1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
64 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
65 4.77 as the kernel support then wasn't solid, having issues in 64bit
66 mode. Now, it's been long enough. Add support for FreeBSD also.
68 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
69 case where the routing stage had gathered several addresses to send to
70 a host before calling the transport for the first, we previously failed
71 to close down TLS in the old transport process before passing the TCP
72 connection to the new process. The new one sent a STARTTLS command
73 which naturally failed, giving a failed delivery and bloating the retry
74 database. Investigation and fix prototype from Wolfgang Breyha.
76 JH/12 Fix check on SMTP command input synchronisation. Previously there were
77 false-negatives in the check that the sender had not preempted a response
78 or prompt from Exim (running as a server), due to that code's lack of
79 awareness of the SMTP input buffering.
81 PP/04 Add commandline_checks_require_admin option.
82 Exim drops privileges sanely, various checks such as -be aren't a
83 security problem, as long as you trust local users with access to their
84 own account. When invoked by services which pass untrusted data to
85 Exim, this might be an issue. Set this option in main configuration
86 AND make fixes to the calling application, such as using `--` to stop
89 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
90 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
91 responses to those, into a single TLS record each way (this usually means
92 a single packet). As a side issue, smtp_enforce_sync now works on TLS
95 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
96 affects you only if you're dancing at the edge of the param size limits.
97 If you are, and this message makes sense to you, then: raise the
98 configured limit or use OpenSSL 1.1. Nothing we can do for older
101 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
102 cmdline element to get the documented default one. Previously it was
105 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
106 are unsuable for non-message-specific reasons - eg. connection timeout,
107 banner-time rejection.
109 JH/16 Fix logging of delivery remote port, when specified by router, under
112 PP/06 Repair manualroute's ability to take options in any order, even if one
113 is the name of a transport.
120 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
121 than -2003 did; needs libidn2 in addition to libidn.
123 JH/02 The path option on a pipe transport is now expanded before use.
125 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
126 Patch provided by "Björn", documentation fix added too.
128 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
129 missing a wire-to-host endian conversion.
131 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
132 close after a BDAT command line could be taken as a following command,
133 giving a synch failure. Fix by only checking for synch immediately
134 before acknowledging the chunk.
136 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
137 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
138 macro. Patches provided by Josh Soref.
140 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
141 Previously we did not; the RFC seems ambiguous and VRFY is not listed
142 by IANA as a service extension. However, John Klensin suggests that we
145 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
146 the dkim code may be unix-mode line endings rather than smtp wire-format
147 CRLF, so prepend a CR to any bare LF.
149 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
150 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
152 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
153 the first were themselves being wrongly included in the feed into dkim
154 processing; with most chunk sizes in use this resulted in an incorrect
155 body hash calculated value.
157 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
158 DKIM signature block, for verification. Although advised against by
159 standards it is specifically not ruled illegal.
161 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
163 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
164 missing a body hash (the bh= tag).
166 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
167 It seems that HAProxy sends the Proxy Protocol information in clear and
168 only then does a TLS startup, so do the same.
170 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
171 TCP connections (such as for Spamd) unless the daemon successfully set
172 Fast Open mode on its listening sockets. This fixes breakage seen on
173 too-old kernels or those not configured for Fast Open, at the cost of
174 requiring both directions being enabled for TFO, and TFO never being used
175 by non-daemon-related Exim processes.
177 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
178 endings, at least on the first header line. Try to canonify any that get
179 past that check, despite the cost.
181 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
182 now limited to an arbitrary five deep, while parsing addresses with the
183 strip_excess_angle_brackets option enabled.
185 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
186 instead leave the unprompted TLS handshake in socket buffer for the
187 TLS library to consume.
189 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
191 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
193 JH/16 Drop variables when they go out of scope. Memory management drops a whole
194 region in one operation, for speed, and this leaves assigned pointers
195 dangling. Add checks run only under the testsuite which checks all
196 variables at a store-reset and panics on a dangling pointer; add code
197 explicitly nulling out all the variables discovered. Fixes one known
198 bug: a transport crash, where a dangling pointer for $sending_ip_address
199 originally assigned in a verify callout, is re-used.
201 PP/06 Drop '.' from @INC in various Perl scripts.
203 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
205 PP/08 Reduce a number of compilation warnings under clang; building with
206 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
207 should be warning-free.
209 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
211 HS/01 Fix portability problems introduced by PP/08 for platforms where
212 realloc(NULL) is not equivalent to malloc() [SunOS et al].
214 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
215 chunk. This allows us to accept broken chunked messages. We need a more
216 general solution here.
218 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
219 already-broken messages in the queue.
221 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
223 JH/19 Fix reference counting bug in routing-generated-address tracking.
229 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
230 supports it and a size is available (ie. the sending peer gave us one).
232 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
233 years of being deprecated). The replacements are the ACLs
234 acl_smtp_mime and acl_not_smtp_mime.
236 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
237 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
238 or even an in-clear connection were permitted. Now, if the host lookup
239 was dnssec and dane was requested then the host is only used if the
240 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
241 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
242 if one fails this test.
243 This means that a poorly-configured remote DNS will make it incommunicado;
244 but it protects against a DNS-interception attack on it.
246 JH/04 Bug 1810: make continued-use of an open smtp transport connection
247 non-noisy when a race steals the message being considered.
249 JH/05 If main configuration option tls_certificate is unset, generate a
250 self-signed certificate for inbound TLS connections.
252 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
253 in rewrites and routers.
255 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
256 and logged a warning sing 4.83; now they are a configuration file error.
258 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
259 (lacking @domain). Apply the same qualification processing as RCPT.
261 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
263 JH/10 Support ${sha256:} applied to a string (as well as the previous
266 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
267 a cutthrough deliver is pending, as we always want to make a connection.
268 This also avoids re-routing the message when later placing the cutthrough
269 connection after a verify cache hit.
270 Do not update it with the verify result either.
272 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
273 when routing results in more than one destination address.
275 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
276 signing (which inhibits the cutthrough capability). Previously only
277 the presence of an option was tested; now an expansion evaluating as
278 empty is permissible (obviously it should depend only on data available
279 when the cutthrough connection is made).
281 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
282 the relevant preceding SMTP command did not note the pipelining mode.
284 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
285 Previously they were not counted.
287 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
288 as one having no matching records. Previously we deferred the message
289 that needed the lookup.
291 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
292 distinguished as "(=".
294 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
295 for missing MX records. Previously it only worked for missing A records.
297 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
299 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
300 after the data-go-ahead and data-ack. Patch from Jason Betts.
302 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
303 even for a "none" policy. Patch from Tony Meyer.
305 JH/22 Fix continued use of a connection for further deliveries. If a port was
306 specified by a router, it must also match for the delivery to be
309 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
310 When one of the recipients of a message was unsuitable for the connection
311 (has no matching addresses), we lost track of needing to mark it
312 deferred. As a result mail would be lost.
314 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
316 JH/25 Decoding ACL controls is now done using a binary search; the source code
317 takes up less space and should be simpler to maintain. Merge the ACL
318 condition decode tables also, with similar effect.
320 JH/26 Fix problem with one_time used on a redirect router which returned the
321 parent address unchanged. A retry would see the parent address marked as
322 delivered, so not attempt the (identical) child. As a result mail would
325 JH/27 Fix a possible security hole, wherein a process operating with the Exim
326 UID can gain a root shell. Credit to http://www.halfdog.net/ for
327 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
330 JH/28 Enable {spool,log} filesystem space and inode checks as default.
331 Main config options check_{log,spool}_{inodes,space} are now
332 100 inodes, 10MB unless set otherwise in the configuration.
334 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
335 Previously it only applied to the main-section connection policy
338 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
340 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
341 by me. Added RFC7919 DH primes as an alternative.
343 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
344 are not in the system include path.
346 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
347 GnuTLS, when a session startup failed (eg because the client disconnected)
348 Exim did stdio operations after fclose. This was exposed by a recent
349 change which nulled out the file handle after the fclose.
351 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
352 signed directly by the cert-signing cert, rather than an intermediate
353 OCSP-signing cert. This is the model used by LetsEncrypt.
355 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
357 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
358 an incoming connection.
360 HS/02 Bug 1802: Do not half-close the connection after sending a request
363 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
364 fallback to "prime256v1".
366 JH/34 SECURITY: Use proper copy of DATA command in error message.
367 Could leak key material. Remotely exploitable. CVE-2016-9963.
373 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
374 and 3.4.4 - once the server is enabled to respond to an OCSP request
375 it does even when not requested, resulting in a stapling non-aware
376 client dropping the TLS connection.
378 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
379 support variable-length bit vectors. No functional change.
381 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
382 The I= interface field on outgoing lines is now after the H= remote
383 host field, same as incoming lines. There is a separate
384 outgoing_interface log selector which allows you to disable the
387 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
388 If not running log_selector +smtp_connection the mainlog would be held
389 open indefinitely after a "too many connections" event, including to a
390 deleted file after a log rotate. Leave the per net connection logging
391 leaving it open for efficiency as that will be quickly detected by the
392 check on the next write.
394 HS/01 Bug 1671: Fix post transport crash.
395 Processing the wait-<transport> messages could crash the delivery
396 process if the message IDs didn't exist for some reason. When
397 using 'split_spool_directory=yes' the construction of the spool
398 file name failed already, exposing the same netto behaviour.
400 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
401 mime_regex ACL conditions.
403 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
404 to DSN fail messages (bounces): remote IP, remote greeting, remote response
405 to HELO, local diagnostic string.
407 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
408 log line to debug. Even when configured with a tls authenticator many
409 client connections are expected to not authenticate in this way, so
410 an authenticate fail is not an error.
412 HS/02 Add the Exim version string to the process info. This way exiwhat
413 gives some more detail about the running daemon.
415 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
416 matter for fast-change records such as DNSBLs.
418 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
419 retry record, even if constant. There may be multiple transports with
420 different interface settings and the retry behaviour needs to be kept
423 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
425 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
427 JH/10 Bug 840: fix log_defer_output option of pipe transport
429 JH/11 Bug 830: use same host for all RCPTS of a message, even under
430 hosts_randomize. This matters a lot when combined with mua_wrapper.
432 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
433 ${quote_pgsql:<string>} operator.
435 JH/13 Bug 1708: avoid misaligned access in cached lookup.
437 JH/14 Change header file name for freeradius-client. Relevant if compiling
438 with Radius support; from the Gentoo tree and checked under Fedora.
440 JH/15 Bug 1712: Introduce $prdr_requested flag variable
442 JH/16 Bug 1714: Permit an empty string as expansion result for transport
443 option transport_filter, meaning no filtering.
445 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
447 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
448 defaults to "*" (all hosts). The variable is now available when not built
449 with TLS, default unset, mainly to enable keeping the testsuite sane.
450 If a server certificate is not supplied (via tls_certificate) an error is
451 logged, and clients will find TLS connections fail on startup. Presumably
452 they will retry in-clear.
453 Packagers of Exim are strongly encouraged to create a server certificate
454 at installation time.
456 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
457 with the $config_file variable.
459 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
460 in transport context, after the attempt, and per-recipient. The latter type
461 is per host attempted. The event data is the error message, and the errno
462 information encodes the lookup type (A vs. MX) used for the (first) host,
463 and the trailing two digits of the smtp 4xx response.
465 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
466 to write to mainlog (or rejectlog, paniclog) in the window between file
467 creation and permissions/ownership being changed. Particularly affects
468 installations where exicyclog is run as root, rather than exim user;
469 result is that the running daemon panics and dies.
471 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
473 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
474 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
475 "pri" and "weight". Note that the previous implicit priority given by the
476 list order is no longer honoured.
478 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
481 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
482 by defining SUPPORT_SOCKS.
484 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
485 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
486 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
487 variables are renamed to proxy_{local,external}_{address,port}.
489 JH/27 Move Internationalisation support from Experimental to mainline, enabled
490 for a build by defining SUPPORT_I18N
492 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
493 of the query string, and make ${quote_redis:} do that quoting.
495 JH/29 Move Events support from Experimental to mainline, enabled by default
496 and removable for a build by defining DISABLE_EVENT.
498 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
500 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
501 cached by the daemon.
503 JH/32 Move Redis support from Experimental to mainline, enabled for a build
504 by defining LOOKUP_REDIS. The libhiredis library is required.
506 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
507 keys are given for lookup.
509 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
510 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
511 only supported when built with TLS support. The PolarSSL SHA routines
512 are still used when the TLS library is too old for convenient support.
514 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
515 openssl_options), for security. OpenSSL forces this from version 1.1.0
516 server-side so match that on older versions.
518 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
519 allocation for $value could be released as the expansion processing
520 concluded, but leaving the global pointer active for it.
522 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
523 and to use the domains and local_parts ACL conditions.
525 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
526 incorrectly not doubled on cutthrough transmission, hence seen as a
527 body-termination at the receiving system - resulting in truncated mails.
528 Commonly the sender saw a TCP-level error, and retransmitted the message
529 via the normal store-and-forward channel. This could result in duplicates
530 received - but deduplicating mailstores were liable to retain only the
531 initial truncated version.
533 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
535 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
537 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
538 we're in there, support oversigning also; bug 1309.
540 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
542 HS/04 Add support for keep_environment and add_environment options.
544 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
545 either intentional arithmetic overflow during PRNG, or testing config-
548 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
549 delivery resulted in actual delivery. Cancel cutthrough before DATA
552 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
553 rejects a recipient: pass the reject to the originator.
555 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
556 Many were false-positives and ignorable, but it's worth fixing the
559 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
560 for the new environment-manipulation done at startup. Move the routines
561 from being local to tls.c to being global via the os.c file.
563 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
564 an extract embedded as result-arg for a map, the first arg for extract
565 is unavailable so we cannot tell if this is a numbered or keyed
566 extraction. Accept either.
572 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
575 JH/02 The smtp transport option "multi_domain" is now expanded.
577 JH/03 The smtp transport now requests PRDR by default, if the server offers
580 JH/04 Certificate name checking on server certificates, when exim is a client,
581 is now done by default. The transport option tls_verify_cert_hostnames
582 can be used to disable this per-host. The build option
583 EXPERIMENTAL_CERTNAMES is withdrawn.
585 JH/05 The value of the tls_verify_certificates smtp transport and main options
586 default to the word "system" to access the system default CA bundle.
587 For GnuTLS, only version 3.0.20 or later.
589 JH/06 Verification of the server certificate for a TLS connection is now tried
590 (but not required) by default. The verification status is now logged by
591 default, for both outbound TLS and client-certificate supplying inbound
594 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
597 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
598 Status Notification (bounce) messages are now MIME format per RFC 3464.
599 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
600 under the control of the dsn_advertise_hosts option, and routers may
601 have a dsn_lasthop option.
603 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
604 default, modifiable by a malware= option. The list separator for
605 the options can now be changed in the usual way. Bug 68.
607 JH/10 The smtp_receive_timeout main option is now expanded before use.
609 JH/11 The incoming_interface log option now also enables logging of the
610 local interface on delivery outgoing connections.
612 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
613 if the interface and destination host and port all match.
615 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
618 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
619 Patch from Andrew Lewis.
621 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
622 now supports optional time-restrictions, weighting, and priority
623 modifiers per server. Patch originally by <rommer@active.by>.
625 JH/16 The spamd_address main option now supports a mixed list of local
626 and remote servers. Remote servers can be IPv6 addresses, and
627 specify a port-range.
629 JH/17 Bug 68: The spamd_address main option now supports an optional
630 timeout value per server.
632 JH/18 Bug 1581: Router and transport options headers_add/remove can
633 now have the list separator specified.
635 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
638 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
641 JH/21 Support for the A6 type of dns record is withdrawn.
643 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
644 rather than the verbs used.
646 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
647 from 255 to 1024 chars.
649 JH/24 Verification callouts now attempt to use TLS by default.
651 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
652 are generic router options now. The defaults didn't change.
654 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
655 Original patch from Alexander Shikoff, worked over by JH.
657 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
658 files if ZCAT_COMMAND is not executable.
660 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
662 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
664 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
665 Normally benign, it bites when the pair was led to by a CNAME;
666 modern usage is to not canonicalize the domain to a CNAME target
667 (and we were inconsistent anyway for A-only vs AAAA+A).
669 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
671 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
672 when evaluating $sender_host_dnssec.
674 JH/31 Check the HELO verification lookup for DNSSEC, adding new
675 $sender_helo_dnssec variable.
677 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
679 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
681 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
683 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
684 documented as working, but never had. Support all but $spam_report.
686 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
687 added for tls authenticator.
689 HS/03 Add perl_taintmode main config option
695 TL/01 When running the test suite, the README says that variables such as
696 no_msglog_check are global and can be placed anywhere in a specific
697 test's script, however it was observed that placement needed to be near
698 the beginning for it to behave that way. Changed the runtest perl
699 script to read through the entire script once to detect and set these
700 variables, reset to the beginning of the script, and then run through
701 the script parsing/test process like normal.
703 TL/02 The BSD's have an arc4random API. One of the functions to induce
704 adding randomness was arc4random_stir(), but it has been removed in
705 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
706 function when detected.
708 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
709 cause callback expansion.
711 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
712 syntax errors in an expansion can be treated as a string instead of
713 logging or causing an error, due to the internal use of bool_lax
714 instead of bool when processing it.
716 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
717 server certificates when making smtp deliveries.
719 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
721 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
723 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
725 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
726 Merged patch from Sebastian Wiedenroth.
728 JH/05 Fix results-pipe from transport process. Several recipients, combined
729 with certificate use, exposed issues where response data items split
730 over buffer boundaries were not parsed properly. This eventually
731 resulted in duplicates being sent. This issue only became common enough
732 to notice due to the introduction of connection certificate information,
733 the item size being so much larger. Found and fixed by Wolfgang Breyha.
735 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
736 size buffer was used, resulting in syntax errors when an expansion
739 JH/07 Add support for directories of certificates when compiled with a GnuTLS
740 version 3.3.6 or later.
742 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
743 is EXPERIMENTAL_EVENT, the main-configuration and transport options
744 both become "event_action", the variables become $event_name, $event_data
745 and $event_defer_errno. There is a new variable $verify_mode, usable in
746 routers, transports and related events. The tls:cert event is now also
747 raised for inbound connections, if the main configuration event_action
750 TL/06 In test suite, disable OCSP for old versions of openssl which contained
751 early OCSP support, but no stapling (appears to be less than 1.0.0).
753 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
754 server certificate names available under the smtp transport option
755 "tls_verify_cert_hostname" now do not permit multi-component wildcard
758 JH/10 Time-related extraction expansions from certificates now use the main
759 option "timezone" setting for output formatting, and are consistent
760 between OpenSSL and GnuTLS compilations. Bug 1541.
762 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
763 encoded parameter in the incoming message. Bug 1558.
765 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
766 include certificate info, eximon was claiming there were spoolfile
769 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
771 JH/14 Log delivery-related information more consistently, using the sequence
772 "H=<name> [<ip>]" wherever possible.
774 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
775 are problematic for Debian distribution, omit them from the release
778 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
780 JH/16 Fix string representation of time values on 64bit time_t architectures.
783 JH/17 Fix a null-indirection in certextract expansions when a nondefault
784 output list separator was used.
789 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
790 checkers that were complaining about end of non-void function with no
793 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
794 This was a regression introduced in 4.83 by another bugfix.
796 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
798 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
799 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
805 TF/01 Correctly close the server side of TLS when forking for delivery.
807 When a message was received over SMTP with TLS, Exim failed to clear up
808 the incoming connection properly after forking off the child process to
809 deliver the message. In some situations the subsequent outgoing
810 delivery connection happened to have the same fd number as the incoming
811 connection previously had. Exim would try to use TLS and fail, logging
812 a "Bad file descriptor" error.
814 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
815 utilities have not been installed.
817 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
818 temporary space as the ACL may create new global variables.
820 TL/01 LDAP support uses per connection or global context settings, depending
821 upon the detected version of the libraries at build time.
823 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
824 to extract and use the src ip:port in logging and expansions as if it
825 were a direct connection from the outside internet. PPv2 support was
826 updated based on HAProxy spec change in May 2014.
828 JH/02 Add ${listextract {number}{list}{success}{fail}}.
830 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
831 Properly escape header and check for NULL return.
833 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
836 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
838 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
839 characters in header names, implemented as a verify condition.
840 Contributed by Michael Fischer v. Mollard.
842 TL/05 Rename SPF condition results err_perm and err_temp to standardized
843 results permerror and temperror. Previous values are deprecated but
844 still accepted. In a future release, err_perm and err_temp will be
845 completely removed, which will be a backward incompatibility if the
846 ACL tests for either of these two old results. Patch contributed by
847 user bes-internal on the mailing list.
849 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
851 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
852 selectors, in both main and reject logs.
854 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
857 JH/07 Add malware type "sock" for talking to simple daemon.
859 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
861 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
862 routers/transports under cutthrough routing.
864 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
865 numbers. Touch up "bool" conditional to keep the same definition.
867 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
869 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
871 JH/12 Expand items in router/transport headers_add or headers_remove lists
872 individually rather than the list as a whole. Bug 1452.
874 Required for reasonable handling of multiple headers_ options when
875 they may be empty; requires that headers_remove items with embedded
876 colons must have them doubled (or the list-separator changed).
878 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
879 view the policy declared in the DMARC record. Currently, $dmarc_status
880 is a combined value of both the record presence and the result of the
883 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
885 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
886 dnslookup router and the smtp transport (applying to the forward
889 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
890 of ldap servers used for a specific lookup. Patch provided by Heiko
893 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
894 New variable $lookup_dnssec_authenticated for observability.
896 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
897 Patch submitted by Lars Timman.
899 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
901 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
902 Requires trusted mode and valid format message id, aborts otherwise.
903 Patch contributed by Heiko Schlichting.
905 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
906 certextract with support for various fields. Bug 1358.
908 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
909 is requested by default, modifiable by smtp transport option
912 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
913 operate on certificate variables to give certificate fingerprints
914 Also new ${sha256:cert_variable}.
916 JH/23 The PRDR feature is moved from being Experimental into the mainline.
918 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
919 Christian Aistleitner.
921 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
923 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
924 file. Patch from Wolfgang Breyha.
926 JH/25 Expand the coverage of the delivery $host and $host_address to
927 client authenticators run in verify callout. Bug 1476.
929 JH/26 Port service names are now accepted for tls_on_connect_ports, to
930 align with daemon_smtp_ports. Bug 72.
932 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
933 support and error reporting did not work properly.
935 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
936 and is readable. Patch from Andrew Colin Kissa.
938 TL/14 Enhance documentation of ${run expansion and how it parses the
939 commandline after expansion, particularly in the case when an
940 unquoted variable expansion results in an empty value.
942 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
944 PP/02 Fix internal collision of T_APL on systems which support RFC3123
945 by renaming away from it. Addresses GH issue 15, reported by
948 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
950 TL/15 SECURITY: prevent double expansion in math comparison functions
951 (can expand unsanitized data). Not remotely exploitable.
958 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
960 PP/02 Make -n do something, by making it not do something.
961 When combined with -bP, the name of an option is not output.
963 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
966 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
967 $sender_host_name and config options to manage this, and basic check
970 PP/05 DSCP support for outbound connections and control modifier for inbound.
972 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
973 (Only plugin which currently uses this is kerberos4, which nobody should
974 be using, but we should make it available and other future plugins might
975 conceivably use it, even though it would break NAT; stuff *should* be
976 using channel bindings instead).
978 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
979 name; added for Sendmail compatibility; requires admin caller.
980 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
981 just ignore it); requires trusted caller.
982 Also parse but ignore: -Ac -Am -X<logfile>
985 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
987 TL/02 Add +smtp_confirmation as a default logging option.
989 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
990 Patch by Magnus Holmgren from 2007-02-20.
992 TL/04 Bugzilla 1281 - Spec typo.
993 Bugzilla 1283 - Spec typo.
994 Bugzilla 1290 - Spec grammar fixes.
996 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
998 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1000 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1001 mailing list by by Dmitry Isaikin.
1003 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1005 JH/02 Support "G" suffix to numbers in ${if comparisons.
1007 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1009 NM/01 Bugzilla 1197 - Spec typo
1010 Bugzilla 1196 - Spec examples corrections
1012 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1014 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1015 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1018 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1019 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1021 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1022 "acl {{name}{arg}...}", and optional args on acl condition
1025 JH/05 Permit multiple router/transport headers_add/remove lines.
1027 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1029 JH/07 Avoid using a waiting database for a single-message-only transport.
1030 Performance patch from Paul Fisher. Bugzilla 1262.
1032 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1035 JH/09 Add $headers_added variable, with content from use of ACL modifier
1036 add_header (but not yet added to the message). Bugzilla 199.
1038 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1039 Pulled from Bugzilla 817 by Wolfgang Breyha.
1041 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1043 (nb: this is the same fix as in Exim 4.80.1)
1045 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1048 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1051 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1053 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1054 advertise SMTP AUTH mechanism to us, instead of a generic
1055 protocol violation error. Also, make Exim more robust to bad
1056 data from the Dovecot auth socket.
1058 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1060 When a queue runner is handling a message, Exim first routes the
1061 recipient addresses, during which it prunes them based on the retry
1062 hints database. After that it attempts to deliver the message to
1063 any remaining recipients. It then updates the hints database using
1066 So if a recipient address works intermittently, it can get repeatedly
1067 deferred at routing time. The retry hints record remains fresh so the
1068 address never reaches the final cutoff time.
1070 This is a fairly common occurrence when a user is bumping up against
1071 their storage quota. Exim had some logic in its local delivery code
1072 to deal with this. However it did not apply to per-recipient defers
1073 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1075 This change adds a proper retry rule check during routing so that the
1076 final cutoff time is checked against the message's age. We only do
1077 this check if there is an address retry record and there is not a
1078 domain retry record; this implies that previous attempts to handle
1079 the address had the retry_use_local_parts option turned on. We use
1080 this as an approximation for the destination being like a local
1081 delivery, as in LMTP.
1083 I suspect this new check makes the old local delivery cutoff check
1084 redundant, but I have not verified this so I left the code in place.
1086 TF/02 Correct gecos expansion when From: is a prefix of the username.
1088 Test 0254 submits a message to Exim with the header
1092 When I ran the test suite under the user fanf2, Exim expanded
1093 the header to contain my full name, whereas it should have added
1094 a Resent-Sender: header. It erroneously treats any prefix of the
1095 username as equal to the username.
1097 This change corrects that bug.
1099 GF/01 DCC debug and logging tidyup
1100 Error conditions log to paniclog rather than rejectlog.
1101 Debug lines prefixed by "DCC: " to remove any ambiguity.
1103 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1105 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1106 Bug spotted by Jeremy Harris; was flawed since initial commit.
1107 Would have resulted in OCSP responses post-SNI triggering an Exim
1108 NULL dereference and crash.
1110 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1112 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1113 Bug detection, analysis and fix by Samuel Thibault.
1114 Bugzilla 1331, Debian bug #698092.
1116 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1118 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1119 Server implementation by Todd Lyons, client by JH.
1120 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1121 config variable "prdr_enable" controls whether the server
1122 advertises the facility. If the client requests PRDR a new
1123 acl_data_smtp_prdr ACL is called once for each recipient, after
1124 the body content is received and before the acl_smtp_data ACL.
1125 The client is controlled by both of: a hosts_try_prdr option
1126 on the smtp transport, and the server advertisement.
1127 Default client logging of deliveries and rejections involving
1128 PRDR are flagged with the string "PRDR".
1130 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1131 fclose(). Diagnosis by Todd Lyons.
1133 PP/17 Update configure.default to handle IPv6 localhost better.
1134 Patch by Alain Williams (plus minor tweaks).
1137 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1138 This is now consistent with GnuTLS, and is now documented: the
1139 previous undocumented portable approach to treating the option as
1140 unset was to force an expansion failure. That still works, and
1141 an empty string is now equivalent.
1143 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1144 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1145 not performing validation itself.
1147 PP/20 Added force_command boolean option to pipe transport.
1148 Patch from Nick Koston, of cPanel Inc.
1150 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1153 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1155 PP/21 Fix eximon continuous updating with timestamped log-files.
1156 Broken in a format-string cleanup in 4.80, missed when I repaired the
1157 other false fix of the same issue.
1158 Report and fix from Heiko Schlichting.
1161 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1162 Report from Prashanth Katuri.
1164 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1165 It's SecureTransport, so affects any MacOS clients which use the
1166 system-integrated TLS libraries, including email clients.
1168 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1169 using a MIME ACL for non-SMTP local injection.
1170 Report and assistance in diagnosis by Warren Baker.
1172 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1174 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1176 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1177 last id that failed so it may be referenced in subsequent ACL's.
1179 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1182 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1183 ldap library initialization, allowing self-signed CA's to be
1184 used. Also properly sets require_cert option later in code by
1185 using NULL (global ldap config) instead of ldap handle (per
1186 session). Bug diagnosis and testing by alxgomz.
1188 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1189 the src/util/ subdirectory.
1191 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1192 renamed to Transport Post Delivery Action by Jeremy Harris, as
1195 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1196 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1197 redis_servers = needs to be configured which will be used by the redis
1198 lookup. Patch from Warren Baker, of The Packet Hub.
1200 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1202 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1203 hostname or reverse DNS when processing a host list. Used suggestions
1204 from multiple comments on this bug.
1206 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1208 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1209 Missed a few lines, added it to make the runtest require no keyboard
1212 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1213 contains upper case chars. Make router use caseful_local_part.
1215 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1216 support when GnuTLS has been built with p11-kit.
1222 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1224 This, or similar/improved, will also be change PP/11 of 4.82.
1230 PP/01 Handle short writes when writing local log-files.
1231 In practice, only affects FreeBSD (8 onwards).
1232 Bugzilla 1053, with thanks to Dmitry Isaikin.
1234 NM/01 Bugzilla 949 - Documentation tweak
1236 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1239 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1241 PP/02 Implemented gsasl authenticator.
1243 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1245 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1246 `pkg-config foo` for cflags/libs.
1248 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1249 with rest of GSASL and with heimdal_gssapi.
1251 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1252 `pkg-config foo` for cflags/libs for the TLS implementation.
1254 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1255 properties get this fed in as external SSF. A number of robustness
1256 and debugging improvements to the cyrus_sasl authenticator.
1258 PP/08 cyrus_sasl server now expands the server_realm option.
1260 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1261 Patch by Jeremy Harris.
1263 PP/10 Added dbmjz lookup type.
1265 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1267 PP/12 MAIL args handles TAB as well as SP, for better interop with
1268 non-compliant senders.
1269 Analysis and variant patch by Todd Lyons.
1271 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1272 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1273 Patch from Dirk Mueller <dmueller@suse.com>
1275 PP/13 tls_peerdn now print-escaped for spool files.
1276 Observed some $tls_peerdn in wild which contained \n, which resulted
1277 in spool file corruption.
1279 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1280 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1281 or write after TLS renegotiation, which otherwise led to messages
1284 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1285 as a tracking header (ie: a signed header comes before the signature).
1286 Patch from Wolfgang Breyha.
1288 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1289 comma-sep list; embedded commas doubled.
1291 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1293 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1295 Report and patch from Dmitry Banschikov.
1297 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1298 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1299 failures. We appear to now support TLS1.1+ with Exim.
1301 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1302 lets Exim select keys and certificates based upon TLS SNI from client.
1303 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1304 before an outbound SMTP session. New log_selector, +tls_sni.
1306 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1307 NULL dereference. Report and patch from Alun Jones.
1309 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1310 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1311 Not seeing resolver debug output on NetBSD, but suspect this is a
1312 resolver implementation change.
1314 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1315 Left warnings. Added "eximon gdb" invocation mode.
1317 PP/21 Defaulting "accept_8bitmime" to true, not false.
1319 PP/22 Added -bw for inetd wait mode support.
1321 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1322 locate the relevant includes and libraries. Made this the default.
1324 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1325 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1327 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1328 This may cause build issues on older platforms.
1330 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1331 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1332 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1333 Added SNI support via GnuTLS too.
1334 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1336 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1338 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1339 Applied second patch from Janne, implementing suggestion to default
1340 multiple-strings-in-record handling to match SPF spec.
1342 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1344 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1345 read-only, out of scope).
1346 Patch from Wolfgang Breyha, report from Stuart Northfield.
1348 PP/29 Fix three issues highlighted by clang analyser static analysis.
1349 Only crash-plausible issue would require the Cambridge-specific
1350 iplookup router and a misconfiguration.
1351 Report from Marcin Mirosław.
1353 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1355 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1356 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1357 As part of this, removing so much warning spew let me fix some minor
1358 real issues in debug logging.
1360 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1361 assignment on my part. Fixed.
1363 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1364 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1365 Janne Snabb (who went above and beyond: thank you).
1367 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1368 string otherwise requires a connection and a bunch more work and it's
1369 relatively easy to get wrong. Should also expose TLS library linkage
1372 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1373 64-bit ${eval} (JH/03).
1375 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1376 GNU libc to support some of the 64-bit stuff, should not lead to
1377 conflicts. Defined before os.h is pulled in, so if a given platform
1378 needs to override this, it can.
1380 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1381 protection layer was required, which is not implemented.
1382 Bugzilla 1254, patch from Wolfgang Breyha.
1384 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1385 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1386 tls_dhparam take prime identifiers. Also unbreak combination of
1387 OpenSSL+DH_params+TLSSNI.
1389 PP/39 Disable SSLv2 by default in OpenSSL support.
1395 PP/01 Solaris build fix for Oracle's LDAP libraries.
1396 Bugzilla 1109, patch from Stephen Usher.
1398 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1400 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1403 TF/02 Fix a couple more cases where we did not log the error message
1404 when unlink() failed. See also change 4.74-TF/03.
1406 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1407 lock up or crash if it happened to be inside a call to libc when it
1408 got a SIGUSR1 from exiwhat.
1410 The SIGUSR1 handler appends the current process status to the process
1411 log which is later printed by exiwhat. It used to use the general
1412 purpose logging code to do this, but several functions it calls are
1413 not safe for signals.
1415 The new output code in the SIGUSR1 handler is specific to the process
1416 log, and simple enough that it's easy to inspect for signal safety.
1417 Removing some special cases also simplifies the general logging code.
1418 Removing the spurious timestamps from the process log simplifies
1421 TF/04 Improved ratelimit ACL condition.
1423 The /noupdate option has been deprecated in favour of /readonly which
1424 has clearer semantics. The /leaky, /strict, and /readonly update modes
1425 are mutually exclusive. The update mode is no longer included in the
1426 database key; it just determines when the database is updated. (This
1427 means that when you upgrade Exim will forget old rate measurements.)
1429 Exim now checks that the per_* options are used with an update mode that
1430 makes sense for the current ACL. For example, when Exim is processing a
1431 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1432 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1433 must specify per_mail/readonly. If you omit the update mode it defaults to
1434 /leaky where that makes sense (as before) or /readonly where required.
1436 The /noupdate option is now undocumented but still supported for
1437 backwards compatibility. It is equivalent to /readonly except that in
1438 ACLs where /readonly is required you may specify /leaky/noupdate or
1439 /strict/noupdate which are treated the same as /readonly.
1441 A useful new feature is the /count= option. This is a generalization
1442 of the per_byte option, so that you can measure the throughput of other
1443 aggregate values. For example, the per_byte option is now equivalent
1444 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1446 The per_rcpt option has been generalized using the /count= mechanism
1447 (though it's more complicated than the per_byte equivalence). When it is
1448 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1449 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1450 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1451 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1452 Note that using per_rcpt with a non-readonly update mode in more than
1453 one ACL will cause the recipients to be double-counted. (The per_mail
1454 and per_byte options don't have this problem.)
1456 The handling of very low rates has changed slightly. If the computed rate
1457 is less than the event's count (usually one) then this event is the first
1458 after a long gap. In this case the rate is set to the same as this event's
1459 count, so that the first message of a spam run is counted properly.
1461 The major new feature is a mechanism for counting the rate of unique
1462 events. The new per_addr option counts the number of different
1463 recipients that someone has sent messages to in the last time period. It
1464 behaves like per_rcpt if all the recipient addresses are different, but
1465 duplicate recipient addresses do not increase the measured rate. Like
1466 the /count= option this is a general mechanism, so the per_addr option
1467 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1468 example, measure the rate that a client uses different sender addresses
1469 with the options per_mail/unique=$sender_address. There are further
1470 details in the main documentation.
1472 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1474 TF/06 Removed a few PCRE remnants.
1476 TF/07 Automatically extract Exim's version number from tags in the git
1477 repository when doing development or release builds.
1479 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1480 Bugzilla 879. Patch from Paul Fisher.
1482 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1483 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1486 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1488 PP/05 Variable $av_failed, true if the AV scanner deferred.
1489 Bugzilla 1078. Patch from John Horne.
1491 PP/06 Stop make process more reliably on build failure.
1492 Bugzilla 1087. Patch from Heiko Schlittermann.
1494 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1495 Bugzilla 1089. Patch from Heiko Schlittermann.
1497 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1498 Bugzilla 1131. Patch from Holger Weiß.
1500 PP/09 Handle IPv6 addresses with SPF.
1501 Bugzilla 860. Patch from Wolfgang Breyha.
1503 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1505 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1508 PP/11 match_* no longer expand right-hand-side by default.
1509 New compile-time build option, EXPAND_LISTMATCH_RHS.
1510 New expansion conditions, "inlist", "inlisti".
1512 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1514 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1516 PP/14 fix log_write() format string regression from TF/03.
1517 Bugzilla 1152. Patch from Dmitry Isaikin.
1523 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1525 PP/02 Harmonised TLS library version reporting; only show if debugging.
1526 Layout now matches that introduced for other libraries in 4.74 PP/03.
1528 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1530 PP/04 New "dns_use_edns0" global option.
1532 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1535 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1536 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1538 TK/01 Updated PolarSSL code to 0.14.2.
1539 Bugzilla 1097. Patch from Andreas Metzler.
1541 PP/07 Catch divide-by-zero in ${eval:...}.
1542 Fixes bugzilla 1102.
1544 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1547 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1548 format-string attack -- SECURITY: remote arbitrary code execution.
1550 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1551 time unintentionally subject to list matching rules, letting the header
1552 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1553 arbitrary string expansion). This allowed for information disclosure.
1555 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1556 INT_MIN/-1 -- value coerced to INT_MAX.
1562 NM/01 Workaround for PCRE version dependency in version reporting
1565 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1566 This fixes portability to compilers other than gcc, notably
1567 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1569 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1570 makefiles for portability to HP-UX and POSIX correctness.
1572 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1573 Also via indented variable definition in the Makefile.
1574 (Debugging by Oliver Heesakkers).
1576 PP/02 Restore caching of spamd results with expanded spamd_address.
1577 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1579 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1580 Improves build reliability. Fix from: Frank Elsner
1582 NM/02 Fix wide character breakage in the rfc2047 coding
1583 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1585 NM/03 Allow underscore in dnslist lookups
1586 Fixes bug 1026. Patch from Graeme Fowler
1588 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1589 Code patches from Adam Ciarcinski of NetBSD.
1591 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1594 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1595 is logged, to avoid truncation. Patch from John Horne.
1597 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1598 Patch from Jakob Hirsch.
1600 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1601 SQL string expansion failure details.
1602 Patch from Andrey Oktyabrski.
1604 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1605 Patch from Simon Arlott.
1607 PP/09 New lookups functionality failed to compile on old gcc which rejects
1608 extern declarations in function scope.
1609 Patch from Oliver Fleischmann
1611 PP/10 Use sig_atomic_t for flags set from signal handlers.
1612 Check getgroups() return and improve debugging.
1613 Fixed developed for diagnosis in bug 927 (which turned out to be
1616 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1617 Patch from Mark Zealey.
1619 PP/12 Bugzilla 1056: Improved spamd server selection.
1620 Patch from Mark Zealey.
1622 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1623 Based on patch from Heiko Schlittermann.
1625 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1626 Patch from Uwe Doering, sign-off by Michael Haardt.
1628 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1629 variable declaration deep within a block. Bug and patch from
1632 PP/15 lookups-Makefile IRIX compatibility coercion.
1634 PP/16 Make DISABLE_DKIM build knob functional.
1636 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1637 Patch by Simon Arlott
1639 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1640 variable argument macros. Our copy now differs from upstream.
1646 TF/01 Failure to get a lock on a hints database can have serious
1647 consequences so log it to the panic log.
1649 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1650 controlled using the smtp_confirmation log selector.
1652 TF/03 Include the error message when we fail to unlink a spool file.
1654 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1655 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1656 for maintaining out-of-tree patches for some time.
1658 PP/01 Bugzilla 139: Documentation and portability issues.
1659 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1660 Handle per-OS dynamic-module compilation flags.
1662 PP/02 Let /dev/null have normal permissions.
1663 The 4.73 fixes were a little too stringent and complained about the
1664 permissions on /dev/null. Exempt it from some checks.
1665 Reported by Andreas M. Kirchwitz.
1667 PP/03 Report version information for many libraries, including
1668 Exim version information for dynamically loaded libraries. Created
1669 version.h, now support a version extension string for distributors
1670 who patch heavily. Dynamic module ABI change.
1672 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1673 privilege escalation vulnerability whereby the Exim run-time user
1674 can cause root to append content of the attacker's choosing to
1677 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1680 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1681 If dropping privileges for untrusted macros, we disabled normal logging
1682 on the basis that it would fail; for the Exim run-time user, this is not
1683 the case, and it resulted in successful deliveries going unlogged.
1684 Fixed. Reported by Andreas Metzler.
1690 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1691 only prepend for the Resent-* case. Fixes regression introduced in
1692 Exim 4.70 by NM/22 for Bugzilla 607.
1694 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1695 increasing numbers of administrators be bitten by this.
1697 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1699 PP/03 Bugzilla 994: added openssl_options main configuration option.
1701 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1703 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1705 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1707 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1708 without a peer certificate, leading to a segfault because of an
1709 assumption that peers always have certificates. Be a little more
1710 paranoid. Problem reported by Martin Tscholak.
1712 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1713 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1714 NB: ClamAV planning to remove STREAM in "middle of 2010".
1715 CL also introduces -bmalware, various -d+acl logging additions and
1716 more caution in buffer sizes.
1718 PP/09 Implemented reverse_ip expansion operator.
1720 PP/10 Bugzilla 937: provide a "debug" ACL control.
1722 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1724 PP/12 Bugzilla 973: Implement --version.
1726 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1728 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1730 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1732 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1733 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1734 ignore trailing whitespace.
1736 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1738 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1741 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1742 Notification from Dr Andrew Aitchison.
1744 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1745 ExtendedDetectionInfo response format.
1746 Notification from John Horne.
1748 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1751 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1752 XSL and documented dependency on system catalogs, with examples of how
1755 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1758 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1759 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1760 configuration file which is writeable by the Exim user or group.
1762 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1763 of configuration files to cover files specified with the -C option if
1764 they are going to be used with root privileges, not just the default
1767 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1768 option (effectively making it always true).
1770 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1771 files to be used while preserving root privileges.
1773 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1774 that rogue child processes cannot use them.
1776 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1777 run-time user, instead of root.
1779 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1780 Exim run-time user without dropping privileges.
1782 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1783 result string, instead of calling string_vformat() twice with the same
1786 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1787 for other users. Others should always drop root privileges if they use
1788 -C on the command line, even for a whitelisted configure file.
1790 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1792 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1798 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1799 $data_path, and $header_path variables; fixed documentation bugs and
1802 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1803 exipick to access non-standard spools, including the "frozen" queue
1806 NM/01 Bugzilla 965: Support mysql stored procedures.
1807 Patch from Alain Williams
1809 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1811 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1812 Patch from Andreas Metzler
1814 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1815 Patch from Kirill Miazine
1817 NM/05 Bugzilla 671: Added umask to procmail example.
1819 JJ/03 installed exipick 20100323.0, fixing doc bug
1821 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1822 directory. Notification and patch from Dan Rosenberg.
1824 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1826 TK/02 Improve log output when DKIM signing operation fails.
1828 MH/01 Treat the transport option dkim_domain as a colon separated
1829 list, not as a single string, and sign the message with each element,
1830 omitting multiple occurences of the same signer.
1832 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1833 Bugzilla 985, 986. Patch by Simon Arlott
1835 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1836 Patch by Simon Arlott
1838 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1839 MBX locking. Notification from Dan Rosenberg.
1845 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1847 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1849 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1851 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1853 NM/04 Bugzilla 915: Flag broken perl installation during build.
1859 TK/01 Added patch by Johannes Berg that expands the main option
1860 "spamd_address" if it starts with a dollar sign.
1862 TK/02 Write list of recipients to X-Envelope-Sender header when building
1863 the mbox-format spool file for content scanning (suggested by Jakob
1866 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1867 (http://www.dcc-servers.net/) support via dccifd. Activated by
1868 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1870 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1871 by Mark Daniel Reidel <mr@df.eu>.
1873 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1874 When building exim an external PCRE library is now needed -
1875 PCRE is a system library on the majority of modern systems.
1876 See entry on PCRE_LIBS in EDITME file.
1878 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1879 conversation. Added nologin parameter to request.
1880 Patch contributed by Kirill Miazine.
1882 TF/01 Do not log submission mode rewrites if they do not change the address.
1884 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1886 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1887 log files in place. Contributed by Roberto Lima.
1889 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1891 TF/03 Bugzilla 615: When checking the local_parts router precondition
1892 after a local_part_suffix or local_part_prefix option, Exim now
1893 does not use the address's named list lookup cache, since this
1894 contains cached lookups for the whole local part.
1896 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1897 Robert Millan. Documentation is in experimental-spec.txt.
1899 TF/04 Bugzilla 668: Fix parallel build (make -j).
1901 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1903 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1904 Patch provided by Jan Srzednicki.
1906 TF/05 Leading white space used to be stripped from $spam_report which
1907 wrecked the formatting. Now it is preserved.
1909 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1910 that they are available at delivery time.
1912 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1914 TF/08 TLS error reporting now respects the incoming_interface and
1915 incoming_port log selectors.
1917 TF/09 Produce a more useful error message if an SMTP transport's hosts
1918 setting expands to an empty string.
1920 NM/06 Bugzilla 744: EXPN did not work under TLS.
1921 Patch provided by Phil Pennock.
1923 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1924 Patch provided by Richard Godbee.
1926 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1927 acl_smtp_notquit, added index entry.
1929 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1930 Patch provided by Eugene Bujak.
1932 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1933 accept(). Patch provided by Maxim Dounin.
1935 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1936 Patch provided by Phil Pennock.
1938 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1940 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1941 Patch provided by Brad "anomie" Jorsch.
1943 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1944 Patch provided by Dean Brooks.
1946 TK/05 Add native DKIM support (does not depend on external libraries).
1948 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1949 Patch provided by Graeme Fowler.
1951 NM/16 Bugzilla 851: Documentation example syntax fix.
1953 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1955 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1958 NM/19 Bugzilla 745: TLS version reporting.
1959 Patch provided by Phil Pennock.
1961 NM/20 Bugzilla 167: bool: condition support.
1962 Patch provided by Phil Pennock.
1964 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1965 clients. Patch provided by Phil Pennock.
1967 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1968 Patch provided by Brad "anomie" Jorsch.
1970 NM/23 Bugzilla 687: Fix misparses in eximstats.
1971 Patch provided by Heiko Schlittermann.
1973 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1974 Patch provided by Heiko Schlittermann.
1976 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1977 plus update to original patch.
1979 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1981 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1982 Patch provided by David Brownlee.
1984 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1986 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1988 NM/30 Bugzilla 888: TLS documentation bugfixes.
1990 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1992 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1993 Unlike the original bugzilla I have changed all shell scripts in src tree.
1995 NM/33 Bugzilla 898: Transport filter timeout fix.
1996 Patch by Todd Rinaldo.
1998 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1999 Patch by Serge Demonchaux.
2001 NM/35 Bugzilla 39: Base64 decode bug fixes.
2002 Patch by Jakob Hirsch.
2004 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2006 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2008 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2010 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2016 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2017 ALT-N's libdkim that I have put here:
2018 http://duncanthrax.net/exim-experimental/
2020 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2021 were called 'true' and it seems that C99 defines that as a reserved
2022 keyword to be used with 'bool' variable types. That means you could
2023 not include C99-style headers which use bools without triggering
2024 build errors in sieve.c.
2026 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2027 as mailq or other aliases. Changed the --help handling significantly
2028 to do whats expected. exim_usage() emits usage/help information.
2030 SC/01 Added the -bylocaldomain option to eximstats.
2032 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2034 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2036 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2042 PH/01 Another patch from the Sieve maintainer.
2044 PH/02 When an IPv6 address is converted to a string for single-key lookup
2045 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2046 dots are used instead of colons so that keys in lsearch files need not
2047 contain colons. This was done some time before quoting was made available
2048 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2049 (notated using the quote facility) so as to distinguish them from IPv4
2050 keys. This meant that lookups for IP addresses in host lists did not work
2051 for iplsearch lookups.
2053 This has been fixed by arranging for IPv6 addresses to be expressed with
2054 colons if the lookup type is iplsearch. This is not incompatible, because
2055 previously such lookups could never work.
2057 The situation is now rather anomalous, since one *can* have colons in
2058 ordinary lsearch keys. However, making the change in all cases is
2059 incompatible and would probably break a number of configurations.
2061 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2064 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2065 conversion specification without a maximum field width, thereby enabling
2066 a rogue spamd server to cause a buffer overflow. While nobody in their
2067 right mind would setup Exim to query an untrusted spamd server, an
2068 attacker that gains access to a server running spamd could potentially
2069 exploit this vulnerability to run arbitrary code as the Exim user.
2071 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2072 $primary_hostname instead of what libspf2 thinks the hosts name is.
2074 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2075 a directory entry by the name of the lookup key. Previously, if a
2076 symlink pointed to a non-existing file or a file in a directory that
2077 Exim lacked permissions to read, a lookup for a key matching that
2078 symlink would fail. Now it is enough that a matching directory entry
2079 exists, symlink or not. (Bugzilla 503.)
2081 PH/03 The body_linecount and body_zerocount variables are now exported in the
2084 PH/04 Added the $dnslist_matched variable.
2086 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2087 This means they are set thereafter only if the connection becomes
2090 PH/06 Added the client_condition to authenticators so that some can be skipped
2091 by clients under certain conditions.
2093 PH/07 The error message for a badly-placed control=no_multiline_responses left
2094 "_responses" off the end of the name.
2096 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2098 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2099 (without spaces) instead of just copying the configuration text.
2101 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2103 PH/11 Added $max_received_linelength.
2105 PH/12 Added +ignore_defer and +include_defer to host lists.
2107 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2108 way in which PCRE > 7.0 is built.
2110 PH/14 Implemented queue_only_load_latch.
2112 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2113 MAIL command. The effect was to mangle the value on 64-bit systems.
2115 PH/16 Another patch from the Sieve maintainer.
2117 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2119 PH/18 If a system quota error occurred while trying to create the file for
2120 a maildir delivery, the message "Mailbox is full" was not appended to the
2121 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2122 only to a quota excession during the actual writing of the file.
2124 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2125 characters?) which causes problems in log lines. The DN values are now
2126 passed through string_printing() before being added to log lines.
2128 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2129 and InterBase are left for another time.)
2131 PH/21 Added message_body_newlines option.
2133 PH/22 Guard against possible overflow in moan_check_errorcopy().
2135 PH/23 POSIX allows open() to be a macro; guard against that.
2137 PH/24 If the recipient of an error message contained an @ in the local part
2138 (suitably quoted, of course), incorrect values were put in $domain and
2139 $local_part during the evaluation of errors_copy.
2145 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2146 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2149 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2150 issue a MAIL command.
2152 PH/02 In an ACL statement such as
2154 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2156 if a client was not listed at all, or was listed with a value other than
2157 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2158 the condition was not true (as it should be), so access was not denied.
2159 The bug was that the ! inversion was incorrectly passed on to the second
2160 item. This has been fixed.
2162 PH/03 Added additional dnslists conditions == and =& which are different from
2163 = and & when the dns lookup returns more than one IP address.
2165 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2166 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2168 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2169 FSYNC, which compiles an option called disable_fsync that allows for
2170 bypassing fsync(). The documentation is heavily laced with warnings.
2172 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2174 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2175 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2176 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2177 including adding "make clean"; (3) Added -fPIC when compiling the test
2178 dynamically loaded module, to get rid of a warning.
2180 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2181 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2182 = 0s. The bug is otherwise harmless.
2184 PH/07 There was a bug in the dovecot authenticator such that the value of
2185 $auth1 could be overwritten, and so not correctly preserved, after a
2186 successful authentication. This usually meant that the value preserved by
2187 the server_setid option was incorrect.
2189 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2191 PH/09 Installed PCRE release 7.0.
2193 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2194 run for batched SMTP input. It is now run at the start of every message
2195 in the batch. While fixing this I discovered that the process information
2196 (output by running exiwhat) was not always getting set for -bs and -bS
2197 input. This is fixed, and it now also says "batched" for BSMTP.
2199 PH/11 Added control=no_pipelining.
2201 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2202 patch, slightly modified), and move the expansion of helo_data till after
2203 the connection is made in the smtp transport (so it can use these
2206 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2208 PH/14 Added log_selector = +pid.
2210 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2212 PH/16 Add ${if forany and ${if forall.
2214 PH/17 Added dsn_from option to vary the From: line in DSNs.
2216 PH/18 Flush SMTP output before performing a callout, unless control =
2217 no_callout_flush is set.
2219 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2220 was true (the default) a successful delivery failed to delete the retry
2221 item, thus causing premature timeout of the address. The bug is now
2224 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2226 PH/21 Long custom messages for fakedefer and fakereject are now split up
2227 into multiline responses in the same way that messages for "deny" and
2228 other ACL rejections are.
2230 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2231 with slight modification.
2233 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2234 draft, changing the syntax and factoring some duplicate code".
2236 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2237 for deliveries of the second and subsequent messages over the same SMTP
2240 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2241 ${reduce, with only minor "tidies".
2243 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2245 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2246 expansion side effects.
2248 PH/27 When a message times out after an over-quota error from an Exim-imposed
2249 quota, the bounce message says "mailbox is full". This message was not
2250 being given when it was a system quota that was exceeded. It now should
2253 MH/03 Made $recipients available in local_scan(). local_scan() already has
2254 better access to the recipient list through recipients_list[], but
2255 $recipients can be useful in postmaster-provided expansion strings.
2257 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2258 in the case of a MAIL command with additional options following the
2259 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2260 were accidentally chopped off.
2262 PH/29 SMTP synchronization checks are implemented when a command is read -
2263 there is a check that no more input is waiting when there shouldn't be
2264 any. However, for some commands, a delay in an ACL can mean that it is
2265 some time before the response is written. In this time, more input might
2266 arrive, invalidly. So now there are extra checks after an ACL has run for
2267 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2268 pipelining has not been advertised.
2270 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2272 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2273 returned in the SMTP response when smtp_return_error_details was false.
2274 This has been fixed.
2276 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2277 instead of the C I/O that was originally supplied, because problems were
2278 reported on Solaris.
2280 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2281 Exim which did not show up earlier: it was assuming that a call to
2282 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2283 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2284 was a macro that became an assignment, so it seemed to work. This has
2285 changed to a proper function call with a void return, hence the compile
2286 error. Exim's code has been fixed.
2288 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2291 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2293 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2295 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2296 because it turns out to be too big or there is a timeout) there is no
2297 "Completed" line in the log. When some messages of this type were
2298 selected by exigrep, they were listed as "not completed". Others were
2299 picked up by some special patterns. I have improved the selection
2300 criteria to be more general.
2302 PH/38 The host_find_failed option in the manualroute router can now be set
2303 to "ignore", to completely ignore a host whose IP address cannot be
2304 found. If all hosts are ignored, the behaviour is controlled by the new
2305 host_all_ignored option.
2307 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2308 homing or because of multiple MX records with /mx) generated more than
2309 one IP address, and the following item turned out to be the local host,
2310 all the secondary addresses of the first item were incorrectly removed
2311 from the list, along with the local host and any following hosts (which
2312 is what is supposed to happen).
2314 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2315 whoever called Exim into the -H file. In the case of the daemon it was
2316 behaving confusingly. When first started, it used values for whoever
2317 started the daemon, but after a SIGHUP it used the Exim user (because it
2318 calls itself on a restart). I have changed the code so that it now always
2321 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2322 message are rejected with the same error (e.g. no authentication or bad
2323 sender address), and a DATA command is nevertheless sent (as can happen
2324 with PIPELINING or a stupid MUA), the error message that was given to the
2325 RCPT commands is included in the rejection of the DATA command. This is
2326 intended to be helpful for MUAs that show only the final error to their
2329 PH/42 Another patch from the Sieve maintainer.
2331 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2332 Eximstats - Fixed some broken HTML links and added missing column headers
2334 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2335 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2337 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2339 PH/43 Yet another patch from the Sieve maintainer.
2341 PH/44 I found a way to check for a TCP/IP connection going away before sending
2342 the response to the final '.' that terminates a message, but only in the
2343 case where the client has not sent further data following the '.'
2344 (unfortunately, this is allowed). However, in many cases there won't be
2345 any further data because there won't be any more messages to send. A call
2346 to select() can be used: if it shows that the input is "ready", there is
2347 either input waiting, or the socket has been closed. An attempt to read
2348 the next input character can distinguish the two cases. Previously, Exim
2349 would have sent an OK response which the client would never have see.
2350 This could lead to message repetition. This fix should cure that, at
2351 least in a lot of common cases.
2353 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2354 advertised in response to EHLO.
2360 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2361 fixed by 4.65/MH/01 (is this a record?) are fixed:
2363 (i) An empty string was always treated as zero by the numeric comparison
2364 operators. This behaviour has been restored.
2366 (ii) It is documented that the numeric comparison operators always treat
2367 their arguments as decimal numbers. This was broken in that numbers
2368 starting with 0 were being interpreted as octal.
2370 While fixing these problems I realized that there was another issue that
2371 hadn't been noticed. Values of message_size_limit (both the global option
2372 and the transport option) were treated as octal if they started with 0.
2373 The documentation was vague. These values are now always treated as
2374 decimal, and I will make that clear in the documentation.
2380 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2381 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2384 MH/01 Don't check that the operands of numeric comparison operators are
2385 integers when their expansion is in "skipping" mode (fixes bug
2386 introduced by 4.64-PH/07).
2388 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2389 child addresses, Exim now panics and dies. Previously, because the count
2390 is held in a short int, deliveries were likely to be lost. As such a
2391 large number of recipients for a single message is ridiculous
2392 (performance will be very, very poor), I have chosen to impose a limit
2393 rather than extend the field.
2399 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2400 leftover -K file (the existence of which was triggered by #402).
2401 While we were at it, introduced process PID as part of the -K
2402 filename. This should rule out race conditions when creating
2405 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2406 processing considerably. Previous code took too long for large mails,
2407 triggering a timeout which in turn triggers #401.
2409 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2410 in the DK code in transports.c. sendfile() is not really portable,
2411 hence the _LINUX specificness.
2413 TF/01 In the add_headers option to the mail command in an Exim filter,
2414 there was a bug that Exim would claim a syntax error in any
2415 header after the first one which had an odd number of characters
2418 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2419 callout verification, Exim cached a "reject" for the entire domain. This
2420 is correct for most verifications, but it is not correct for a recipient
2421 verification with use_sender or use_postmaster set, because in that case
2422 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2423 case of MAIL FROM:<> rejection from other early rejections (e.g.
2424 rejection of HELO). When verifying a recipient using a non-null MAIL
2425 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2426 Whatever the result of the callout, the value of the domain cache is
2427 left unchanged (for any other kind of callout, getting as far as trying
2428 RCPT means that the domain itself is ok).
2430 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2433 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2434 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2435 session, the daemon ignores ECONNECT errors and logs others; it now
2436 ignores EPIPE as well.
2438 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2439 (quoted-printable decoding).
2441 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2442 later the small subsequent patch to fix an introduced bug.
2444 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2446 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2448 PH/08 An error is now given if message_size_limit is specified negative.
2450 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2451 to be given (somewhat) arbitrary names.
2453 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2456 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2457 miscellaneous code fixes
2459 PH/10 Added the log_reject_target ACL modifier to specify where to log
2462 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2463 hostname. This is wrong, because it relates to the incoming message (and
2464 probably the interface on which it is arriving) and not to the outgoing
2465 callout (which could be using a different interface). This has been
2466 changed to use the value of the helo_data option from the smtp transport
2467 instead - this is what is used when a message is actually being sent. If
2468 there is no remote transport (possible with a router that sets up host
2469 addresses), $smtp_active_hostname is used.
2471 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2472 tweaks were necessary in order to get it to work (see also 21 below):
2473 (a) The code assumed that strncpy() returns a negative number on buffer
2474 overflow, which isn't the case. Replaced with Exim's string_format()
2476 (b) There were several signed/unsigned issues. I just did the minimum
2477 hacking in of casts. There is scope for a larger refactoring.
2478 (c) The code used strcasecmp() which is not a standard C function.
2479 Replaced with Exim's strcmpic() function.
2480 (d) The code set only $1; it now sets $auth1 as well.
2481 (e) A simple test gave the error "authentication client didn't specify
2482 service in request". It would seem that Dovecot has changed its
2483 interface. Fortunately there's a specification; I followed it and
2484 changed what the client sends and it appears to be working now.
2486 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2489 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2490 address A is aliased to B and C, where B exists and C does not. Without
2491 -v the output is "A verified" because verification stops after a
2492 successful redirection if more than one address is generated. However,
2493 with -v the child addresses are also verified. Exim was outputting "A
2494 failed to verify" and then showing the successful verification for C,
2495 with its parentage. It now outputs "B failed to verify", showing B's
2496 parentage before showing the successful verification of C.
2498 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2499 look up a TXT record in a specific list after matching in a combined
2502 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2503 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2504 they consult the DNS. I had assumed they would set it the way they
2505 wanted; and indeed my experiments on Linux seem to show that in some
2506 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2507 To be on the safe side, however, I have now made the interface to
2508 host_find_byname() similar to host_find_bydns(), with an argument
2509 containing the DNS resolver options. The host_find_byname() function now
2510 sets these options at its start, just as host_find_bydns() does. The smtp
2511 transport options dns_qualify_single and dns_search_parents are passed to
2512 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2513 of host_find_byname() use the default settings of RES_DEFNAMES
2514 (qualify_single) but not RES_DNSRCH (search_parents).
2516 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2517 spool_read_header() do less string testing, by means of a preliminary
2518 switch on the second character of optional "-foo" lines. (This is
2519 overdue, caused by the large number of possibilities that now exist.
2520 Originally there were few.) While I was there, I also converted the
2521 str(n)cmp tests so they don't re-test the leading "-" and the first
2522 character, in the hope this might squeeze out yet more improvement.
2524 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2525 flag allowing group syntax was set by the header_syntax check but not
2526 turned off, possible causing trouble later; (2) The flag was not being
2527 set at all for the header_verify test, causing "group"-style headers to
2528 be rejected. I have now set it in this case, and also caused header_
2529 verify to ignore an empty address taken from a group. While doing this, I
2530 came across some other cases where the code for allowing group syntax
2531 while scanning a header line wasn't quite right (mostly, not resetting
2532 the flag correctly in the right place). These bugs could have caused
2533 trouble for malformed header lines. I hope it is now all correct.
2535 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2536 with the "reply" argument non-NULL. The code, however (which originally
2537 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2538 but it didn't always do it. This confused somebody who was copying the
2539 code for some other use. I have removed all the tests.
2541 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2542 feature that was used to support insecure browsers during the U.S. crypto
2543 embargo. It requires special client support, and Exim is probably the
2544 only MTA that supported it -- and would never use it because real RSA is
2545 always available. This code has been removed, because it had the bad
2546 effect of slowing Exim down by computing (never used) parameters for the
2547 RSA_EXPORT functionality.
2549 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2550 authenticator to fail if there's a tab character in the incoming data
2551 (there should never be unless someone is messing about, as it's supposed
2552 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2553 if the connection is using TLS or if the remote IP is the same as the
2554 local IP, and the "valid-client-cert option" if a client certificate has
2557 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2558 authenticators. This can be used for authorization after authentication
2559 succeeds. (In the case of plaintext, it servers for both authentication
2562 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2563 if any retry times were supplied.
2565 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2566 connection, where there is no client IP address to check. In this
2567 situation, the verify now always succeeds.
2569 PH/25 Applied John Jetmore's -Mset patch.
2571 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2573 PH/27 In a string expansion for a processed (not raw) header when multiple
2574 headers of the same name were present, leading whitespace was being
2575 removed from all of them, but trailing whitespace was being removed only
2576 from the last one. Now trailing whitespace is removed from each header
2577 before concatenation. Completely empty headers in a concatenation (as
2578 before) are ignored.
2580 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2581 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2583 PH/29 [Removed. This was a change that I later backed out, and forgot to
2584 correct the ChangeLog entry (that I had efficiently created) before
2585 committing the later change.]
2587 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2588 address errors (4xx response to RCPT) over the same connection as other
2589 messages routed to the same hosts. Such deliveries are always "forced",
2590 so retry times are not inspected. This resulted in far too many retries
2591 for the affected addresses. The effect occurred only when there were more
2592 hosts than the hosts_max_try setting in the smtp transport when it had
2593 the 4xx errors. Those hosts that it had tried were not added to the list
2594 of hosts for which the message was waiting, so if all were tried, there
2595 was no problem. Two fixes have been applied:
2597 (i) If there are any address or message errors in an SMTP delivery, none
2598 of the hosts (tried or untried) are now added to the list of hosts
2599 for which the message is waiting, so the message should not be a
2600 candidate for sending over the same connection that was used for a
2601 successful delivery of some other message. This seems entirely
2602 reasonable: after all the message is NOT "waiting for some host".
2603 This is so "obvious" that I'm not sure why it wasn't done
2604 previously. Hope I haven't missed anything, but it can't do any
2605 harm, as the worst effect is to miss an optimization.
2607 (ii) If, despite (i), such a delivery is accidentally attempted, the
2608 routing retry time is respected, so at least it doesn't keep
2609 hammering the server.
2611 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2612 in ${readsocket because some servers need this prod.
2614 PH/32 Added some extra debug output when updating a wait-xxx database.
2616 PH/33 The hint "could be header name not terminated by colon", which has been
2617 given for certain expansion errors for a long time, was not being given
2618 for the ${if def:h_colon_omitted{... case.
2620 PH/34 The spec says: "With one important exception, whenever a domain list is
2621 being scanned, $domain contains the subject domain." There was at least
2622 one case where this was not true.
2624 PH/35 The error "getsockname() failed: connection reset by peer" was being
2625 written to the panic log as well as the main log, but it isn't really
2626 panic-worthy as it just means the connection died rather early on. I have
2627 removed the panic log writing for the ECONNRESET error when getsockname()
2630 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2631 runs only) independently of the message's sender address. This meant
2632 that, if the 4xx error was in fact related to the sender, a different
2633 message to the same recipient with a different sender could confuse
2634 things. In particular, this can happen when sending to a greylisting
2635 server, but other circumstances could also provoke similar problems.
2636 I have changed the default so that the retry time for these errors is now
2637 based a combination of the sender and recipient addresses. This change
2638 can be overridden by setting address_retry_include_sender=false in the
2641 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2642 remote server are returned as part of bounce messages. This was not
2643 happening for LMTP over a pipe (the lmtp transport), but now it is the
2644 same for both kinds of LMTP.
2646 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2647 in header lines that were in fact CNAMEs. This is no longer the case.
2649 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2650 and queue runs started by the daemon processed all messages. This has
2651 been fixed so that -R and -S can now usefully be given with -q<time>.
2653 PH/40 Import PCRE release 6.7 (fixes some bugs).
2655 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2657 PH/42 Give an error if -q is specified more than once.
2659 PH/43 Renamed the variables $interface_address and $interface_port as
2660 $received_ip_address and $received_port, to make it clear that these
2661 values apply to message reception, and not to the outgoing interface when
2662 a message is delivered. (The old names remain recognized, of course.)
2664 PH/44 There was no timeout on the connect() call when using a Unix domain
2665 socket in the ${readsocket expansion. There now is.
2667 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2668 be meaningful with "accept".
2670 SC/01 Eximstats V1.43
2671 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2673 SC/02 Eximstats V1.44
2674 Use a glob alias rather than an array ref in the generated
2675 parser. This improves both readability and performance.
2677 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2678 Collect SpamAssassin and rejection statistics.
2679 Don't display local sender or destination tables unless
2680 there is data to show.
2681 Added average volumes into the top table text output.
2683 SC/04 Eximstats V1.46
2684 Collect data on the number of addresses (recipients)
2685 as well as the number of messages.
2687 SC/05 Eximstats V1.47
2688 Added 'Message too big' to the list of mail rejection
2689 reasons (thanks to Marco Gaiarin).
2691 SC/06 Eximstats V1.48
2692 Mainlog lines which have GMT offsets and are too short to
2693 have a flag are now skipped.
2695 SC/07 Eximstats V1.49 (Alain Williams)
2696 Added the -emptyok flag.
2698 SC/08 Eximstats V1.50
2699 Fixes for obtaining the IP address from reject messages.
2701 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2702 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2703 whitespace changes from 4.64-PH/27
2705 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2708 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2709 are found, allow negative numbers in numeric criteria)
2711 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2713 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2714 to match changes made in 4.64-PH/43
2716 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2718 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2719 "domain" is actually an IP address, and force a failure. This locks out
2720 those revolvers/nameservers that support "A-for-A" lookups, in
2721 contravention of the specifications.
2723 PH/48 When a host name was looked up from an IP address, and the subsequent
2724 forward lookup of the name timed out, the host name was left in
2725 $sender_host_name, contrary to the specification.
2727 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2728 restricted to single-key lookups, Exim was not diagnosing an error if
2729 * or *@ was used with a query-style lookup.
2731 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2733 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2734 long ago, but noone interested enough thought of it. Let's just say that
2735 the "1.1" means that there are some new functions that weren't there at
2736 some point in the past.
2738 PH/51 Error processing for expansion failure of helo_data from an smtp
2739 transport during callout processing was broken.
2741 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2742 tested/used via the -bh/-bhc/-bs options.
2744 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2745 bug, fixed in subsequent PCRE releases).
2747 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2748 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2750 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2756 SC/01 Use a glob alias rather than an array ref in eximstats generated
2757 parser. This improves both readability and performance.
2759 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2760 Don't display local sender or destination tables in eximstats unless
2761 there is data to show.
2762 Added average volumes into the eximstats top table text output.
2764 SC/03 Collect data on the number of addresses (recipients) as well
2765 as the number of messages in eximstats.
2767 TF/01 Correct an error in the documentation for the redirect router. Exim
2768 does (usually) call initgroups() when daemonizing.
2770 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2771 with consistent privilege compared to when running as a daemon.
2773 TF/03 Note in the spec that $authenticated_id is not set for local
2774 submissions from trusted users.
2776 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2777 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2779 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2780 by adding some example configuration directives to the default
2781 configuration file. A little bit of work is required to uncomment the
2782 directives and define how usernames and passwords are checked, but
2783 there is now a framework to start from.
2785 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2786 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2787 without this. I don't know how relevant this is to other LDAP libraries.
2789 PH/02 Add the verb name to the "unknown ACL verb" error.
2791 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2793 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2795 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2796 directory not expanded when it should be if an expanded home directory
2797 was set for the address (which is overridden by the transport).
2799 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2802 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2803 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2804 because it is too late at that time, and has no effect.
2806 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2807 security issue with \' (bugzilla #107). I could not use the
2808 PQescapeStringConn() function, because it needs a PGconn value as one of
2811 PH/08 When testing addresses using -bt, indicate those final addresses that
2812 are duplicates that would not cause an additional delivery. At least one
2813 person was confused, thinking that -bt output corresponded to deliveries.
2814 (Suppressing duplicates isn't a good idea as you lose the information
2815 about possibly different redirections that led to the duplicates.)
2817 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2818 systems where poll() doesn't work, in particular OS X.
2820 PH/10 Added more information to debugging output for retry time not reached.
2822 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2823 operations in malware.c.
2825 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2828 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2829 syslog_duplication set false, log lines that would normally be written
2830 both the the main log and to the reject log were not written to syslog at
2833 PH/14 In the default configuration, change the use of "message" in ACL warn
2834 statements to "add_header".
2836 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2837 not followed by a command (e.g. "seen endif").
2839 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2840 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2843 PH/17 Added extra conditions to the default value of delay_warning_condition
2847 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2848 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2849 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2852 The Auto-Submitted: and various List- headers are standardised, whereas I
2853 don't think Precedence: ever was.
2855 PH/18 Refactored debugging code in route_finduser() to show more information,
2856 in particular, the error code if getpwnam() issues one.
2858 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2859 This is apparently needed in addition to the PH/07 change above to avoid
2860 any possible encoding problems.
2862 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2863 but not after initializing Perl.
2865 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2866 output them only if debugging. By default they are written stderr,
2867 apparently, which is not desirable.
2869 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2872 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2875 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2877 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2878 authenticated or an ident call has been made. Suppress the default
2879 values for $authenticated_id and $authenticated_sender (but permit -oMai
2880 and -oMas) when testing with -bh.
2882 PH/24 Re-jigged the order of the tests in the default configuration so that the
2883 tests for valid domains and recipients precede the DNS black list and CSA
2884 tests, on the grounds that those ones are more expensive.
2886 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2887 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2888 This bug exists in every version of Exim that I still have, right back to
2891 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2892 However, an attempt to turn on case-sensitivity in a regex key by
2893 including (?-i) didn't work because the subject string was already
2894 lowercased, and the effects were non-intuitive. It turns out that a
2895 one-line patch can be used to allow (?-i) to work as expected.
2901 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2902 other effects) broke the use of negated acl sub-conditions.
2904 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2907 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2908 "Deny" causes Exim to reject the incoming connection with a 554 error.
2909 Unfortunately, if there is a major crisis, such as a disk failure,
2910 tcp-wrappers gives "deny", whereas what one would like would be some
2911 kind of temporary error. A kludge has been added to help with this.
2912 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2913 554 error is used if errno is still zero or contains ENOENT (which occurs
2914 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2917 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2919 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2920 errors. Otherwise a message that provokes a temporary error (when other
2921 messages do not) can cause a whole host to time out.
2923 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2924 addresses were routed directly to files or pipes from a redirect router.
2925 File deliveries just didn't batch; pipe deliveries might have suffered
2928 PH/06 A failure to get a lock for a hints database would erroneously always say
2929 "Failed to get write lock", even when it was really a read lock.
2931 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2932 of 0600. This has been changed to use the value of the lockfile_mode
2933 option (which defaults to 0600).
2935 PH/08 Applied small patch from the Sieve maintainer.
2937 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2938 folder from quota calculations, a direct delivery into this folder messed
2939 up the contents of the maildirsize file. This was because the regex was
2940 used only to exclude .Trash (or whatever) when the size of the mailbox
2941 was calculated. There was no check that a delivery was happening into an
2942 excluded directory. This bug has been fixed by ignoring all quota
2943 processing for deliveries into excluded directories.
2945 PH/10 Added the maildirfolder_create_regex option to appendfile.
2951 PH/01 The code for finding all the local interface addresses on a FreeBSD
2952 system running IPv6 was broken. This may well have applied to all BSD
2953 systems, as well as to others that have similar system calls. The broken
2954 code found IPv4 interfaces correctly, but gave incorrect values for the
2955 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2956 that it would not match correctly against @[] and not recognize the IPv6
2959 PH/02 The ipliteral router was not recognizing addresses of the form user@
2960 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2962 PH/03 Added disable_ipv6.
2964 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2965 decoded form, because it is most often used to construct To: headers
2966 lines in autoreplies, and the decoded form may well be syntactically
2967 invalid. However, $reply_address has leading white space removed, and all
2968 newlines turned into spaces so that the autoreply transport does not
2971 PH/05 If group was specified without a user on a router, and no group or user
2972 was specified on a transport, the group from the router was ignored.
2974 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2975 for visible compile-time settings that can be used to change these
2976 numbers, for those that want even more. Backwards compatibility with old
2977 spool files has been maintained. However, going back to a previous Exim
2978 release will lost any variables that are in spool files.
2980 PH/07 Two small changes when running in the test harness: increase delay when
2981 passing a TCP/IP connection to a new process, in case the original
2982 process has to generate a bounce, and remove special handling of
2983 127.0.0.2 (sic), which is no longer necessary.
2985 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2986 be the same on different OS.
2988 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2991 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2992 whether --show-vars was specified or not
2994 JJ/02 exipick: Added support for new ACL variable spool format introduced
2997 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2998 syntactically invalid From: or Reply-to: line, and a filter used this to
2999 generate an autoreply, and therefore failed to obtain an address for the
3000 autoreply, Exim could try to deliver to a non-existent relative file
3001 name, causing unrelated and misleading errors. What now happens is that
3002 it logs this as a hard delivery error, but does not attempt to create a
3005 PH/11 The exinext utility has a -C option for testing purposes, but although
3006 the given file was scanned by exinext itself; it wasn't being passed on
3007 when Exim was called.
3009 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3010 an end-of-file indication when reading a command response.
3012 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3013 compiled. In many other places in Exim, IPv6 addresses are always
3014 recognized, so I have changed this. It also means that IPv4 domain
3015 literals of the form [IPV4:n.n.n.n] are now always recognized.
3017 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3018 used if the router is not running as root, for example, when verifying at
3019 ACL time, or when using -bh. The debugging output from this situation was
3020 non-existent - all you got was a failure to exec. I have made two
3023 (a) Failures to set uid/gid, the current directory, or a process leader
3024 in a subprocess such as that created by queryprogram now generate
3025 suitable debugging output when -d is set.
3027 (b) The queryprogram router detects when it is not running as root,
3028 outputs suitable debugging information if -d is set, and then runs
3029 the subprocess without attempting to change uid/gid.
3031 PH/15 Minor change to Makefile for building test_host (undocumented testing
3034 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3035 additional section of a DNS packet that returns MX or SRV records.
3036 Instead, it always explicitly searches for A/AAAA records. This avoids
3037 major problems that occur when a DNS server includes only records of one
3038 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3039 fixed another bug: if SRV records were looked up and the corresponding
3040 address records were *not* found in the additional section, the port
3041 values from the SRV records were lost.
3043 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3044 using the correct key (the original address) when searching the retry
3045 rules in order to find which one to use for generating the retry hint.
3047 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3048 adding the default one. Similarly, if it contains a Reply-To: header, the
3049 errors_reply_to option, if set, is not used.
3051 PH/19 When calculating a retry time, Exim used to measure the "time since
3052 failure" by looking at the "first failed" field in the retry record. Now
3053 it does not use this if it is later than than the arrival time of the
3054 message. Instead it uses the arrival time. This makes for better
3055 behaviour in cases where some deliveries succeed, thus re-setting the
3056 "first failed" field. An example is a quota failure for a huge message
3057 when small messages continue to be delivered. Without this change, the
3058 "time since failure" will always be short, possible causing more frequent
3059 delivery attempts for the huge message than are intended.
3060 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3062 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3063 $1, $2, $3) because the numerical variables can be reset during some
3064 expansion items (e.g. "match"), thereby losing the authentication data.
3066 PH/21 Make -bV show the size of off_t variables so that the test suite can
3067 decide whether to run tests for quotas > 2G.
3069 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3070 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3071 filecount value is greater than 2G or if a quota value is greater than 2G
3072 on a system where the size of off_t is not greater than 4, a panic error
3075 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3076 never match. The debug and -bh output now contains an explicit error
3077 message indicating a malformed IPv4 address or mask.
3079 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3080 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3081 PH/23 above applies.
3083 PH/25 Do not write to syslog when running in the test harness. The only
3084 occasion when this arises is a failure to open the main or panic logs
3085 (for which there is an explicit test).
3087 PH/26 Added the /no_tell option to "control=freeze".
3089 PH/27 If a host name lookup failed very early in a connection, for example, if
3090 the IP address matched host_lookup and the reverse lookup yielded a name
3091 that did not have a forward lookup, an error message of the form "no IP
3092 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3093 could be logged. Now it outputs the IP address instead of "NULL".
3095 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3096 allows the sender and the authenticated sender to be set when
3097 submitting a message from within Exim. Since child_open_exim() is
3098 documented for local_scan(), the new function should be too.
3100 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3101 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3102 results in an empty string is now treated as unset.
3104 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3106 PH/31 Added sender_verify_fail logging option.
3108 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3109 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3110 tidied the source and removed it altogether.
3112 PH/33 When a queue run was abandoned because the load average was too high, a
3113 log line was always written; now it is written only if the queue_run log
3114 selector is set. In addition, the log line for abandonment now contains
3115 information about the queue run such as the pid. This is always present
3116 in "start" and "stop" lines but was omitted from the "abandon" line.
3118 PH/34 Omit spaces between a header name and the colon in the error message that
3119 is given when verify = headers_syntax fails (if there are lots of them,
3120 the message gets confusing).
3122 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3123 names, as there are now some PTR records that contain slashes. This check
3124 is only to protect against broken name servers that fall over on strange
3125 characters, so the fact that it applies to all lookups doesn't matter.
3127 PH/36 Now that the new test suite is complete, we can remove some of the
3128 special code in Exim that was needed for the old test suite. For example,
3129 sorting DNS records because real resolvers return them in an arbitrary
3130 order. The new test suite's fake resolver always returns records in the
3133 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3134 bounces) except when queue_only is set, to avoid logging races between
3135 the different processes.
3137 PH/38 Panic-die if .include specifies a non-absolute path.
3139 PH/39 A tweak to the "H" retry rule from its user.
3141 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3142 a label. They prevented compilation on older perls.
3144 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3145 a warning to be raised on newish perls.
3147 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3148 on queue. Changes to match documented behaviour of showing count of
3149 messages matching specified criteria.
3151 PH/40 Changed the default ident timeout from 30s to 5s.
3153 PH/41 Added support for the use of login_cap features, on those BSD systems
3154 that have them, for controlling the resources used by pipe deliveries.
3156 PH/42 The content-scanning code uses fopen() to create files in which to put
3157 message data. Previously it was not paying any attention to the mode of
3158 the files. Exim runs with umask(0) because the rest of the code creates
3159 files with open(), and sets the required mode explicitly. Thus, these
3160 files were ending up world-writeable. This was not a big issue, because,
3161 being within the spool directory, they were not world-accessible. I have
3162 created a function called modefopen, which takes an additional mode
3163 argument. It sets umask(777), creates the file, chmods it to the required
3164 mode, then resets the umask. All the relevant calls to fopen() in the
3165 content scanning code have been changed to use this function.
3167 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3168 to 24 hours. This avoids potential overflow problems when processing G
3169 and H retry rules. I suspect nobody ever tinkers with this value.
3171 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3173 PH/45 When the plaintext authenticator is running as a client, the server's
3174 challenges are checked to ensure they are valid base64 strings. By
3175 default, the authentication attempt is cancelled if an invalid string is
3176 received. Setting client_ignore_invalid_base64 true ignores these errors.
3177 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3178 they are received. Thus, the responses can be made to depend on the
3179 challenges. If an invalid string is ignored, an empty string is placed in
3182 PH/46 Messages that are created by the autoreply transport now contains a
3183 References: header, in accordance with RFCs 2822 and 3834.
3185 PH/47 Added authenticated_sender_force to the smtp transport.
3187 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3189 PH/49 Installed latest patch from the Sieve maintainer.
3191 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3192 was also set, the appendfile transport was unnecessarily scanning a
3193 directory of message files (e.g. for maildir delivery) to find the count
3194 of files (along with the size), even though it did not need this
3195 information. It now does the scan only if it needs to find either the
3196 size of the count of files.
3198 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3200 PH/52 Two bugs concerned with error handling when the smtp transport is
3203 (i) Exim was not creating retry information for temporary errors given
3204 for individual recipients after the DATA command when the smtp transport
3205 was used in LMTP mode. This meant that they could be retried too
3206 frequently, and not timed out correctly.
3208 (ii) Exim was setting the flag that allows error details to be returned
3209 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3210 recipients that were returned after the DATA command.
3212 PH/53 This is related to PH/52, but is more general: for any failing address,
3213 when detailed error information was permitted to be returned to the
3214 sender, but the error was temporary, then after the final timeout, only
3215 "retry timeout exceeded" was returned. Now it returns the full error as
3216 well as "retry timeout exceeded".
3218 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3219 do this, and (what is worse) MTAs that accept it.
3221 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3222 will now be deprecated.
3224 PH/56 New os.c-cygwin from the Cygwin maintainer.
3226 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3227 formats (previously only available in exim formats via -bpr, -bpru,
3228 and -bpra. Now also available in native and exiqgrep formats)
3230 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3231 with very large, slow to parse queues
3233 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3235 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3237 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3238 responses to authentication challenges, though it was showing the
3239 challenges; (ii) I've removed the CR characters from the debug output for
3242 PH/58 Allow for the insertion of a newline as well as a space when a string
3243 is turned into more than one encoded-word during RFC 2047 encoding. The
3244 Sieve code now uses this.
3246 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3247 data_4xx, lost_connection, tls_required.
3249 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3250 message was being sent as an SMTP response.
3252 PH/61 Add -l and -k options to exicyclog.
3254 PH/62 When verifying, if an address was redirected to one new address, so that
3255 verification continued, and the new address failed or deferred after
3256 having set something in $address_data, the value of $address_data was not
3257 passed back to the ACL. This was different to the case when no
3258 redirection occurred. The value is now passed back in both cases.
3260 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3261 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3262 use login_cap.h, so on its own it isn't the distinguishing feature. The
3263 new name refers directly to the setclassresources() function.
3265 PH/65 Added configuration files for NetBSD3.
3267 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3269 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3270 is preferred over IPv4.
3272 PH/68 The bounce_return_message and bounce_return_body options were not being
3273 honoured for bounces generated during the reception of non-SMTP messages.
3274 In particular, this applied to messages rejected by the ACL. This bug has
3275 been fixed. However, if bounce_return_message is true and bounce_return_
3276 body is false, the headers that are returned for a non-SMTP message
3277 include only those that have been read before the error was detected.
3278 (In the case of an ACL rejection, they have all been read.)
3280 PH/69 The HTML version of the specification is now built in a directory called
3281 spec_html instead of spec.html, because the latter looks like a path with
3282 a MIME-type, and this confuses some software.
3284 PH/70 Catch two compiler warnings in sieve.c.
3286 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3287 function verify_get_ident() calls ip_connect() to connect a socket, but
3288 if the "connect()" function timed out, ip_connect() used to close the
3289 socket. However, verify_get_ident() also closes the socket later, and in
3290 between Exim writes to the log, which may get opened at this point. When
3291 the socket was closed in ip_connect(), the log could get the same file
3292 descriptor number as the socket. This naturally causes chaos. The fix is
3293 not to close the socket in ip_connect(); the socket should be closed by
3294 the function that creates it. There was only one place in the code where
3295 this was missing, in the iplookup router, which I don't think anybody now
3296 uses, but I've fixed it anyway.
3298 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3299 well as to direct DNS lookups. Otherwise the handling of names in host
3300 lists is inconsistent and therefore confusing.
3306 PH/01 Two changes to the default runtime configuration:
3308 (1) Move the checks for relay_from_hosts and authenticated clients from
3309 after to before the (commented out) DNS black list checks.
3311 (2) Add control=submission to the relay_from_hosts and authenticated
3312 clients checks, on the grounds that messages accepted by these
3313 statements are most likely to be submissions.
3315 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3317 (1) Generate an error if the third argument for the ${prvs expansion is
3320 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3323 (3) Reset the variables that are obtained from the first argument of
3324 ${prvscheck and used in the second argument before leaving the code,
3325 because their memory is reclaimed, so using them afterwards may do
3328 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3329 one (it's much easier than Tom thought :-).
3331 (5) Because of (4), we can now allow for the use of $prvscheck_result
3332 inside the third argument.
3334 PH/03 For some reason, the default setting of PATH when running a command from
3335 a pipe transport was just "/usr/bin". I have changed it to
3338 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3339 anything to be listed in the output from -bV.
3341 PH/05 When a filter generated an autoreply, the entire To: header line was
3342 quoted in the delivery log line, like this:
3344 => >A.N.Other <ano@some.domain> <original@ddress> ...
3346 This has been changed so that it extracts the operative address. There
3347 may be more than one such address. If so, they are comma-separated, like
3350 => >ano@some.domain,ona@other.domain <original@ddress> ...
3352 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3353 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3354 not being looked up in the rDNS to get a host name, Exim was showing the
3355 IP address twice in Received: lines, even though the IP addresses were
3356 identical. For example:
3358 Received: from [1.2.3.4] (helo=[1.2.3.4])
3360 However, if the real host name was known, it was omitting the HELO data
3361 if it matched the actual IP address. This has been tidied up so that it
3362 doesn't show the same IP address twice.
3364 PH/07 When both +timestamp and +memory debugging was on, the value given by
3365 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3366 called by the debug printing, thereby overwriting the timestamp buffer.
3367 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3369 PH/08 When the original message was included in an autoreply transport, it
3370 always said "this is a copy of the message, including all the headers",
3371 even if body_only or headers_only was set. It now gives an appropriate
3374 PH/09 Applied a patch from the Sieve maintainer which:
3376 o fixes some comments
3377 o adds the (disabled) notify extension core
3378 o adds some debug output for the result of if/elsif tests
3379 o points to the current vacation draft in the documentation
3380 and documents the missing references header update
3384 o fixes a bug in processing the envelope test (when testing
3385 multiple envelope elements, the last element determined the
3388 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3389 Electronic Mail") by including:
3391 Auto-submitted: auto-generated
3393 in the messages that it generates (bounce messages and others, such as
3394 warnings). In the case of bounce messages for non-SMTP messages, there was
3395 also a typo: it was using "Auto_submitted" (underscore instead of
3396 hyphen). Since every message generated by Exim is necessarily in response
3397 to another message, thes have all been changed to:
3399 Auto-Submitted: auto-replied
3401 in accordance with these statements in the RFC:
3403 The auto-replied keyword:
3405 - SHOULD be used on messages sent in direct response to another
3406 message by an automatic process,
3408 - MUST NOT be used on manually-generated messages,
3410 - MAY be used on Delivery Status Notifications (DSNs) and Message
3411 Disposition Notifications (MDNs),
3413 - MUST NOT be used on messages generated by automatic or periodic
3414 processes, except for messages which are automatic responses to
3417 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3418 to the default Received: header definition.
3420 PH/12 Added log selector acl_warn_skipped (default on).
3422 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3423 variables because (a) they are in the wrong storage pool and (b) even if
3424 they were copied, it wouldn't work properly because of the caching.
3426 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3427 checking when decoding. Apparently there are clients that generate
3428 overlong encoded strings. Why am I not surprised?
3430 PH/15 If the first argument of "${if match_address" was not empty, but did not
3431 contain an "@" character, Exim crashed. Now it writes a panic log message
3432 and treats the condition as false.
3434 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3436 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3437 extension "envelope-auth". The code is finished and in agreement with
3438 other implementations, but there is no documentation so far and in fact,
3439 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3440 not changing the active code.
3442 Print executed "if" and "elsif" statements when debugging is used. This
3443 helps a great deal to understand what a filter does.
3445 Document more things not specified clearly in RFC3028. I had all this
3446 sorted out, when out of a sudden new issues came to my mind. Oops."
3448 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3451 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3452 canonical form (as documented). However, after a host name lookup from
3453 the IP address, check_host() was doing a simple string comparison with
3454 addresses acquired from the DNS when checking that the found name did
3455 have the original IP as one of its addresses. Since any found IPv6
3456 addresses are likely to be in abbreviated form, the comparison could
3457 fail. Luckily, there already exists a function for doing the comparison
3458 by converting both addresses to binary, so now that is used instead of
3459 the text comparison.
3461 PH/20 There was another similar case to PH/19, when a complete host name was
3462 given in a host list; looking up its IP address could give an abbreviated
3463 form, whereas the current host's name might or might not be abbreviated.
3464 The same fix has been applied.
3470 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3471 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3474 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3475 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3477 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3479 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3480 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3481 Also, for "Auto-Submitted", treat anything other than "no" as
3482 non-personal, in accordance with RFC 3834. (Previously it treated
3483 anything starting "auto-" as non-personal.)
3485 TF/01 The control=submission/name=... option had a problem with syntax
3486 errors if the name included a slash character. The /name= option
3487 now slurps the rest of the string, so it can include any characters
3488 but it must come last in the list of options (after /sender_retain
3491 PH/05 Some modifications to the interface to the fake nameserver for the new
3499 TK/01 Added the "success_on_redirect" address verification option. See
3500 NewStuff for rationale and an example.
3502 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3504 PH/02 Patch to exigrep to allow it to work on syslog lines.
3506 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3507 fread() to skip over the body file's header line, because in Cygwin the
3508 header line is locked and is inaccessible.
3510 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3511 co-exist for some time) to make it clear that it is the Exim ID that is
3512 referenced, not the Message-ID: header line.
3514 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3515 string_format() function, because snprintf() does not exist on all
3518 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3519 ${stat: expansion item.
3521 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3522 protocol synchronization error", to keep the pedants happy.
3524 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3525 well as for IRIX systems, when gcc is being used. See the host.c source
3528 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3530 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3533 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3534 transport and to the smtp transport in LMTP mode.
3536 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3538 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3539 run a filter in a subprocess. This could lead to confusion in subsequent
3540 lookups in the parent process. There should also be a search_tidyup() at
3541 the end of the subprocess.
3543 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3544 only if the host matched helo_try_verify_hosts, which caused the
3545 verification to occur when the EHLO/HELO command was issued. The ACL just
3546 tested the remembered result. Now, if a previous verification attempt has
3547 not happened, "verify = helo" does it there and then.
3549 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3551 TK/03 Fix log output including CR from clamd.
3553 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3554 exist provoked a memory error which could cause a segfault.
3556 PH/15 Installed PCRE 6.2
3558 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3560 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3561 of the problem. Specifically, suggested +O2 rather than +O1 for the
3564 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3566 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3567 record, the port was not correctly logged when the outgoing_port log
3568 selector was set (it logged the transort's default port).
3570 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3571 fallback_hosts, and "hosts" in the smtp transport.
3573 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3574 host errors such as "Connection refused".
3576 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3577 authentication with radiusclient 0.4.9:
3579 - Error returned from rc_read_config was caught wrongly
3580 - Username/password not passed on to radius server due to wrong length.
3582 The presumption is that some radiusclient API changes for 4.51/PH/17
3583 were not taken care of correctly. The code is still untested by me (my
3584 Linux distribution still has 0.3.2 of radiusclient), but it was
3585 contributed by a Radius user.
3587 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3588 expanding the "port" option of the smtp transport.
3590 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3591 while reading a MIME header. Thanks to Tom Hughes for a patch.
3593 PH/24 Include config.h inside local_scan.h so that configuration settings are
3596 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3597 that in an ACL for RCPT (for example), you can examine exactly what was
3600 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3601 commands, but it was not correctly comparing the address with the actual
3602 client host address. Thus, it would show the EHLO address in Received:
3603 header lines when this was not necessary.
3605 PH/27 Added the % operator to ${eval:}.
3607 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3608 it should be ignoring failures (because with -C, for example, it has lost
3609 privilege). It wasn't ignoring creation failures other than "already
3612 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3615 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3616 because an input line was too long, either on its own, or by virtue of
3617 too many continuations, the temporary file was not being removed, and the
3618 return code was incorrect.
3620 PH/31 Missing "BOOL" in function definition in filtertest.c.
3622 PH/32 Applied Sieve patches from the maintainer.
3624 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3626 PH/33 Added "verify = not_blind".
3628 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3629 Local/Makefile (with some defaults set). These are used in built scripts
3630 such as exicyclog, but they have never been used in the exim_install
3631 script (though there are many overriding facilities there). I have
3632 arranged that the exim_install script now takes note of these two
3635 PH/35 Installed configuration files for Dragonfly.
3637 PH/36 When a locally submitted message by a trusted user did not contain a
3638 From: header, and the sender address was obtained from -f or from an SMTP
3639 MAIL command, and the trusted user did not use -F to supply a sender
3640 name, $originator_name was incorrectly used when constructing a From:
3641 header. Furthermore, $originator_name was used for submission mode
3642 messages from external hosts without From: headers in a similar way,
3643 which is clearly wrong.
3645 PH/37 Added control=suppress_local_fixups.
3647 PH/38 When log_selector = +received_sender was set, and the addition of the
3648 sender made the log line's construction buffer exactly full, or one byte
3649 less than full, an overflow happened when the terminating "\n" was
3652 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3653 when the result of a list match is failure because a DNS lookup failed.
3655 PH/40 RM_COMMAND is now used in the building process.
3657 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3658 the "build-* directories that it finds.
3660 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3661 address in a domain literal was a prefix of an interface address.
3663 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3664 when verifying a sender address, unless rewrite_headers is false.
3666 PH/44 Wrote a long comment about why errors_to addresses are verified as
3667 recipients, not senders.
3669 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3670 the ratelimit ACL was added.
3672 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3674 PH/46 Added extra information about PostgreSQL errors to the error string.
3676 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3677 avoiding the need to install special zones in a real server. This is
3678 backwards compatible; if it can't find the fake resolver, it drops back.
3679 Thus, both old and new test suites can be run.
3681 TF/02 Added util/ratelimit.pl
3683 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3684 clock is set back in time.
3686 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3687 Candler <B.Candler@pobox.com>.
3689 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3690 OK for addresses that are the result of redirections.
3692 PH/48 A number of further additions for the benefit of the new test suite,
3693 including a fake gethostbyname() that interfaces to the fake DNS resolver
3696 TF/06 The fix for widen_domains has also been applied to qualify_single and
3697 search_parents which are the other dnslookup options that can cause
3700 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3703 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3705 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3706 service" if it encountered an MX record with an empty target hostname.
3707 The message is now "an MX or SRV record indicated no SMTP service".
3709 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3710 if the DNS of the sending site is misconfigured. This is quite a
3711 common situation. This change restores the behaviour of treating a
3712 helo verification defer as a failure.
3714 PH/51 If self=fail was set on a router, the bounce message did not include the
3715 actual error message.
3721 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3723 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3724 command itself ended in error, the underlying message about the transport
3725 filter timeout was being overwritten with the pipe command error. Now the
3726 underlying error message should be appended to the second error message.
3728 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3730 PH/02 Reduce the amount of output that "make" produces by default. Full output
3731 can still be requested.
3733 PH/03 The warning log line about a condition test deferring for a "warn" verb
3734 was being output only once per connection, rather than after each
3735 occurrence (because it was using the same function as for successful
3736 "warn" verbs). This seems wrong, so I have changed it.
3738 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3739 it should not have, which might have caused a crash in the right
3740 circumstances, but probably never did.
3742 PH/04 Installed a modified version of Tony Finch's patch to make submission
3743 mode fix the return path as well as the Sender: header line, and to
3744 add a /name= option so that you can make the user's friendly name appear
3747 TF/03 Added the control = fakedefer ACL modifier.
3749 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3750 Mark Lowes for thorough testing.
3752 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3754 TK/03 Merged latest SRS patch from Miles Wilton.
3756 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3757 with the definition in sysexits.h (which is #included earlier).
3758 Fortunately, Exim does not actually use EX_OK. The code used to try to
3759 preserve the sysexits.h value, by assuming that macro definitions were
3760 scanned for macro replacements. I have been disabused of this notion,
3761 so now the code just undefines EX_OK before #including unistd.h.
3763 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3764 in the smtp transport. When a block could not be written in a single
3765 write() function, the timeout was being re-applied to each part-write.
3766 This seems wrong - if the receiver was accepting one byte at a time it
3767 would take for ever. The timeout is now adjusted when this happens. It
3768 doesn't have to be particularly precise.
3770 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3771 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3773 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3774 without a domain if the check to <postmaster@domain> fails.
3776 SC/01 Eximstats: added -xls and the ability to specify output files
3777 (patch written by Frank Heydlauf).
3779 SC/02 Eximstats: use FileHandles for outputting results.
3781 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3783 SC/04 Eximstats: fixed display of large numbers with -nvr option
3785 SC/05 Eximstats: fixed merging of reports with empty tables.
3787 SC/06 Eximstats: added the -include_original_destination flag
3789 SC/07 Eximstats: removed tabs and trailing whitespace.
3791 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3793 TK/06 MBOX spool code: Add real "From " MBOX separator line
3794 so the .eml file is really in mbox format (even though
3795 most programs do not really care). Patch from Alex Miller.
3797 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3798 The latter is generated from $received_to and is only set if the
3799 message has one envelope recipient. SA can use these headers,
3800 obviously out-of-the-box. Patch from Alex Miller.
3802 PH/08 The ${def test on a variable was returning false if the variable's
3803 value was "0", contrary to what the specification has always said!
3804 The result should be true unless the variable is empty.
3806 PH/09 The syntax error of a character other than { following "${if
3807 def:variable_name" (after optional whitespace) was not being diagnosed.
3808 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3809 accidental colon was present, for example, could give incorrect results.
3811 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3812 result is used (not including appendfile, where other changes are about
3815 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3816 This involved changing a lot of size variables from int to off_t. It
3817 should work with maildirs and everything.
3819 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3820 spamd dying while we are connected to it.
3822 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3825 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3826 function for BDB 4.3.
3828 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3830 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3831 the sources for PCRE have been re-arranged and more files are now
3834 PH/15 The code I had for printing potentially long long variables in PH/11
3835 above was not the best (it lost precision). The length of off_t variables
3836 is now inspected at build time, and an appropriate printing format (%ld
3837 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3838 to be "long long int" or "long int". This is needed for the internal
3839 formatting function string_vformat().
3841 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3842 the configuration file to be ":syslog", then the script "guesses" where
3843 the logs files are, rather than using the compiled in default. In our
3844 case the guess is not the same as the compiled default, so the script
3845 suddenly stopped working when I started to use syslog. The patch checks
3846 to see if log_file_path is "". If so, it attempts to read it from exim
3847 with no configuration file to get the compiled in version, before it
3848 falls back to the previous guessing code."
3850 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3851 implementing BATV in an Exim configuration. See NewStuff for the gory
3854 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3855 Makefile that are specific to HP-UX.
3857 PH/18 If the "use_postmaster" option was set for a recipient callout together
3858 with the "random" option, the postmaster address was used as the MAIL
3859 FROM address for the random test, but not for the subsequent recipient
3860 test. It is now used for both.
3862 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3863 patch removes a few documentation additions to RFC 3028, because the
3864 latest draft now contains them. It adds the new en;ascii-case comparator
3865 and a new error check for 8bit text in MIME parts. Comparator and
3866 require names are now matched exactly. I enabled the subaddress
3867 extension, but it is not well tested yet (read: it works for me)."
3869 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3870 rework some of the code of TK/09 above to avoid the hardwired use of
3871 "%lld" and "long long". Replaced the call to snprintf() with a call to
3874 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3875 records point to non-existent hosts", "retry timeout exceeded", and
3876 "retry time not reached for any host after a long failure period".
3878 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3879 experimental DomainKeys support:
3881 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3882 (2) On an error such as an illegally used "control", the wrong name for
3883 the control was given.
3885 These problems did NOT occur unless DomainKeys support was compiled.
3887 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3889 PH/24 Added ${if match_ip condition.
3891 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3892 databases so that it will be absolutely obvious if a crash occurs in the
3893 DB library. This is a regular occurrence (often caused by mis-matched
3896 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3897 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3898 picked up on a user's system that detects such things. There doesn't seem
3899 to be a gcc warning option for this - only an attribute that has to be
3900 put on the function's prototype. It seems that in Fedora Core 4 they have
3901 set this on a number of new functions. No doubt there will be more in due
3904 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3905 specify a transport. However, if an address that was verified by such a
3906 router was the subject of a callout, Exim crashed because it tried to
3907 read the rcpt_include_affixes from the non-existent transport. Now it
3908 just assumes that the setting of that option is false. This bug was
3909 introduced by 4.51/PH/31.
3911 PH/28 Changed -d+all to exclude +memory, because that information is very
3912 rarely of interest, but it makes the output a lot bigger. People tend to
3913 do -d+all out of habit.
3915 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3916 code in os-type was giving problems when libc.so lives in lib64, like on
3919 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3920 aren't the modern standard, and it seems that some systems' include files
3921 don't always have them. Exim was already checking for some of the newer
3922 ones like T_AAAA, and defining it itself. I've added checks for all the
3923 record types that Exim uses.
3925 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3926 not automatically generating a new one, as it is supposed to. This
3927 prevented TLS from working. If the file did exist, but contained invalid
3928 data, a new version was generated, as expected. It was only the case of a
3929 non-existent file that was broken.
3931 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3932 with a change in libdomainkeys > 0.64.
3934 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3935 from DNS. If the selector record carries the flag, it now has
3936 precedence over the domain-wide flag.
3938 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3940 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3941 the use of a port name that isn't defined in /etc/services) occurred, the
3942 message was deferred as in a normal delivery, and thus remained on the
3943 spool, instead of being failed because of the mua_wrapper setting. This
3944 is now fixed, and I tidied up some of the mua_wrapper messages at the
3947 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3948 the messages in a hash of arrays rather than using individual hashes.
3949 This is a bit cleaner and results in dramatic memory savings, albeit
3950 at a slight CPU cost.
3952 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3953 as requested by Marc Sherman.
3955 SC/10 Eximstats: added histograms for user specified patterns as requested
3958 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3960 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3961 fopen() in the content-scanning modules that did not already have it.
3967 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3968 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3970 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3972 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3974 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3975 to test Sieve filters that use "vacation".
3977 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3978 that changes the way the GnuTLS parameters are stored in the cache file.
3979 The new format can be generated externally. For backward compatibility,
3980 if the data in the cache doesn't make sense, Exim assumes it has read an
3981 old-format file, and it generates new data and writes a new file. This
3982 means that you can't go back to an older release without removing the
3985 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3986 work if there are any delivery delays because "one_time" forces the
3987 parent to be marked "delivered", so its unseen clone is never tried
3988 again. For this reason, Exim now forbids the simultaneous setting of
3991 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3992 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3993 entry if you want to know the details. The fix, however, neglected to
3994 consider the case when local delivery batching is involved. The test for
3995 "previously delivered" was not happening when checking to see if an
3996 address could be batched with a previous (undelivered) one; under
3997 certain circumstances this could lead to multiple deliveries to the same
4000 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4001 in its include files, and this causes problems building Exim.
4003 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4004 header_syntax) but Exim was just ignoring anything given after a slash.
4005 In particular, this caused confusion with an attempt to use "verify =
4006 reverse_host_lookup/defer_ok". An error is now given when options are
4007 supplied for verify items that do not have them. (Maybe reverse_host_
4008 lookup should have a defer_ok option, but that's a different point.)
4010 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4011 defined by RFC 821) to 2048, because there were problems with some AUTH
4012 commands, and RFC 1869 says the size should be increased for extended
4013 SMTP commands that take arguments.
4015 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4018 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4019 "unknown" error; now it says that the functionality isn't in the binary.
4021 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4022 an address' error message when a string expansion fails (syntax or
4023 whatever). Otherwise the password may appear in the log. Following change
4024 PH/42 below, there is no longer a chance of it appearing in a bounce
4027 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4029 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4030 of its addresses was ever tried. (Bugzilla bug #2.)
4032 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4033 the result incorrectly in the debug output. (It correctly added a newline
4034 to what was transported.)
4036 TF/01 Added $received_time.
4038 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4039 commented out examples of how to interface to a virus scanner and to
4040 SpamAssassin. Also added commented examples of av_scanner and
4041 spamd_address settings.
4043 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4044 and controls are allowed in which ACLs. There were a couple of minor
4045 errors. Some of the entries in the conditions table (which is a table of
4046 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4047 negation of where the condition IS allowed.
4049 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4051 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4052 header file does not have a version number, so I've had to invent a new
4053 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4054 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4055 radiusclient), but it was contributed by a Radius user.
4057 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4058 files or directories, for OpenSSL.
4060 PH/19 When an Exim process that is running as root has to create an Exim log
4061 file, it does so in a subprocess that runs as exim:exim so as to get the
4062 ownership right at creation (otherwise, other Exim processes might see
4063 the file with the wrong ownership). There was no test for failure of this
4064 fork() call, which would lead to the process getting stuck as it waited
4065 for a non-existent subprocess. Forks do occasionally fail when resources
4066 run out. I reviewed all the other calls to fork(); they all seem to check
4069 PH/20 When checking for unexpected SMTP input at connect time (before writing
4070 the banner), Exim was not dealing correctly with a non-positive return
4071 from the read() function. If the client had disconnected by this time,
4072 the result was a log entry for a synchronization error with an empty
4073 string after "input=" when read() returned zero. If read() returned -1
4074 (an event I could not check), uninitialized data bytes were printed.
4075 There were reports of junk text (parts of files, etc) appearing after
4078 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4080 PH/22 Added support for macro redefinition, and (re)definition in between
4081 driver and ACL definitions.
4083 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4084 forgetting to use the resulting value; it was using the unexpanded value.
4086 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4087 hadn't been configured. The fix is from Juergen Kreileder, who
4088 understands it better than I do:
4090 "Here's what I see happening with three configured cyrus_sasl
4091 authenticators configured (plain, login, cram-md5):
4093 On startup auth_cyrus_sasl_init() gets called for each of these.
4094 This means three calls to sasl_listmech() without a specified mech_list.
4095 => SASL tests which mechs of all available mechs actually work
4096 => three warnings about OTP not working
4097 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4099 With the patch, sasl_listmech() also gets called three times. But now
4100 SASL's mech_list option is set to the server_mech specified in the the
4101 authenticator. Or in other words, the answer from sasl_listmech()
4102 gets limited to just the mech you're testing for (which is different
4104 => the return list contains just 'plain' or 'login', 'cram-md5' or
4105 nothing depending on the value of ob->server_mech.
4107 I've just tested the patch: Authentication still works fine,
4108 unavailable mechs specified in the exim configuration are still
4109 caught, and the auth.log warnings about OTP are gone."
4111 PH/25 When debugging is enabled, the contents of the command line are added
4112 to the debugging output, even when log_selector=+arguments is not
4115 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4116 answer is "GNU", and only if the return is "GNU/something" is the answer
4119 PH/27 $acl_verify_message is now set immediately after the failure of a
4120 verification in an ACL, and so is available in subsequent modifiers. In
4121 particular, the message can be preserved by coding like this:
4123 warn !verify = sender
4124 set acl_m0 = $acl_verify_message
4126 Previously, $acl_verify_message was set only while expanding "message"
4127 and "log_message" when a very denied access.
4129 PH/28 Modified OS/os.c-Linux with
4131 -#ifndef OS_LOAD_AVERAGE
4132 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4134 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4135 nomenclature these days.)
4137 PH/29 Installed patch from the Sieve maintainer that adds the options
4138 sieve_useraddress and sieve_subaddress to the redirect router.
4140 PH/30 In these circumstances:
4141 . Two addresses routed to the same list of hosts;
4142 . First host does not offer TLS;
4143 . First host accepts first address;
4144 . First host gives temporary error to second address;
4145 . Second host offers TLS and a TLS session is established;
4146 . Second host accepts second address.
4147 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4148 and peerdn, if requested) that were in fact used only for the second
4151 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4152 attention to any local part prefix or suffix that was matched by the
4153 router that accepted the address. It now behaves in the same way as it
4154 does for delivery: the affixes are removed from the local part unless
4155 rcpt_include_affixes is set on the transport.
4157 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4158 timeout during the DATA phase of an incoming message.
4160 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4161 applied a patch sanctioned by the Sieve maintainer.
4163 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4164 the uid or gid is negative. A case of a negative gid caused this to be
4165 noticed. The fix allows for either to be negative.
4167 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4168 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4171 PH/36 The change PH/12 above was broken. Fixed it.
4173 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4174 the grounds that routing the same address twice would always produce the
4175 same answer. This might have been true once, but it is certainly no
4176 longer true now. Routing a child address may depend on the previous
4177 routing that produced that child. Some complicated redirection strategies
4178 went wrong when messages had multiple recipients, and made Exim's
4179 behaviour dependent on the order in which the addresses were given.
4181 I have moved the duplicate checking until after the routing is complete.
4182 Exim scans the addresses that are assigned to local and remote
4183 transports, and removes any duplicates. This means that more work will be
4184 done, as duplicates will always all be routed, but duplicates are
4185 presumably rare, so I don't expect this is of any significance.
4187 For deliveries to pipes, files, and autoreplies, the duplicate checking
4188 still happens during the routing process, since they are not going to be
4191 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4192 It corrects a timeout issue with spamd. This is Ian's comment: "The
4193 background is that sometimes spamd either never reads data from a
4194 connection it has accepted, or it never writes response data. The exiscan
4195 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4196 blindly assumes that writes won't block so it may never time out."
4198 PH/39 Allow G after quota size as well as K and M.
4200 PH/40 The value set for $authenticated_id in an authenticator may not contain
4201 binary zeroes or newlines because the value is written to log lines and
4202 to spool files. There was no check on this. Now the value is run through
4203 the string_printing() function so that such characters are converted to
4204 printable escape sequences.
4206 PH/41 $message_linecount is a new variable that contains the total number of
4207 lines in the message. Compare $body_linecount, which is the count for the
4210 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4211 bounce and delay warning messages, except in certain special cases, which
4214 (a) An SMTP error message from a remote host;
4215 (b) A message specified in a :fail: redirection;
4216 (c) A message specified in a "fail" command in a system filter;
4217 (d) A message specified in a FAIL return from the queryprogram router;
4218 (e) A message specified by the cannot_route_message router option.
4220 In these cases only, Exim does include the error details in bounce and
4221 warning messages. There are also a few cases where bland messages such
4222 as "unrouteable address" or "local delivery error" are given.
4224 PH/43 $value is now also set for the "else" part of a ${run expansion.
4226 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4227 being worked on, but at least Exim now implements the latest version to
4230 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4231 process to complete was treated as a delivery failure, a timeout while
4232 writing the message to the pipe was logged, but erroneously treated as a
4233 successful delivery. Such timeouts include transport filter timeouts. For
4234 consistency with the overall process timeout, these timeouts are now
4235 treated as errors, giving rise to delivery failures by default. However,
4236 there is now a new Boolean option for the pipe transport called
4237 timeout_defer, which, if set TRUE, converts the failures into defers for
4238 both kinds of timeout. A transport filter timeout is now identified in
4241 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4242 systems where "make" and "gmake" are different, calling "gmake" at top
4243 level broke things. I've arranged for the value of $(MAKE) to be passed
4244 from the Makefile to this script so that it can call the same version of
4248 A note about Exim versions 4.44 and 4.50
4249 ----------------------------------------
4251 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4252 changes of various kinds. As a consequence, a big documentation update was
4253 needed. This delayed the release for rather longer than seemed good, especially
4254 in the light of a couple of (minor) security issues. Therefore, the changes
4255 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4256 release. So 4.44 and 4.50 are in effect two different branches that both start
4259 I have left the 4.50 change log unchanged; it contains all the changes since
4260 4.43. The change log for 4.44 is below; many of its items are identical to
4261 those for 4.50. This seems to be the most sensible way to preserve the
4262 historical information.
4268 1. Minor wording change to the doc/README.SIEVE file.
4270 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4271 computation of the current number of files was incorrect.
4273 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4274 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4277 4. Give more explanation in the error message when the command for a transport
4278 filter fails to execute.
4280 5. There are several places where Exim runs a non-Exim command in a
4281 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4282 was being done only for the command run by the queryprogram router. It is
4283 now done for all such subprocesses. The other cases are: ${run, transport
4284 filters, and the commands run by the lmtp and pipe transports.
4286 6. Added CONFIGURE_GROUP build-time option.
4288 7. Some older OS have a limit of 256 on the maximum number of file
4289 descriptors. Exim was using setrlimit() to set 1000 as a large value
4290 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4291 systems. I've change it so that if it can't get 1000, it tries for 256.
4293 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4294 was an oversight, and furthermore, ever since the addition of extra
4295 controls (e.g. 4.43/32), the checks on when to allow different forms of
4296 "control" were broken. There should now be diagnostics for all cases when a
4297 control that does not make sense is encountered.
4299 9. Added the /retain_sender option to "control=submission".
4301 10. $recipients is now available in the predata ACL (oversight).
4303 11. Tidy the search cache before the fork to do a delivery from a message
4304 received from the command line. Otherwise the child will trigger a lookup
4305 failure and thereby defer the delivery if it tries to use (for example) a
4306 cached ldap connection that the parent has called unbind on.
4308 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4309 of $address_data from the recipient verification was clobbered by the
4310 sender verification.
4312 13. The value of address_data from a sender verification is now available in
4313 $sender_address_data in subsequent conditions in the ACL statement.
4315 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4317 15. Added a new option "connect=<time>" to callout options, to set a different
4320 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4321 was its contents. (It was OK if the option was not defined at all.)
4323 17. A "Completed" log line is now written for messages that are removed from
4324 the spool by the -Mrm option.
4326 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4327 information about exactly what failed.
4329 19. Added -dd to debug only the daemon process.
4331 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4332 handles timeouts, both on the server side and network timeouts. Renamed the
4333 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4335 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4336 It is now set to "smtps".
4338 22. $host_address is now set to the target address during the checking of
4339 ignore_target_hosts.
4341 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4342 being passed; this would have caused $sender_host_name to have been used if
4343 matching the list had actually called for a host name (not very likely,
4344 since this list is usually IP addresses). A host name is now passed as
4347 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4348 code that specifies a non-restarting handler (typically sigaction() in
4349 modern systems) in an attempt to fix a rare and obscure crash bug.
4351 25. Narrowed the window for a race in the daemon that could cause it to ignore
4352 SIGCHLD signals. This is not a major problem, because they are used only to
4353 wake it up if nothing else does.
4355 26. A malformed maildirsize file could cause Exim to calculate negative values
4356 for the mailbox size or file count. Odd effects could occur as a result.
4357 The maildirsize information is now recalculated if the size or filecount
4360 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4361 support for a long time. Removed HAVE_SYS_VFS_H.
4363 28. Installed the latest version of exipick from John Jetmore.
4365 29. In an address list, if the pattern was not a regular expression, an empty
4366 subject address (from a bounce message) matched only if the pattern was an
4367 empty string. Non-empty patterns were not even tested. This was the wrong
4368 because it is perfectly reasonable to use an empty address as part of a
4369 database query. An empty address is now tested by patterns that are
4370 lookups. However, all the other forms of pattern expect the subject to
4371 contain a local part and a domain, and therefore, for them, an empty
4372 address still always fails if the pattern is not itself empty.
4374 30. Exim went into a mad DNS loop when attempting to do a callout where the
4375 host was specified on an smtp transport, and looking it up yielded more
4376 than one IP address.
4378 31. Re-factored the code for checking spool and log partition space into a
4379 function that finds that data and another that does the check. The former
4380 is then used to implement four new variables: $spool_space, $log_space,
4381 $spool_inodes, and $log_inodes.
4383 32. The RFC2047 encoding function was originally intended for short strings
4384 such as real names; it was not keeping to the 75-character limit for
4385 encoded words that the RFC imposes. It now respects the limit, and
4386 generates multiple encoded words if necessary. To be on the safe side, I
4387 have increased the buffer size for the ${rfc2047: expansion operator from
4390 33. It is now permitted to omit both strings after an "if" condition; if the
4391 condition is true, the result is "true". As before, when the second string
4392 is omitted, a false condition yields an empty string. This makes it less
4393 cumbersome to write custom ACL and router conditions.
4395 34. Failure to deliver a bounce message always caused it to be frozen, even if
4396 there was an errors_to setting on the router. The errors_to setting is now
4399 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4400 canonical form (fully expanded) before being placed in
4401 $sender_host_address.
4403 36. The table in the code that translates DNS record types into text (T_A to
4404 "A" for instance) was missing entries for NS and CNAME. It is just possible
4405 that this could have caused confusion if both these types were looked up
4406 for the same domain, because the text type is used as part of Exim's
4407 per-process caching. But the chance of anyone hitting this buglet seems
4410 37. The dnsdb lookup has been extended in a number of ways.
4412 (1) There is a new type, "zns", which walks up the domain tree until it
4413 finds some nameserver records. It should be used with care.
4415 (2) There is a new type, "mxh", which is like "mx" except that it returns
4416 just the host names, not the priorities.
4418 (3) It is now possible to give a list of domains (or IP addresses) to be
4419 looked up. The behaviour when one of the lookups defers can be
4420 controlled by a keyword.
4422 (4) It is now possible to specify the separator character for use when
4423 multiple records are returned.
4425 38. The dnslists ACL condition has been extended: it is now possible to supply
4426 a list of IP addresses and/or domains to be looked up in a particular DNS
4429 39. Added log_selector=+queue_time_overall.
4431 40. When running the queue in the test harness, wait just a tad after forking a
4432 delivery process, to get repeatability of debugging output.
4434 41. Include certificate and key file names in error message when GnuTLS fails
4435 to set them up, because the GnuTLS error message doesn't include the name
4436 of the failing file when there is a problem reading it.
4438 42. Allow both -bf and -bF in the same test run.
4440 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4442 44. The "Exiscan patch" is now merged into the mainline Exim source.
4444 45. Sometimes the final signoff response after QUIT could fail to get
4445 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4446 < 0 before doing a fflush(). This bug looks as though it goes back to the
4447 introduction of TLS in release 3.20, but "sometimes" must have been rare
4448 because the tests only now provoked it.
4450 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4451 (this can affect the format of dates).
4453 47. exim_tidydb, when checking for the continued existence of a message for
4454 which it has found a message-specific retry record, was not finding
4455 messages that were in split spool directories. Consequently, it was
4456 deleting retry records that should have stayed in existence.
4458 48. Steve fixed some bugs in eximstats.
4460 49. The SPA authentication driver was not abandoning authentication and moving
4461 on to the next authenticator when an expansion was forced to fail,
4462 contradicting the general specification for all authenticators. Instead it
4463 was generating a temporary error. It now behaves as specified.
4465 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4466 (the order specifies the preference for clients). The order is now AES256,
4467 AES128, 3DES, ARCFOUR128.
4469 51. Small patch to Sieve code - explicitly set From: when generating an
4472 52. Exim crashed if a remote delivery caused a very long error message to be
4473 recorded - for instance if somebody sent an entire SpamAssassin report back
4474 as a large number of 550 error lines. This bug was coincidentally fixed by
4475 increasing the size of one of Exim's internal buffers (big_buffer) that
4476 happened as part of the Exiscan merge. However, to be on the safe side, I
4477 have made the code more robust (and fixed the comments that describe what
4480 53. Now that there can be additional text after "Completed" in log lines (if
4481 the queue_time_overall log selector is set), a one-byte patch to exigrep
4482 was needed to allow it to recognize "Completed" as not the last thing in
4485 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4486 patch that reportedly fixes this has been added. I am not expert enough to
4487 create a test for it. This is what the patch creator wrote:
4489 "I found a little strange behaviour of ldap code when working with
4490 Windows 2003 AD Domain, where users was placed in more than one
4491 Organization Units. When I tried to give exim partial DN, the exit code
4492 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4493 But simultaneously result of request was absolutely normal ldap result,
4494 so I produce this patch..."
4496 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4497 REFERENCE, so I have modified the code to exclude the patch when that macro
4500 55. Some experimental protocols are using DNS PTR records for new purposes. The
4501 keys for these records are domain names, not reversed IP addresses. The
4502 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4503 leaves it alone. Component reversal etc. now happens only for IP addresses.
4506 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4508 57. Double the size of the debug message buffer (to 2048) so that more of very
4509 long debug lines gets shown.
4511 58. The exicyclog utility now does better if the number of log files to keep
4512 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4514 59. Two changes related to the smtp_active_hostname option:
4516 (1) $smtp_active_hostname is now available as a variable.
4517 (2) The default for smtp_banner uses $smtp_active_hostname instead
4518 of $primary_hostname.
4520 60. The host_aton() function is supposed to be passed a string that is known
4521 to be a valid IP address. However, in the case of IPv6 addresses, it was
4522 not checking this. This is a hostage to fortune. Exim now panics and dies
4523 if the condition is not met. A case was found where this could be provoked
4524 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4525 components; fortuitously, this particular loophole had already been fixed
4526 by change 4.50/55 above.
4528 If there are any other similar loopholes, the new check in host_aton()
4529 itself should stop them being exploited. The report I received stated that
4530 data on the command line could provoke the exploit when Exim was running as
4531 exim, but did not say which command line option was involved. All I could
4532 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4533 running as the user.
4536 61. There was a buffer overflow vulnerability in the SPA authentication code
4537 (which came originally from the Samba project). I have added a test to the
4538 spa_base64_to_bits() function which I hope fixes it.
4541 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4542 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4544 63. The daemon start-up calls getloadavg() while still root for those OS that
4545 need the first call to be done as root, but it missed one case: when
4546 deliver_queue_load_max is set with deliver_drop_privilege. This is
4547 necessary for the benefit of the queue runner, because there is no re-exec
4548 when deliver_drop_privilege is set.
4550 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4551 This has been fixed.
4553 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4554 was in use, was not putting the data itself into the right store pool;
4555 consequently, it could be overwritten for a subsequent message in the same
4556 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4559 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4561 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4562 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4563 the function were treating the return as a boolean value, which happened to
4564 work because 0=false and not-0=true, but is not correct code.
4566 68. The host_aton() function was not handling scoped IPv6 addresses (those
4567 with, for example, "%eth0" on the end) correctly.
4569 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4570 negated items (that is, ~something) in unsigned ints. Some compilers
4571 apparently mutter when there is no cast.
4573 70. If an address verification called from an ACL failed, and did not produce a
4574 user-specific message (i.e. there was only a "system" message), nothing was
4575 put in $acl_verify_message. In this situation, it now puts the system
4578 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4579 session; change 4.31/43 added the unwanted input to the log line - except
4580 that it did not do this in the start of session case. It now does.
4582 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4583 This is wrong and can cause the other end to generate a synchronization
4584 error if it is another Exim or anything else that does the synchronization
4585 check. A QUIT command is no longer sent after a timeout.
4587 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4588 during host lookups.
4590 74. The defer_ok option of callout verification was not working if it was used
4591 when verifying addresses in header lines, that is, for this case:
4593 verify = header_sender/callout=defer_ok
4595 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4596 those file descriptors could be used for SMTP connections. If anything
4597 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4598 could be sent to the SMTP client, causing chaos. The daemon now opens
4599 stdin, stdout, and stderr to /dev/null when it puts itself into the
4602 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4603 log by default. The user can override this with suitable Perl magic.
4605 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4606 the log message when discard triggers, was not working for the DATA ACL or
4607 for the non-SMTP ACL.
4609 78. Error message wording change in sieve.c.
4611 79. If smtp_accept_max_per_host was set, the number of connections could be
4612 restricted to fewer than expected, because the daemon was trying to set up
4613 a new connection before checking whether the processes handling previous
4614 connections had finished. The check for completed processes is now done
4615 earlier. On busy systems, this bug wouldn't be noticed because something
4616 else would have woken the daemon, and it would have reaped the completed
4619 80. If a message was submitted locally by a user whose login name contained one
4620 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4621 It caused a spool format error. I have fixed the spool reading code. A
4622 related problem was that the "from" clause in the Received: line became
4623 illegal because of the space(s). It is now covered by ${quote_local_part.
4625 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4628 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4630 83. Patch from Sieve maintainer to fix unterminated string problem in
4631 "vacation" handling.
4633 84. Some minor changes to the Linux configuration files to help with other
4634 OS variants using glibc.
4636 85. One more patch for Sieve to update vacation handling to latest spec.
4639 ----------------------------------------------------
4640 See the note above about the 4.44 and 4.50 releases.
4641 ----------------------------------------------------
4647 1. Change 4.43/35 introduced a bug that caused file counts to be
4648 incorrectly computed when quota_filecount was set in an appendfile
4651 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4652 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4655 3. Give more explanation in the error message when the command for a transport
4656 filter fails to execute.
4658 4. There are several places where Exim runs a non-Exim command in a
4659 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4660 was being done only for the command run by the queryprogram router. It is
4661 now done for all such subprocesses. The other cases are: ${run, transport
4662 filters, and the commands run by the lmtp and pipe transports.
4664 5. Some older OS have a limit of 256 on the maximum number of file
4665 descriptors. Exim was using setrlimit() to set 1000 as a large value
4666 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4667 systems. I've change it so that if it can't get 1000, it tries for 256.
4669 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4670 was an oversight, and furthermore, ever since the addition of extra
4671 controls (e.g. 4.43/32), the checks on when to allow different forms of
4672 "control" were broken. There should now be diagnostics for all cases when a
4673 control that does not make sense is encountered.
4675 7. $recipients is now available in the predata ACL (oversight).
4677 8. Tidy the search cache before the fork to do a delivery from a message
4678 received from the command line. Otherwise the child will trigger a lookup
4679 failure and thereby defer the delivery if it tries to use (for example) a
4680 cached ldap connection that the parent has called unbind on.
4682 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4683 of $address_data from the recipient verification was clobbered by the
4684 sender verification.
4686 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4687 was its contents. (It was OK if the option was not defined at all.)
4689 11. A "Completed" log line is now written for messages that are removed from
4690 the spool by the -Mrm option.
4692 12. $host_address is now set to the target address during the checking of
4693 ignore_target_hosts.
4695 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4696 being passed; this would have caused $sender_host_name to have been used if
4697 matching the list had actually called for a host name (not very likely,
4698 since this list is usually IP addresses). A host name is now passed as
4701 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4702 code that specifies a non-restarting handler (typically sigaction() in
4703 modern systems) in an attempt to fix a rare and obscure crash bug.
4705 15. Narrowed the window for a race in the daemon that could cause it to ignore
4706 SIGCHLD signals. This is not a major problem, because they are used only to
4707 wake it up if nothing else does.
4709 16. A malformed maildirsize file could cause Exim to calculate negative values
4710 for the mailbox size or file count. Odd effects could occur as a result.
4711 The maildirsize information is now recalculated if the size or filecount
4714 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4715 support for a long time. Removed HAVE_SYS_VFS_H.
4717 18. Updated exipick to current release from John Jetmore.
4719 19. Allow an empty sender to be matched against a lookup in an address list.
4720 Previously the only cases considered were a regular expression, or an
4723 20. Exim went into a mad DNS lookup loop when doing a callout where the
4724 host was specified on the transport, if the DNS lookup yielded more than
4727 21. The RFC2047 encoding function was originally intended for short strings
4728 such as real names; it was not keeping to the 75-character limit for
4729 encoded words that the RFC imposes. It now respects the limit, and
4730 generates multiple encoded words if necessary. To be on the safe side, I
4731 have increased the buffer size for the ${rfc2047: expansion operator from
4734 22. Failure to deliver a bounce message always caused it to be frozen, even if
4735 there was an errors_to setting on the router. The errors_to setting is now
4738 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4739 canonical form (fully expanded) before being placed in
4740 $sender_host_address.
4742 24. Updated eximstats to version 1.33
4744 25. Include certificate and key file names in error message when GnuTLS fails
4745 to set them up, because the GnuTLS error message doesn't include the name
4746 of the failing file when there is a problem reading it.
4748 26. Expand error message when OpenSSL has problems setting up cert/key files.
4751 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4752 (this can affect the format of dates).
4754 28. exim_tidydb, when checking for the continued existence of a message for
4755 which it has found a message-specific retry record, was not finding
4756 messages that were in split spool directories. Consequently, it was
4757 deleting retry records that should have stayed in existence.
4759 29. eximstats updated to version 1.35
4760 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4761 1.35 - bugfix such that pie charts by volume are generated correctly
4763 30. The SPA authentication driver was not abandoning authentication and moving
4764 on to the next authenticator when an expansion was forced to fail,
4765 contradicting the general specification for all authenticators. Instead it
4766 was generating a temporary error. It now behaves as specified.
4768 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4769 (the order specifies the preference for clients). The order is now AES256,
4770 AES128, 3DES, ARCFOUR128.
4772 31. Small patch to Sieve code - explicitly set From: when generating an
4775 32. Exim crashed if a remote delivery caused a very long error message to be
4776 recorded - for instance if somebody sent an entire SpamAssassin report back
4777 as a large number of 550 error lines. This bug was coincidentally fixed by
4778 increasing the size of one of Exim's internal buffers (big_buffer) that
4779 happened as part of the Exiscan merge. However, to be on the safe side, I
4780 have made the code more robust (and fixed the comments that describe what
4783 33. Some experimental protocols are using DNS PTR records for new purposes. The
4784 keys for these records are domain names, not reversed IP addresses. The
4785 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4786 leaves it alone. Component reversal etc. now happens only for IP addresses.
4789 34. The host_aton() function is supposed to be passed a string that is known
4790 to be a valid IP address. However, in the case of IPv6 addresses, it was
4791 not checking this. This is a hostage to fortune. Exim now panics and dies
4792 if the condition is not met. A case was found where this could be provoked
4793 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4794 components; fortuitously, this particular loophole had already been fixed
4795 by change 4.50/55 or 4.44/33 above.
4797 If there are any other similar loopholes, the new check in host_aton()
4798 itself should stop them being exploited. The report I received stated that
4799 data on the command line could provoke the exploit when Exim was running as
4800 exim, but did not say which command line option was involved. All I could
4801 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4802 running as the user.
4805 35. There was a buffer overflow vulnerability in the SPA authentication code
4806 (which came originally from the Samba project). I have added a test to the
4807 spa_base64_to_bits() function which I hope fixes it.
4810 36. The daemon start-up calls getloadavg() while still root for those OS that
4811 need the first call to be done as root, but it missed one case: when
4812 deliver_queue_load_max is set with deliver_drop_privilege. This is
4813 necessary for the benefit of the queue runner, because there is no re-exec
4814 when deliver_drop_privilege is set.
4816 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4817 was in use, was not putting the data itself into the right store pool;
4818 consequently, it could be overwritten for a subsequent message in the same
4819 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4822 38. Sometimes the final signoff response after QUIT could fail to get
4823 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4824 < 0 before doing a fflush(). This bug looks as though it goes back to the
4825 introduction of TLS in release 3.20, but "sometimes" must have been rare
4826 because the tests only now provoked it.
4832 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4833 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4834 Now it returns an integer. A number of calls were still expecting a T/F
4835 return. Fortuitously, in all cases, the tests worked in OK situations,
4836 which is the norm. However, things would have gone wrong on any write
4837 failures on the smtp file descriptor. This function is used when sending
4838 messages over SMTP and also when doing verify callouts.
4840 2. When Exim is called to do synchronous delivery of a locally submitted
4841 message (the -odf or -odi options), it no longer closes stderr before doing
4844 3. Implemented the mua_wrapper option.
4846 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4848 5. Implemented the functions header_remove(), header_testname(),
4849 header_add_at_position(), and receive_remove_recipient(), and exported them
4852 6. If an ACL "warn" statement specified the addition of headers, Exim already
4853 inserted X-ACL-Warn: at the start if there was no header name. However, it
4854 was not making this test for the second and subsequent header lines if
4855 there were newlines in the string. This meant that an invalid header could
4856 be inserted if Exim was badly configured.
4858 7. Allow an ACL "warn" statement to add header lines at the start or after all
4859 the Received: headers, as well as at the end.
4861 8. Added the rcpt_4xx retry error code.
4863 9. Added postmaster_mailfrom=xxx to callout verification option.
4865 10. Added mailfrom=xxxx to the callout verification option, for verify=
4868 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4869 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4871 12. Inserted some casts to stop certain compilers warning when using pointer
4872 differences as field lengths or precisions in printf-type calls (mostly
4873 affecting debugging statements).
4875 13. Added optional readline() support for -be (dynamically loaded).
4877 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4878 same clock tick as a message's arrival, so that its received time was the
4879 same as the "first fail" time on the retry record, and that message
4880 remained on the queue past the ultimate address timeout, every queue runner
4881 would try a delivery (because it was past the ultimate address timeout) but
4882 after another failure, the ultimate address timeout, which should have then
4883 bounced the address, did not kick in. This was a "< instead of <=" error;
4884 in most cases the first failure would have been in the next clock tick
4885 after the received time, and all would be well.
4887 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4888 being recognized when the domain list was tested by the match_domain
4889 condition in an expansion string.
4891 16. Added the ${str2b64: operator.
4893 17. Exim was always calling setrlimit() to set a large limit for the number of
4894 processes, without checking whether the existing limit was already
4895 adequate. (It did check for the limit on file descriptors.) Furthermore,
4896 errors from getrlimit() and setrlimit() were being ignored. Now they are
4897 logged to the main and panic logs, but Exim does carry on, to try to do its
4898 job under whatever limits there are.
4900 18. Imported PCRE 5.0.
4902 19. Trivial typo in log message " temporarily refused connection" (the leading
4905 20. If the log selector return_path_on_delivery was set and an address was
4906 redirected to /dev/null, the delivery process crashed because it assumed
4907 that a return path would always be set for a "successful" delivery. In this
4908 case, the whole delivery is bypassed as an optimization, and therefore no
4911 21. Internal re-arrangement: the function for sending a challenge and reading
4912 a response while authentication was assuming a zero-terminated challenge
4913 string. It's now changed to take a pointer and a length, to allow for
4914 binary data in such strings.
4916 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4918 23. Exim was not respecting finduser_retries when seeking the login of the
4919 uid under which it was called; it was always trying 10 times. (The default
4920 setting of finduser_retries is zero.) Also, it was sleeping after the final
4921 failure, which is pointless.
4923 24. Implemented tls_on_connect_ports.
4925 25. Implemented acl_smtp_predata.
4927 26. If the domain in control=submission is set empty, Exim assumes that the
4928 authenticated id is a complete email address when it generates From: or
4929 Sender: header lines.
4931 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4932 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4933 chown and chgrp in /bin and hostname in /usr/bin.
4935 28. Exim was keeping the "process log" file open after each use, just as it
4936 does for the main log. This opens the possibility of it remaining open for
4937 long periods when the USR1 signal hits a daemon. Occasional processlog
4938 errors were reported, that could have been caused by this. Anyway, it seems
4939 much more sensible not to leave this file open at all, so that is what now
4942 29. The long-running daemon process does not normally write to the log once it
4943 has entered its main loop, and it closes the log before doing so. This is
4944 so that log files can straightforwardly be renamed and moved. However,
4945 there are a couple of unusual error situations where the daemon does write
4946 log entries, and I had neglected to close the log afterwards.
4948 30. The text of an SMTP error response that was received during a remote
4949 delivery was being truncated at 512 bytes. This is too short for some of
4950 the long messages that one sometimes sees. I've increased the limit to
4953 31. It is now possible to make retry rules that apply only when a message has a
4954 specific sender, in particular, an empty sender.
4956 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4957 it possible to be selective about when SMTP synchronization is enforced.
4959 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4961 32. Implemented hosts_connection_nolog.
4963 33. Added an ACL for QUIT.
4965 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4968 35. Added mailbox_size and mailbox_filecount to appendfile.
4970 36. Added control = no_multiline_responses to ACLs.
4972 37. There was a bug in the logic of the code that waits for the clock to tick
4973 in the case where the clock went backwards by a substantial amount such
4974 that the microsecond fraction of "now" was more than the microsecond
4975 fraction of "then" (but the whole seconds number was less).
4977 38. Added support for the libradius Radius client library this is found on
4978 FreeBSD (previously only the radiusclient library was supported).
4984 1. When certain lookups returned multiple values in the form name=value, the
4985 quoting of the values was not always being done properly. Specifically:
4986 (a) If the value started with a double quote, but contained no whitespace,
4988 (b) If the value contained whitespace other than a space character (i.e.
4989 tabs or newlines or carriage returns) it was not quoted.
4990 This fix has been applied to the mysql and pgsql lookups by writing a
4991 separate quoting function and calling it from the lookup code. The fix
4992 should probably also be applied to nisplus, ibase and oracle lookups, but
4993 since I cannot test any of those, I have not disturbed their existing code.
4995 2. A hit in the callout cache for a specific address caused a log line with no
4996 reason for rejecting RCPT. Now it says "Previous (cached) callout
4997 verification failure".
4999 3. There was an off-by-one bug in the queryprogram router. An over-long
5000 return line was truncated at 256 instead of 255 characters, thereby
5001 overflowing its buffer with the terminating zero. As well as fixing this, I
5002 have increased the buffer size to 1024 (and made a note to document this).
5004 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5005 when Exim is waiting for an SMTP response from a remote server, Exim
5006 restarts its select() call on the socket, thereby resetting its timeout.
5007 This is not a problem when such interrupts are rare. Somebody set up a cron
5008 job to run exiwhat every 2 minutes, which is less than the normal select()
5009 timeout (5 or 10 minutes). This meant that the select() timeout never
5010 kicked in because it was always reset. I have fixed this by comparing the
5011 time when an interrupt arrives with the time at the start of the first call
5012 to select(). If more time than the timeout has elapsed, the interrupt is
5013 treated as a timeout.
5015 5. Some internal re-factoring in preparation for the addition of Sieve
5016 extensions (by MH). In particular, the "personal" test is moved to a
5017 separate function, and given an option for scanning Cc: and Bcc: (which is
5018 not set for Exim filters).
5020 6. When Exim created an email address using the login of the caller as the
5021 local part (e.g. when creating a From: or Sender: header line), it was not
5022 quoting the local part when it contained special characters such as @.
5024 7. Installed new OpenBSD configuration files.
5026 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5027 try to make them clearer.
5029 9. Callout options, other than the timeout value, were being ignored when
5030 verifying sender addresses in header lines. For example, when using
5032 verify = header_sender/callout=no_cache
5034 the cache was (incorrectly) being used.
5036 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5037 only the Cygwin environment.
5039 11. When return_path_on_delivery was set as a log selector, if different remote
5040 addresses in the same message used different return paths and parallel
5041 remote delivery occurred, the wrong values would sometimes be logged.
5042 (Whenever a remote delivery process finished, the return path value from
5043 the most recently started remote delivery process was logged.)
5045 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5046 lines when AUTH and/or TLS are in use. This is the "received protocol"
5047 field. Exim used to use "asmtp" for authenticated SMTP, without any
5048 indication (in the protocol name) for TLS use. Now it follows the RFC and
5049 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5050 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5051 names appear in log lines as well as in Received: header lines.
5053 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5054 extensions, and comparison tests, and to fix some bugs.
5056 14. Changes to the "personal" filter test:
5058 (1) The test was buggy in that it was just doing the equivalent of
5059 "contains" tests on header lines. For example, if a user's address was
5060 anne@some.where, the "personal" test would incorrectly be true for
5062 To: susanne@some.where
5064 This test is now done by extracting each address from the header in turn,
5065 and checking the entire address. Other tests that are part of "personal"
5066 are now done using regular expressions (for example, to check local parts
5067 of addresses in From: header lines).
5069 (2) The list of non-personal local parts in From: addresses has been
5070 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5071 taken from the Sieve specification recommendations.
5073 (3) If the message contains any header line starting with "List-" it is
5074 treated as non-personal.
5076 (4) The test for "circular" in the Subject: header line has been removed
5077 because it now seems ill-conceived.
5079 15. Minor typos in src/EDITME comments corrected.
5081 16. Installed latest exipick from John Jetmore.
5083 17. If headers_add on a router specified a text string that was too long for
5084 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5085 of string_sprintf() is now avoided.
5087 18. $message_body_size was not set (it was always zero) when running the DATA
5088 ACL and the local_scan() function.
5090 19. For the "mail" command in an Exim filter, no default was being set for
5091 the once_repeat time, causing a random time value to be used if "once" was
5092 specified. (If the value happened to be <= 0, no repeat happened.) The
5093 default is now 0s, meaning "never repeat". The "vacation" command was OK
5094 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5095 (I found it when inspecting the code).
5097 20. There is now an overall timeout for performing a callout verification. It
5098 defaults to 4 times the callout timeout, which applies to individual SMTP
5099 commands during the callout. The overall timeout applies when there is more
5100 than one host that can be tried. The timeout is checked before trying the
5101 next host. This prevents very long delays if there are a large number of
5102 hosts and all are timing out (e.g. when the network connections are timing
5103 out). The value of the overall timeout can be changed by specifying an
5104 additional sub-option for "callout", called "maxwait". For example:
5106 verify = sender/callout=5s,maxwait=20s
5108 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5109 to the end before writing, but this should make it even safer).
5111 22. Exim was forgetting that it had advertised PIPELINING for the second and
5112 subsequent messages on an SMTP connection. It was also not resetting its
5113 memory on STARTTLS and an internal HELO.
5115 23. When Exim logs an SMTP synchronization error within a session, it now
5116 records whether PIPELINING has been advertised or not.
5118 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5119 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5120 rather than long int.
5122 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5124 26. Added the never_mail option to autoreply.
5130 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5131 crash if the getsockname() call failed; this can happen if a connection is
5132 closed very soon after it is established. The problem was simply in the
5133 order in which certain operations were done, causing Exim to try to write
5134 to the SMTP stream before it had set up the file descriptor. The bug has
5135 been fixed by making things happen in the correct order.
5141 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5142 before the connection was closed, thus losing the rejection response.
5144 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5145 some early Solaris releases, but causes trouble in current releases where
5146 socklen_t is defined.
5148 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5151 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5154 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5155 that releasing the top of it at the end releases what was used for sub-
5156 expansions (unless the block got too big). However, discard this block if
5157 the first thing is a variable or header, so that we can use its block when
5158 it is dynamic (useful for very large $message_headers, for example).
5160 6. Lookups now cache *every* query, not just the most recent. A new, separate
5161 store pool is used for this. It can be recovered when all lookup caches are
5162 flushed. Lookups now release memory at the end of their result strings.
5163 This has involved some general refactoring of the lookup sources.
5165 7. Some code has been added to the store_xxx() functions to reduce the amount
5166 of flapping under certain conditions.
5168 8. log_incoming_interface used to affect only the <= reception log lines. Now
5169 it causes the local interface and port to be added to several more SMTP log
5170 lines, for example "SMTP connection from", and rejection lines.
5172 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5174 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5176 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5177 could be overwritten at the end of the current message (or the start of a
5178 new message if it was set in a HELO ACL). The value is now preserved for
5179 the duration of the SMTP connection.
5181 12. If a transport had a headers_rewrite setting, and a matching header line
5182 contained an unqualified address, that address was qualified, even if it
5183 did not match any rewriting rules. The underlying bug was that the values
5184 of the flags that permit the existence of unqualified sender and recipient
5185 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5186 non-local messages, and by -bnq for local messages) were not being
5187 preserved with the message after it was received.
5189 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5190 "next input=" as part of the text comprising the host identity instead of
5191 the correct text. The code was using the same buffer for two different
5192 strings. However, depending on which order the printing function evaluated
5193 its arguments, the bug did not always show up. Under Linux, for example, my
5194 test suite worked just fine.
5196 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5197 doesn't work with some older versions of Perl. It has been changed to "my",
5198 which in any case is probably the better facility to use.
5200 15. A really picky compiler found some instances of statements for creating
5201 error messages that either had too many or two few arguments for the format
5204 16. The size of the buffer for calls to the DNS resolver has been increased
5205 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5206 for addresses that have a lot of PTR records. This alleviates a problem; it
5207 does not fully solve it.
5209 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5210 buffer now truncates the list and logs the incident, which is the same
5211 action as happens when Exim is looking up a host name and its aliases.
5212 Previously in this situation something unpredictable would happen;
5213 sometimes it was "internal error: store_reset failed".
5215 18. If a server dropped the connection unexpectedly when an Exim client was
5216 using GnuTLS and trying to read a response, the client delivery process
5217 crashed while trying to generate an error log message.
5219 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5220 string, for example:
5222 warn message = H1: something\nH2: something
5224 the text was added as a single header line from Exim's point of view
5225 though it ended up OK in the delivered message. However, searching for the
5226 second and subsequent header lines using $h_h2: did not work. This has been
5227 fixed. Similarly, if a system filter added multiple headers in this way,
5228 the routers could not see them.
5230 20. Expanded the error message when iplsearch is called with an invalid key to
5231 suggest using net-iplsearch in a host list.
5233 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5234 ACLs are no longer actually imposed (and a message to that effect is
5237 22. If a "gecos" field in a passwd entry contained escaped characters, in
5238 particular, if it contained a \" sequence, Exim got it wrong when building
5239 a From: or a Sender: header from that name. A second bug also caused
5240 incorrect handling when an unquoted " was present following a character
5241 that needed quoting.
5243 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5244 was not being matched caselessly.
5246 24. Arranged for all hyphens in the exim.8 source to be escaped with
5249 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5250 properly. Recipient callout cache records were still being keyed to include
5251 the sender, even when use_sender was set false. This led to far more
5252 callouts that were necessary. The sender is no longer included in the key
5253 when use_sender is false.
5255 26. Added "control = submission" modifier to ACLs.
5257 27. Added the ${base62d: operator to decode base 62 numbers.
5259 28. dnsdb lookups can now access SRV records.
5261 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5262 the configuration file.
5264 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5265 (-v) mode. This makes the output for a verbose queue run more intelligible.
5267 31. Added a use_postmaster feature to recipient callouts.
5269 32. Added the $body_zerocount variable, containing the number of binary zero
5270 bytes in the message body.
5272 33. The time of last modification of the "new" subdirectory is now used as the
5273 "mailbox time last read" when there is a quota error for a maildir
5276 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5278 35. Added +ignore_unknown as a special item in host lists.
5280 36. Code for decoding IPv6 addresses in host lists is now included, even if
5281 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5282 address was recognized as an IP address, but was then not correctly decoded
5283 into binary, causing unexpected and incorrect effects when compared with
5290 1. Very minor rewording of debugging text in manualroute to say "list of
5291 hosts" instead of "hostlist".
5293 2. If verify=header_syntax was set, and a header line with an unqualified
5294 address (no domain) and a large number of spaces between the end of the
5295 name and the colon was received, the reception process suffered a buffer
5296 overflow, and (when I tested it) crashed. This was caused by some obsolete
5297 code that should have been removed. The fix is to remove it!
5299 3. When running in the test harness, delay a bit after writing a bounce
5300 message to get a bit more predictability in the log output.
5302 4. Added a call to search_tidyup() just before forking a reception process. In
5303 theory, someone could use a lookup in the expansion of smtp_accept_max_
5304 per_host which, without the tidyup, could leave open a database connection.
5306 5. Added the variables $recipient_data and $sender_data which get set from a
5307 lookup success in an ACL "recipients" or "senders" condition, or a router
5308 "senders" option, similar to $domain_data and $local_part_data.
5310 6. Moved the writing of debug_print from before to after the "senders" test
5313 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5314 problems for message scanning, either using a data ACL, or using
5315 local_scan() because the Received: header was not generated till after they
5316 were called (in order to set the time as the time of reception completion).
5317 I have revised the way this works. The header is now generated after the
5318 body is received, but before the ACL or local_scan() are called. After they
5319 are run, the timestamp in the header is updated.
5325 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5326 before starting a queue runner without re-exec. This happened only when
5327 deliver_drop_privilege was set or when the Exim user was set to root. The
5328 effect of the bug was that timeouts during subsequent deliveries caused
5329 crashes instead of being properly handled. The handler is now left at its
5330 default (and expected) setting.
5332 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5333 message, again when deliver_drop_privilege is set or Exim is run as root.
5334 The bug described in (1) was not present in this case, but the tidying up
5335 of the other signals was missing. I have made the two cases consistent.
5337 3. The ignore_target_hosts setting on a manualroute router was being ignored
5338 for hosts that were looked up using the /MX notation.
5340 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5343 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5344 operated on the sender address. After changing the $sender_address to <>
5345 for the sender address verify, Exim was re-instated it as the original
5346 (before rewriting) address, but remembering that it had rewritten it, so it
5347 wasn't rewriting it again. This bug also had the effect of breaking the
5348 sender address verification caching when the sender address was rewritten.
5350 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5351 This has been changed so that if the ip literal address matches
5352 ignore_target_hosts, the router declines.
5354 7. Added expansion conditions match_domain, match_address, and match_local_
5355 part (NOT match_host).
5357 8. The placeholder for the Received: header didn't have a length field set.
5359 9. Added code to Exim itself and to exim_lock to test for a specific race
5360 condition that could lead to file corruption when using MBX delivery. The
5361 issue is with the lockfile that is created in /tmp. If this file is removed
5362 after a process has opened it but before that process has acquired a lock,
5363 there is the potential for a second process to recreate the file and also
5364 acquire a lock. This could lead to two Exim processes writing to the file
5365 at the same time. The added code performs the same test as UW imapd; it
5366 checks after acquiring the lock that its file descriptor still refers to
5367 the same named file.
5369 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5370 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5373 11. Added the smtp_active_hostname option. If used, this will typically be made
5374 to depend on the incoming interface address. Because $interface_address is
5375 not set up until the daemon has forked a reception process, error responses
5376 that can happen earlier (such as "too many connections") no longer contain
5379 12. If an expansion in a condition on a "warn" statement fails because a lookup
5380 defers, the "warn" statement is abandoned, and the next ACL statement is
5381 processed. Previously this caused the whole ACL to be aborted.
5383 13. Added the iplsearch lookup type.
5385 14. Added ident_timeout as a log selector.
5387 15. Added tls_certificate_verified as a log selector.
5389 16. Added a global option tls_require_ciphers (compare the smtp transport
5390 option of the same name). This controls incoming TLS connections.
5392 17. I finally figured out how to make tls_require_ciphers do a similar thing
5393 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5394 before starting the TLS session.
5396 18. Tabs are now shown as \t in -bP output.
5398 19. If the log selector return_path_on_delivery was set, Exim crashed when
5399 bouncing a message because it had too many Received: header lines.
5401 20. If two routers both had headers_remove settings, and the first one included
5402 a superfluous trailing colon, the final name in the first list and the
5403 first name in the second list were incorrectly joined into one item (with a
5404 colon in the middle).
5410 1. Added -C and -D options to the exinext utility, mainly to make it easier
5411 to include in the automated testing, but these could be helpful when
5412 multiple configurations are in use.
5414 2. The exinext utility was not formatting the output nicely when there was
5415 an alternate port involved in the retry record key, nor when there was a
5416 message id as well (for retries that were specific to a specific message
5417 and a specific host). It was also confused by IPv6 addresses, because of
5418 the additional colons they contain. I have fixed the IPv4 problem, and
5419 patched it up to do a reasonable job for IPv6.
5421 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5422 delivery, the log line now contains "pipelined" if PIPELINING was used.
5424 4. An SMTP transport process used to panic and die if the bind() call to set
5425 an explicit outgoing interface failed. This has been changed; it is now
5426 treated in the same way as a connect() failure.
5428 5. A reference to $sender_host_name in the part of a conditional expansion
5429 that was being skipped was still causing a DNS lookup. This no longer
5432 6. The def: expansion condition was not recognizing references to header lines
5433 that used bh_ and bheader_.
5435 7. Added the _cache feature to named lists.
5437 8. The code for checking quota_filecount in the appendfile transport was
5438 allowing one more file than it should have been.
5440 9. For compatibility with Sendmail, the command line option
5448 and sets the incoming protocol and host name (for trusted callers). The
5449 host name and its colon can be omitted when only the protocol is to be set.
5450 Note the Exim already has two private options, -pd and -ps, that refer to
5451 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5452 "s", but I don't think that's a major issue.
5454 10. A number of refactoring changes to the code, none of which should affect
5457 (a) The number of logging options was getting close to filling up the
5458 32-bit word that was used as a bit map. I have split them into two classes:
5459 those that are passed in the argument to log_write(), and those that are
5460 only ever tested independently outside of that function. These are now in
5461 separate 32-bit words, so there is plenty of room for expansion again.
5462 There is no change in the user interface or the logging behaviour.
5464 (b) When building, for example, log lines, the code previously used a
5465 macro that called string_cat() twice, in order to add two strings. This is
5466 not really sufficiently general. Furthermore, there was one instance where
5467 it was actually wrong because one of the argument was used twice, and in
5468 one call a function was used. (As it happened, calling the function twice
5469 did not affect the overall behaviour.) The macro has been replaced by a
5470 function that can join an arbitrary number of extra strings onto a growing
5473 (c) The code for expansion conditions now uses a table and a binary chop
5474 instead of a serial search (which was left over from when there were very
5475 few conditions). Also, it now recognizes conditions like "pam" even when
5476 the relevant support is not compiled in: a suitably worded error message is
5477 given if an attempt is made to use such a condition.
5479 11. Added ${time_interval:xxxxx}.
5481 12. A bug was causing one of the ddress fields not to be passed back correctly
5482 from remote delivery subprocesses. The field in question was not being
5483 subsequently used, so this caused to problems in practice.
5485 13. Added new log selectors queue_time and deliver_time.
5487 14. Might have fixed a bug in maildirsizefile handling that threw up
5488 "unexpected character" debug warnings, and recalculated the data
5489 unnecessarily. In any case, I expanded the warning message to give more
5492 15. Added the message "Restricted characters in address" to the statements in
5493 the default ACL that block characters like @ and % in local parts.
5495 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5496 Three changes have been made:
5498 (a) There was a serious bug; a negative response to MAIL caused the whole
5499 recipient domain to be cached as invalid, thereby blocking all messages
5500 to all local parts at the same domain, from all senders. This bug has
5501 been fixed. The domain is no longer cached after a negative response to
5502 MAIL if the sender used is not empty.
5504 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5507 (c) A new callout option, "use_sender" has been added for people who want
5508 the modified behaviour.
5514 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5517 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5518 indeed breaks things for older releases.
5520 3. Added additional logging to the case where there is a problem reading data
5521 from a filter that is running in a subprocess using a pipe, in order to
5522 try to track down a specific problem.
5524 4. Testing facility fudge: when running in the test harness and attempting
5525 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5526 getting "No route to host". Convert this to a timeout.
5528 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5531 6. Some OS don't have socklen_t but use size_t instead. This affects the
5532 fifth argument of getsockopt() amongst other things. This is now
5533 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5534 can be set for individual OS. I have set it for SunOS5, OSF1, and
5535 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5536 some earlier ones do not.
5538 7. Change 4.30/15 was not doing the test caselessly.
5540 8. The standard form for an IPv6 address literal was being rejected by address
5541 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5542 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5543 this, as well as the form without the "IPv6" on the front (but only when
5544 address literals are enabled, of course).
5546 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5548 10. Exim crashed if a message with an empty sender address specified by -f
5549 encountered a router with an errors_to setting. This could be provoked only
5550 by a command such as
5554 where an empty string was supplied; "<>" did not hit this bug.
5556 11. Installed PCRE release 4.5.
5558 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5559 remained set. It is now erased.
5561 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5562 times from message ids (which are base 36 rather than the normal 62).
5564 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5565 were being counted as actual protocol errors, and logged if the log
5566 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5567 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5568 and DATA following a set of rejected RCPTs do not count as protocol errors.
5569 In other words, Exim assumes they were pipelined, though this may not
5570 actually be the case. Of course, in all cases the client gets an
5571 appropriate error code.
5573 15. If a lookup fails in an ACL condition, a message about the failure may
5574 be available; it is used if testing the ACL cannot continue, because most
5575 such messages specify what the cause of the deferral is. However, some
5576 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5577 that caused an old message to be retained and used if a later statement
5578 caused a defer, replacing the real cause of the deferral.
5580 16. If an IP address had so many PTR records that the DNS lookup buffer
5581 was not large enough to hold them, Exim could crash while trying to process
5582 the truncated data. It now detects and logs this case.
5584 17. Further to 4.21/58, another change has been made: if (and only if) the
5585 first line of a message (the first header line) ends with CRLF, a bare LF
5586 in a subsequent header line has a space inserted after it, so as not to
5587 terminate the header.
5589 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5590 unnecessarily, used atoi() instead of strtol(), and didn't check the
5591 termination when getting file sizes from file names by regex.
5593 19. Completely re-implemented the support for maildirsize files, in the light
5594 of a number of problems with the previous contributed implementation
5595 (4.30/29). In particular:
5597 . If the quota is zero, the maildirsize file is maintained, but no quota is
5600 . If the maildir directory does not exist, it is created before any attempt
5601 to write a maildirsize file.
5603 . The quota value in the file is just a cache; if the quota is changed in
5604 the transport, the new value overrides.
5606 . A regular expression is available for excluding directories from the
5609 20. The autoreply transport checks the characters in options that define the
5610 message's headers; it allows continued headers, but it was checking with
5611 isspace() after an embedded newline instead of explicitly looking for a
5614 21. If all the "regular" hosts to which an address was routed had passed their
5615 expiry times, and had not reached their retry times, the address was
5616 bounced, even if fallback hosts were defined. Now Exim should go on to try
5619 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5620 equivalent code in the SMTP transport. Some hosts send humungous responses
5621 to HELO/EHLO, more than 1024 it seems.
5623 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5624 gives compiler warnings in some environments. I've now done it "properly",
5627 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5628 (because of problems with the built-in one) was declared to return uschar *
5629 instead of char *, causing compiler failure.
5631 25. Fixed a file descriptor leak when processing alias/forward files.
5633 26. Fixed a minor format string issue in dbfn.c.
5635 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5637 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5638 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5641 29. When a sender address is verified, it is cached, to save repeating the test
5642 when there is more than one recipient in a message. However, when the
5643 verification involves a callout, it is possible for different callout
5644 options to be set for different recipients. It is too complicated to keep
5645 track of this in the cache, so now Exim always runs a verification when a
5646 callout is required, relying on the callout cache for the optimization.
5647 The overhead is duplication of the address routing, but this should not be
5650 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5651 to be verified with callout=postmaster, and the main callout worked but the
5652 postmaster check failed, the verification correctly failed. However, if a
5653 subsequent RCPT command asked for sender verification *without* the
5654 postmaster check, incorrect caching caused this verification also to fail,
5657 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5658 it was not caching the DNS options (qualify_single, search_parents) that
5659 were used when the lookup failed. A subsequent lookup with different
5660 options therefore always gave the same answer, though there were cases
5661 where it should not have. (Example: a "domains = !$mx_any" option on a
5662 dnslookup router: the "domains" option is always processed without any
5663 widening, but the router might have qualify_single set.) Now Exim uses the
5664 cached value only when the same options are set.
5666 32. Added John Jetmore's "exipick" utility to the distribution.
5668 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5669 than a timeout (e.g. a certificate is required, and is not provided), an
5670 Exim server now closes the connection immediately. Previously it waited for
5671 the client to close - but if the client is SSL, it seems that they each
5672 wait for each other, leading to a delay before one of them times out.
5674 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5675 maintained 0.8.x compatibility because I don't think many are using it, and
5676 it is clearly obsolete.
5678 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5681 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5682 client certificate was expired. A simple patch fixes this, though I don't
5683 understand the full logic of why the verify callback is called multiple
5686 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5687 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5688 which causes problems with some clients (such as the Certicom SSL Plus
5689 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5690 disables the coutermeasure allowing Eudora to connect."
5692 38. Exim was not checking that a write() to a log file succeeded. This could
5693 lead to Bad Things if a log got too big, in particular if it hit a file
5694 size limit. Exim now panics and dies if it cannot write to a log file, just
5695 as it does if it cannot open a log file.
5697 39. Modified OS/Makefile-Linux so that it now contains
5699 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5701 The two -D definitions ensure that Exim is compiled with large file
5702 support, which makes it possible to handle log files that are bigger than
5705 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5706 instance) a domain was checked against a named list that involved a lookup,
5707 causing $domain_data to be set, then another domain was checked against the
5708 same list, then the first domain was re-checked, the value of $domain_data
5709 after the final check could be wrong. In particular, if the second check
5710 failed, it could be set empty. This bug probably also applied to
5713 41. The strip_trailing_dot option was not being applied to the address given
5714 with the -f command-line option.
5716 42. The code for reading a message's header from the spool was incrementing
5717 $received_count, but never initializing it. This meant that the value was
5718 incorrect (doubled) while delivering a message in the same process in which
5719 it was received. In the most common configuration of Exim, this never
5720 happens - a fresh exec is done - but it can happen when
5721 deliver_drop_privilege is set.
5723 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5724 it now includes up to 150 characters of the unexpected data in the log
5727 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5728 and building data strings. The size of both of these buffers was 10 000
5729 bytes - far larger than anybody would *ever* want, thought I. Needless to
5730 say, somebody hit the limit. I have increased the maximum line length to
5731 20 000 and the maximum data length of concatenated lines to 100 000. I have
5732 also fixed two bugs, because there was no checking on these buffers. Tsk,
5733 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5734 buffer is too small.
5736 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5737 lsearch lookups. Now it does.
5739 46. When parsing a route_list item in a manualroute router, a fixed-length
5740 buffer was used for the list of hosts. I made this 1024 bytes long,
5741 thinking that nobody would ever have a list of hosts that long. Wrong.
5742 Somebody had a whole pile of complicated expansion conditions, and the
5743 string was silently truncated, leading to an expansion error. It turns out
5744 that it is easier to change to an unlimited length (owing to other changes
5745 that have happened since this code was originally written) than to build
5746 structure for giving a limitation error. The length of the item that
5747 expands into the list of hosts is now unlimited.
5749 47. The lsearch lookup could not handle data where the length of text line was
5750 more than 4095 characters. Such lines were truncated, leading to shortened
5751 data being returned. It should now handle lines of any length.
5753 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5754 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5757 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5758 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5759 no longer happens in comment lines. A list of replacements is now placed
5760 at the head of all of the source files, except those whose only change is
5761 to replace PERL_COMMAND in the very first #! line.
5763 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5764 messages on the queue, with a bottom-up merge sort, using code contributed
5765 by Michael Haardt. This should make operations like -bp somewhat faster on
5766 large queues. It won't affect queue runners, except when queue_run_in_order
5769 51. Installed eximstats 1.31 in the distribution.
5771 52. Added support for SRV lookups to the dnslookup router.
5773 53. If an ACL referred to $message_body or $message_body_end, the value was not
5774 reset for any messages that followed in the same SMTP session.
5776 54. The store-handling optimization for building very long strings was not
5777 differentiating between the different store pools. I don't think this
5778 actually made any difference in practice, but I've tidied it.
5780 55. While running the routers to verify a sender address, $sender_address
5781 was still set to the sender address. This is wrong, because when routing to
5782 send a bounce to the sender, it would be empty. Therefore, I have changed
5783 it so that, while verifying a sender address, $sender_address is set to <>.
5784 (There is no change to what happens when verifying a recipient address.)
5786 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5787 target A or AAAA records (if not already returned) without resetting the
5788 qualify_single or search_parents options of the DNS resolver. These are
5789 inappropriate in this case because the targets of MX and SRV records must
5790 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5791 target that, when qualified, matched something in the local domain. These
5792 two options are now turned off when doing these lookups.
5794 57. It seems that at least some releases of Reiserfs (which does not have the
5795 concept of a fixed number of inodes) returns zero and not -1 for the
5796 number of available inodes. This interacted badly with check_spool_inodes,
5797 which assumed that -1 was the "no such thing" setting. What I have done is
5798 to check that the total number of inodes is greater than zero before doing
5799 the test of how many are available.
5801 58. When a "warn" ACL statement has a log_message modifier, the message is
5802 remembered, and not repeated. This is to avoid a lot of repetition when a
5803 message has many recipients that cause the same warning to be written.
5804 However, Exim was preserving the list of already written lines for an
5805 entire SMTP session, which doesn't seem right. The memory is now reset if a
5806 new message is started.
5808 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5809 debugging output unless log_rewrite was also set.
5811 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5812 of (int)(handle) when we know that handle contains (void *)(-1).
5814 61. The Exim daemon panic-logs an error return when it closes the incoming
5815 connection. However "connection reset by peer" seems to be common, and
5816 isn't really an error worthy of noting specially, so that particular error
5819 62. When Exim is trying to find all the local interfaces, it used to panic and
5820 die if the ioctl to get the interface flags failed. However, it seems that
5821 on at least one OS (Solaris 9) it is possible to have an interface that is
5822 included in the list of interfaces, but for which you get a failure error
5823 for this call. This happens when the interface is not "plumbed" into a
5824 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5825 failure of the "get flags" call assumes that the interface is down.
5827 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5828 makes life easier for people who are doing arithmetic on fields extracted
5829 from dates, where you often get leading zeros that should not be
5830 interpreted as octal.
5832 64. Added qualify_domain to the redirect router, to override the global
5835 65. If a pathologically long header line contained very many addresses (the
5836 report of this problem mentioned 10 000) and each of them was rewritten,
5837 Exim could use up a very large amount of memory. (It kept on making new
5838 copies of the header line as it rewrote, and never released the old ones.)
5839 At the expense of a bit more processing, the header rewriting function has
5840 been changed so that it no longer eats memory in this way.
5842 66. The generation of the Received: header has been moved from the time that a
5843 message starts to be received, to the time that it finishes. The timestamp
5844 in the Received: header should now be very close to that of the <= log
5845 line. There are two side-effects of this change:
5847 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5848 logged header lines no longer include the local Received: line, because
5849 it has not yet been created. The same applies to a copy of the message
5850 that is returned to a non-SMTP sender when a message is rejected.
5852 (b) When a filter file is tested using -bf, no additional Received: header
5853 is added to the test message. After some thought, I decided that this
5856 This change does not affect the value of $received_for. It is still set
5857 after address rewriting, but before local_scan() is called.
5859 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5861 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5862 gave an unhelpful panic error message, and a defer error. I have managed to
5863 change this behaviour so that it now rejects any supplied certificate,
5864 which seems right, as the list of acceptable certificates is empty.
5866 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5867 gave an unhelpful defer error. I have not managed to make this reject any
5868 supplied certificates, but the error message it gives is "no certificate
5869 supplied", which is not helpful.
5871 70. exigrep's output now also includes lines that are not associated with any
5872 message, but which match the given pattern. Implemented by a patch from
5873 Martin Sluka, which also tidied up the Perl a bit.
5875 71. Recipient callout verification, like sender verification, was using <> in
5876 the MAIL FROM command. This isn't really the right thing, since the actual
5877 sender may affect whether the remote host accepts the recipient or not. I
5878 have changed it to use the actual sender in the callout; this means that
5879 the cache record is now keyed on a recipient/sender pair, not just the
5880 recipient address. There doesn't seem to be a real danger of callout loops,
5881 since a callout by the remote host to check the sender would use <>.
5882 [SEE ABOVE: changed after hitting problems.]
5884 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5885 temporary errors. However, in the case of such a code being given after
5886 the end of a data transmission (i.e. after ".") Exim was failing to write
5887 a retry record for the message. (Yes, there was some broken host that was
5888 actually sending 8xx at this point.)
5890 73. An unknown lookup type in a host list could cause Exim to panic-die when
5891 the list was checked. (An example that provoked this was putting <; in the
5892 middle of a list instead of at the start.) If this happened during a DATA
5893 ACL check, a -D file could be left lying around. This kind of configuration
5894 error no longer causes Exim to die; instead it causes a defer error. The
5895 incident is still logged to the main and panic logs.
5897 74. Buglet left over from Exim 3 conversion. The message "too many messages
5898 in one connection" was written to the rejectlog but not the mainlog, except
5899 when address rewriting (yes!) was being logged.
5901 75. Added write_rejectlog option.
5903 76. When a system filter was run not as root (that is, when system_filter_user
5904 was set), the values of the $n variables were not being returned to the
5905 main process; thus, they were not subsequently available in the $sn
5908 77. Added +return_path_on_delivery log selector.
5910 78. A connection timeout was being treated differently from recipients deferred
5911 when testing hosts_max_try with a message that was older than the host's
5912 retry timeout. (The host should not be counted, thus allowing all hosts to
5913 be tried at least once before bouncing.) This may have been the cause of an
5914 occasionally reported bug whereby a message would remain on the queue
5915 longer than the retry timeout, but would be bounced if a delivery was
5916 forced. I say "may" because I never totally pinned down the problem;
5917 setting up timeout/retry tests is difficult. See also the next item.
5919 79. The ultimate address timeout was not being applied to errors that involved
5920 a combination of host plus message (for example, a timeout on a MAIL
5921 command). When an address resolved to a number of possible hosts, and they
5922 were not all tried for each delivery (e.g. because of hosts_max_try), a
5923 message could remain on the queue longer than the retry timeout.
5925 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5928 81. Fixed an obscure SMTP outgoing bug which required at least the following
5929 conditions: (a) there was another message waiting for the same server;
5930 (b) the server returned 5xx to all RCPT commands in the first message so
5931 that the message was not completed; (c) the server dropped the connection
5932 or gave a negative response to the RSET that Exim sends to abort the
5933 transaction. The observed case was a dropped connection after DATA that had
5934 been sent in pipelining mode. That is, the server had advertised PIPELINING
5935 but was not implementing it correctly. The effect of the bug was incorrect
5936 behaviour, such as trying another host, and this could lead to a crash.
5942 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5943 and daemon.c were passed as pointers to ints; they should have been
5944 pointers to socklen_t variables (which are typically unsigned ints).
5946 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5949 3. Fixed a really odd bug that affected only the testing scheme; patching a
5950 certain fixed string in the binary changed the value of another string that
5951 happened to be identical to the end of the original first string.
5953 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5954 name", it returns that address as the IP address. On some operating
5955 systems (e.g. Solaris), it also passes back the IP address string as the
5956 "host name". However, on others (e.g. Linux), it passes back an empty
5957 string. Exim wasn't checking for this, and was changing the host name to an
5958 empty string, assuming it had been canonicalized.
5960 5. Although rare, it is permitted to have more than one PTR record for a given
5961 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5962 all the names associated with an address, because they do in Solaris.
5963 However, it seems that they do not in Linux for data that comes from the
5964 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5965 I found this out when I moved to a new Linux workstation and tried to run
5966 the Exim test suite.
5968 To get round this problem I have changed the code so that it now does its
5969 own call to the DNS to look up PTR records when searching for a host name.
5970 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5971 addresses that are only in /etc/hosts are still found.
5973 This behaviour is, however, controlled by an option called host_lookup_
5974 order, which defaults to "bydns:byaddr". If people want to use the other
5975 order, or indeed, just use one or the other means of lookup, they can
5976 specify it in this variable.
5978 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5979 some operating systems, this comes back from gethostbyaddr() as an empty
5980 string, and this is what Exim used to test for. However, it seems that in
5981 other systems, "." is yielded. Exim now tests for this case too.
5983 7. The values of check_spool_space and check_log_space are now held internally
5984 as a number of kilobytes instead of an absolute number of bytes. If a
5985 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5986 kilobyte. This means that much larger values can be stored.
5988 8. Exim monitor: an attempt to get the action menu when not actually pointing
5989 at a message produces an empty menu entitled "No message selected". This
5990 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5991 no entries in it ("Shell widget menu has zero width and/or height"). So I
5992 have added a single, blank menu entry in this case.
5994 9. Added ${quote_local_part.
5996 10. MIME decoding is now applied to the contents of Subject: header lines when
5999 11. Now that a reference to $sender_host_address automatically causes a reverse
6000 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6001 host lookup before query-style lookups in lists that might use this
6002 variable. This has therefore been abolished, and the "net-" prefix is no
6003 longer necessary for query-style lookups.
6005 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6006 have been a typo for LFLAGS, so it has been changed.
6008 13. The install script calls Exim with "-C /dev/null" in order to find the
6009 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6010 to be output. However, since Exim outputs its version number before the
6011 error, it didn't break the script. It just looked ugly. I fixed this by
6012 always allowing "-C /dev/null" if the caller is root.
6014 14. Ignore overlarge ACL variable number when reading spool file - insurance
6015 against a later release with more variables having written the file.
6017 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6018 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6019 this, as well as the form without the "IPv6" on the front.
6021 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6022 OS/Makefile-Darwin file.
6024 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6025 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6027 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6028 variable. However, some OS use ints for inodes. I've added cast to long int
6029 to get rid of the compiler warning.
6031 19. I had forgotten to lock out "/../" in configuration file names when
6032 ALT_CONFIG_PREFIX was set.
6034 20. Routers used for verification do not need to specify transports. However,
6035 if such a router generated a host list, and callout was configured, Exim
6036 crashed, because it could not find a port number from the (non-existent)
6037 transport. It now assumes port 25 in this circumstance.
6039 21. Added the -t option to exigrep.
6041 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6042 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6043 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6044 them is set, LOOKUP_LSEARCH is forced.
6046 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6048 24. Added sender and host information to the "rejected by local_scan()" log
6049 line; previously there was no indication of these.
6051 25. Added .include_if_exists.
6053 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6054 Linux. It turns out that not all file systems support this. Apparently some
6055 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6056 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6057 not supported on the file descriptor, is now ignored when Exim is trying to
6058 sync a directory. This applies only to Linux.
6060 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6062 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6063 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6066 29. Added support for maildirsize files from supplied patch (modified a bit).
6068 30. The use of :fail: followed by an empty string could lead Exim to respond to
6069 sender verification failures with (e.g.):
6071 550 Verification failed for <xxx>
6072 550 Sender verify failed
6074 where the first response line was missing the '-' that indicates it is not
6075 the final line of the response.
6077 31. The loop for finding the name of the user that called Exim had a hardwired
6078 limit of 10; it now uses the value of finduser_retries, which is used for
6079 all other user lookups.
6081 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6084 33. Exim was neglecting to zero errno before one call of strtol() when
6085 expanding a string and expecting an integer value. On some systems this
6086 resulted in spurious "integer overflow" errors. Also, it was casting the
6087 result into an int without checking.
6089 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6090 did not work. The code looks as if it has *never* worked, though it appears
6091 to have been documented since at least release 1.62. I have made it work.
6093 35. The "timeout_DNS" error in retry rules, also documented since at least
6094 1.62, also never worked. As it isn't clear exactly what this means, and
6095 clearly it isn't a major issue, I have abolished the feature by treating it
6096 as "timeout", and writing a warning to the main and panic logs.
6098 36. The display of retry rules for -brt wasn't always showing the error code
6101 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6102 timeout_connect_A, timeout_connect_MX.
6104 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6105 to the empty sender.
6107 39. The daemon was not analysing the content of -oX till after it had closed
6108 stderr and disconnected from the controlling terminal. This meant that any
6109 syntax errors were only noted on the panic log, and the return code from
6110 the command was 0. By re-arranging the code a little, I've made the
6111 decoding happen first, so such errors now appear on stderr, and the return
6112 code is 1. However, the actual setting up of the sockets still happens in
6113 the disconnected process, so errors there are still only recorded on the
6116 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6117 connections (as happens on some IP stacks) was logged at start up time as
6118 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6119 it from "IPv6 and IPv4", which means that two separate sockets are being
6122 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6123 says whether AF_INET or AF_INET6 was passed as an argument.
6125 42. Exiwhat output was messed up when time zones were included in log
6128 43. Exiwhat now gives more information about the daemon's listening ports,
6129 and whether -tls-on-connect was used.
6131 44. The "port" option of the smtp transport is now expanded.
6133 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6134 silently ignored. Now an error message is written to the main and panic
6137 46. There's a new ACL modifier called "logwrite" which writes to a log file
6138 as soon as it is encountered.
6140 47. Added $local_user_uid and $local_user_gid at routing time.
6142 48. Exim crashed when trying to verify a sender address that was being
6145 49. Exim was recognizing only a space character after ".include". It now also
6146 recognizes a tab character.
6148 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6149 extracting the relevant information from the specification. The man page no
6150 longer contains scrambled data for the -d option, and I've added a section
6151 at the front about calling Exim under different names.
6153 51. Added "extra_headers" argument to the "mail" command in filter files.
6155 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6158 53. Installed eximstats 1.29.
6160 54. Added transport_filter_timeout as a generic transport option.
6162 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6163 Cc: header lines. This was required by RFC 822, but it not required by RFC
6166 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6167 incoming messages that did not have them. Now it does so only if the
6168 message originates locally, that is, if there is no associated remote host
6169 address. When Resent- header lines are present, this applies to the Resent-
6170 lines rather than the non-Resent- lines.
6172 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6173 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6175 58. Messages for configuration errors now include the name of the main
6176 configuration file - useful now that there may be more than one file in a
6177 list (.included file names were always shown).
6179 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6180 for those rare installations that do not start the daemon as root or run it
6181 setuid root. I've cut out the call to initgroups() if the daemon is not
6184 60. The Exim user and group can now be bound into the binary as text strings
6185 that are looked up at the start of Exim's processing.
6187 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6189 62. Added $mailstore_basename variable.
6191 63. Installed patch to sieve.c from Michael Haardt.
6193 64. When Exim failed to open the panic log after failing to open the main log,
6194 the original message it was trying to log was written to stderr and debug
6195 output, but if they were not available (the usual case in production), it
6196 was lost. Now it is written to syslog before the two lines that record the
6197 failures to open the logs.
6199 65. Users' Exim filters run in subprocesses under the user's uid. It is
6200 possible for a "deliver" command or an alias in a "personal" command to
6201 provoke an address rewrite. If logging of address rewriting is configured,
6202 this fails because the process is not running as root or exim. There may be
6203 a better way of dealing with this, but for the moment (because 4.30 needs
6204 to be released), I have disabled address rewrite logging when running a
6205 filter in a non-root, non-exim process.
6211 1. The buildconfig auxiliary program wasn't quoting the value set for
6212 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6213 not defined. This bug was masked in 4.22 by the effect that was fixed in
6216 2. Some messages that were rejected after a message id was allocated were
6217 shown as "incomplete" by exigrep. It no longer does this for messages that
6218 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6220 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6221 have allow_domain_literals set, the ID did not get logged in the <= line.
6222 Domain literals are now always recognized in Message-ID: header lines.
6224 4. The first argument for a ${extract expansion item is the key name or field
6225 number. Leading and trailing spaces in this item were not being ignored,
6226 causing some misleading effects.
6228 5. When deliver_drop_privilege was set, single queue runner processes started
6229 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6230 same command in the process it spins off) were not dropping privilege.
6232 6. When the daemon running as "exim" started a queue runner, it always
6233 re-executed Exim in the spun-off process. This is a waste of effort when
6234 deliver_drop_privilege is set. The new process now just calls the
6235 queue-runner function directly.
6241 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6244 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6245 ignored. Though the use of -oP was forcing the writing of a pid file, it
6246 was always written to the default place.
6248 3. If the message "no IP address found for host xxxx" is generated during
6249 incoming verification, it is now followed by identification of the incoming
6250 connection (so you can more easily find what provoked it).
6252 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6254 5. Added some features to "harden" Exim a bit more against certain attacks:
6256 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6257 be put in Local/Makefile. This is like the never_users runtime option,
6258 but it cannot be overridden. The default setting is "root".
6260 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6261 prefix string with which any file named in a -C command line option
6264 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6265 is retained for -C and -D only if the caller of Exim is root. Without
6266 it, the exim user may also use -C and -D and retain privilege.
6268 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6269 command line option is disabled.
6271 6. Macro names set by the -D option must start with an upper case letter, just
6272 like macro names defined in the configuration file.
6274 7. Added "dereference=" facility to LDAP.
6276 8. Two instances of the typo "uknown" in the source files are fixed.
6278 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6279 the Configure-Makefile script screwed up while processing it.
6281 10. Incorporated PCRE 4.4.
6283 11. The SMTP synchronization check was not operating right at the start of an
6284 SMTP session. For example, it could not catch a HELO sent before the client
6285 waited for the greeting. There is now a check for outstanding input at the
6286 point when the greeting is written. Because of the duplex, asynchronous
6287 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6288 way, but not yet received, when the check is performed.
6290 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6291 on TCP/IP sockets, because this apparently causes some broken clients to
6294 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6295 unchanged) from the Cygwin maintainer.
6297 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6298 was supported instead of testing for mbx. Effectively a typo.
6300 15. The spa authenticator server code was not checking that the input it
6301 received was valid base64.
6303 16. The debug output line for the "set" modifier in ACLs was not showing the
6304 name of the variable that was being set.
6306 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6308 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6309 lookup on its own. Something else was needed to trigger the lookup. For
6310 example, a match in host_lookup or the need for a host name in a host list.
6311 Now, if $sender_host_name is referenced and the host name has not yet been
6312 looked up, a lookup is performed. If the lookup fails, the variable remains
6313 empty, and $host_lookup_failed is set to "1".
6315 19. Added "eqi" as a case-independent comparison operator.
6317 20. The saslauthd authentication condition could segfault if neither service
6318 nor realm was specified.
6320 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6321 error message that was logged was misleading, and incoming SMTP
6322 connections were dropped. The message is now more accurate, and temporary
6323 errors are given to SMTP connections.
6325 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6326 (except RSET) with a 421 error, until QUIT is received. However, it was
6327 failing to send a response to QUIT.
6329 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6330 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6331 ACL is now run after the helo_try_verify_hosts code.
6333 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6334 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6335 case-independent, so other case variants are also recognized). Apparently
6336 some systems use these upper case variants.
6338 25. If more than two messages were waiting for the same host, and a transport
6339 filter was specified for the transport, Exim sent two messages over the
6340 same TCP/IP connection, and then failed with "socket operation on non-
6341 socket" when it tried to send the third.
6343 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6345 27. The extern definition of crypt16() in expand.c was not being excluded when
6346 the OS had its own crypt16() function.
6348 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6349 as a preferred synonym for return_size_limit, both as an option and as an
6352 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6354 30. Changed the default configuration ACL to relax the local part checking rule
6355 for addresses that are not in any local domains. For these addresses,
6356 slashes and pipe symbols are allowed within local parts, but the sequence
6357 /../ is explicitly forbidden.
6359 31. SPA server authentication was not clearing the challenge buffer before
6362 32. log_message in a "warn" ACL statement was writing to the reject log as
6363 well as to the main log, which contradicts the documentation and doesn't
6364 seem right (because no rejection is happening). So I have stopped it.
6366 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6367 However, I am unable to do any testing of this.
6369 34. Fixed an infelicity in the appendfile transport. When checking directories
6370 for a mailbox, to see if any needed to be created, it was accidentally
6371 using path names with one or more superfluous leading slashes; tracing
6372 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6374 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6375 contents are added to the log line that is written for every discarded
6376 recipient. (Previously a log_message setting was ignored.)
6378 36. The ${quote: operator now quotes the string if it is empty.
6380 37. The install script runs exim in order to find its version number. If for
6381 some reason other than non-existence or emptiness, which it checks, it
6382 could not run './exim', it was installing it with an empty version number,
6383 i.e. as "exim-". This error state is now caught, and the installation is
6386 38. An argument was missing from the function that creates an error message
6387 when Exim fails to connect to the socket for saslauthd authentication.
6388 This could cause Exim to crash, or give a corrupted message.
6390 39. Added isip, isip4, and isip6 to ${if conditions.
6392 40. The ACL variables $acl_xx are now saved with the message, and can be
6393 accessed later in routers, transports, and filters.
6395 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6396 strings in the file are not string-expanded.
6398 42. If a MAIL command specified a SIZE value that was too large to fit into an
6399 int variable, the check against message_size_limit failed. Such values are
6400 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6401 day this will have to be increased, but I don't think I want to be around
6402 when emails are that large.
6409 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6410 iconv() is not standard in FreeBSD.
6412 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6413 IPv6 enabled. The observed symptom was a segmentation fault on return from
6414 the function os_common_find_running_interfaces() in src/os.c.
6416 3. In the check_special_case() function in daemon.c I had used "errno" as an
6417 argument name, which causes warnings on some systems. This was basically a
6418 typo, since it was named "eno" in the comments!
6420 4. The code that waits for the clock to tick (at a resolution of some fraction
6421 of a second) so as to ensure message-id uniqueness was always waiting for
6422 at least one whole tick, when it could have waited for less. [This is
6423 almost certainly not relevant at current processor speeds, where it is
6424 unlikely to ever wait at all. But we try to future-proof.]
6426 5. The function that sleeps for a time interval that includes fractions of a
6427 second contained a race. It did not block SIGALRM between setting the
6428 timer, and suspending (a couple of lines later). If the interval was short
6429 and the sigsuspend() was delayed until after it had expired, the suspension
6430 never ended. On busy systems this could lead to processes getting stuck for
6433 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6434 process, before it forks any delivery processes. The open lookup caching
6435 mechanism meant that the open file or database connection was passed into
6436 the delivery process. The problem was that delivery processes always tidy
6437 up cached lookup data. This could cause a problem for the next delivery
6438 process started by the queue runner, because the external queue runner
6439 process does not know about the closure. So the next delivery process
6440 still has data in the lookup cache. In the case of a file lookup, there was
6441 no problem because closing a file descriptor in a subprocess doesn't affect
6442 the parent. However, if the lookup was caching a connection to a database,
6443 the connection was closed, and the second delivery process was likely to
6444 see errors such as "PGSQL: query failed: server closed the connection
6445 unexpectedly". The problem has been fixed by closing all cached lookups
6446 in a queue runner before running a delivery process.
6448 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6449 seem to have the "const" qualifier which it has on other OS. I've
6452 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6453 *of the same type* (client or server) with the same public name that an
6454 error should be diagnosed.
6456 9. When Exim looked up a host name for an IP address, but failed to find the
6457 original IP address when looking up the host name (a safety check), it
6458 output the message "<ip address> does not match any IP for NULL", which was
6459 confusing, to say the least. The bug was that the host name should have
6460 appeared instead of "NULL".
6462 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6463 user that is built into the binary, and the -C or -D options is used, root
6464 privilege is dropped before the configuration file is read. In addition,
6465 logging is switched to stderr instead of the normal log files. If the
6466 configuration then re-defines the Exim user, the unprivileged environment
6467 is probably not what is expected, so Exim logs a panic warning message (but
6470 However, if deliver_drop_privilege is set, the unprivileged state may well
6471 be exactly what is intended, so the warning has been cut out in that case,
6472 and Exim is allowed to try to write to its normal log files.
6478 1. smtp_return_error_details was not giving details for temporary sender
6479 or receiver verification errors.
6481 2. Diagnose a configuration error if two authenticators have the same public
6484 3. Exim used not to create the message log file for a message until the first
6485 delivery attempt. This could be confusing when incoming messages were held
6486 for policy or load reasons. The message log file is now created at the time
6487 the message is received, and an initial "Received" line is written to it.
6489 4. The automatically generated man page for command line options had a minor
6490 bug that caused no ill effects; however, a more serious problem was that
6491 the procedure for building the man page automatically didn't always
6492 operate. Consequently, release 4.20 contains an out-of-date version. This
6493 shouldn't happen again.
6495 5. When building Exim with embedded Perl support, the script that builds the
6496 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6497 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6499 6. The freeze_tell option was not being used for messages that were frozen on
6500 arrival, either by an ACL or by local_scan().
6502 7. Added the smtp_incomplete_transaction log selector.
6504 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6505 was accepting AUTH without a new EHLO.
6507 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6508 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6511 10. Logging of TCP/IP connections (when configured) now happens in the main
6512 daemon process instead of the child process, so that the TCP/IP connection
6513 count is more accurate (but it can never be perfect).
6515 11. The use of "drop" in a nested ACL was not being handled correctly in the
6516 outer ACL. Now, if condition failure induced by the nested "drop" causes
6517 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6518 or "require"), the connection is dropped.
6520 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6521 that yield "discard" can now be used with an "accept" or a "discard" verb,
6522 but an error is generated for any others (because I can't see a useful way
6523 to define what should happen).
6525 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6526 are now processed in the same way as lines in the Exim configuration file.
6527 In particular, continuation lines are supported.
6529 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6531 15. Added -ti meaning -t -i.
6533 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6534 domains, and warn by logging if others are found.
6536 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6537 returned from GIFCONF when Exim is trying to find the list of interfaces on
6538 a host. The code in os.c has been modified to copy each ifreq to an aligned
6539 structure in all cases.
6541 Also, in some cases, the returned ifreq's were being copied to a 'struct
6542 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6543 means the last couple of bytes of an IPv6 address could be chopped if the
6544 ifreq contained only a normal sockaddr (14 bytes storage).
6546 18. Named domain lists were not supported in the hosts_treat_as_local option.
6547 An entry such as +xxxx was not recognized, and was treated as a literal
6550 19. Ensure that header lines added by a DATA ACL are included in the reject log
6551 if the ACL subsequently rejects the message.
6553 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6554 MD5 (which is deprecated).
6556 21. When testing a filter file using -bf, Exim was writing a message when it
6557 took the sender from a "From " line in the message, but it was not doing so
6558 when it took $return_path from a Return-Path: header line. It now does.
6560 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6561 with a valid header line field name (a series of printing characters
6562 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6564 23. Changed "disc" in the source to "disk" to conform to the documentation and
6565 the book and for uniformity.
6567 24. Ignore Sendmail's -Ooption=value command line item.
6569 25. When execve() failed while trying to run a command in a pipe transport,
6570 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6571 could be confused with a return value of 69 from the command itself. This
6572 has been changed to 127, the value the shell returns if it is asked to run
6573 a non-existent command. The wording for the related log line suggests a
6574 non-existent command as the problem.
6576 26. If received_header_text expands to an empty string, do not add a Received:
6577 header line to the message. (Well, it adds a token one on the spool, but
6578 marks it "old" so that it doesn't get used or transmitted.)
6580 27. Installed eximstats 1.28 (addition of -nt option).
6582 28. There was no check for failure on the call to getsockname() in the daemon
6583 code. This can fail if there is a shortage of resources on the system, with
6584 ENOMEM, for example. A temporary error is now given on failure.
6586 29. Contrary to the C standard, it seems that in some environments, the
6587 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6588 program. Exim now does this explicitly; it affects the formatting of
6589 timestamps using strftime().
6591 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6592 complaints. I've now initialized all the variables, to avoid this.
6594 32. Header lines added by a system filter were not being "seen" during
6595 transport-time rewrites.
6597 33. The info_callback() function passed to OpenSSL is set up with type void
6598 (*)(SSL *, int, int), as described somewhere. However, when calling the
6599 function (actually a macro) that sets it up, the type void(*)() is
6600 expected. I've put in a cast to prevent warnings from picky compilers.
6602 34. If a DNS black list lookup found a CNAME record, but there were no A
6603 records associated with the domain it pointed at, Exim crashed.
6605 35. If a DNS black list lookup returned more than one A record, Exim ignored
6606 all but the first. It now scans all returned addresses if a particular IP
6607 value is being sought. In this situation, the contents of the
6608 $dnslist_value variable are a list of all the addresses, separated by a
6611 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6612 to accept a host name and all its aliases if the forward lookup for any of
6613 them yielded the IP address of the incoming connection. Now it accepts only
6614 those names whose forward lookup yields the correct IP address. Any other
6615 names are discarded. This closes a loophole whereby a rogue DNS
6616 administrator could create reverse DNS records to break through a
6617 wildcarded host restriction in an ACL.
6619 37. If a user filter or a system filter that ran in a subprocess used any of
6620 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6621 the wrong values were passed to the pipe command ($thisaddress had the
6622 value of $0, $0 had the value of $1, etc). This bug was introduced by
6623 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6625 38. Improved the line breaking for long SMTP error messages from ACLs.
6626 Previously, if there was no break point between 40 and 75 characters, Exim
6627 left the rest of the message alone. Two changes have been made: (a) I've
6628 reduced the minimum length to 35 characters; (b) if it can't find a break
6629 point between 35 and 75 characters, it looks ahead and uses the first one
6630 that it finds. This may give the occasional overlong line, but at least the
6631 remaining text gets split now.
6633 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6634 file descriptors might be low, and that setting 1000 would always raise it.
6635 It turns out that in some environments, the limit is already over 1000 and
6636 that lowering it causes trouble. So now Exim takes care not to decrease it.
6638 40. When delivering a message, the value of $return_path is set to $sender_
6639 address at the start of routing (routers may change the value). By an
6640 oversight, this default was not being set up when an address was tested by
6641 -bt or -bv, which affected the outcome if any router or filter referred to
6644 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6645 log only when "message" or "log_message" are set. However, if one of the
6646 conditions was an address verification, or a call to a nested ACL, the
6647 messages generated by the underlying test were being passed through. This
6648 no longer happens. The underlying message is available in $acl_verify_
6649 message for both "message" and "log_message" expansions, so it can be
6650 passed through if needed.
6652 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6653 new expansion $bh_ to give the encoded byte string without charset
6654 translation. Translation happens only if iconv() is available; HAVE_ICONV
6655 indicates this at build time. HEADERS_CHARSET gives the charset to
6656 translate to; headers_charset can change it in the configuration, and
6657 "headers charset" can change it in an individual filter file.
6659 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6660 that creates RFC 2047 encoded "words" labels them as that charset instead
6661 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6662 expansion operator; (ii) when Exim creates a From: line for a local
6663 message; (iii) when a header line is rewritten to include a "phrase" part.
6665 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6666 buggy, causing it to skip the first lines of messages whose message ID
6667 ended in 'D'. This would not have bitten before Exim release 4.14, because
6668 message IDs were unlikely to end in 'D' before then. The effect was to have
6669 incorrect size information for certain domains.
6671 45. #include "config.h" was missing at the start of the crypt16.c module. This
6672 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6675 46. If there was a timeout during a "random" callout check, Exim treated it as
6676 a failure of the random address, and carried on sending RSET and the real
6677 address. If the delay was just some slowness somewhere, the response to the
6678 original RCPT would be taken as a response to RSET and so on, causing
6679 mayhem of various kinds.
6681 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6682 when I implemented it. It didn't allow for the fact that some option values
6683 may legitimately be negative (e.g. size_addition), and it didn't even do
6684 the right test for positive values.
6686 48. Domain names in DNS records are case-independent. Exim always looks them up
6687 in lower case. Some resolvers return domain names in exactly the case they
6688 appear in the zone file, that is, they may contain uppercase letters. Not
6689 all resolvers do this - some return always lower case. Exim was treating a
6690 change of case by a resolver as a change of domain, similar to a widening
6691 of a domain abbreviation. This triggered its re-routing code and so it was
6692 trying to route what was effectively the same domain again. This normally
6693 caused routing to fail (because the router wouldn't handle the domain
6694 twice). Now Exim checks for this case specially, and just changes the
6695 casing of the domain that it ultimately uses when it transmits the message
6698 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6701 50. If a filter generated a file delivery with a non-absolute name (possible if
6702 no home directory exists for the router), the forbid_file option was not
6705 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6706 the existing equality matching.
6708 52. Exim was using ints instead of ino_t variables in some places where it was
6709 dealing with inode numbers.
6711 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6712 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6713 called TMPDIR, and if it finds it is different, it changes its value.
6715 54. The smtp_printf() function is now made available to local_scan() so
6716 additional output lines can be written before returning. There is also an
6717 smtp_fflush() function to enable the detection of a dropped connection.
6718 The variables smtp_input and smtp_batched_input are exported to
6721 55. Changed the default runtime configuration: the message "Unknown user"
6722 has been removed from the ACL, and instead placed on the localuser router,
6723 using the cannot_route_message feature. This means that any verification
6724 failures that generate their own messages won't get overridden. Similarly,
6725 the "Unrouteable address" message that was in the ACL for unverifiable
6726 relay addresses has also been removed.
6728 56. Added hosts_avoid_esmtp to the smtp transport.
6730 57. The exicyclog script was not checking for the esoteric option
6731 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6732 will work only if exicyclog is run under the appropriate euid.
6734 58. Following a discussion on the list, the rules by which Exim recognises line
6735 endings on incoming messages have been changed. The -dropcr and drop_cr
6736 options are now no-ops, retained only for backwards compatibility. The
6737 following line terminators are recognized: LF CRLF CR. However, special
6738 processing applies to CR:
6740 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6741 nor a local message in the state where . is a terminator.
6743 (ii) If a bare CR is encountered in a header line, an extra space is added
6744 after the line terminator so as not to end the header. The reasoning
6745 behind this is that bare CRs in header lines are most likely either
6746 to be mistakes, or people trying to play silly games.
6748 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6749 was being incorrectly given as 18 bytes larger than it should have been.
6750 This is a VOB (very old bug).
6752 60. This may never have affected anything current, but just in case it has:
6753 When the local host is found other than at the start of a list of hosts,
6754 the local host, those with the same MX, and any that follow, are discarded.
6755 When the list in question was part of a longer list of hosts, the following
6756 hosts (not currently being processed) were also being discarded. This no
6757 longer happens. I'm not sure if this situation could ever has previously
6760 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6763 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6764 header. This is something that is recommended in a new Internet Draft, and
6765 is something that is documented as being done by Sendmail. There are two
6766 possible values. For messages generated by the autoreply transport, Exim
6769 Auto-Submitted: auto-replied
6771 whereas for all other generated messages (e.g. bounces) it adds
6773 Auto-Submitted: auto-generated
6775 63. The "personal" condition in filters now includes a test for the
6776 Auto-Submitted: header. If it contains the string "auto-" the message it
6777 not considered personal.
6779 64. Added rcpt_include_affixes as a generic transport option.
6781 65. Added queue_only_override (default true).
6783 66. Added the syslog_duplication option.
6785 67. If what should have been the first header line of a message consisted of
6786 a space followed by a colon, Exim was mis-interpreting it as a header line.
6787 It isn't of course - it is syntactically invalid and should therefore be
6788 treated as the start of the message body. The misbehaviour could have
6789 caused a number of strange effects, including loss of data in subsequent
6790 header lines, and spool format errors.
6792 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6793 client host had authenticated. This control can now be exercised by an ACL
6794 for more flexibility.
6796 69. By default, callouts do not happen when testing with -bh. There is now a
6797 variant, -bhc, which does actually run the callout code, including
6798 consulting and updating the callout cache.
6800 70. Added support for saslauthd authentication, courtesy of Alexander
6803 71. If statvfs() failed on the spool or log directories while checking their
6804 size for availability, Exim confusingly gave the error "space shortage".
6805 Furthermore, in debugging mode it crashed with a floating point exception.
6806 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6807 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6808 serious problem, Exim now writes to the main and panic logs when this
6809 happens, with details of the failure. It then refuses to accept the
6810 incoming message, giving the message "spool directory problem" or "log
6811 directory problem" with a 421 code for SMTP messages.
6813 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6814 0, 1, and 2 exist, because some OS complain for execs without them (see
6815 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6816 descriptors. However, the code omitted to check that the open succeeded,
6817 causing mysterious errors if for some reason the permissions on /dev/null
6818 got screwed. Now Exim writes a message to the main and panic logs, and
6819 bombs out if it can't open /dev/null.
6821 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6822 interact so that it is all more flexible. It is supposed to remain
6823 backwards compatible. Also added extra_local_interfaces.
6825 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6826 to bomb out with an assertion failure - to the client this appears as a
6827 connection drop. This problem occurs in the part of the code that was taken
6828 from the Samba project. Fortunately, the assertion is in a very simple
6829 function, so I have fixed this by reproducing the function inline in the
6830 one place where it is called, and arranging for authentication to fail
6831 instead of killing the process with assert().
6833 75. The SPA client code was not working when the server requested OEM rather
6834 than Unicode encoding.
6836 76. Added code to make require_files with a specific uid setting more usable in
6837 the case where statting the file as root fails - usually a non-root-mounted
6838 NFS file system. When this happens and the failure is EACCES, Exim now
6839 forks a subprocess and does the per-uid checking as the relevant uid.
6841 77. Added process_log_path.
6843 78. If log_file_path was not explicitly set, a setting of check_log_space or
6844 check_log_inodes was ignored.
6846 79. If a space check for the spool or log partitions fails, the incident is now
6847 logged. Of course, in the latter case the data may get lost...
6849 80. Added the %p formatting code to string_format() so that it can be used to
6850 print addresses in debug_print(). Adjusted all the address printing in the
6851 debugging in store.c to use %p rather than %d.
6853 81. There was a concern that a line of code in smtp_in.c could overflow a
6854 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6855 initially expressed, the concern was not well-founded, because trailing
6856 spaces are removed early. However, if the trailing spaces were followed by
6857 a NULL, they did not get removed, so the overflow was possible. Two fixes
6860 (a) I re-wrote the offending code in a cleaner fashion.
6861 (b) If an incoming SMTP command contains a NULL character, it is rejected
6864 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6865 runs initgroups(), so that if the Exim user is in any additional groups,
6866 they will be used during message reception.
6872 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob