipient callout cache records were still being keyed to include
5211 the sender, even when use_sender was set false. This led to far more
5212 callouts that were necessary. The sender is no longer included in the key
5213 when use_sender is false.
5215 26. Added "control = submission" modifier to ACLs.
5217 27. Added the ${base62d: operator to decode base 62 numbers.
5219 28. dnsdb lookups can now access SRV records.
5221 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5222 the configuration file.
5224 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5225 (-v) mode. This makes the output for a verbose queue run more intelligible.
5227 31. Added a use_postmaster feature to recipient callouts.
5229 32. Added the $body_zerocount variable, containing the number of binary zero
5230 bytes in the message body.
5232 33. The time of last modification of the "new" subdirectory is now used as the
5233 "mailbox time last read" when there is a quota error for a maildir
5236 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5238 35. Added +ignore_unknown as a special item in host lists.
5240 36. Code for decoding IPv6 addresses in host lists is now included, even if
5241 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5242 address was recognized as an IP address, but was then not correctly decoded
5243 into binary, causing unexpected and incorrect effects when compared with
5250 1. Very minor rewording of debugging text in manualroute to say "list of
5251 hosts" instead of "hostlist".
5253 2. If verify=header_syntax was set, and a header line with an unqualified
5254 address (no domain) and a large number of spaces between the end of the
5255 name and the colon was received, the reception process suffered a buffer
5256 overflow, and (when I tested it) crashed. This was caused by some obsolete
5257 code that should have been removed. The fix is to remove it!
5259 3. When running in the test harness, delay a bit after writing a bounce
5260 message to get a bit more predictability in the log output.
5262 4. Added a call to search_tidyup() just before forking a reception process. In
5263 theory, someone could use a lookup in the expansion of smtp_accept_max_
5264 per_host which, without the tidyup, could leave open a database connection.
5266 5. Added the variables $recipient_data and $sender_data which get set from a
5267 lookup success in an ACL "recipients" or "senders" condition, or a router
5268 "senders" option, similar to $domain_data and $local_part_data.
5270 6. Moved the writing of debug_print from before to after the "senders" test
5273 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5274 problems for message scanning, either using a data ACL, or using
5275 local_scan() because the Received: header was not generated till after they
5276 were called (in order to set the time as the time of reception completion).
5277 I have revised the way this works. The header is now generated after the
5278 body is received, but before the ACL or local_scan() are called. After they
5279 are run, the timestamp in the header is updated.
5285 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5286 before starting a queue runner without re-exec. This happened only when
5287 deliver_drop_privilege was set or when the Exim user was set to root. The
5288 effect of the bug was that timeouts during subsequent deliveries caused
5289 crashes instead of being properly handled. The handler is now left at its
5290 default (and expected) setting.
5292 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5293 message, again when deliver_drop_privilege is set or Exim is run as root.
5294 The bug described in (1) was not present in this case, but the tidying up
5295 of the other signals was missing. I have made the two cases consistent.
5297 3. The ignore_target_hosts setting on a manualroute router was being ignored
5298 for hosts that were looked up using the /MX notation.
5300 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5303 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5304 operated on the sender address. After changing the $sender_address to <>
5305 for the sender address verify, Exim was re-instated it as the original
5306 (before rewriting) address, but remembering that it had rewritten it, so it
5307 wasn't rewriting it again. This bug also had the effect of breaking the
5308 sender address verification caching when the sender address was rewritten.
5310 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5311 This has been changed so that if the ip literal address matches
5312 ignore_target_hosts, the router declines.
5314 7. Added expansion conditions match_domain, match_address, and match_local_
5315 part (NOT match_host).
5317 8. The placeholder for the Received: header didn't have a length field set.
5319 9. Added code to Exim itself and to exim_lock to test for a specific race
5320 condition that could lead to file corruption when using MBX delivery. The
5321 issue is with the lockfile that is created in /tmp. If this file is removed
5322 after a process has opened it but before that process has acquired a lock,
5323 there is the potential for a second process to recreate the file and also
5324 acquire a lock. This could lead to two Exim processes writing to the file
5325 at the same time. The added code performs the same test as UW imapd; it
5326 checks after acquiring the lock that its file descriptor still refers to
5327 the same named file.
5329 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5330 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5333 11. Added the smtp_active_hostname option. If used, this will typically be made
5334 to depend on the incoming interface address. Because $interface_address is
5335 not set up until the daemon has forked a reception process, error responses
5336 that can happen earlier (such as "too many connections") no longer contain
5339 12. If an expansion in a condition on a "warn" statement fails because a lookup
5340 defers, the "warn" statement is abandoned, and the next ACL statement is
5341 processed. Previously this caused the whole ACL to be aborted.
5343 13. Added the iplsearch lookup type.
5345 14. Added ident_timeout as a log selector.
5347 15. Added tls_certificate_verified as a log selector.
5349 16. Added a global option tls_require_ciphers (compare the smtp transport
5350 option of the same name). This controls incoming TLS connections.
5352 17. I finally figured out how to make tls_require_ciphers do a similar thing
5353 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5354 before starting the TLS session.
5356 18. Tabs are now shown as \t in -bP output.
5358 19. If the log selector return_path_on_delivery was set, Exim crashed when
5359 bouncing a message because it had too many Received: header lines.
5361 20. If two routers both had headers_remove settings, and the first one included
5362 a superfluous trailing colon, the final name in the first list and the
5363 first name in the second list were incorrectly joined into one item (with a
5364 colon in the middle).
5370 1. Added -C and -D options to the exinext utility, mainly to make it easier
5371 to include in the automated testing, but these could be helpful when
5372 multiple configurations are in use.
5374 2. The exinext utility was not formatting the output nicely when there was
5375 an alternate port involved in the retry record key, nor when there was a
5376 message id as well (for retries that were specific to a specific message
5377 and a specific host). It was also confused by IPv6 addresses, because of
5378 the additional colons they contain. I have fixed the IPv4 problem, and
5379 patched it up to do a reasonable job for IPv6.
5381 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5382 delivery, the log line now contains "pipelined" if PIPELINING was used.
5384 4. An SMTP transport process used to panic and die if the bind() call to set
5385 an explicit outgoing interface failed. This has been changed; it is now
5386 treated in the same way as a connect() failure.
5388 5. A reference to $sender_host_name in the part of a conditional expansion
5389 that was being skipped was still causing a DNS lookup. This no longer
5392 6. The def: expansion condition was not recognizing references to header lines
5393 that used bh_ and bheader_.
5395 7. Added the _cache feature to named lists.
5397 8. The code for checking quota_filecount in the appendfile transport was
5398 allowing one more file than it should have been.
5400 9. For compatibility with Sendmail, the command line option
5408 and sets the incoming protocol and host name (for trusted callers). The
5409 host name and its colon can be omitted when only the protocol is to be set.
5410 Note the Exim already has two private options, -pd and -ps, that refer to
5411 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5412 "s", but I don't think that's a major issue.
5414 10. A number of refactoring changes to the code, none of which should affect
5417 (a) The number of logging options was getting close to filling up the
5418 32-bit word that was used as a bit map. I have split them into two classes:
5419 those that are passed in the argument to log_write(), and those that are
5420 only ever tested independently outside of that function. These are now in
5421 separate 32-bit words, so there is plenty of room for expansion again.
5422 There is no change in the user interface or the logging behaviour.
5424 (b) When building, for example, log lines, the code previously used a
5425 macro that called string_cat() twice, in order to add two strings. This is
5426 not really sufficiently general. Furthermore, there was one instance where
5427 it was actually wrong because one of the argument was used twice, and in
5428 one call a function was used. (As it happened, calling the function twice
5429 did not affect the overall behaviour.) The macro has been replaced by a
5430 function that can join an arbitrary number of extra strings onto a growing
5433 (c) The code for expansion conditions now uses a table and a binary chop
5434 instead of a serial search (which was left over from when there were very
5435 few conditions). Also, it now recognizes conditions like "pam" even when
5436 the relevant support is not compiled in: a suitably worded error message is
5437 given if an attempt is made to use such a condition.
5439 11. Added ${time_interval:xxxxx}.
5441 12. A bug was causing one of the ddress fields not to be passed back correctly
5442 from remote delivery subprocesses. The field in question was not being
5443 subsequently used, so this caused to problems in practice.
5445 13. Added new log selectors queue_time and deliver_time.
5447 14. Might have fixed a bug in maildirsizefile handling that threw up
5448 "unexpected character" debug warnings, and recalculated the data
5449 unnecessarily. In any case, I expanded the warning message to give more
5452 15. Added the message "Restricted characters in address" to the statements in
5453 the default ACL that block characters like @ and % in local parts.
5455 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5456 Three changes have been made:
5458 (a) There was a serious bug; a negative response to MAIL caused the whole
5459 recipient domain to be cached as invalid, thereby blocking all messages
5460 to all local parts at the same domain, from all senders. This bug has
5461 been fixed. The domain is no longer cached after a negative response to
5462 MAIL if the sender used is not empty.
5464 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5467 (c) A new callout option, "use_sender" has been added for people who want
5468 the modified behaviour.
5474 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5477 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5478 indeed breaks things for older releases.
5480 3. Added additional logging to the case where there is a problem reading data
5481 from a filter that is running in a subprocess using a pipe, in order to
5482 try to track down a specific problem.
5484 4. Testing facility fudge: when running in the test harness and attempting
5485 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5486 getting "No route to host". Convert this to a timeout.
5488 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5491 6. Some OS don't have socklen_t but use size_t instead. This affects the
5492 fifth argument of getsockopt() amongst other things. This is now
5493 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5494 can be set for individual OS. I have set it for SunOS5, OSF1, and
5495 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5496 some earlier ones do not.
5498 7. Change 4.30/15 was not doing the test caselessly.
5500 8. The standard form for an IPv6 address literal was being rejected by address
5501 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5502 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5503 this, as well as the form without the "IPv6" on the front (but only when
5504 address literals are enabled, of course).
5506 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5508 10. Exim crashed if a message with an empty sender address specified by -f
5509 encountered a router with an errors_to setting. This could be provoked only
5510 by a command such as
5514 where an empty string was supplied; "<>" did not hit this bug.
5516 11. Installed PCRE release 4.5.
5518 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5519 remained set. It is now erased.
5521 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5522 times from message ids (which are base 36 rather than the normal 62).
5524 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5525 were being counted as actual protocol errors, and logged if the log
5526 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5527 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5528 and DATA following a set of rejected RCPTs do not count as protocol errors.
5529 In other words, Exim assumes they were pipelined, though this may not
5530 actually be the case. Of course, in all cases the client gets an
5531 appropriate error code.
5533 15. If a lookup fails in an ACL condition, a message about the failure may
5534 be available; it is used if testing the ACL cannot continue, because most
5535 such messages specify what the cause of the deferral is. However, some
5536 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5537 that caused an old message to be retained and used if a later statement
5538 caused a defer, replacing the real cause of the deferral.
5540 16. If an IP address had so many PTR records that the DNS lookup buffer
5541 was not large enough to hold them, Exim could crash while trying to process
5542 the truncated data. It now detects and logs this case.
5544 17. Further to 4.21/58, another change has been made: if (and only if) the
5545 first line of a message (the first header line) ends with CRLF, a bare LF
5546 in a subsequent header line has a space inserted after it, so as not to
5547 terminate the header.
5549 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5550 unnecessarily, used atoi() instead of strtol(), and didn't check the
5551 termination when getting file sizes from file names by regex.
5553 19. Completely re-implemented the support for maildirsize files, in the light
5554 of a number of problems with the previous contributed implementation
5555 (4.30/29). In particular:
5557 . If the quota is zero, the maildirsize file is maintained, but no quota is
5560 . If the maildir directory does not exist, it is created before any attempt
5561 to write a maildirsize file.
5563 . The quota value in the file is just a cache; if the quota is changed in
5564 the transport, the new value overrides.
5566 . A regular expression is available for excluding directories from the
5569 20. The autoreply transport checks the characters in options that define the
5570 message's headers; it allows continued headers, but it was checking with
5571 isspace() after an embedded newline instead of explicitly looking for a
5574 21. If all the "regular" hosts to which an address was routed had passed their
5575 expiry times, and had not reached their retry times, the address was
5576 bounced, even if fallback hosts were defined. Now Exim should go on to try
5579 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5580 equivalent code in the SMTP transport. Some hosts send humungous responses
5581 to HELO/EHLO, more than 1024 it seems.
5583 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5584 gives compiler warnings in some environments. I've now done it "properly",
5587 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5588 (because of problems with the built-in one) was declared to return uschar *
5589 instead of char *, causing compiler failure.
5591 25. Fixed a file descriptor leak when processing alias/forward files.
5593 26. Fixed a minor format string issue in dbfn.c.
5595 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5597 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5598 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5601 29. When a sender address is verified, it is cached, to save repeating the test
5602 when there is more than one recipient in a message. However, when the
5603 verification involves a callout, it is possible for different callout
5604 options to be set for different recipients. It is too complicated to keep
5605 track of this in the cache, so now Exim always runs a verification when a
5606 callout is required, relying on the callout cache for the optimization.
5607 The overhead is duplication of the address routing, but this should not be
5610 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5611 to be verified with callout=postmaster, and the main callout worked but the
5612 postmaster check failed, the verification correctly failed. However, if a
5613 subsequent RCPT command asked for sender verification *without* the
5614 postmaster check, incorrect caching caused this verification also to fail,
5617 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5618 it was not caching the DNS options (qualify_single, search_parents) that
5619 were used when the lookup failed. A subsequent lookup with different
5620 options therefore always gave the same answer, though there were cases
5621 where it should not have. (Example: a "domains = !$mx_any" option on a
5622 dnslookup router: the "domains" option is always processed without any
5623 widening, but the router might have qualify_single set.) Now Exim uses the
5624 cached value only when the same options are set.
5626 32. Added John Jetmore's "exipick" utility to the distribution.
5628 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5629 than a timeout (e.g. a certificate is required, and is not provided), an
5630 Exim server now closes the connection immediately. Previously it waited for
5631 the client to close - but if the client is SSL, it seems that they each
5632 wait for each other, leading to a delay before one of them times out.
5634 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5635 maintained 0.8.x compatibility because I don't think many are using it, and
5636 it is clearly obsolete.
5638 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5641 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5642 client certificate was expired. A simple patch fixes this, though I don't
5643 understand the full logic of why the verify callback is called multiple
5646 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5647 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5648 which causes problems with some clients (such as the Certicom SSL Plus
5649 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5650 disables the coutermeasure allowing Eudora to connect."
5652 38. Exim was not checking that a write() to a log file succeeded. This could
5653 lead to Bad Things if a log got too big, in particular if it hit a file
5654 size limit. Exim now panics and dies if it cannot write to a log file, just
5655 as it does if it cannot open a log file.
5657 39. Modified OS/Makefile-Linux so that it now contains
5659 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5661 The two -D definitions ensure that Exim is compiled with large file
5662 support, which makes it possible to handle log files that are bigger than
5665 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5666 instance) a domain was checked against a named list that involved a lookup,
5667 causing $domain_data to be set, then another domain was checked against the
5668 same list, then the first domain was re-checked, the value of $domain_data
5669 after the final check could be wrong. In particular, if the second check
5670 failed, it could be set empty. This bug probably also applied to
5673 41. The strip_trailing_dot option was not being applied to the address given
5674 with the -f command-line option.
5676 42. The code for reading a message's header from the spool was incrementing
5677 $received_count, but never initializing it. This meant that the value was
5678 incorrect (doubled) while delivering a message in the same process in which
5679 it was received. In the most common configuration of Exim, this never
5680 happens - a fresh exec is done - but it can happen when
5681 deliver_drop_privilege is set.
5683 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5684 it now includes up to 150 characters of the unexpected data in the log
5687 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5688 and building data strings. The size of both of these buffers was 10 000
5689 bytes - far larger than anybody would *ever* want, thought I. Needless to
5690 say, somebody hit the limit. I have increased the maximum line length to
5691 20 000 and the maximum data length of concatenated lines to 100 000. I have
5692 also fixed two bugs, because there was no checking on these buffers. Tsk,
5693 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5694 buffer is too small.
5696 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5697 lsearch lookups. Now it does.
5699 46. When parsing a route_list item in a manualroute router, a fixed-length
5700 buffer was used for the list of hosts. I made this 1024 bytes long,
5701 thinking that nobody would ever have a list of hosts that long. Wrong.
5702 Somebody had a whole pile of complicated expansion conditions, and the
5703 string was silently truncated, leading to an expansion error. It turns out
5704 that it is easier to change to an unlimited length (owing to other changes
5705 that have happened since this code was originally written) than to build
5706 structure for giving a limitation error. The length of the item that
5707 expands into the list of hosts is now unlimited.
5709 47. The lsearch lookup could not handle data where the length of text line was
5710 more than 4095 characters. Such lines were truncated, leading to shortened
5711 data being returned. It should now handle lines of any length.
5713 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5714 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5717 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5718 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5719 no longer happens in comment lines. A list of replacements is now placed
5720 at the head of all of the source files, except those whose only change is
5721 to replace PERL_COMMAND in the very first #! line.
5723 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5724 messages on the queue, with a bottom-up merge sort, using code contributed
5725 by Michael Haardt. This should make operations like -bp somewhat faster on
5726 large queues. It won't affect queue runners, except when queue_run_in_order
5729 51. Installed eximstats 1.31 in the distribution.
5731 52. Added support for SRV lookups to the dnslookup router.
5733 53. If an ACL referred to $message_body or $message_body_end, the value was not
5734 reset for any messages that followed in the same SMTP session.
5736 54. The store-handling optimization for building very long strings was not
5737 differentiating between the different store pools. I don't think this
5738 actually made any difference in practice, but I've tidied it.
5740 55. While running the routers to verify a sender address, $sender_address
5741 was still set to the sender address. This is wrong, because when routing to
5742 send a bounce to the sender, it would be empty. Therefore, I have changed
5743 it so that, while verifying a sender address, $sender_address is set to <>.
5744 (There is no change to what happens when verifying a recipient address.)
5746 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5747 target A or AAAA records (if not already returned) without resetting the
5748 qualify_single or search_parents options of the DNS resolver. These are
5749 inappropriate in this case because the targets of MX and SRV records must
5750 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5751 target that, when qualified, matched something in the local domain. These
5752 two options are now turned off when doing these lookups.
5754 57. It seems that at least some releases of Reiserfs (which does not have the
5755 concept of a fixed number of inodes) returns zero and not -1 for the
5756 number of available inodes. This interacted badly with check_spool_inodes,
5757 which assumed that -1 was the "no such thing" setting. What I have done is
5758 to check that the total number of inodes is greater than zero before doing
5759 the test of how many are available.
5761 58. When a "warn" ACL statement has a log_message modifier, the message is
5762 remembered, and not repeated. This is to avoid a lot of repetition when a
5763 message has many recipients that cause the same warning to be written.
5764 However, Exim was preserving the list of already written lines for an
5765 entire SMTP session, which doesn't seem right. The memory is now reset if a
5766 new message is started.
5768 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5769 debugging output unless log_rewrite was also set.
5771 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5772 of (int)(handle) when we know that handle contains (void *)(-1).
5774 61. The Exim daemon panic-logs an error return when it closes the incoming
5775 connection. However "connection reset by peer" seems to be common, and
5776 isn't really an error worthy of noting specially, so that particular error
5779 62. When Exim is trying to find all the local interfaces, it used to panic and
5780 die if the ioctl to get the interface flags failed. However, it seems that
5781 on at least one OS (Solaris 9) it is possible to have an interface that is
5782 included in the list of interfaces, but for which you get a failure error
5783 for this call. This happens when the interface is not "plumbed" into a
5784 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5785 failure of the "get flags" call assumes that the interface is down.
5787 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5788 makes life easier for people who are doing arithmetic on fields extracted
5789 from dates, where you often get leading zeros that should not be
5790 interpreted as octal.
5792 64. Added qualify_domain to the redirect router, to override the global
5795 65. If a pathologically long header line contained very many addresses (the
5796 report of this problem mentioned 10 000) and each of them was rewritten,
5797 Exim could use up a very large amount of memory. (It kept on making new
5798 copies of the header line as it rewrote, and never released the old ones.)
5799 At the expense of a bit more processing, the header rewriting function has
5800 been changed so that it no longer eats memory in this way.
5802 66. The generation of the Received: header has been moved from the time that a
5803 message starts to be received, to the time that it finishes. The timestamp
5804 in the Received: header should now be very close to that of the <= log
5805 line. There are two side-effects of this change:
5807 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5808 logged header lines no longer include the local Received: line, because
5809 it has not yet been created. The same applies to a copy of the message
5810 that is returned to a non-SMTP sender when a message is rejected.
5812 (b) When a filter file is tested using -bf, no additional Received: header
5813 is added to the test message. After some thought, I decided that this
5816 This change does not affect the value of $received_for. It is still set
5817 after address rewriting, but before local_scan() is called.
5819 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5821 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5822 gave an unhelpful panic error message, and a defer error. I have managed to
5823 change this behaviour so that it now rejects any supplied certificate,
5824 which seems right, as the list of acceptable certificates is empty.
5826 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5827 gave an unhelpful defer error. I have not managed to make this reject any
5828 supplied certificates, but the error message it gives is "no certificate
5829 supplied", which is not helpful.
5831 70. exigrep's output now also includes lines that are not associated with any
5832 message, but which match the given pattern. Implemented by a patch from
5833 Martin Sluka, which also tidied up the Perl a bit.
5835 71. Recipient callout verification, like sender verification, was using <> in
5836 the MAIL FROM command. This isn't really the right thing, since the actual
5837 sender may affect whether the remote host accepts the recipient or not. I
5838 have changed it to use the actual sender in the callout; this means that
5839 the cache record is now keyed on a recipient/sender pair, not just the
5840 recipient address. There doesn't seem to be a real danger of callout loops,
5841 since a callout by the remote host to check the sender would use <>.
5842 [SEE ABOVE: changed after hitting problems.]
5844 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5845 temporary errors. However, in the case of such a code being given after
5846 the end of a data transmission (i.e. after ".") Exim was failing to write
5847 a retry record for the message. (Yes, there was some broken host that was
5848 actually sending 8xx at this point.)
5850 73. An unknown lookup type in a host list could cause Exim to panic-die when
5851 the list was checked. (An example that provoked this was putting <; in the
5852 middle of a list instead of at the start.) If this happened during a DATA
5853 ACL check, a -D file could be left lying around. This kind of configuration
5854 error no longer causes Exim to die; instead it causes a defer error. The
5855 incident is still logged to the main and panic logs.
5857 74. Buglet left over from Exim 3 conversion. The message "too many messages
5858 in one connection" was written to the rejectlog but not the mainlog, except
5859 when address rewriting (yes!) was being logged.
5861 75. Added write_rejectlog option.
5863 76. When a system filter was run not as root (that is, when system_filter_user
5864 was set), the values of the $n variables were not being returned to the
5865 main process; thus, they were not subsequently available in the $sn
5868 77. Added +return_path_on_delivery log selector.
5870 78. A connection timeout was being treated differently from recipients deferred
5871 when testing hosts_max_try with a message that was older than the host's
5872 retry timeout. (The host should not be counted, thus allowing all hosts to
5873 be tried at least once before bouncing.) This may have been the cause of an
5874 occasionally reported bug whereby a message would remain on the queue
5875 longer than the retry timeout, but would be bounced if a delivery was
5876 forced. I say "may" because I never totally pinned down the problem;
5877 setting up timeout/retry tests is difficult. See also the next item.
5879 79. The ultimate address timeout was not being applied to errors that involved
5880 a combination of host plus message (for example, a timeout on a MAIL
5881 command). When an address resolved to a number of possible hosts, and they
5882 were not all tried for each delivery (e.g. because of hosts_max_try), a
5883 message could remain on the queue longer than the retry timeout.
5885 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5888 81. Fixed an obscure SMTP outgoing bug which required at least the following
5889 conditions: (a) there was another message waiting for the same server;
5890 (b) the server returned 5xx to all RCPT commands in the first message so
5891 that the message was not completed; (c) the server dropped the connection
5892 or gave a negative response to the RSET that Exim sends to abort the
5893 transaction. The observed case was a dropped connection after DATA that had
5894 been sent in pipelining mode. That is, the server had advertised PIPELINING
5895 but was not implementing it correctly. The effect of the bug was incorrect
5896 behaviour, such as trying another host, and this could lead to a crash.
5902 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5903 and daemon.c were passed as pointers to ints; they should have been
5904 pointers to socklen_t variables (which are typically unsigned ints).
5906 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5909 3. Fixed a really odd bug that affected only the testing scheme; patching a
5910 certain fixed string in the binary changed the value of another string that
5911 happened to be identical to the end of the original first string.
5913 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5914 name", it returns that address as the IP address. On some operating
5915 systems (e.g. Solaris), it also passes back the IP address string as the
5916 "host name". However, on others (e.g. Linux), it passes back an empty
5917 string. Exim wasn't checking for this, and was changing the host name to an
5918 empty string, assuming it had been canonicalized.
5920 5. Although rare, it is permitted to have more than one PTR record for a given
5921 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5922 all the names associated with an address, because they do in Solaris.
5923 However, it seems that they do not in Linux for data that comes from the
5924 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5925 I found this out when I moved to a new Linux workstation and tried to run
5926 the Exim test suite.
5928 To get round this problem I have changed the code so that it now does its
5929 own call to the DNS to look up PTR records when searching for a host name.
5930 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5931 addresses that are only in /etc/hosts are still found.
5933 This behaviour is, however, controlled by an option called host_lookup_
5934 order, which defaults to "bydns:byaddr". If people want to use the other
5935 order, or indeed, just use one or the other means of lookup, they can
5936 specify it in this variable.
5938 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5939 some operating systems, this comes back from gethostbyaddr() as an empty
5940 string, and this is what Exim used to test for. However, it seems that in
5941 other systems, "." is yielded. Exim now tests for this case too.
5943 7. The values of check_spool_space and check_log_space are now held internally
5944 as a number of kilobytes instead of an absolute number of bytes. If a
5945 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5946 kilobyte. This means that much larger values can be stored.
5948 8. Exim monitor: an attempt to get the action menu when not actually pointing
5949 at a message produces an empty menu entitled "No message selected". This
5950 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5951 no entries in it ("Shell widget menu has zero width and/or height"). So I
5952 have added a single, blank menu entry in this case.
5954 9. Added ${quote_local_part.
5956 10. MIME decoding is now applied to the contents of Subject: header lines when
5959 11. Now that a reference to $sender_host_address automatically causes a reverse
5960 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5961 host lookup before query-style lookups in lists that might use this
5962 variable. This has therefore been abolished, and the "net-" prefix is no
5963 longer necessary for query-style lookups.
5965 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5966 have been a typo for LFLAGS, so it has been changed.
5968 13. The install script calls Exim with "-C /dev/null" in order to find the
5969 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5970 to be output. However, since Exim outputs its version number before the
5971 error, it didn't break the script. It just looked ugly. I fixed this by
5972 always allowing "-C /dev/null" if the caller is root.
5974 14. Ignore overlarge ACL variable number when reading spool file - insurance
5975 against a later release with more variables having written the file.
5977 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5978 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5979 this, as well as the form without the "IPv6" on the front.
5981 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5982 OS/Makefile-Darwin file.
5984 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5985 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5987 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5988 variable. However, some OS use ints for inodes. I've added cast to long int
5989 to get rid of the compiler warning.
5991 19. I had forgotten to lock out "/../" in configuration file names when
5992 ALT_CONFIG_PREFIX was set.
5994 20. Routers used for verification do not need to specify transports. However,
5995 if such a router generated a host list, and callout was configured, Exim
5996 crashed, because it could not find a port number from the (non-existent)
5997 transport. It now assumes port 25 in this circumstance.
5999 21. Added the -t option to exigrep.
6001 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6002 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6003 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6004 them is set, LOOKUP_LSEARCH is forced.
6006 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6008 24. Added sender and host information to the "rejected by local_scan()" log
6009 line; previously there was no indication of these.
6011 25. Added .include_if_exists.
6013 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6014 Linux. It turns out that not all file systems support this. Apparently some
6015 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6016 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6017 not supported on the file descriptor, is now ignored when Exim is trying to
6018 sync a directory. This applies only to Linux.
6020 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6022 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6023 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6026 29. Added support for maildirsize files from supplied patch (modified a bit).
6028 30. The use of :fail: followed by an empty string could lead Exim to respond to
6029 sender verification failures with (e.g.):
6031 550 Verification failed for <xxx>
6032 550 Sender verify failed
6034 where the first response line was missing the '-' that indicates it is not
6035 the final line of the response.
6037 31. The loop for finding the name of the user that called Exim had a hardwired
6038 limit of 10; it now uses the value of finduser_retries, which is used for
6039 all other user lookups.
6041 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6044 33. Exim was neglecting to zero errno before one call of strtol() when
6045 expanding a string and expecting an integer value. On some systems this
6046 resulted in spurious "integer overflow" errors. Also, it was casting the
6047 result into an int without checking.
6049 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6050 did not work. The code looks as if it has *never* worked, though it appears
6051 to have been documented since at least release 1.62. I have made it work.
6053 35. The "timeout_DNS" error in retry rules, also documented since at least
6054 1.62, also never worked. As it isn't clear exactly what this means, and
6055 clearly it isn't a major issue, I have abolished the feature by treating it
6056 as "timeout", and writing a warning to the main and panic logs.
6058 36. The display of retry rules for -brt wasn't always showing the error code
6061 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6062 timeout_connect_A, timeout_connect_MX.
6064 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6065 to the empty sender.
6067 39. The daemon was not analysing the content of -oX till after it had closed
6068 stderr and disconnected from the controlling terminal. This meant that any
6069 syntax errors were only noted on the panic log, and the return code from
6070 the command was 0. By re-arranging the code a little, I've made the
6071 decoding happen first, so such errors now appear on stderr, and the return
6072 code is 1. However, the actual setting up of the sockets still happens in
6073 the disconnected process, so errors there are still only recorded on the
6076 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6077 connections (as happens on some IP stacks) was logged at start up time as
6078 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6079 it from "IPv6 and IPv4", which means that two separate sockets are being
6082 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6083 says whether AF_INET or AF_INET6 was passed as an argument.
6085 42. Exiwhat output was messed up when time zones were included in log
6088 43. Exiwhat now gives more information about the daemon's listening ports,
6089 and whether -tls-on-connect was used.
6091 44. The "port" option of the smtp transport is now expanded.
6093 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6094 silently ignored. Now an error message is written to the main and panic
6097 46. There's a new ACL modifier called "logwrite" which writes to a log file
6098 as soon as it is encountered.
6100 47. Added $local_user_uid and $local_user_gid at routing time.
6102 48. Exim crashed when trying to verify a sender address that was being
6105 49. Exim was recognizing only a space character after ".include". It now also
6106 recognizes a tab character.
6108 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6109 extracting the relevant information from the specification. The man page no
6110 longer contains scrambled data for the -d option, and I've added a section
6111 at the front about calling Exim under different names.
6113 51. Added "extra_headers" argument to the "mail" command in filter files.
6115 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6118 53. Installed eximstats 1.29.
6120 54. Added transport_filter_timeout as a generic transport option.
6122 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6123 Cc: header lines. This was required by RFC 822, but it not required by RFC
6126 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6127 incoming messages that did not have them. Now it does so only if the
6128 message originates locally, that is, if there is no associated remote host
6129 address. When Resent- header lines are present, this applies to the Resent-
6130 lines rather than the non-Resent- lines.
6132 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6133 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6135 58. Messages for configuration errors now include the name of the main
6136 configuration file - useful now that there may be more than one file in a
6137 list (.included file names were always shown).
6139 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6140 for those rare installations that do not start the daemon as root or run it
6141 setuid root. I've cut out the call to initgroups() if the daemon is not
6144 60. The Exim user and group can now be bound into the binary as text strings
6145 that are looked up at the start of Exim's processing.
6147 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6149 62. Added $mailstore_basename variable.
6151 63. Installed patch to sieve.c from Michael Haardt.
6153 64. When Exim failed to open the panic log after failing to open the main log,
6154 the original message it was trying to log was written to stderr and debug
6155 output, but if they were not available (the usual case in production), it
6156 was lost. Now it is written to syslog before the two lines that record the
6157 failures to open the logs.
6159 65. Users' Exim filters run in subprocesses under the user's uid. It is
6160 possible for a "deliver" command or an alias in a "personal" command to
6161 provoke an address rewrite. If logging of address rewriting is configured,
6162 this fails because the process is not running as root or exim. There may be
6163 a better way of dealing with this, but for the moment (because 4.30 needs
6164 to be released), I have disabled address rewrite logging when running a
6165 filter in a non-root, non-exim process.
6171 1. The buildconfig auxiliary program wasn't quoting the value set for
6172 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6173 not defined. This bug was masked in 4.22 by the effect that was fixed in
6176 2. Some messages that were rejected after a message id was allocated were
6177 shown as "incomplete" by exigrep. It no longer does this for messages that
6178 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6180 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6181 have allow_domain_literals set, the ID did not get logged in the <= line.
6182 Domain literals are now always recognized in Message-ID: header lines.
6184 4. The first argument for a ${extract expansion item is the key name or field
6185 number. Leading and trailing spaces in this item were not being ignored,
6186 causing some misleading effects.
6188 5. When deliver_drop_privilege was set, single queue runner processes started
6189 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6190 same command in the process it spins off) were not dropping privilege.
6192 6. When the daemon running as "exim" started a queue runner, it always
6193 re-executed Exim in the spun-off process. This is a waste of effort when
6194 deliver_drop_privilege is set. The new process now just calls the
6195 queue-runner function directly.
6201 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6204 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6205 ignored. Though the use of -oP was forcing the writing of a pid file, it
6206 was always written to the default place.
6208 3. If the message "no IP address found for host xxxx" is generated during
6209 incoming verification, it is now followed by identification of the incoming
6210 connection (so you can more easily find what provoked it).
6212 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6214 5. Added some features to "harden" Exim a bit more against certain attacks:
6216 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6217 be put in Local/Makefile. This is like the never_users runtime option,
6218 but it cannot be overridden. The default setting is "root".
6220 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6221 prefix string with which any file named in a -C command line option
6224 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6225 is retained for -C and -D only if the caller of Exim is root. Without
6226 it, the exim user may also use -C and -D and retain privilege.
6228 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6229 command line option is disabled.
6231 6. Macro names set by the -D option must start with an upper case letter, just
6232 like macro names defined in the configuration file.
6234 7. Added "dereference=" facility to LDAP.
6236 8. Two instances of the typo "uknown" in the source files are fixed.
6238 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6239 the Configure-Makefile script screwed up while processing it.
6241 10. Incorporated PCRE 4.4.
6243 11. The SMTP synchronization check was not operating right at the start of an
6244 SMTP session. For example, it could not catch a HELO sent before the client
6245 waited for the greeting. There is now a check for outstanding input at the
6246 point when the greeting is written. Because of the duplex, asynchronous
6247 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6248 way, but not yet received, when the check is performed.
6250 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6251 on TCP/IP sockets, because this apparently causes some broken clients to
6254 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6255 unchanged) from the Cygwin maintainer.
6257 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6258 was supported instead of testing for mbx. Effectively a typo.
6260 15. The spa authenticator server code was not checking that the input it
6261 received was valid base64.
6263 16. The debug output line for the "set" modifier in ACLs was not showing the
6264 name of the variable that was being set.
6266 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6268 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6269 lookup on its own. Something else was needed to trigger the lookup. For
6270 example, a match in host_lookup or the need for a host name in a host list.
6271 Now, if $sender_host_name is referenced and the host name has not yet been
6272 looked up, a lookup is performed. If the lookup fails, the variable remains
6273 empty, and $host_lookup_failed is set to "1".
6275 19. Added "eqi" as a case-independent comparison operator.
6277 20. The saslauthd authentication condition could segfault if neither service
6278 nor realm was specified.
6280 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6281 error message that was logged was misleading, and incoming SMTP
6282 connections were dropped. The message is now more accurate, and temporary
6283 errors are given to SMTP connections.
6285 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6286 (except RSET) with a 421 error, until QUIT is received. However, it was
6287 failing to send a response to QUIT.
6289 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6290 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6291 ACL is now run after the helo_try_verify_hosts code.
6293 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6294 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6295 case-independent, so other case variants are also recognized). Apparently
6296 some systems use these upper case variants.
6298 25. If more than two messages were waiting for the same host, and a transport
6299 filter was specified for the transport, Exim sent two messages over the
6300 same TCP/IP connection, and then failed with "socket operation on non-
6301 socket" when it tried to send the third.
6303 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6305 27. The extern definition of crypt16() in expand.c was not being excluded when
6306 the OS had its own crypt16() function.
6308 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6309 as a preferred synonym for return_size_limit, both as an option and as an
6312 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6314 30. Changed the default configuration ACL to relax the local part checking rule
6315 for addresses that are not in any local domains. For these addresses,
6316 slashes and pipe symbols are allowed within local parts, but the sequence
6317 /../ is explicitly forbidden.
6319 31. SPA server authentication was not clearing the challenge buffer before
6322 32. log_message in a "warn" ACL statement was writing to the reject log as
6323 well as to the main log, which contradicts the documentation and doesn't
6324 seem right (because no rejection is happening). So I have stopped it.
6326 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6327 However, I am unable to do any testing of this.
6329 34. Fixed an infelicity in the appendfile transport. When checking directories
6330 for a mailbox, to see if any needed to be created, it was accidentally
6331 using path names with one or more superfluous leading slashes; tracing
6332 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6334 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6335 contents are added to the log line that is written for every discarded
6336 recipient. (Previously a log_message setting was ignored.)
6338 36. The ${quote: operator now quotes the string if it is empty.
6340 37. The install script runs exim in order to find its version number. If for
6341 some reason other than non-existence or emptiness, which it checks, it
6342 could not run './exim', it was installing it with an empty version number,
6343 i.e. as "exim-". This error state is now caught, and the installation is
6346 38. An argument was missing from the function that creates an error message
6347 when Exim fails to connect to the socket for saslauthd authentication.
6348 This could cause Exim to crash, or give a corrupted message.
6350 39. Added isip, isip4, and isip6 to ${if conditions.
6352 40. The ACL variables $acl_xx are now saved with the message, and can be
6353 accessed later in routers, transports, and filters.
6355 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6356 strings in the file are not string-expanded.
6358 42. If a MAIL command specified a SIZE value that was too large to fit into an
6359 int variable, the check against message_size_limit failed. Such values are
6360 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6361 day this will have to be increased, but I don't think I want to be around
6362 when emails are that large.
6369 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6370 iconv() is not standard in FreeBSD.
6372 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6373 IPv6 enabled. The observed symptom was a segmentation fault on return from
6374 the function os_common_find_running_interfaces() in src/os.c.
6376 3. In the check_special_case() function in daemon.c I had used "errno" as an
6377 argument name, which causes warnings on some systems. This was basically a
6378 typo, since it was named "eno" in the comments!
6380 4. The code that waits for the clock to tick (at a resolution of some fraction
6381 of a second) so as to ensure message-id uniqueness was always waiting for
6382 at least one whole tick, when it could have waited for less. [This is
6383 almost certainly not relevant at current processor speeds, where it is
6384 unlikely to ever wait at all. But we try to future-proof.]
6386 5. The function that sleeps for a time interval that includes fractions of a
6387 second contained a race. It did not block SIGALRM between setting the
6388 timer, and suspending (a couple of lines later). If the interval was short
6389 and the sigsuspend() was delayed until after it had expired, the suspension
6390 never ended. On busy systems this could lead to processes getting stuck for
6393 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6394 process, before it forks any delivery processes. The open lookup caching
6395 mechanism meant that the open file or database connection was passed into
6396 the delivery process. The problem was that delivery processes always tidy
6397 up cached lookup data. This could cause a problem for the next delivery
6398 process started by the queue runner, because the external queue runner
6399 process does not know about the closure. So the next delivery process
6400 still has data in the lookup cache. In the case of a file lookup, there was
6401 no problem because closing a file descriptor in a subprocess doesn't affect
6402 the parent. However, if the lookup was caching a connection to a database,
6403 the connection was closed, and the second delivery process was likely to
6404 see errors such as "PGSQL: query failed: server closed the connection
6405 unexpectedly". The problem has been fixed by closing all cached lookups
6406 in a queue runner before running a delivery process.
6408 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6409 seem to have the "const" qualifier which it has on other OS. I've
6412 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6413 *of the same type* (client or server) with the same public name that an
6414 error should be diagnosed.
6416 9. When Exim looked up a host name for an IP address, but failed to find the
6417 original IP address when looking up the host name (a safety check), it
6418 output the message "<ip address> does not match any IP for NULL", which was
6419 confusing, to say the least. The bug was that the host name should have
6420 appeared instead of "NULL".
6422 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6423 user that is built into the binary, and the -C or -D options is used, root
6424 privilege is dropped before the configuration file is read. In addition,
6425 logging is switched to stderr instead of the normal log files. If the
6426 configuration then re-defines the Exim user, the unprivileged environment
6427 is probably not what is expected, so Exim logs a panic warning message (but
6430 However, if deliver_drop_privilege is set, the unprivileged state may well
6431 be exactly what is intended, so the warning has been cut out in that case,
6432 and Exim is allowed to try to write to its normal log files.
6438 1. smtp_return_error_details was not giving details for temporary sender
6439 or receiver verification errors.
6441 2. Diagnose a configuration error if two authenticators have the same public
6444 3. Exim used not to create the message log file for a message until the first
6445 delivery attempt. This could be confusing when incoming messages were held
6446 for policy or load reasons. The message log file is now created at the time
6447 the message is received, and an initial "Received" line is written to it.
6449 4. The automatically generated man page for command line options had a minor
6450 bug that caused no ill effects; however, a more serious problem was that
6451 the procedure for building the man page automatically didn't always
6452 operate. Consequently, release 4.20 contains an out-of-date version. This
6453 shouldn't happen again.
6455 5. When building Exim with embedded Perl support, the script that builds the
6456 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6457 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6459 6. The freeze_tell option was not being used for messages that were frozen on
6460 arrival, either by an ACL or by local_scan().
6462 7. Added the smtp_incomplete_transaction log selector.
6464 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6465 was accepting AUTH without a new EHLO.
6467 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6468 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6471 10. Logging of TCP/IP connections (when configured) now happens in the main
6472 daemon process instead of the child process, so that the TCP/IP connection
6473 count is more accurate (but it can never be perfect).
6475 11. The use of "drop" in a nested ACL was not being handled correctly in the
6476 outer ACL. Now, if condition failure induced by the nested "drop" causes
6477 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6478 or "require"), the connection is dropped.
6480 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6481 that yield "discard" can now be used with an "accept" or a "discard" verb,
6482 but an error is generated for any others (because I can't see a useful way
6483 to define what should happen).
6485 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6486 are now processed in the same way as lines in the Exim configuration file.
6487 In particular, continuation lines are supported.
6489 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6491 15. Added -ti meaning -t -i.
6493 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6494 domains, and warn by logging if others are found.
6496 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6497 returned from GIFCONF when Exim is trying to find the list of interfaces on
6498 a host. The code in os.c has been modified to copy each ifreq to an aligned
6499 structure in all cases.
6501 Also, in some cases, the returned ifreq's were being copied to a 'struct
6502 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6503 means the last couple of bytes of an IPv6 address could be chopped if the
6504 ifreq contained only a normal sockaddr (14 bytes storage).
6506 18. Named domain lists were not supported in the hosts_treat_as_local option.
6507 An entry such as +xxxx was not recognized, and was treated as a literal
6510 19. Ensure that header lines added by a DATA ACL are included in the reject log
6511 if the ACL subsequently rejects the message.
6513 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6514 MD5 (which is deprecated).
6516 21. When testing a filter file using -bf, Exim was writing a message when it
6517 took the sender from a "From " line in the message, but it was not doing so
6518 when it took $return_path from a Return-Path: header line. It now does.
6520 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6521 with a valid header line field name (a series of printing characters
6522 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6524 23. Changed "disc" in the source to "disk" to conform to the documentation and
6525 the book and for uniformity.
6527 24. Ignore Sendmail's -Ooption=value command line item.
6529 25. When execve() failed while trying to run a command in a pipe transport,
6530 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6531 could be confused with a return value of 69 from the command itself. This
6532 has been changed to 127, the value the shell returns if it is asked to run
6533 a non-existent command. The wording for the related log line suggests a
6534 non-existent command as the problem.
6536 26. If received_header_text expands to an empty string, do not add a Received:
6537 header line to the message. (Well, it adds a token one on the spool, but
6538 marks it "old" so that it doesn't get used or transmitted.)
6540 27. Installed eximstats 1.28 (addition of -nt option).
6542 28. There was no check for failure on the call to getsockname() in the daemon
6543 code. This can fail if there is a shortage of resources on the system, with
6544 ENOMEM, for example. A temporary error is now given on failure.
6546 29. Contrary to the C standard, it seems that in some environments, the
6547 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6548 program. Exim now does this explicitly; it affects the formatting of
6549 timestamps using strftime().
6551 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6552 complaints. I've now initialized all the variables, to avoid this.
6554 32. Header lines added by a system filter were not being "seen" during
6555 transport-time rewrites.
6557 33. The info_callback() function passed to OpenSSL is set up with type void
6558 (*)(SSL *, int, int), as described somewhere. However, when calling the
6559 function (actually a macro) that sets it up, the type void(*)() is
6560 expected. I've put in a cast to prevent warnings from picky compilers.
6562 34. If a DNS black list lookup found a CNAME record, but there were no A
6563 records associated with the domain it pointed at, Exim crashed.
6565 35. If a DNS black list lookup returned more than one A record, Exim ignored
6566 all but the first. It now scans all returned addresses if a particular IP
6567 value is being sought. In this situation, the contents of the
6568 $dnslist_value variable are a list of all the addresses, separated by a
6571 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6572 to accept a host name and all its aliases if the forward lookup for any of
6573 them yielded the IP address of the incoming connection. Now it accepts only
6574 those names whose forward lookup yields the correct IP address. Any other
6575 names are discarded. This closes a loophole whereby a rogue DNS
6576 administrator could create reverse DNS records to break through a
6577 wildcarded host restriction in an ACL.
6579 37. If a user filter or a system filter that ran in a subprocess used any of
6580 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6581 the wrong values were passed to the pipe command ($thisaddress had the
6582 value of $0, $0 had the value of $1, etc). This bug was introduced by
6583 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6585 38. Improved the line breaking for long SMTP error messages from ACLs.
6586 Previously, if there was no break point between 40 and 75 characters, Exim
6587 left the rest of the message alone. Two changes have been made: (a) I've
6588 reduced the minimum length to 35 characters; (b) if it can't find a break
6589 point between 35 and 75 characters, it looks ahead and uses the first one
6590 that it finds. This may give the occasional overlong line, but at least the
6591 remaining text gets split now.
6593 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6594 file descriptors might be low, and that setting 1000 would always raise it.
6595 It turns out that in some environments, the limit is already over 1000 and
6596 that lowering it causes trouble. So now Exim takes care not to decrease it.
6598 40. When delivering a message, the value of $return_path is set to $sender_
6599 address at the start of routing (routers may change the value). By an
6600 oversight, this default was not being set up when an address was tested by
6601 -bt or -bv, which affected the outcome if any router or filter referred to
6604 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6605 log only when "message" or "log_message" are set. However, if one of the
6606 conditions was an address verification, or a call to a nested ACL, the
6607 messages generated by the underlying test were being passed through. This
6608 no longer happens. The underlying message is available in $acl_verify_
6609 message for both "message" and "log_message" expansions, so it can be
6610 passed through if needed.
6612 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6613 new expansion $bh_ to give the encoded byte string without charset
6614 translation. Translation happens only if iconv() is available; HAVE_ICONV
6615 indicates this at build time. HEADERS_CHARSET gives the charset to
6616 translate to; headers_charset can change it in the configuration, and
6617 "headers charset" can change it in an individual filter file.
6619 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6620 that creates RFC 2047 encoded "words" labels them as that charset instead
6621 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6622 expansion operator; (ii) when Exim creates a From: line for a local
6623 message; (iii) when a header line is rewritten to include a "phrase" part.
6625 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6626 buggy, causing it to skip the first lines of messages whose message ID
6627 ended in 'D'. This would not have bitten before Exim release 4.14, because
6628 message IDs were unlikely to end in 'D' before then. The effect was to have
6629 incorrect size information for certain domains.
6631 45. #include "config.h" was missing at the start of the crypt16.c module. This
6632 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6635 46. If there was a timeout during a "random" callout check, Exim treated it as
6636 a failure of the random address, and carried on sending RSET and the real
6637 address. If the delay was just some slowness somewhere, the response to the
6638 original RCPT would be taken as a response to RSET and so on, causing
6639 mayhem of various kinds.
6641 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6642 when I implemented it. It didn't allow for the fact that some option values
6643 may legitimately be negative (e.g. size_addition), and it didn't even do
6644 the right test for positive values.
6646 48. Domain names in DNS records are case-independent. Exim always looks them up
6647 in lower case. Some resolvers return domain names in exactly the case they
6648 appear in the zone file, that is, they may contain uppercase letters. Not
6649 all resolvers do this - some return always lower case. Exim was treating a
6650 change of case by a resolver as a change of domain, similar to a widening
6651 of a domain abbreviation. This triggered its re-routing code and so it was
6652 trying to route what was effectively the same domain again. This normally
6653 caused routing to fail (because the router wouldn't handle the domain
6654 twice). Now Exim checks for this case specially, and just changes the
6655 casing of the domain that it ultimately uses when it transmits the message
6658 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6661 50. If a filter generated a file delivery with a non-absolute name (possible if
6662 no home directory exists for the router), the forbid_file option was not
6665 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6666 the existing equality matching.
6668 52. Exim was using ints instead of ino_t variables in some places where it was
6669 dealing with inode numbers.
6671 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6672 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6673 called TMPDIR, and if it finds it is different, it changes its value.
6675 54. The smtp_printf() function is now made available to local_scan() so
6676 additional output lines can be written before returning. There is also an
6677 smtp_fflush() function to enable the detection of a dropped connection.
6678 The variables smtp_input and smtp_batched_input are exported to
6681 55. Changed the default runtime configuration: the message "Unknown user"
6682 has been removed from the ACL, and instead placed on the localuser router,
6683 using the cannot_route_message feature. This means that any verification
6684 failures that generate their own messages won't get overridden. Similarly,
6685 the "Unrouteable address" message that was in the ACL for unverifiable
6686 relay addresses has also been removed.
6688 56. Added hosts_avoid_esmtp to the smtp transport.
6690 57. The exicyclog script was not checking for the esoteric option
6691 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6692 will work only if exicyclog is run under the appropriate euid.
6694 58. Following a discussion on the list, the rules by which Exim recognises line
6695 endings on incoming messages have been changed. The -dropcr and drop_cr
6696 options are now no-ops, retained only for backwards compatibility. The
6697 following line terminators are recognized: LF CRLF CR. However, special
6698 processing applies to CR:
6700 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6701 nor a local message in the state where . is a terminator.
6703 (ii) If a bare CR is encountered in a header line, an extra space is added
6704 after the line terminator so as not to end the header. The reasoning
6705 behind this is that bare CRs in header lines are most likely either
6706 to be mistakes, or people trying to play silly games.
6708 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6709 was being incorrectly given as 18 bytes larger than it should have been.
6710 This is a VOB (very old bug).
6712 60. This may never have affected anything current, but just in case it has:
6713 When the local host is found other than at the start of a list of hosts,
6714 the local host, those with the same MX, and any that follow, are discarded.
6715 When the list in question was part of a longer list of hosts, the following
6716 hosts (not currently being processed) were also being discarded. This no
6717 longer happens. I'm not sure if this situation could ever has previously
6720 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6723 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6724 header. This is something that is recommended in a new Internet Draft, and
6725 is something that is documented as being done by Sendmail. There are two
6726 possible values. For messages generated by the autoreply transport, Exim
6729 Auto-Submitted: auto-replied
6731 whereas for all other generated messages (e.g. bounces) it adds
6733 Auto-Submitted: auto-generated
6735 63. The "personal" condition in filters now includes a test for the
6736 Auto-Submitted: header. If it contains the string "auto-" the message it
6737 not considered personal.
6739 64. Added rcpt_include_affixes as a generic transport option.
6741 65. Added queue_only_override (default true).
6743 66. Added the syslog_duplication option.
6745 67. If what should have been the first header line of a message consisted of
6746 a space followed by a colon, Exim was mis-interpreting it as a header line.
6747 It isn't of course - it is syntactically invalid and should therefore be
6748 treated as the start of the message body. The misbehaviour could have
6749 caused a number of strange effects, including loss of data in subsequent
6750 header lines, and spool format errors.
6752 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6753 client host had authenticated. This control can now be exercised by an ACL
6754 for more flexibility.
6756 69. By default, callouts do not happen when testing with -bh. There is now a
6757 variant, -bhc, which does actually run the callout code, including
6758 consulting and updating the callout cache.
6760 70. Added support for saslauthd authentication, courtesy of Alexander
6763 71. If statvfs() failed on the spool or log directories while checking their
6764 size for availability, Exim confusingly gave the error "space shortage".
6765 Furthermore, in debugging mode it crashed with a floating point exception.
6766 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6767 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6768 serious problem, Exim now writes to the main and panic logs when this
6769 happens, with details of the failure. It then refuses to accept the
6770 incoming message, giving the message "spool directory problem" or "log
6771 directory problem" with a 421 code for SMTP messages.
6773 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6774 0, 1, and 2 exist, because some OS complain for execs without them (see
6775 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6776 descriptors. However, the code omitted to check that the open succeeded,
6777 causing mysterious errors if for some reason the permissions on /dev/null
6778 got screwed. Now Exim writes a message to the main and panic logs, and
6779 bombs out if it can't open /dev/null.
6781 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6782 interact so that it is all more flexible. It is supposed to remain
6783 backwards compatible. Also added extra_local_interfaces.
6785 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6786 to bomb out with an assertion failure - to the client this appears as a
6787 connection drop. This problem occurs in the part of the code that was taken
6788 from the Samba project. Fortunately, the assertion is in a very simple
6789 function, so I have fixed this by reproducing the function inline in the
6790 one place where it is called, and arranging for authentication to fail
6791 instead of killing the process with assert().
6793 75. The SPA client code was not working when the server requested OEM rather
6794 than Unicode encoding.
6796 76. Added code to make require_files with a specific uid setting more usable in
6797 the case where statting the file as root fails - usually a non-root-mounted
6798 NFS file system. When this happens and the failure is EACCES, Exim now
6799 forks a subprocess and does the per-uid checking as the relevant uid.
6801 77. Added process_log_path.
6803 78. If log_file_path was not explicitly set, a setting of check_log_space or
6804 check_log_inodes was ignored.
6806 79. If a space check for the spool or log partitions fails, the incident is now
6807 logged. Of course, in the latter case the data may get lost...
6809 80. Added the %p formatting code to string_format() so that it can be used to
6810 print addresses in debug_print(). Adjusted all the address printing in the
6811 debugging in store.c to use %p rather than %d.
6813 81. There was a concern that a line of code in smtp_in.c could overflow a
6814 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6815 initially expressed, the concern was not well-founded, because trailing
6816 spaces are removed early. However, if the trailing spaces were followed by
6817 a NULL, they did not get removed, so the overflow was possible. Two fixes
6820 (a) I re-wrote the offending code in a cleaner fashion.
6821 (b) If an incoming SMTP command contains a NULL character, it is rejected
6824 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6825 runs initgroups(), so that if the Exim user is in any additional groups,
6826 they will be used during message reception.
6832 The change log for 4.20 and earlier releases has been archived.