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.
116 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
122 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
123 than -2003 did; needs libidn2 in addition to libidn.
125 JH/02 The path option on a pipe transport is now expanded before use.
127 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
128 Patch provided by "Björn", documentation fix added too.
130 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
131 missing a wire-to-host endian conversion.
133 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
134 close after a BDAT command line could be taken as a following command,
135 giving a synch failure. Fix by only checking for synch immediately
136 before acknowledging the chunk.
138 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
139 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
140 macro. Patches provided by Josh Soref.
142 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
143 Previously we did not; the RFC seems ambiguous and VRFY is not listed
144 by IANA as a service extension. However, John Klensin suggests that we
147 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
148 the dkim code may be unix-mode line endings rather than smtp wire-format
149 CRLF, so prepend a CR to any bare LF.
151 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
152 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
154 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
155 the first were themselves being wrongly included in the feed into dkim
156 processing; with most chunk sizes in use this resulted in an incorrect
157 body hash calculated value.
159 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
160 DKIM signature block, for verification. Although advised against by
161 standards it is specifically not ruled illegal.
163 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
165 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
166 missing a body hash (the bh= tag).
168 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
169 It seems that HAProxy sends the Proxy Protocol information in clear and
170 only then does a TLS startup, so do the same.
172 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
173 TCP connections (such as for Spamd) unless the daemon successfully set
174 Fast Open mode on its listening sockets. This fixes breakage seen on
175 too-old kernels or those not configured for Fast Open, at the cost of
176 requiring both directions being enabled for TFO, and TFO never being used
177 by non-daemon-related Exim processes.
179 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
180 endings, at least on the first header line. Try to canonify any that get
181 past that check, despite the cost.
183 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
184 now limited to an arbitrary five deep, while parsing addresses with the
185 strip_excess_angle_brackets option enabled.
187 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
188 instead leave the unprompted TLS handshake in socket buffer for the
189 TLS library to consume.
191 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
193 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
195 JH/16 Drop variables when they go out of scope. Memory management drops a whole
196 region in one operation, for speed, and this leaves assigned pointers
197 dangling. Add checks run only under the testsuite which checks all
198 variables at a store-reset and panics on a dangling pointer; add code
199 explicitly nulling out all the variables discovered. Fixes one known
200 bug: a transport crash, where a dangling pointer for $sending_ip_address
201 originally assigned in a verify callout, is re-used.
203 PP/06 Drop '.' from @INC in various Perl scripts.
205 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
207 PP/08 Reduce a number of compilation warnings under clang; building with
208 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
209 should be warning-free.
211 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
213 HS/01 Fix portability problems introduced by PP/08 for platforms where
214 realloc(NULL) is not equivalent to malloc() [SunOS et al].
216 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
217 chunk. This allows us to accept broken chunked messages. We need a more
218 general solution here.
220 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
221 already-broken messages in the queue.
223 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
225 JH/19 Fix reference counting bug in routing-generated-address tracking.
231 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
232 supports it and a size is available (ie. the sending peer gave us one).
234 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
235 years of being deprecated). The replacements are the ACLs
236 acl_smtp_mime and acl_not_smtp_mime.
238 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
239 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
240 or even an in-clear connection were permitted. Now, if the host lookup
241 was dnssec and dane was requested then the host is only used if the
242 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
243 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
244 if one fails this test.
245 This means that a poorly-configured remote DNS will make it incommunicado;
246 but it protects against a DNS-interception attack on it.
248 JH/04 Bug 1810: make continued-use of an open smtp transport connection
249 non-noisy when a race steals the message being considered.
251 JH/05 If main configuration option tls_certificate is unset, generate a
252 self-signed certificate for inbound TLS connections.
254 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
255 in rewrites and routers.
257 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
258 and logged a warning sing 4.83; now they are a configuration file error.
260 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
261 (lacking @domain). Apply the same qualification processing as RCPT.
263 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
265 JH/10 Support ${sha256:} applied to a string (as well as the previous
268 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
269 a cutthrough deliver is pending, as we always want to make a connection.
270 This also avoids re-routing the message when later placing the cutthrough
271 connection after a verify cache hit.
272 Do not update it with the verify result either.
274 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
275 when routing results in more than one destination address.
277 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
278 signing (which inhibits the cutthrough capability). Previously only
279 the presence of an option was tested; now an expansion evaluating as
280 empty is permissible (obviously it should depend only on data available
281 when the cutthrough connection is made).
283 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
284 the relevant preceding SMTP command did not note the pipelining mode.
286 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
287 Previously they were not counted.
289 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
290 as one having no matching records. Previously we deferred the message
291 that needed the lookup.
293 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
294 distinguished as "(=".
296 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
297 for missing MX records. Previously it only worked for missing A records.
299 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
301 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
302 after the data-go-ahead and data-ack. Patch from Jason Betts.
304 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
305 even for a "none" policy. Patch from Tony Meyer.
307 JH/22 Fix continued use of a connection for further deliveries. If a port was
308 specified by a router, it must also match for the delivery to be
311 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
312 When one of the recipients of a message was unsuitable for the connection
313 (has no matching addresses), we lost track of needing to mark it
314 deferred. As a result mail would be lost.
316 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
318 JH/25 Decoding ACL controls is now done using a binary search; the source code
319 takes up less space and should be simpler to maintain. Merge the ACL
320 condition decode tables also, with similar effect.
322 JH/26 Fix problem with one_time used on a redirect router which returned the
323 parent address unchanged. A retry would see the parent address marked as
324 delivered, so not attempt the (identical) child. As a result mail would
327 JH/27 Fix a possible security hole, wherein a process operating with the Exim
328 UID can gain a root shell. Credit to http://www.halfdog.net/ for
329 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
332 JH/28 Enable {spool,log} filesystem space and inode checks as default.
333 Main config options check_{log,spool}_{inodes,space} are now
334 100 inodes, 10MB unless set otherwise in the configuration.
336 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
337 Previously it only applied to the main-section connection policy
340 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
342 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
343 by me. Added RFC7919 DH primes as an alternative.
345 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
346 are not in the system include path.
348 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
349 GnuTLS, when a session startup failed (eg because the client disconnected)
350 Exim did stdio operations after fclose. This was exposed by a recent
351 change which nulled out the file handle after the fclose.
353 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
354 signed directly by the cert-signing cert, rather than an intermediate
355 OCSP-signing cert. This is the model used by LetsEncrypt.
357 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
359 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
360 an incoming connection.
362 HS/02 Bug 1802: Do not half-close the connection after sending a request
365 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
366 fallback to "prime256v1".
368 JH/34 SECURITY: Use proper copy of DATA command in error message.
369 Could leak key material. Remotely exploitable. CVE-2016-9963.
375 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
376 and 3.4.4 - once the server is enabled to respond to an OCSP request
377 it does even when not requested, resulting in a stapling non-aware
378 client dropping the TLS connection.
380 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
381 support variable-length bit vectors. No functional change.
383 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
384 The I= interface field on outgoing lines is now after the H= remote
385 host field, same as incoming lines. There is a separate
386 outgoing_interface log selector which allows you to disable the
389 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
390 If not running log_selector +smtp_connection the mainlog would be held
391 open indefinitely after a "too many connections" event, including to a
392 deleted file after a log rotate. Leave the per net connection logging
393 leaving it open for efficiency as that will be quickly detected by the
394 check on the next write.
396 HS/01 Bug 1671: Fix post transport crash.
397 Processing the wait-<transport> messages could crash the delivery
398 process if the message IDs didn't exist for some reason. When
399 using 'split_spool_directory=yes' the construction of the spool
400 file name failed already, exposing the same netto behaviour.
402 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
403 mime_regex ACL conditions.
405 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
406 to DSN fail messages (bounces): remote IP, remote greeting, remote response
407 to HELO, local diagnostic string.
409 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
410 log line to debug. Even when configured with a tls authenticator many
411 client connections are expected to not authenticate in this way, so
412 an authenticate fail is not an error.
414 HS/02 Add the Exim version string to the process info. This way exiwhat
415 gives some more detail about the running daemon.
417 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
418 matter for fast-change records such as DNSBLs.
420 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
421 retry record, even if constant. There may be multiple transports with
422 different interface settings and the retry behaviour needs to be kept
425 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
427 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
429 JH/10 Bug 840: fix log_defer_output option of pipe transport
431 JH/11 Bug 830: use same host for all RCPTS of a message, even under
432 hosts_randomize. This matters a lot when combined with mua_wrapper.
434 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
435 ${quote_pgsql:<string>} operator.
437 JH/13 Bug 1708: avoid misaligned access in cached lookup.
439 JH/14 Change header file name for freeradius-client. Relevant if compiling
440 with Radius support; from the Gentoo tree and checked under Fedora.
442 JH/15 Bug 1712: Introduce $prdr_requested flag variable
444 JH/16 Bug 1714: Permit an empty string as expansion result for transport
445 option transport_filter, meaning no filtering.
447 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
449 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
450 defaults to "*" (all hosts). The variable is now available when not built
451 with TLS, default unset, mainly to enable keeping the testsuite sane.
452 If a server certificate is not supplied (via tls_certificate) an error is
453 logged, and clients will find TLS connections fail on startup. Presumably
454 they will retry in-clear.
455 Packagers of Exim are strongly encouraged to create a server certificate
456 at installation time.
458 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
459 with the $config_file variable.
461 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
462 in transport context, after the attempt, and per-recipient. The latter type
463 is per host attempted. The event data is the error message, and the errno
464 information encodes the lookup type (A vs. MX) used for the (first) host,
465 and the trailing two digits of the smtp 4xx response.
467 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
468 to write to mainlog (or rejectlog, paniclog) in the window between file
469 creation and permissions/ownership being changed. Particularly affects
470 installations where exicyclog is run as root, rather than exim user;
471 result is that the running daemon panics and dies.
473 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
475 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
476 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
477 "pri" and "weight". Note that the previous implicit priority given by the
478 list order is no longer honoured.
480 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
483 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
484 by defining SUPPORT_SOCKS.
486 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
487 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
488 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
489 variables are renamed to proxy_{local,external}_{address,port}.
491 JH/27 Move Internationalisation support from Experimental to mainline, enabled
492 for a build by defining SUPPORT_I18N
494 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
495 of the query string, and make ${quote_redis:} do that quoting.
497 JH/29 Move Events support from Experimental to mainline, enabled by default
498 and removable for a build by defining DISABLE_EVENT.
500 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
502 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
503 cached by the daemon.
505 JH/32 Move Redis support from Experimental to mainline, enabled for a build
506 by defining LOOKUP_REDIS. The libhiredis library is required.
508 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
509 keys are given for lookup.
511 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
512 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
513 only supported when built with TLS support. The PolarSSL SHA routines
514 are still used when the TLS library is too old for convenient support.
516 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
517 openssl_options), for security. OpenSSL forces this from version 1.1.0
518 server-side so match that on older versions.
520 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
521 allocation for $value could be released as the expansion processing
522 concluded, but leaving the global pointer active for it.
524 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
525 and to use the domains and local_parts ACL conditions.
527 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
528 incorrectly not doubled on cutthrough transmission, hence seen as a
529 body-termination at the receiving system - resulting in truncated mails.
530 Commonly the sender saw a TCP-level error, and retransmitted the message
531 via the normal store-and-forward channel. This could result in duplicates
532 received - but deduplicating mailstores were liable to retain only the
533 initial truncated version.
535 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
537 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
539 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
540 we're in there, support oversigning also; bug 1309.
542 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
544 HS/04 Add support for keep_environment and add_environment options.
546 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
547 either intentional arithmetic overflow during PRNG, or testing config-
550 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
551 delivery resulted in actual delivery. Cancel cutthrough before DATA
554 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
555 rejects a recipient: pass the reject to the originator.
557 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
558 Many were false-positives and ignorable, but it's worth fixing the
561 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
562 for the new environment-manipulation done at startup. Move the routines
563 from being local to tls.c to being global via the os.c file.
565 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
566 an extract embedded as result-arg for a map, the first arg for extract
567 is unavailable so we cannot tell if this is a numbered or keyed
568 extraction. Accept either.
574 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
577 JH/02 The smtp transport option "multi_domain" is now expanded.
579 JH/03 The smtp transport now requests PRDR by default, if the server offers
582 JH/04 Certificate name checking on server certificates, when exim is a client,
583 is now done by default. The transport option tls_verify_cert_hostnames
584 can be used to disable this per-host. The build option
585 EXPERIMENTAL_CERTNAMES is withdrawn.
587 JH/05 The value of the tls_verify_certificates smtp transport and main options
588 default to the word "system" to access the system default CA bundle.
589 For GnuTLS, only version 3.0.20 or later.
591 JH/06 Verification of the server certificate for a TLS connection is now tried
592 (but not required) by default. The verification status is now logged by
593 default, for both outbound TLS and client-certificate supplying inbound
596 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
599 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
600 Status Notification (bounce) messages are now MIME format per RFC 3464.
601 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
602 under the control of the dsn_advertise_hosts option, and routers may
603 have a dsn_lasthop option.
605 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
606 default, modifiable by a malware= option. The list separator for
607 the options can now be changed in the usual way. Bug 68.
609 JH/10 The smtp_receive_timeout main option is now expanded before use.
611 JH/11 The incoming_interface log option now also enables logging of the
612 local interface on delivery outgoing connections.
614 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
615 if the interface and destination host and port all match.
617 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
620 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
621 Patch from Andrew Lewis.
623 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
624 now supports optional time-restrictions, weighting, and priority
625 modifiers per server. Patch originally by <rommer@active.by>.
627 JH/16 The spamd_address main option now supports a mixed list of local
628 and remote servers. Remote servers can be IPv6 addresses, and
629 specify a port-range.
631 JH/17 Bug 68: The spamd_address main option now supports an optional
632 timeout value per server.
634 JH/18 Bug 1581: Router and transport options headers_add/remove can
635 now have the list separator specified.
637 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
640 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
643 JH/21 Support for the A6 type of dns record is withdrawn.
645 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
646 rather than the verbs used.
648 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
649 from 255 to 1024 chars.
651 JH/24 Verification callouts now attempt to use TLS by default.
653 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
654 are generic router options now. The defaults didn't change.
656 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
657 Original patch from Alexander Shikoff, worked over by JH.
659 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
660 files if ZCAT_COMMAND is not executable.
662 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
664 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
666 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
667 Normally benign, it bites when the pair was led to by a CNAME;
668 modern usage is to not canonicalize the domain to a CNAME target
669 (and we were inconsistent anyway for A-only vs AAAA+A).
671 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
673 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
674 when evaluating $sender_host_dnssec.
676 JH/31 Check the HELO verification lookup for DNSSEC, adding new
677 $sender_helo_dnssec variable.
679 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
681 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
683 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
685 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
686 documented as working, but never had. Support all but $spam_report.
688 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
689 added for tls authenticator.
691 HS/03 Add perl_taintmode main config option
697 TL/01 When running the test suite, the README says that variables such as
698 no_msglog_check are global and can be placed anywhere in a specific
699 test's script, however it was observed that placement needed to be near
700 the beginning for it to behave that way. Changed the runtest perl
701 script to read through the entire script once to detect and set these
702 variables, reset to the beginning of the script, and then run through
703 the script parsing/test process like normal.
705 TL/02 The BSD's have an arc4random API. One of the functions to induce
706 adding randomness was arc4random_stir(), but it has been removed in
707 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
708 function when detected.
710 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
711 cause callback expansion.
713 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
714 syntax errors in an expansion can be treated as a string instead of
715 logging or causing an error, due to the internal use of bool_lax
716 instead of bool when processing it.
718 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
719 server certificates when making smtp deliveries.
721 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
723 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
725 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
727 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
728 Merged patch from Sebastian Wiedenroth.
730 JH/05 Fix results-pipe from transport process. Several recipients, combined
731 with certificate use, exposed issues where response data items split
732 over buffer boundaries were not parsed properly. This eventually
733 resulted in duplicates being sent. This issue only became common enough
734 to notice due to the introduction of connection certificate information,
735 the item size being so much larger. Found and fixed by Wolfgang Breyha.
737 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
738 size buffer was used, resulting in syntax errors when an expansion
741 JH/07 Add support for directories of certificates when compiled with a GnuTLS
742 version 3.3.6 or later.
744 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
745 is EXPERIMENTAL_EVENT, the main-configuration and transport options
746 both become "event_action", the variables become $event_name, $event_data
747 and $event_defer_errno. There is a new variable $verify_mode, usable in
748 routers, transports and related events. The tls:cert event is now also
749 raised for inbound connections, if the main configuration event_action
752 TL/06 In test suite, disable OCSP for old versions of openssl which contained
753 early OCSP support, but no stapling (appears to be less than 1.0.0).
755 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
756 server certificate names available under the smtp transport option
757 "tls_verify_cert_hostname" now do not permit multi-component wildcard
760 JH/10 Time-related extraction expansions from certificates now use the main
761 option "timezone" setting for output formatting, and are consistent
762 between OpenSSL and GnuTLS compilations. Bug 1541.
764 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
765 encoded parameter in the incoming message. Bug 1558.
767 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
768 include certificate info, eximon was claiming there were spoolfile
771 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
773 JH/14 Log delivery-related information more consistently, using the sequence
774 "H=<name> [<ip>]" wherever possible.
776 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
777 are problematic for Debian distribution, omit them from the release
780 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
782 JH/16 Fix string representation of time values on 64bit time_t architectures.
785 JH/17 Fix a null-indirection in certextract expansions when a nondefault
786 output list separator was used.
791 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
792 checkers that were complaining about end of non-void function with no
795 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
796 This was a regression introduced in 4.83 by another bugfix.
798 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
800 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
801 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
807 TF/01 Correctly close the server side of TLS when forking for delivery.
809 When a message was received over SMTP with TLS, Exim failed to clear up
810 the incoming connection properly after forking off the child process to
811 deliver the message. In some situations the subsequent outgoing
812 delivery connection happened to have the same fd number as the incoming
813 connection previously had. Exim would try to use TLS and fail, logging
814 a "Bad file descriptor" error.
816 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
817 utilities have not been installed.
819 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
820 temporary space as the ACL may create new global variables.
822 TL/01 LDAP support uses per connection or global context settings, depending
823 upon the detected version of the libraries at build time.
825 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
826 to extract and use the src ip:port in logging and expansions as if it
827 were a direct connection from the outside internet. PPv2 support was
828 updated based on HAProxy spec change in May 2014.
830 JH/02 Add ${listextract {number}{list}{success}{fail}}.
832 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
833 Properly escape header and check for NULL return.
835 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
838 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
840 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
841 characters in header names, implemented as a verify condition.
842 Contributed by Michael Fischer v. Mollard.
844 TL/05 Rename SPF condition results err_perm and err_temp to standardized
845 results permerror and temperror. Previous values are deprecated but
846 still accepted. In a future release, err_perm and err_temp will be
847 completely removed, which will be a backward incompatibility if the
848 ACL tests for either of these two old results. Patch contributed by
849 user bes-internal on the mailing list.
851 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
853 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
854 selectors, in both main and reject logs.
856 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
859 JH/07 Add malware type "sock" for talking to simple daemon.
861 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
863 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
864 routers/transports under cutthrough routing.
866 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
867 numbers. Touch up "bool" conditional to keep the same definition.
869 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
871 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
873 JH/12 Expand items in router/transport headers_add or headers_remove lists
874 individually rather than the list as a whole. Bug 1452.
876 Required for reasonable handling of multiple headers_ options when
877 they may be empty; requires that headers_remove items with embedded
878 colons must have them doubled (or the list-separator changed).
880 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
881 view the policy declared in the DMARC record. Currently, $dmarc_status
882 is a combined value of both the record presence and the result of the
885 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
887 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
888 dnslookup router and the smtp transport (applying to the forward
891 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
892 of ldap servers used for a specific lookup. Patch provided by Heiko
895 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
896 New variable $lookup_dnssec_authenticated for observability.
898 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
899 Patch submitted by Lars Timman.
901 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
903 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
904 Requires trusted mode and valid format message id, aborts otherwise.
905 Patch contributed by Heiko Schlichting.
907 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
908 certextract with support for various fields. Bug 1358.
910 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
911 is requested by default, modifiable by smtp transport option
914 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
915 operate on certificate variables to give certificate fingerprints
916 Also new ${sha256:cert_variable}.
918 JH/23 The PRDR feature is moved from being Experimental into the mainline.
920 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
921 Christian Aistleitner.
923 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
925 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
926 file. Patch from Wolfgang Breyha.
928 JH/25 Expand the coverage of the delivery $host and $host_address to
929 client authenticators run in verify callout. Bug 1476.
931 JH/26 Port service names are now accepted for tls_on_connect_ports, to
932 align with daemon_smtp_ports. Bug 72.
934 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
935 support and error reporting did not work properly.
937 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
938 and is readable. Patch from Andrew Colin Kissa.
940 TL/14 Enhance documentation of ${run expansion and how it parses the
941 commandline after expansion, particularly in the case when an
942 unquoted variable expansion results in an empty value.
944 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
946 PP/02 Fix internal collision of T_APL on systems which support RFC3123
947 by renaming away from it. Addresses GH issue 15, reported by
950 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
952 TL/15 SECURITY: prevent double expansion in math comparison functions
953 (can expand unsanitized data). Not remotely exploitable.
960 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
962 PP/02 Make -n do something, by making it not do something.
963 When combined with -bP, the name of an option is not output.
965 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
968 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
969 $sender_host_name and config options to manage this, and basic check
972 PP/05 DSCP support for outbound connections and control modifier for inbound.
974 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
975 (Only plugin which currently uses this is kerberos4, which nobody should
976 be using, but we should make it available and other future plugins might
977 conceivably use it, even though it would break NAT; stuff *should* be
978 using channel bindings instead).
980 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
981 name; added for Sendmail compatibility; requires admin caller.
982 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
983 just ignore it); requires trusted caller.
984 Also parse but ignore: -Ac -Am -X<logfile>
987 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
989 TL/02 Add +smtp_confirmation as a default logging option.
991 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
992 Patch by Magnus Holmgren from 2007-02-20.
994 TL/04 Bugzilla 1281 - Spec typo.
995 Bugzilla 1283 - Spec typo.
996 Bugzilla 1290 - Spec grammar fixes.
998 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1000 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1002 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1003 mailing list by by Dmitry Isaikin.
1005 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1007 JH/02 Support "G" suffix to numbers in ${if comparisons.
1009 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1011 NM/01 Bugzilla 1197 - Spec typo
1012 Bugzilla 1196 - Spec examples corrections
1014 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1016 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1017 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1020 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1021 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1023 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1024 "acl {{name}{arg}...}", and optional args on acl condition
1027 JH/05 Permit multiple router/transport headers_add/remove lines.
1029 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1031 JH/07 Avoid using a waiting database for a single-message-only transport.
1032 Performance patch from Paul Fisher. Bugzilla 1262.
1034 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1037 JH/09 Add $headers_added variable, with content from use of ACL modifier
1038 add_header (but not yet added to the message). Bugzilla 199.
1040 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1041 Pulled from Bugzilla 817 by Wolfgang Breyha.
1043 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1045 (nb: this is the same fix as in Exim 4.80.1)
1047 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1050 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1053 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1055 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1056 advertise SMTP AUTH mechanism to us, instead of a generic
1057 protocol violation error. Also, make Exim more robust to bad
1058 data from the Dovecot auth socket.
1060 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1062 When a queue runner is handling a message, Exim first routes the
1063 recipient addresses, during which it prunes them based on the retry
1064 hints database. After that it attempts to deliver the message to
1065 any remaining recipients. It then updates the hints database using
1068 So if a recipient address works intermittently, it can get repeatedly
1069 deferred at routing time. The retry hints record remains fresh so the
1070 address never reaches the final cutoff time.
1072 This is a fairly common occurrence when a user is bumping up against
1073 their storage quota. Exim had some logic in its local delivery code
1074 to deal with this. However it did not apply to per-recipient defers
1075 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1077 This change adds a proper retry rule check during routing so that the
1078 final cutoff time is checked against the message's age. We only do
1079 this check if there is an address retry record and there is not a
1080 domain retry record; this implies that previous attempts to handle
1081 the address had the retry_use_local_parts option turned on. We use
1082 this as an approximation for the destination being like a local
1083 delivery, as in LMTP.
1085 I suspect this new check makes the old local delivery cutoff check
1086 redundant, but I have not verified this so I left the code in place.
1088 TF/02 Correct gecos expansion when From: is a prefix of the username.
1090 Test 0254 submits a message to Exim with the header
1094 When I ran the test suite under the user fanf2, Exim expanded
1095 the header to contain my full name, whereas it should have added
1096 a Resent-Sender: header. It erroneously treats any prefix of the
1097 username as equal to the username.
1099 This change corrects that bug.
1101 GF/01 DCC debug and logging tidyup
1102 Error conditions log to paniclog rather than rejectlog.
1103 Debug lines prefixed by "DCC: " to remove any ambiguity.
1105 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1107 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1108 Bug spotted by Jeremy Harris; was flawed since initial commit.
1109 Would have resulted in OCSP responses post-SNI triggering an Exim
1110 NULL dereference and crash.
1112 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1114 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1115 Bug detection, analysis and fix by Samuel Thibault.
1116 Bugzilla 1331, Debian bug #698092.
1118 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1120 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1121 Server implementation by Todd Lyons, client by JH.
1122 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1123 config variable "prdr_enable" controls whether the server
1124 advertises the facility. If the client requests PRDR a new
1125 acl_data_smtp_prdr ACL is called once for each recipient, after
1126 the body content is received and before the acl_smtp_data ACL.
1127 The client is controlled by both of: a hosts_try_prdr option
1128 on the smtp transport, and the server advertisement.
1129 Default client logging of deliveries and rejections involving
1130 PRDR are flagged with the string "PRDR".
1132 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1133 fclose(). Diagnosis by Todd Lyons.
1135 PP/17 Update configure.default to handle IPv6 localhost better.
1136 Patch by Alain Williams (plus minor tweaks).
1139 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1140 This is now consistent with GnuTLS, and is now documented: the
1141 previous undocumented portable approach to treating the option as
1142 unset was to force an expansion failure. That still works, and
1143 an empty string is now equivalent.
1145 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1146 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1147 not performing validation itself.
1149 PP/20 Added force_command boolean option to pipe transport.
1150 Patch from Nick Koston, of cPanel Inc.
1152 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1155 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1157 PP/21 Fix eximon continuous updating with timestamped log-files.
1158 Broken in a format-string cleanup in 4.80, missed when I repaired the
1159 other false fix of the same issue.
1160 Report and fix from Heiko Schlichting.
1163 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1164 Report from Prashanth Katuri.
1166 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1167 It's SecureTransport, so affects any MacOS clients which use the
1168 system-integrated TLS libraries, including email clients.
1170 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1171 using a MIME ACL for non-SMTP local injection.
1172 Report and assistance in diagnosis by Warren Baker.
1174 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1176 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1178 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1179 last id that failed so it may be referenced in subsequent ACL's.
1181 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1184 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1185 ldap library initialization, allowing self-signed CA's to be
1186 used. Also properly sets require_cert option later in code by
1187 using NULL (global ldap config) instead of ldap handle (per
1188 session). Bug diagnosis and testing by alxgomz.
1190 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1191 the src/util/ subdirectory.
1193 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1194 renamed to Transport Post Delivery Action by Jeremy Harris, as
1197 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1198 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1199 redis_servers = needs to be configured which will be used by the redis
1200 lookup. Patch from Warren Baker, of The Packet Hub.
1202 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1204 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1205 hostname or reverse DNS when processing a host list. Used suggestions
1206 from multiple comments on this bug.
1208 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1210 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1211 Missed a few lines, added it to make the runtest require no keyboard
1214 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1215 contains upper case chars. Make router use caseful_local_part.
1217 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1218 support when GnuTLS has been built with p11-kit.
1224 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1226 This, or similar/improved, will also be change PP/11 of 4.82.
1232 PP/01 Handle short writes when writing local log-files.
1233 In practice, only affects FreeBSD (8 onwards).
1234 Bugzilla 1053, with thanks to Dmitry Isaikin.
1236 NM/01 Bugzilla 949 - Documentation tweak
1238 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1241 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1243 PP/02 Implemented gsasl authenticator.
1245 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1247 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1248 `pkg-config foo` for cflags/libs.
1250 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1251 with rest of GSASL and with heimdal_gssapi.
1253 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1254 `pkg-config foo` for cflags/libs for the TLS implementation.
1256 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1257 properties get this fed in as external SSF. A number of robustness
1258 and debugging improvements to the cyrus_sasl authenticator.
1260 PP/08 cyrus_sasl server now expands the server_realm option.
1262 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1263 Patch by Jeremy Harris.
1265 PP/10 Added dbmjz lookup type.
1267 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1269 PP/12 MAIL args handles TAB as well as SP, for better interop with
1270 non-compliant senders.
1271 Analysis and variant patch by Todd Lyons.
1273 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1274 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1275 Patch from Dirk Mueller <dmueller@suse.com>
1277 PP/13 tls_peerdn now print-escaped for spool files.
1278 Observed some $tls_peerdn in wild which contained \n, which resulted
1279 in spool file corruption.
1281 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1282 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1283 or write after TLS renegotiation, which otherwise led to messages
1286 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1287 as a tracking header (ie: a signed header comes before the signature).
1288 Patch from Wolfgang Breyha.
1290 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1291 comma-sep list; embedded commas doubled.
1293 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1295 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1297 Report and patch from Dmitry Banschikov.
1299 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1300 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1301 failures. We appear to now support TLS1.1+ with Exim.
1303 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1304 lets Exim select keys and certificates based upon TLS SNI from client.
1305 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1306 before an outbound SMTP session. New log_selector, +tls_sni.
1308 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1309 NULL dereference. Report and patch from Alun Jones.
1311 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1312 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1313 Not seeing resolver debug output on NetBSD, but suspect this is a
1314 resolver implementation change.
1316 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1317 Left warnings. Added "eximon gdb" invocation mode.
1319 PP/21 Defaulting "accept_8bitmime" to true, not false.
1321 PP/22 Added -bw for inetd wait mode support.
1323 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1324 locate the relevant includes and libraries. Made this the default.
1326 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1327 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1329 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1330 This may cause build issues on older platforms.
1332 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1333 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1334 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1335 Added SNI support via GnuTLS too.
1336 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1338 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1340 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1341 Applied second patch from Janne, implementing suggestion to default
1342 multiple-strings-in-record handling to match SPF spec.
1344 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1346 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1347 read-only, out of scope).
1348 Patch from Wolfgang Breyha, report from Stuart Northfield.
1350 PP/29 Fix three issues highlighted by clang analyser static analysis.
1351 Only crash-plausible issue would require the Cambridge-specific
1352 iplookup router and a misconfiguration.
1353 Report from Marcin Mirosław.
1355 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1357 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1358 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1359 As part of this, removing so much warning spew let me fix some minor
1360 real issues in debug logging.
1362 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1363 assignment on my part. Fixed.
1365 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1366 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1367 Janne Snabb (who went above and beyond: thank you).
1369 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1370 string otherwise requires a connection and a bunch more work and it's
1371 relatively easy to get wrong. Should also expose TLS library linkage
1374 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1375 64-bit ${eval} (JH/03).
1377 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1378 GNU libc to support some of the 64-bit stuff, should not lead to
1379 conflicts. Defined before os.h is pulled in, so if a given platform
1380 needs to override this, it can.
1382 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1383 protection layer was required, which is not implemented.
1384 Bugzilla 1254, patch from Wolfgang Breyha.
1386 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1387 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1388 tls_dhparam take prime identifiers. Also unbreak combination of
1389 OpenSSL+DH_params+TLSSNI.
1391 PP/39 Disable SSLv2 by default in OpenSSL support.
1397 PP/01 Solaris build fix for Oracle's LDAP libraries.
1398 Bugzilla 1109, patch from Stephen Usher.
1400 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1402 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1405 TF/02 Fix a couple more cases where we did not log the error message
1406 when unlink() failed. See also change 4.74-TF/03.
1408 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1409 lock up or crash if it happened to be inside a call to libc when it
1410 got a SIGUSR1 from exiwhat.
1412 The SIGUSR1 handler appends the current process status to the process
1413 log which is later printed by exiwhat. It used to use the general
1414 purpose logging code to do this, but several functions it calls are
1415 not safe for signals.
1417 The new output code in the SIGUSR1 handler is specific to the process
1418 log, and simple enough that it's easy to inspect for signal safety.
1419 Removing some special cases also simplifies the general logging code.
1420 Removing the spurious timestamps from the process log simplifies
1423 TF/04 Improved ratelimit ACL condition.
1425 The /noupdate option has been deprecated in favour of /readonly which
1426 has clearer semantics. The /leaky, /strict, and /readonly update modes
1427 are mutually exclusive. The update mode is no longer included in the
1428 database key; it just determines when the database is updated. (This
1429 means that when you upgrade Exim will forget old rate measurements.)
1431 Exim now checks that the per_* options are used with an update mode that
1432 makes sense for the current ACL. For example, when Exim is processing a
1433 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1434 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1435 must specify per_mail/readonly. If you omit the update mode it defaults to
1436 /leaky where that makes sense (as before) or /readonly where required.
1438 The /noupdate option is now undocumented but still supported for
1439 backwards compatibility. It is equivalent to /readonly except that in
1440 ACLs where /readonly is required you may specify /leaky/noupdate or
1441 /strict/noupdate which are treated the same as /readonly.
1443 A useful new feature is the /count= option. This is a generalization
1444 of the per_byte option, so that you can measure the throughput of other
1445 aggregate values. For example, the per_byte option is now equivalent
1446 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1448 The per_rcpt option has been generalized using the /count= mechanism
1449 (though it's more complicated than the per_byte equivalence). When it is
1450 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1451 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1452 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1453 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1454 Note that using per_rcpt with a non-readonly update mode in more than
1455 one ACL will cause the recipients to be double-counted. (The per_mail
1456 and per_byte options don't have this problem.)
1458 The handling of very low rates has changed slightly. If the computed rate
1459 is less than the event's count (usually one) then this event is the first
1460 after a long gap. In this case the rate is set to the same as this event's
1461 count, so that the first message of a spam run is counted properly.
1463 The major new feature is a mechanism for counting the rate of unique
1464 events. The new per_addr option counts the number of different
1465 recipients that someone has sent messages to in the last time period. It
1466 behaves like per_rcpt if all the recipient addresses are different, but
1467 duplicate recipient addresses do not increase the measured rate. Like
1468 the /count= option this is a general mechanism, so the per_addr option
1469 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1470 example, measure the rate that a client uses different sender addresses
1471 with the options per_mail/unique=$sender_address. There are further
1472 details in the main documentation.
1474 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1476 TF/06 Removed a few PCRE remnants.
1478 TF/07 Automatically extract Exim's version number from tags in the git
1479 repository when doing development or release builds.
1481 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1482 Bugzilla 879. Patch from Paul Fisher.
1484 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1485 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1488 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1490 PP/05 Variable $av_failed, true if the AV scanner deferred.
1491 Bugzilla 1078. Patch from John Horne.
1493 PP/06 Stop make process more reliably on build failure.
1494 Bugzilla 1087. Patch from Heiko Schlittermann.
1496 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1497 Bugzilla 1089. Patch from Heiko Schlittermann.
1499 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1500 Bugzilla 1131. Patch from Holger Weiß.
1502 PP/09 Handle IPv6 addresses with SPF.
1503 Bugzilla 860. Patch from Wolfgang Breyha.
1505 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1507 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1510 PP/11 match_* no longer expand right-hand-side by default.
1511 New compile-time build option, EXPAND_LISTMATCH_RHS.
1512 New expansion conditions, "inlist", "inlisti".
1514 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1516 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1518 PP/14 fix log_write() format string regression from TF/03.
1519 Bugzilla 1152. Patch from Dmitry Isaikin.
1525 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1527 PP/02 Harmonised TLS library version reporting; only show if debugging.
1528 Layout now matches that introduced for other libraries in 4.74 PP/03.
1530 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1532 PP/04 New "dns_use_edns0" global option.
1534 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1537 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1538 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1540 TK/01 Updated PolarSSL code to 0.14.2.
1541 Bugzilla 1097. Patch from Andreas Metzler.
1543 PP/07 Catch divide-by-zero in ${eval:...}.
1544 Fixes bugzilla 1102.
1546 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1549 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1550 format-string attack -- SECURITY: remote arbitrary code execution.
1552 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1553 time unintentionally subject to list matching rules, letting the header
1554 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1555 arbitrary string expansion). This allowed for information disclosure.
1557 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1558 INT_MIN/-1 -- value coerced to INT_MAX.
1564 NM/01 Workaround for PCRE version dependency in version reporting
1567 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1568 This fixes portability to compilers other than gcc, notably
1569 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1571 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1572 makefiles for portability to HP-UX and POSIX correctness.
1574 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1575 Also via indented variable definition in the Makefile.
1576 (Debugging by Oliver Heesakkers).
1578 PP/02 Restore caching of spamd results with expanded spamd_address.
1579 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1581 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1582 Improves build reliability. Fix from: Frank Elsner
1584 NM/02 Fix wide character breakage in the rfc2047 coding
1585 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1587 NM/03 Allow underscore in dnslist lookups
1588 Fixes bug 1026. Patch from Graeme Fowler
1590 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1591 Code patches from Adam Ciarcinski of NetBSD.
1593 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1596 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1597 is logged, to avoid truncation. Patch from John Horne.
1599 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1600 Patch from Jakob Hirsch.
1602 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1603 SQL string expansion failure details.
1604 Patch from Andrey Oktyabrski.
1606 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1607 Patch from Simon Arlott.
1609 PP/09 New lookups functionality failed to compile on old gcc which rejects
1610 extern declarations in function scope.
1611 Patch from Oliver Fleischmann
1613 PP/10 Use sig_atomic_t for flags set from signal handlers.
1614 Check getgroups() return and improve debugging.
1615 Fixed developed for diagnosis in bug 927 (which turned out to be
1618 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1619 Patch from Mark Zealey.
1621 PP/12 Bugzilla 1056: Improved spamd server selection.
1622 Patch from Mark Zealey.
1624 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1625 Based on patch from Heiko Schlittermann.
1627 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1628 Patch from Uwe Doering, sign-off by Michael Haardt.
1630 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1631 variable declaration deep within a block. Bug and patch from
1634 PP/15 lookups-Makefile IRIX compatibility coercion.
1636 PP/16 Make DISABLE_DKIM build knob functional.
1638 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1639 Patch by Simon Arlott
1641 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1642 variable argument macros. Our copy now differs from upstream.
1648 TF/01 Failure to get a lock on a hints database can have serious
1649 consequences so log it to the panic log.
1651 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1652 controlled using the smtp_confirmation log selector.
1654 TF/03 Include the error message when we fail to unlink a spool file.
1656 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1657 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1658 for maintaining out-of-tree patches for some time.
1660 PP/01 Bugzilla 139: Documentation and portability issues.
1661 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1662 Handle per-OS dynamic-module compilation flags.
1664 PP/02 Let /dev/null have normal permissions.
1665 The 4.73 fixes were a little too stringent and complained about the
1666 permissions on /dev/null. Exempt it from some checks.
1667 Reported by Andreas M. Kirchwitz.
1669 PP/03 Report version information for many libraries, including
1670 Exim version information for dynamically loaded libraries. Created
1671 version.h, now support a version extension string for distributors
1672 who patch heavily. Dynamic module ABI change.
1674 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1675 privilege escalation vulnerability whereby the Exim run-time user
1676 can cause root to append content of the attacker's choosing to
1679 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1682 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1683 If dropping privileges for untrusted macros, we disabled normal logging
1684 on the basis that it would fail; for the Exim run-time user, this is not
1685 the case, and it resulted in successful deliveries going unlogged.
1686 Fixed. Reported by Andreas Metzler.
1692 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1693 only prepend for the Resent-* case. Fixes regression introduced in
1694 Exim 4.70 by NM/22 for Bugzilla 607.
1696 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1697 increasing numbers of administrators be bitten by this.
1699 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1701 PP/03 Bugzilla 994: added openssl_options main configuration option.
1703 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1705 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1707 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1709 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1710 without a peer certificate, leading to a segfault because of an
1711 assumption that peers always have certificates. Be a little more
1712 paranoid. Problem reported by Martin Tscholak.
1714 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1715 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1716 NB: ClamAV planning to remove STREAM in "middle of 2010".
1717 CL also introduces -bmalware, various -d+acl logging additions and
1718 more caution in buffer sizes.
1720 PP/09 Implemented reverse_ip expansion operator.
1722 PP/10 Bugzilla 937: provide a "debug" ACL control.
1724 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1726 PP/12 Bugzilla 973: Implement --version.
1728 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1730 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1732 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1734 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1735 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1736 ignore trailing whitespace.
1738 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1740 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1743 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1744 Notification from Dr Andrew Aitchison.
1746 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1747 ExtendedDetectionInfo response format.
1748 Notification from John Horne.
1750 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1753 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1754 XSL and documented dependency on system catalogs, with examples of how
1757 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1760 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1761 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1762 configuration file which is writeable by the Exim user or group.
1764 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1765 of configuration files to cover files specified with the -C option if
1766 they are going to be used with root privileges, not just the default
1769 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1770 option (effectively making it always true).
1772 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1773 files to be used while preserving root privileges.
1775 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1776 that rogue child processes cannot use them.
1778 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1779 run-time user, instead of root.
1781 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1782 Exim run-time user without dropping privileges.
1784 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1785 result string, instead of calling string_vformat() twice with the same
1788 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1789 for other users. Others should always drop root privileges if they use
1790 -C on the command line, even for a whitelisted configure file.
1792 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1794 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1800 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1801 $data_path, and $header_path variables; fixed documentation bugs and
1804 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1805 exipick to access non-standard spools, including the "frozen" queue
1808 NM/01 Bugzilla 965: Support mysql stored procedures.
1809 Patch from Alain Williams
1811 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1813 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1814 Patch from Andreas Metzler
1816 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1817 Patch from Kirill Miazine
1819 NM/05 Bugzilla 671: Added umask to procmail example.
1821 JJ/03 installed exipick 20100323.0, fixing doc bug
1823 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1824 directory. Notification and patch from Dan Rosenberg.
1826 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1828 TK/02 Improve log output when DKIM signing operation fails.
1830 MH/01 Treat the transport option dkim_domain as a colon separated
1831 list, not as a single string, and sign the message with each element,
1832 omitting multiple occurences of the same signer.
1834 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1835 Bugzilla 985, 986. Patch by Simon Arlott
1837 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1838 Patch by Simon Arlott
1840 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1841 MBX locking. Notification from Dan Rosenberg.
1847 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1849 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1851 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1853 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1855 NM/04 Bugzilla 915: Flag broken perl installation during build.
1861 TK/01 Added patch by Johannes Berg that expands the main option
1862 "spamd_address" if it starts with a dollar sign.
1864 TK/02 Write list of recipients to X-Envelope-Sender header when building
1865 the mbox-format spool file for content scanning (suggested by Jakob
1868 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1869 (http://www.dcc-servers.net/) support via dccifd. Activated by
1870 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1872 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1873 by Mark Daniel Reidel <mr@df.eu>.
1875 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1876 When building exim an external PCRE library is now needed -
1877 PCRE is a system library on the majority of modern systems.
1878 See entry on PCRE_LIBS in EDITME file.
1880 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1881 conversation. Added nologin parameter to request.
1882 Patch contributed by Kirill Miazine.
1884 TF/01 Do not log submission mode rewrites if they do not change the address.
1886 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1888 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1889 log files in place. Contributed by Roberto Lima.
1891 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1893 TF/03 Bugzilla 615: When checking the local_parts router precondition
1894 after a local_part_suffix or local_part_prefix option, Exim now
1895 does not use the address's named list lookup cache, since this
1896 contains cached lookups for the whole local part.
1898 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1899 Robert Millan. Documentation is in experimental-spec.txt.
1901 TF/04 Bugzilla 668: Fix parallel build (make -j).
1903 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1905 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1906 Patch provided by Jan Srzednicki.
1908 TF/05 Leading white space used to be stripped from $spam_report which
1909 wrecked the formatting. Now it is preserved.
1911 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1912 that they are available at delivery time.
1914 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1916 TF/08 TLS error reporting now respects the incoming_interface and
1917 incoming_port log selectors.
1919 TF/09 Produce a more useful error message if an SMTP transport's hosts
1920 setting expands to an empty string.
1922 NM/06 Bugzilla 744: EXPN did not work under TLS.
1923 Patch provided by Phil Pennock.
1925 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1926 Patch provided by Richard Godbee.
1928 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1929 acl_smtp_notquit, added index entry.
1931 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1932 Patch provided by Eugene Bujak.
1934 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1935 accept(). Patch provided by Maxim Dounin.
1937 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1938 Patch provided by Phil Pennock.
1940 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1942 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1943 Patch provided by Brad "anomie" Jorsch.
1945 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1946 Patch provided by Dean Brooks.
1948 TK/05 Add native DKIM support (does not depend on external libraries).
1950 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1951 Patch provided by Graeme Fowler.
1953 NM/16 Bugzilla 851: Documentation example syntax fix.
1955 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1957 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1960 NM/19 Bugzilla 745: TLS version reporting.
1961 Patch provided by Phil Pennock.
1963 NM/20 Bugzilla 167: bool: condition support.
1964 Patch provided by Phil Pennock.
1966 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1967 clients. Patch provided by Phil Pennock.
1969 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1970 Patch provided by Brad "anomie" Jorsch.
1972 NM/23 Bugzilla 687: Fix misparses in eximstats.
1973 Patch provided by Heiko Schlittermann.
1975 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1976 Patch provided by Heiko Schlittermann.
1978 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1979 plus update to original patch.
1981 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1983 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1984 Patch provided by David Brownlee.
1986 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1988 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1990 NM/30 Bugzilla 888: TLS documentation bugfixes.
1992 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1994 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1995 Unlike the original bugzilla I have changed all shell scripts in src tree.
1997 NM/33 Bugzilla 898: Transport filter timeout fix.
1998 Patch by Todd Rinaldo.
2000 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2001 Patch by Serge Demonchaux.
2003 NM/35 Bugzilla 39: Base64 decode bug fixes.
2004 Patch by Jakob Hirsch.
2006 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2008 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2010 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2012 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2018 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2019 ALT-N's libdkim that I have put here:
2020 http://duncanthrax.net/exim-experimental/
2022 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2023 were called 'true' and it seems that C99 defines that as a reserved
2024 keyword to be used with 'bool' variable types. That means you could
2025 not include C99-style headers which use bools without triggering
2026 build errors in sieve.c.
2028 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2029 as mailq or other aliases. Changed the --help handling significantly
2030 to do whats expected. exim_usage() emits usage/help information.
2032 SC/01 Added the -bylocaldomain option to eximstats.
2034 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2036 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2038 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2044 PH/01 Another patch from the Sieve maintainer.
2046 PH/02 When an IPv6 address is converted to a string for single-key lookup
2047 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2048 dots are used instead of colons so that keys in lsearch files need not
2049 contain colons. This was done some time before quoting was made available
2050 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2051 (notated using the quote facility) so as to distinguish them from IPv4
2052 keys. This meant that lookups for IP addresses in host lists did not work
2053 for iplsearch lookups.
2055 This has been fixed by arranging for IPv6 addresses to be expressed with
2056 colons if the lookup type is iplsearch. This is not incompatible, because
2057 previously such lookups could never work.
2059 The situation is now rather anomalous, since one *can* have colons in
2060 ordinary lsearch keys. However, making the change in all cases is
2061 incompatible and would probably break a number of configurations.
2063 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2066 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2067 conversion specification without a maximum field width, thereby enabling
2068 a rogue spamd server to cause a buffer overflow. While nobody in their
2069 right mind would setup Exim to query an untrusted spamd server, an
2070 attacker that gains access to a server running spamd could potentially
2071 exploit this vulnerability to run arbitrary code as the Exim user.
2073 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2074 $primary_hostname instead of what libspf2 thinks the hosts name is.
2076 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2077 a directory entry by the name of the lookup key. Previously, if a
2078 symlink pointed to a non-existing file or a file in a directory that
2079 Exim lacked permissions to read, a lookup for a key matching that
2080 symlink would fail. Now it is enough that a matching directory entry
2081 exists, symlink or not. (Bugzilla 503.)
2083 PH/03 The body_linecount and body_zerocount variables are now exported in the
2086 PH/04 Added the $dnslist_matched variable.
2088 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2089 This means they are set thereafter only if the connection becomes
2092 PH/06 Added the client_condition to authenticators so that some can be skipped
2093 by clients under certain conditions.
2095 PH/07 The error message for a badly-placed control=no_multiline_responses left
2096 "_responses" off the end of the name.
2098 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2100 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2101 (without spaces) instead of just copying the configuration text.
2103 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2105 PH/11 Added $max_received_linelength.
2107 PH/12 Added +ignore_defer and +include_defer to host lists.
2109 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2110 way in which PCRE > 7.0 is built.
2112 PH/14 Implemented queue_only_load_latch.
2114 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2115 MAIL command. The effect was to mangle the value on 64-bit systems.
2117 PH/16 Another patch from the Sieve maintainer.
2119 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2121 PH/18 If a system quota error occurred while trying to create the file for
2122 a maildir delivery, the message "Mailbox is full" was not appended to the
2123 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2124 only to a quota excession during the actual writing of the file.
2126 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2127 characters?) which causes problems in log lines. The DN values are now
2128 passed through string_printing() before being added to log lines.
2130 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2131 and InterBase are left for another time.)
2133 PH/21 Added message_body_newlines option.
2135 PH/22 Guard against possible overflow in moan_check_errorcopy().
2137 PH/23 POSIX allows open() to be a macro; guard against that.
2139 PH/24 If the recipient of an error message contained an @ in the local part
2140 (suitably quoted, of course), incorrect values were put in $domain and
2141 $local_part during the evaluation of errors_copy.
2147 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2148 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2151 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2152 issue a MAIL command.
2154 PH/02 In an ACL statement such as
2156 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2158 if a client was not listed at all, or was listed with a value other than
2159 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2160 the condition was not true (as it should be), so access was not denied.
2161 The bug was that the ! inversion was incorrectly passed on to the second
2162 item. This has been fixed.
2164 PH/03 Added additional dnslists conditions == and =& which are different from
2165 = and & when the dns lookup returns more than one IP address.
2167 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2168 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2170 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2171 FSYNC, which compiles an option called disable_fsync that allows for
2172 bypassing fsync(). The documentation is heavily laced with warnings.
2174 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2176 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2177 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2178 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2179 including adding "make clean"; (3) Added -fPIC when compiling the test
2180 dynamically loaded module, to get rid of a warning.
2182 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2183 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2184 = 0s. The bug is otherwise harmless.
2186 PH/07 There was a bug in the dovecot authenticator such that the value of
2187 $auth1 could be overwritten, and so not correctly preserved, after a
2188 successful authentication. This usually meant that the value preserved by
2189 the server_setid option was incorrect.
2191 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2193 PH/09 Installed PCRE release 7.0.
2195 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2196 run for batched SMTP input. It is now run at the start of every message
2197 in the batch. While fixing this I discovered that the process information
2198 (output by running exiwhat) was not always getting set for -bs and -bS
2199 input. This is fixed, and it now also says "batched" for BSMTP.
2201 PH/11 Added control=no_pipelining.
2203 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2204 patch, slightly modified), and move the expansion of helo_data till after
2205 the connection is made in the smtp transport (so it can use these
2208 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2210 PH/14 Added log_selector = +pid.
2212 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2214 PH/16 Add ${if forany and ${if forall.
2216 PH/17 Added dsn_from option to vary the From: line in DSNs.
2218 PH/18 Flush SMTP output before performing a callout, unless control =
2219 no_callout_flush is set.
2221 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2222 was true (the default) a successful delivery failed to delete the retry
2223 item, thus causing premature timeout of the address. The bug is now
2226 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2228 PH/21 Long custom messages for fakedefer and fakereject are now split up
2229 into multiline responses in the same way that messages for "deny" and
2230 other ACL rejections are.
2232 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2233 with slight modification.
2235 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2236 draft, changing the syntax and factoring some duplicate code".
2238 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2239 for deliveries of the second and subsequent messages over the same SMTP
2242 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2243 ${reduce, with only minor "tidies".
2245 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2247 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2248 expansion side effects.
2250 PH/27 When a message times out after an over-quota error from an Exim-imposed
2251 quota, the bounce message says "mailbox is full". This message was not
2252 being given when it was a system quota that was exceeded. It now should
2255 MH/03 Made $recipients available in local_scan(). local_scan() already has
2256 better access to the recipient list through recipients_list[], but
2257 $recipients can be useful in postmaster-provided expansion strings.
2259 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2260 in the case of a MAIL command with additional options following the
2261 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2262 were accidentally chopped off.
2264 PH/29 SMTP synchronization checks are implemented when a command is read -
2265 there is a check that no more input is waiting when there shouldn't be
2266 any. However, for some commands, a delay in an ACL can mean that it is
2267 some time before the response is written. In this time, more input might
2268 arrive, invalidly. So now there are extra checks after an ACL has run for
2269 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2270 pipelining has not been advertised.
2272 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2274 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2275 returned in the SMTP response when smtp_return_error_details was false.
2276 This has been fixed.
2278 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2279 instead of the C I/O that was originally supplied, because problems were
2280 reported on Solaris.
2282 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2283 Exim which did not show up earlier: it was assuming that a call to
2284 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2285 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2286 was a macro that became an assignment, so it seemed to work. This has
2287 changed to a proper function call with a void return, hence the compile
2288 error. Exim's code has been fixed.
2290 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2293 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2295 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2297 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2298 because it turns out to be too big or there is a timeout) there is no
2299 "Completed" line in the log. When some messages of this type were
2300 selected by exigrep, they were listed as "not completed". Others were
2301 picked up by some special patterns. I have improved the selection
2302 criteria to be more general.
2304 PH/38 The host_find_failed option in the manualroute router can now be set
2305 to "ignore", to completely ignore a host whose IP address cannot be
2306 found. If all hosts are ignored, the behaviour is controlled by the new
2307 host_all_ignored option.
2309 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2310 homing or because of multiple MX records with /mx) generated more than
2311 one IP address, and the following item turned out to be the local host,
2312 all the secondary addresses of the first item were incorrectly removed
2313 from the list, along with the local host and any following hosts (which
2314 is what is supposed to happen).
2316 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2317 whoever called Exim into the -H file. In the case of the daemon it was
2318 behaving confusingly. When first started, it used values for whoever
2319 started the daemon, but after a SIGHUP it used the Exim user (because it
2320 calls itself on a restart). I have changed the code so that it now always
2323 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2324 message are rejected with the same error (e.g. no authentication or bad
2325 sender address), and a DATA command is nevertheless sent (as can happen
2326 with PIPELINING or a stupid MUA), the error message that was given to the
2327 RCPT commands is included in the rejection of the DATA command. This is
2328 intended to be helpful for MUAs that show only the final error to their
2331 PH/42 Another patch from the Sieve maintainer.
2333 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2334 Eximstats - Fixed some broken HTML links and added missing column headers
2336 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2337 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2339 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2341 PH/43 Yet another patch from the Sieve maintainer.
2343 PH/44 I found a way to check for a TCP/IP connection going away before sending
2344 the response to the final '.' that terminates a message, but only in the
2345 case where the client has not sent further data following the '.'
2346 (unfortunately, this is allowed). However, in many cases there won't be
2347 any further data because there won't be any more messages to send. A call
2348 to select() can be used: if it shows that the input is "ready", there is
2349 either input waiting, or the socket has been closed. An attempt to read
2350 the next input character can distinguish the two cases. Previously, Exim
2351 would have sent an OK response which the client would never have see.
2352 This could lead to message repetition. This fix should cure that, at
2353 least in a lot of common cases.
2355 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2356 advertised in response to EHLO.
2362 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2363 fixed by 4.65/MH/01 (is this a record?) are fixed:
2365 (i) An empty string was always treated as zero by the numeric comparison
2366 operators. This behaviour has been restored.
2368 (ii) It is documented that the numeric comparison operators always treat
2369 their arguments as decimal numbers. This was broken in that numbers
2370 starting with 0 were being interpreted as octal.
2372 While fixing these problems I realized that there was another issue that
2373 hadn't been noticed. Values of message_size_limit (both the global option
2374 and the transport option) were treated as octal if they started with 0.
2375 The documentation was vague. These values are now always treated as
2376 decimal, and I will make that clear in the documentation.
2382 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2383 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2386 MH/01 Don't check that the operands of numeric comparison operators are
2387 integers when their expansion is in "skipping" mode (fixes bug
2388 introduced by 4.64-PH/07).
2390 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2391 child addresses, Exim now panics and dies. Previously, because the count
2392 is held in a short int, deliveries were likely to be lost. As such a
2393 large number of recipients for a single message is ridiculous
2394 (performance will be very, very poor), I have chosen to impose a limit
2395 rather than extend the field.
2401 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2402 leftover -K file (the existence of which was triggered by #402).
2403 While we were at it, introduced process PID as part of the -K
2404 filename. This should rule out race conditions when creating
2407 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2408 processing considerably. Previous code took too long for large mails,
2409 triggering a timeout which in turn triggers #401.
2411 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2412 in the DK code in transports.c. sendfile() is not really portable,
2413 hence the _LINUX specificness.
2415 TF/01 In the add_headers option to the mail command in an Exim filter,
2416 there was a bug that Exim would claim a syntax error in any
2417 header after the first one which had an odd number of characters
2420 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2421 callout verification, Exim cached a "reject" for the entire domain. This
2422 is correct for most verifications, but it is not correct for a recipient
2423 verification with use_sender or use_postmaster set, because in that case
2424 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2425 case of MAIL FROM:<> rejection from other early rejections (e.g.
2426 rejection of HELO). When verifying a recipient using a non-null MAIL
2427 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2428 Whatever the result of the callout, the value of the domain cache is
2429 left unchanged (for any other kind of callout, getting as far as trying
2430 RCPT means that the domain itself is ok).
2432 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2435 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2436 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2437 session, the daemon ignores ECONNECT errors and logs others; it now
2438 ignores EPIPE as well.
2440 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2441 (quoted-printable decoding).
2443 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2444 later the small subsequent patch to fix an introduced bug.
2446 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2448 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2450 PH/08 An error is now given if message_size_limit is specified negative.
2452 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2453 to be given (somewhat) arbitrary names.
2455 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2458 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2459 miscellaneous code fixes
2461 PH/10 Added the log_reject_target ACL modifier to specify where to log
2464 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2465 hostname. This is wrong, because it relates to the incoming message (and
2466 probably the interface on which it is arriving) and not to the outgoing
2467 callout (which could be using a different interface). This has been
2468 changed to use the value of the helo_data option from the smtp transport
2469 instead - this is what is used when a message is actually being sent. If
2470 there is no remote transport (possible with a router that sets up host
2471 addresses), $smtp_active_hostname is used.
2473 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2474 tweaks were necessary in order to get it to work (see also 21 below):
2475 (a) The code assumed that strncpy() returns a negative number on buffer
2476 overflow, which isn't the case. Replaced with Exim's string_format()
2478 (b) There were several signed/unsigned issues. I just did the minimum
2479 hacking in of casts. There is scope for a larger refactoring.
2480 (c) The code used strcasecmp() which is not a standard C function.
2481 Replaced with Exim's strcmpic() function.
2482 (d) The code set only $1; it now sets $auth1 as well.
2483 (e) A simple test gave the error "authentication client didn't specify
2484 service in request". It would seem that Dovecot has changed its
2485 interface. Fortunately there's a specification; I followed it and
2486 changed what the client sends and it appears to be working now.
2488 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2491 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2492 address A is aliased to B and C, where B exists and C does not. Without
2493 -v the output is "A verified" because verification stops after a
2494 successful redirection if more than one address is generated. However,
2495 with -v the child addresses are also verified. Exim was outputting "A
2496 failed to verify" and then showing the successful verification for C,
2497 with its parentage. It now outputs "B failed to verify", showing B's
2498 parentage before showing the successful verification of C.
2500 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2501 look up a TXT record in a specific list after matching in a combined
2504 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2505 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2506 they consult the DNS. I had assumed they would set it the way they
2507 wanted; and indeed my experiments on Linux seem to show that in some
2508 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2509 To be on the safe side, however, I have now made the interface to
2510 host_find_byname() similar to host_find_bydns(), with an argument
2511 containing the DNS resolver options. The host_find_byname() function now
2512 sets these options at its start, just as host_find_bydns() does. The smtp
2513 transport options dns_qualify_single and dns_search_parents are passed to
2514 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2515 of host_find_byname() use the default settings of RES_DEFNAMES
2516 (qualify_single) but not RES_DNSRCH (search_parents).
2518 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2519 spool_read_header() do less string testing, by means of a preliminary
2520 switch on the second character of optional "-foo" lines. (This is
2521 overdue, caused by the large number of possibilities that now exist.
2522 Originally there were few.) While I was there, I also converted the
2523 str(n)cmp tests so they don't re-test the leading "-" and the first
2524 character, in the hope this might squeeze out yet more improvement.
2526 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2527 flag allowing group syntax was set by the header_syntax check but not
2528 turned off, possible causing trouble later; (2) The flag was not being
2529 set at all for the header_verify test, causing "group"-style headers to
2530 be rejected. I have now set it in this case, and also caused header_
2531 verify to ignore an empty address taken from a group. While doing this, I
2532 came across some other cases where the code for allowing group syntax
2533 while scanning a header line wasn't quite right (mostly, not resetting
2534 the flag correctly in the right place). These bugs could have caused
2535 trouble for malformed header lines. I hope it is now all correct.
2537 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2538 with the "reply" argument non-NULL. The code, however (which originally
2539 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2540 but it didn't always do it. This confused somebody who was copying the
2541 code for some other use. I have removed all the tests.
2543 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2544 feature that was used to support insecure browsers during the U.S. crypto
2545 embargo. It requires special client support, and Exim is probably the
2546 only MTA that supported it -- and would never use it because real RSA is
2547 always available. This code has been removed, because it had the bad
2548 effect of slowing Exim down by computing (never used) parameters for the
2549 RSA_EXPORT functionality.
2551 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2552 authenticator to fail if there's a tab character in the incoming data
2553 (there should never be unless someone is messing about, as it's supposed
2554 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2555 if the connection is using TLS or if the remote IP is the same as the
2556 local IP, and the "valid-client-cert option" if a client certificate has
2559 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2560 authenticators. This can be used for authorization after authentication
2561 succeeds. (In the case of plaintext, it servers for both authentication
2564 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2565 if any retry times were supplied.
2567 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2568 connection, where there is no client IP address to check. In this
2569 situation, the verify now always succeeds.
2571 PH/25 Applied John Jetmore's -Mset patch.
2573 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2575 PH/27 In a string expansion for a processed (not raw) header when multiple
2576 headers of the same name were present, leading whitespace was being
2577 removed from all of them, but trailing whitespace was being removed only
2578 from the last one. Now trailing whitespace is removed from each header
2579 before concatenation. Completely empty headers in a concatenation (as
2580 before) are ignored.
2582 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2583 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2585 PH/29 [Removed. This was a change that I later backed out, and forgot to
2586 correct the ChangeLog entry (that I had efficiently created) before
2587 committing the later change.]
2589 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2590 address errors (4xx response to RCPT) over the same connection as other
2591 messages routed to the same hosts. Such deliveries are always "forced",
2592 so retry times are not inspected. This resulted in far too many retries
2593 for the affected addresses. The effect occurred only when there were more
2594 hosts than the hosts_max_try setting in the smtp transport when it had
2595 the 4xx errors. Those hosts that it had tried were not added to the list
2596 of hosts for which the message was waiting, so if all were tried, there
2597 was no problem. Two fixes have been applied:
2599 (i) If there are any address or message errors in an SMTP delivery, none
2600 of the hosts (tried or untried) are now added to the list of hosts
2601 for which the message is waiting, so the message should not be a
2602 candidate for sending over the same connection that was used for a
2603 successful delivery of some other message. This seems entirely
2604 reasonable: after all the message is NOT "waiting for some host".
2605 This is so "obvious" that I'm not sure why it wasn't done
2606 previously. Hope I haven't missed anything, but it can't do any
2607 harm, as the worst effect is to miss an optimization.
2609 (ii) If, despite (i), such a delivery is accidentally attempted, the
2610 routing retry time is respected, so at least it doesn't keep
2611 hammering the server.
2613 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2614 in ${readsocket because some servers need this prod.
2616 PH/32 Added some extra debug output when updating a wait-xxx database.
2618 PH/33 The hint "could be header name not terminated by colon", which has been
2619 given for certain expansion errors for a long time, was not being given
2620 for the ${if def:h_colon_omitted{... case.
2622 PH/34 The spec says: "With one important exception, whenever a domain list is
2623 being scanned, $domain contains the subject domain." There was at least
2624 one case where this was not true.
2626 PH/35 The error "getsockname() failed: connection reset by peer" was being
2627 written to the panic log as well as the main log, but it isn't really
2628 panic-worthy as it just means the connection died rather early on. I have
2629 removed the panic log writing for the ECONNRESET error when getsockname()
2632 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2633 runs only) independently of the message's sender address. This meant
2634 that, if the 4xx error was in fact related to the sender, a different
2635 message to the same recipient with a different sender could confuse
2636 things. In particular, this can happen when sending to a greylisting
2637 server, but other circumstances could also provoke similar problems.
2638 I have changed the default so that the retry time for these errors is now
2639 based a combination of the sender and recipient addresses. This change
2640 can be overridden by setting address_retry_include_sender=false in the
2643 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2644 remote server are returned as part of bounce messages. This was not
2645 happening for LMTP over a pipe (the lmtp transport), but now it is the
2646 same for both kinds of LMTP.
2648 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2649 in header lines that were in fact CNAMEs. This is no longer the case.
2651 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2652 and queue runs started by the daemon processed all messages. This has
2653 been fixed so that -R and -S can now usefully be given with -q<time>.
2655 PH/40 Import PCRE release 6.7 (fixes some bugs).
2657 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2659 PH/42 Give an error if -q is specified more than once.
2661 PH/43 Renamed the variables $interface_address and $interface_port as
2662 $received_ip_address and $received_port, to make it clear that these
2663 values apply to message reception, and not to the outgoing interface when
2664 a message is delivered. (The old names remain recognized, of course.)
2666 PH/44 There was no timeout on the connect() call when using a Unix domain
2667 socket in the ${readsocket expansion. There now is.
2669 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2670 be meaningful with "accept".
2672 SC/01 Eximstats V1.43
2673 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2675 SC/02 Eximstats V1.44
2676 Use a glob alias rather than an array ref in the generated
2677 parser. This improves both readability and performance.
2679 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2680 Collect SpamAssassin and rejection statistics.
2681 Don't display local sender or destination tables unless
2682 there is data to show.
2683 Added average volumes into the top table text output.
2685 SC/04 Eximstats V1.46
2686 Collect data on the number of addresses (recipients)
2687 as well as the number of messages.
2689 SC/05 Eximstats V1.47
2690 Added 'Message too big' to the list of mail rejection
2691 reasons (thanks to Marco Gaiarin).
2693 SC/06 Eximstats V1.48
2694 Mainlog lines which have GMT offsets and are too short to
2695 have a flag are now skipped.
2697 SC/07 Eximstats V1.49 (Alain Williams)
2698 Added the -emptyok flag.
2700 SC/08 Eximstats V1.50
2701 Fixes for obtaining the IP address from reject messages.
2703 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2704 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2705 whitespace changes from 4.64-PH/27
2707 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2710 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2711 are found, allow negative numbers in numeric criteria)
2713 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2715 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2716 to match changes made in 4.64-PH/43
2718 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2720 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2721 "domain" is actually an IP address, and force a failure. This locks out
2722 those revolvers/nameservers that support "A-for-A" lookups, in
2723 contravention of the specifications.
2725 PH/48 When a host name was looked up from an IP address, and the subsequent
2726 forward lookup of the name timed out, the host name was left in
2727 $sender_host_name, contrary to the specification.
2729 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2730 restricted to single-key lookups, Exim was not diagnosing an error if
2731 * or *@ was used with a query-style lookup.
2733 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2735 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2736 long ago, but noone interested enough thought of it. Let's just say that
2737 the "1.1" means that there are some new functions that weren't there at
2738 some point in the past.
2740 PH/51 Error processing for expansion failure of helo_data from an smtp
2741 transport during callout processing was broken.
2743 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2744 tested/used via the -bh/-bhc/-bs options.
2746 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2747 bug, fixed in subsequent PCRE releases).
2749 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2750 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2752 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2758 SC/01 Use a glob alias rather than an array ref in eximstats generated
2759 parser. This improves both readability and performance.
2761 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2762 Don't display local sender or destination tables in eximstats unless
2763 there is data to show.
2764 Added average volumes into the eximstats top table text output.
2766 SC/03 Collect data on the number of addresses (recipients) as well
2767 as the number of messages in eximstats.
2769 TF/01 Correct an error in the documentation for the redirect router. Exim
2770 does (usually) call initgroups() when daemonizing.
2772 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2773 with consistent privilege compared to when running as a daemon.
2775 TF/03 Note in the spec that $authenticated_id is not set for local
2776 submissions from trusted users.
2778 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2779 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2781 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2782 by adding some example configuration directives to the default
2783 configuration file. A little bit of work is required to uncomment the
2784 directives and define how usernames and passwords are checked, but
2785 there is now a framework to start from.
2787 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2788 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2789 without this. I don't know how relevant this is to other LDAP libraries.
2791 PH/02 Add the verb name to the "unknown ACL verb" error.
2793 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2795 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2797 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2798 directory not expanded when it should be if an expanded home directory
2799 was set for the address (which is overridden by the transport).
2801 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2804 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2805 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2806 because it is too late at that time, and has no effect.
2808 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2809 security issue with \' (bugzilla #107). I could not use the
2810 PQescapeStringConn() function, because it needs a PGconn value as one of
2813 PH/08 When testing addresses using -bt, indicate those final addresses that
2814 are duplicates that would not cause an additional delivery. At least one
2815 person was confused, thinking that -bt output corresponded to deliveries.
2816 (Suppressing duplicates isn't a good idea as you lose the information
2817 about possibly different redirections that led to the duplicates.)
2819 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2820 systems where poll() doesn't work, in particular OS X.
2822 PH/10 Added more information to debugging output for retry time not reached.
2824 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2825 operations in malware.c.
2827 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2830 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2831 syslog_duplication set false, log lines that would normally be written
2832 both the the main log and to the reject log were not written to syslog at
2835 PH/14 In the default configuration, change the use of "message" in ACL warn
2836 statements to "add_header".
2838 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2839 not followed by a command (e.g. "seen endif").
2841 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2842 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2845 PH/17 Added extra conditions to the default value of delay_warning_condition
2849 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2850 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2851 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2854 The Auto-Submitted: and various List- headers are standardised, whereas I
2855 don't think Precedence: ever was.
2857 PH/18 Refactored debugging code in route_finduser() to show more information,
2858 in particular, the error code if getpwnam() issues one.
2860 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2861 This is apparently needed in addition to the PH/07 change above to avoid
2862 any possible encoding problems.
2864 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2865 but not after initializing Perl.
2867 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2868 output them only if debugging. By default they are written stderr,
2869 apparently, which is not desirable.
2871 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2874 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2877 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2879 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2880 authenticated or an ident call has been made. Suppress the default
2881 values for $authenticated_id and $authenticated_sender (but permit -oMai
2882 and -oMas) when testing with -bh.
2884 PH/24 Re-jigged the order of the tests in the default configuration so that the
2885 tests for valid domains and recipients precede the DNS black list and CSA
2886 tests, on the grounds that those ones are more expensive.
2888 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2889 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2890 This bug exists in every version of Exim that I still have, right back to
2893 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2894 However, an attempt to turn on case-sensitivity in a regex key by
2895 including (?-i) didn't work because the subject string was already
2896 lowercased, and the effects were non-intuitive. It turns out that a
2897 one-line patch can be used to allow (?-i) to work as expected.
2903 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2904 other effects) broke the use of negated acl sub-conditions.
2906 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2909 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2910 "Deny" causes Exim to reject the incoming connection with a 554 error.
2911 Unfortunately, if there is a major crisis, such as a disk failure,
2912 tcp-wrappers gives "deny", whereas what one would like would be some
2913 kind of temporary error. A kludge has been added to help with this.
2914 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2915 554 error is used if errno is still zero or contains ENOENT (which occurs
2916 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2919 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2921 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2922 errors. Otherwise a message that provokes a temporary error (when other
2923 messages do not) can cause a whole host to time out.
2925 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2926 addresses were routed directly to files or pipes from a redirect router.
2927 File deliveries just didn't batch; pipe deliveries might have suffered
2930 PH/06 A failure to get a lock for a hints database would erroneously always say
2931 "Failed to get write lock", even when it was really a read lock.
2933 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2934 of 0600. This has been changed to use the value of the lockfile_mode
2935 option (which defaults to 0600).
2937 PH/08 Applied small patch from the Sieve maintainer.
2939 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2940 folder from quota calculations, a direct delivery into this folder messed
2941 up the contents of the maildirsize file. This was because the regex was
2942 used only to exclude .Trash (or whatever) when the size of the mailbox
2943 was calculated. There was no check that a delivery was happening into an
2944 excluded directory. This bug has been fixed by ignoring all quota
2945 processing for deliveries into excluded directories.
2947 PH/10 Added the maildirfolder_create_regex option to appendfile.
2953 PH/01 The code for finding all the local interface addresses on a FreeBSD
2954 system running IPv6 was broken. This may well have applied to all BSD
2955 systems, as well as to others that have similar system calls. The broken
2956 code found IPv4 interfaces correctly, but gave incorrect values for the
2957 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2958 that it would not match correctly against @[] and not recognize the IPv6
2961 PH/02 The ipliteral router was not recognizing addresses of the form user@
2962 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2964 PH/03 Added disable_ipv6.
2966 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2967 decoded form, because it is most often used to construct To: headers
2968 lines in autoreplies, and the decoded form may well be syntactically
2969 invalid. However, $reply_address has leading white space removed, and all
2970 newlines turned into spaces so that the autoreply transport does not
2973 PH/05 If group was specified without a user on a router, and no group or user
2974 was specified on a transport, the group from the router was ignored.
2976 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2977 for visible compile-time settings that can be used to change these
2978 numbers, for those that want even more. Backwards compatibility with old
2979 spool files has been maintained. However, going back to a previous Exim
2980 release will lost any variables that are in spool files.
2982 PH/07 Two small changes when running in the test harness: increase delay when
2983 passing a TCP/IP connection to a new process, in case the original
2984 process has to generate a bounce, and remove special handling of
2985 127.0.0.2 (sic), which is no longer necessary.
2987 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2988 be the same on different OS.
2990 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2993 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2994 whether --show-vars was specified or not
2996 JJ/02 exipick: Added support for new ACL variable spool format introduced
2999 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3000 syntactically invalid From: or Reply-to: line, and a filter used this to
3001 generate an autoreply, and therefore failed to obtain an address for the
3002 autoreply, Exim could try to deliver to a non-existent relative file
3003 name, causing unrelated and misleading errors. What now happens is that
3004 it logs this as a hard delivery error, but does not attempt to create a
3007 PH/11 The exinext utility has a -C option for testing purposes, but although
3008 the given file was scanned by exinext itself; it wasn't being passed on
3009 when Exim was called.
3011 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3012 an end-of-file indication when reading a command response.
3014 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3015 compiled. In many other places in Exim, IPv6 addresses are always
3016 recognized, so I have changed this. It also means that IPv4 domain
3017 literals of the form [IPV4:n.n.n.n] are now always recognized.
3019 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3020 used if the router is not running as root, for example, when verifying at
3021 ACL time, or when using -bh. The debugging output from this situation was
3022 non-existent - all you got was a failure to exec. I have made two
3025 (a) Failures to set uid/gid, the current directory, or a process leader
3026 in a subprocess such as that created by queryprogram now generate
3027 suitable debugging output when -d is set.
3029 (b) The queryprogram router detects when it is not running as root,
3030 outputs suitable debugging information if -d is set, and then runs
3031 the subprocess without attempting to change uid/gid.
3033 PH/15 Minor change to Makefile for building test_host (undocumented testing
3036 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3037 additional section of a DNS packet that returns MX or SRV records.
3038 Instead, it always explicitly searches for A/AAAA records. This avoids
3039 major problems that occur when a DNS server includes only records of one
3040 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3041 fixed another bug: if SRV records were looked up and the corresponding
3042 address records were *not* found in the additional section, the port
3043 values from the SRV records were lost.
3045 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3046 using the correct key (the original address) when searching the retry
3047 rules in order to find which one to use for generating the retry hint.
3049 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3050 adding the default one. Similarly, if it contains a Reply-To: header, the
3051 errors_reply_to option, if set, is not used.
3053 PH/19 When calculating a retry time, Exim used to measure the "time since
3054 failure" by looking at the "first failed" field in the retry record. Now
3055 it does not use this if it is later than than the arrival time of the
3056 message. Instead it uses the arrival time. This makes for better
3057 behaviour in cases where some deliveries succeed, thus re-setting the
3058 "first failed" field. An example is a quota failure for a huge message
3059 when small messages continue to be delivered. Without this change, the
3060 "time since failure" will always be short, possible causing more frequent
3061 delivery attempts for the huge message than are intended.
3062 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3064 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3065 $1, $2, $3) because the numerical variables can be reset during some
3066 expansion items (e.g. "match"), thereby losing the authentication data.
3068 PH/21 Make -bV show the size of off_t variables so that the test suite can
3069 decide whether to run tests for quotas > 2G.
3071 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3072 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3073 filecount value is greater than 2G or if a quota value is greater than 2G
3074 on a system where the size of off_t is not greater than 4, a panic error
3077 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3078 never match. The debug and -bh output now contains an explicit error
3079 message indicating a malformed IPv4 address or mask.
3081 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3082 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3083 PH/23 above applies.
3085 PH/25 Do not write to syslog when running in the test harness. The only
3086 occasion when this arises is a failure to open the main or panic logs
3087 (for which there is an explicit test).
3089 PH/26 Added the /no_tell option to "control=freeze".
3091 PH/27 If a host name lookup failed very early in a connection, for example, if
3092 the IP address matched host_lookup and the reverse lookup yielded a name
3093 that did not have a forward lookup, an error message of the form "no IP
3094 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3095 could be logged. Now it outputs the IP address instead of "NULL".
3097 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3098 allows the sender and the authenticated sender to be set when
3099 submitting a message from within Exim. Since child_open_exim() is
3100 documented for local_scan(), the new function should be too.
3102 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3103 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3104 results in an empty string is now treated as unset.
3106 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3108 PH/31 Added sender_verify_fail logging option.
3110 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3111 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3112 tidied the source and removed it altogether.
3114 PH/33 When a queue run was abandoned because the load average was too high, a
3115 log line was always written; now it is written only if the queue_run log
3116 selector is set. In addition, the log line for abandonment now contains
3117 information about the queue run such as the pid. This is always present
3118 in "start" and "stop" lines but was omitted from the "abandon" line.
3120 PH/34 Omit spaces between a header name and the colon in the error message that
3121 is given when verify = headers_syntax fails (if there are lots of them,
3122 the message gets confusing).
3124 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3125 names, as there are now some PTR records that contain slashes. This check
3126 is only to protect against broken name servers that fall over on strange
3127 characters, so the fact that it applies to all lookups doesn't matter.
3129 PH/36 Now that the new test suite is complete, we can remove some of the
3130 special code in Exim that was needed for the old test suite. For example,
3131 sorting DNS records because real resolvers return them in an arbitrary
3132 order. The new test suite's fake resolver always returns records in the
3135 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3136 bounces) except when queue_only is set, to avoid logging races between
3137 the different processes.
3139 PH/38 Panic-die if .include specifies a non-absolute path.
3141 PH/39 A tweak to the "H" retry rule from its user.
3143 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3144 a label. They prevented compilation on older perls.
3146 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3147 a warning to be raised on newish perls.
3149 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3150 on queue. Changes to match documented behaviour of showing count of
3151 messages matching specified criteria.
3153 PH/40 Changed the default ident timeout from 30s to 5s.
3155 PH/41 Added support for the use of login_cap features, on those BSD systems
3156 that have them, for controlling the resources used by pipe deliveries.
3158 PH/42 The content-scanning code uses fopen() to create files in which to put
3159 message data. Previously it was not paying any attention to the mode of
3160 the files. Exim runs with umask(0) because the rest of the code creates
3161 files with open(), and sets the required mode explicitly. Thus, these
3162 files were ending up world-writeable. This was not a big issue, because,
3163 being within the spool directory, they were not world-accessible. I have
3164 created a function called modefopen, which takes an additional mode
3165 argument. It sets umask(777), creates the file, chmods it to the required
3166 mode, then resets the umask. All the relevant calls to fopen() in the
3167 content scanning code have been changed to use this function.
3169 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3170 to 24 hours. This avoids potential overflow problems when processing G
3171 and H retry rules. I suspect nobody ever tinkers with this value.
3173 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3175 PH/45 When the plaintext authenticator is running as a client, the server's
3176 challenges are checked to ensure they are valid base64 strings. By
3177 default, the authentication attempt is cancelled if an invalid string is
3178 received. Setting client_ignore_invalid_base64 true ignores these errors.
3179 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3180 they are received. Thus, the responses can be made to depend on the
3181 challenges. If an invalid string is ignored, an empty string is placed in
3184 PH/46 Messages that are created by the autoreply transport now contains a
3185 References: header, in accordance with RFCs 2822 and 3834.
3187 PH/47 Added authenticated_sender_force to the smtp transport.
3189 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3191 PH/49 Installed latest patch from the Sieve maintainer.
3193 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3194 was also set, the appendfile transport was unnecessarily scanning a
3195 directory of message files (e.g. for maildir delivery) to find the count
3196 of files (along with the size), even though it did not need this
3197 information. It now does the scan only if it needs to find either the
3198 size of the count of files.
3200 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3202 PH/52 Two bugs concerned with error handling when the smtp transport is
3205 (i) Exim was not creating retry information for temporary errors given
3206 for individual recipients after the DATA command when the smtp transport
3207 was used in LMTP mode. This meant that they could be retried too
3208 frequently, and not timed out correctly.
3210 (ii) Exim was setting the flag that allows error details to be returned
3211 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3212 recipients that were returned after the DATA command.
3214 PH/53 This is related to PH/52, but is more general: for any failing address,
3215 when detailed error information was permitted to be returned to the
3216 sender, but the error was temporary, then after the final timeout, only
3217 "retry timeout exceeded" was returned. Now it returns the full error as
3218 well as "retry timeout exceeded".
3220 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3221 do this, and (what is worse) MTAs that accept it.
3223 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3224 will now be deprecated.
3226 PH/56 New os.c-cygwin from the Cygwin maintainer.
3228 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3229 formats (previously only available in exim formats via -bpr, -bpru,
3230 and -bpra. Now also available in native and exiqgrep formats)
3232 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3233 with very large, slow to parse queues
3235 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3237 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3239 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3240 responses to authentication challenges, though it was showing the
3241 challenges; (ii) I've removed the CR characters from the debug output for
3244 PH/58 Allow for the insertion of a newline as well as a space when a string
3245 is turned into more than one encoded-word during RFC 2047 encoding. The
3246 Sieve code now uses this.
3248 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3249 data_4xx, lost_connection, tls_required.
3251 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3252 message was being sent as an SMTP response.
3254 PH/61 Add -l and -k options to exicyclog.
3256 PH/62 When verifying, if an address was redirected to one new address, so that
3257 verification continued, and the new address failed or deferred after
3258 having set something in $address_data, the value of $address_data was not
3259 passed back to the ACL. This was different to the case when no
3260 redirection occurred. The value is now passed back in both cases.
3262 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3263 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3264 use login_cap.h, so on its own it isn't the distinguishing feature. The
3265 new name refers directly to the setclassresources() function.
3267 PH/65 Added configuration files for NetBSD3.
3269 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3271 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3272 is preferred over IPv4.
3274 PH/68 The bounce_return_message and bounce_return_body options were not being
3275 honoured for bounces generated during the reception of non-SMTP messages.
3276 In particular, this applied to messages rejected by the ACL. This bug has
3277 been fixed. However, if bounce_return_message is true and bounce_return_
3278 body is false, the headers that are returned for a non-SMTP message
3279 include only those that have been read before the error was detected.
3280 (In the case of an ACL rejection, they have all been read.)
3282 PH/69 The HTML version of the specification is now built in a directory called
3283 spec_html instead of spec.html, because the latter looks like a path with
3284 a MIME-type, and this confuses some software.
3286 PH/70 Catch two compiler warnings in sieve.c.
3288 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3289 function verify_get_ident() calls ip_connect() to connect a socket, but
3290 if the "connect()" function timed out, ip_connect() used to close the
3291 socket. However, verify_get_ident() also closes the socket later, and in
3292 between Exim writes to the log, which may get opened at this point. When
3293 the socket was closed in ip_connect(), the log could get the same file
3294 descriptor number as the socket. This naturally causes chaos. The fix is
3295 not to close the socket in ip_connect(); the socket should be closed by
3296 the function that creates it. There was only one place in the code where
3297 this was missing, in the iplookup router, which I don't think anybody now
3298 uses, but I've fixed it anyway.
3300 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3301 well as to direct DNS lookups. Otherwise the handling of names in host
3302 lists is inconsistent and therefore confusing.
3308 PH/01 Two changes to the default runtime configuration:
3310 (1) Move the checks for relay_from_hosts and authenticated clients from
3311 after to before the (commented out) DNS black list checks.
3313 (2) Add control=submission to the relay_from_hosts and authenticated
3314 clients checks, on the grounds that messages accepted by these
3315 statements are most likely to be submissions.
3317 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3319 (1) Generate an error if the third argument for the ${prvs expansion is
3322 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3325 (3) Reset the variables that are obtained from the first argument of
3326 ${prvscheck and used in the second argument before leaving the code,
3327 because their memory is reclaimed, so using them afterwards may do
3330 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3331 one (it's much easier than Tom thought :-).
3333 (5) Because of (4), we can now allow for the use of $prvscheck_result
3334 inside the third argument.
3336 PH/03 For some reason, the default setting of PATH when running a command from
3337 a pipe transport was just "/usr/bin". I have changed it to
3340 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3341 anything to be listed in the output from -bV.
3343 PH/05 When a filter generated an autoreply, the entire To: header line was
3344 quoted in the delivery log line, like this:
3346 => >A.N.Other <ano@some.domain> <original@ddress> ...
3348 This has been changed so that it extracts the operative address. There
3349 may be more than one such address. If so, they are comma-separated, like
3352 => >ano@some.domain,ona@other.domain <original@ddress> ...
3354 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3355 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3356 not being looked up in the rDNS to get a host name, Exim was showing the
3357 IP address twice in Received: lines, even though the IP addresses were
3358 identical. For example:
3360 Received: from [1.2.3.4] (helo=[1.2.3.4])
3362 However, if the real host name was known, it was omitting the HELO data
3363 if it matched the actual IP address. This has been tidied up so that it
3364 doesn't show the same IP address twice.
3366 PH/07 When both +timestamp and +memory debugging was on, the value given by
3367 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3368 called by the debug printing, thereby overwriting the timestamp buffer.
3369 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3371 PH/08 When the original message was included in an autoreply transport, it
3372 always said "this is a copy of the message, including all the headers",
3373 even if body_only or headers_only was set. It now gives an appropriate
3376 PH/09 Applied a patch from the Sieve maintainer which:
3378 o fixes some comments
3379 o adds the (disabled) notify extension core
3380 o adds some debug output for the result of if/elsif tests
3381 o points to the current vacation draft in the documentation
3382 and documents the missing references header update
3386 o fixes a bug in processing the envelope test (when testing
3387 multiple envelope elements, the last element determined the
3390 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3391 Electronic Mail") by including:
3393 Auto-submitted: auto-generated
3395 in the messages that it generates (bounce messages and others, such as
3396 warnings). In the case of bounce messages for non-SMTP messages, there was
3397 also a typo: it was using "Auto_submitted" (underscore instead of
3398 hyphen). Since every message generated by Exim is necessarily in response
3399 to another message, thes have all been changed to:
3401 Auto-Submitted: auto-replied
3403 in accordance with these statements in the RFC:
3405 The auto-replied keyword:
3407 - SHOULD be used on messages sent in direct response to another
3408 message by an automatic process,
3410 - MUST NOT be used on manually-generated messages,
3412 - MAY be used on Delivery Status Notifications (DSNs) and Message
3413 Disposition Notifications (MDNs),
3415 - MUST NOT be used on messages generated by automatic or periodic
3416 processes, except for messages which are automatic responses to
3419 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3420 to the default Received: header definition.
3422 PH/12 Added log selector acl_warn_skipped (default on).
3424 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3425 variables because (a) they are in the wrong storage pool and (b) even if
3426 they were copied, it wouldn't work properly because of the caching.
3428 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3429 checking when decoding. Apparently there are clients that generate
3430 overlong encoded strings. Why am I not surprised?
3432 PH/15 If the first argument of "${if match_address" was not empty, but did not
3433 contain an "@" character, Exim crashed. Now it writes a panic log message
3434 and treats the condition as false.
3436 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3438 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3439 extension "envelope-auth". The code is finished and in agreement with
3440 other implementations, but there is no documentation so far and in fact,
3441 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3442 not changing the active code.
3444 Print executed "if" and "elsif" statements when debugging is used. This
3445 helps a great deal to understand what a filter does.
3447 Document more things not specified clearly in RFC3028. I had all this
3448 sorted out, when out of a sudden new issues came to my mind. Oops."
3450 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3453 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3454 canonical form (as documented). However, after a host name lookup from
3455 the IP address, check_host() was doing a simple string comparison with
3456 addresses acquired from the DNS when checking that the found name did
3457 have the original IP as one of its addresses. Since any found IPv6
3458 addresses are likely to be in abbreviated form, the comparison could
3459 fail. Luckily, there already exists a function for doing the comparison
3460 by converting both addresses to binary, so now that is used instead of
3461 the text comparison.
3463 PH/20 There was another similar case to PH/19, when a complete host name was
3464 given in a host list; looking up its IP address could give an abbreviated
3465 form, whereas the current host's name might or might not be abbreviated.
3466 The same fix has been applied.
3472 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3473 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3476 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3477 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3479 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3481 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3482 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3483 Also, for "Auto-Submitted", treat anything other than "no" as
3484 non-personal, in accordance with RFC 3834. (Previously it treated
3485 anything starting "auto-" as non-personal.)
3487 TF/01 The control=submission/name=... option had a problem with syntax
3488 errors if the name included a slash character. The /name= option
3489 now slurps the rest of the string, so it can include any characters
3490 but it must come last in the list of options (after /sender_retain
3493 PH/05 Some modifications to the interface to the fake nameserver for the new
3501 TK/01 Added the "success_on_redirect" address verification option. See
3502 NewStuff for rationale and an example.
3504 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3506 PH/02 Patch to exigrep to allow it to work on syslog lines.
3508 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3509 fread() to skip over the body file's header line, because in Cygwin the
3510 header line is locked and is inaccessible.
3512 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3513 co-exist for some time) to make it clear that it is the Exim ID that is
3514 referenced, not the Message-ID: header line.
3516 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3517 string_format() function, because snprintf() does not exist on all
3520 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3521 ${stat: expansion item.
3523 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3524 protocol synchronization error", to keep the pedants happy.
3526 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3527 well as for IRIX systems, when gcc is being used. See the host.c source
3530 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3532 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3535 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3536 transport and to the smtp transport in LMTP mode.
3538 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3540 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3541 run a filter in a subprocess. This could lead to confusion in subsequent
3542 lookups in the parent process. There should also be a search_tidyup() at
3543 the end of the subprocess.
3545 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3546 only if the host matched helo_try_verify_hosts, which caused the
3547 verification to occur when the EHLO/HELO command was issued. The ACL just
3548 tested the remembered result. Now, if a previous verification attempt has
3549 not happened, "verify = helo" does it there and then.
3551 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3553 TK/03 Fix log output including CR from clamd.
3555 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3556 exist provoked a memory error which could cause a segfault.
3558 PH/15 Installed PCRE 6.2
3560 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3562 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3563 of the problem. Specifically, suggested +O2 rather than +O1 for the
3566 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3568 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3569 record, the port was not correctly logged when the outgoing_port log
3570 selector was set (it logged the transort's default port).
3572 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3573 fallback_hosts, and "hosts" in the smtp transport.
3575 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3576 host errors such as "Connection refused".
3578 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3579 authentication with radiusclient 0.4.9:
3581 - Error returned from rc_read_config was caught wrongly
3582 - Username/password not passed on to radius server due to wrong length.
3584 The presumption is that some radiusclient API changes for 4.51/PH/17
3585 were not taken care of correctly. The code is still untested by me (my
3586 Linux distribution still has 0.3.2 of radiusclient), but it was
3587 contributed by a Radius user.
3589 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3590 expanding the "port" option of the smtp transport.
3592 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3593 while reading a MIME header. Thanks to Tom Hughes for a patch.
3595 PH/24 Include config.h inside local_scan.h so that configuration settings are
3598 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3599 that in an ACL for RCPT (for example), you can examine exactly what was
3602 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3603 commands, but it was not correctly comparing the address with the actual
3604 client host address. Thus, it would show the EHLO address in Received:
3605 header lines when this was not necessary.
3607 PH/27 Added the % operator to ${eval:}.
3609 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3610 it should be ignoring failures (because with -C, for example, it has lost
3611 privilege). It wasn't ignoring creation failures other than "already
3614 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3617 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3618 because an input line was too long, either on its own, or by virtue of
3619 too many continuations, the temporary file was not being removed, and the
3620 return code was incorrect.
3622 PH/31 Missing "BOOL" in function definition in filtertest.c.
3624 PH/32 Applied Sieve patches from the maintainer.
3626 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3628 PH/33 Added "verify = not_blind".
3630 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3631 Local/Makefile (with some defaults set). These are used in built scripts
3632 such as exicyclog, but they have never been used in the exim_install
3633 script (though there are many overriding facilities there). I have
3634 arranged that the exim_install script now takes note of these two
3637 PH/35 Installed configuration files for Dragonfly.
3639 PH/36 When a locally submitted message by a trusted user did not contain a
3640 From: header, and the sender address was obtained from -f or from an SMTP
3641 MAIL command, and the trusted user did not use -F to supply a sender
3642 name, $originator_name was incorrectly used when constructing a From:
3643 header. Furthermore, $originator_name was used for submission mode
3644 messages from external hosts without From: headers in a similar way,
3645 which is clearly wrong.
3647 PH/37 Added control=suppress_local_fixups.
3649 PH/38 When log_selector = +received_sender was set, and the addition of the
3650 sender made the log line's construction buffer exactly full, or one byte
3651 less than full, an overflow happened when the terminating "\n" was
3654 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3655 when the result of a list match is failure because a DNS lookup failed.
3657 PH/40 RM_COMMAND is now used in the building process.
3659 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3660 the "build-* directories that it finds.
3662 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3663 address in a domain literal was a prefix of an interface address.
3665 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3666 when verifying a sender address, unless rewrite_headers is false.
3668 PH/44 Wrote a long comment about why errors_to addresses are verified as
3669 recipients, not senders.
3671 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3672 the ratelimit ACL was added.
3674 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3676 PH/46 Added extra information about PostgreSQL errors to the error string.
3678 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3679 avoiding the need to install special zones in a real server. This is
3680 backwards compatible; if it can't find the fake resolver, it drops back.
3681 Thus, both old and new test suites can be run.
3683 TF/02 Added util/ratelimit.pl
3685 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3686 clock is set back in time.
3688 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3689 Candler <B.Candler@pobox.com>.
3691 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3692 OK for addresses that are the result of redirections.
3694 PH/48 A number of further additions for the benefit of the new test suite,
3695 including a fake gethostbyname() that interfaces to the fake DNS resolver
3698 TF/06 The fix for widen_domains has also been applied to qualify_single and
3699 search_parents which are the other dnslookup options that can cause
3702 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3705 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3707 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3708 service" if it encountered an MX record with an empty target hostname.
3709 The message is now "an MX or SRV record indicated no SMTP service".
3711 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3712 if the DNS of the sending site is misconfigured. This is quite a
3713 common situation. This change restores the behaviour of treating a
3714 helo verification defer as a failure.
3716 PH/51 If self=fail was set on a router, the bounce message did not include the
3717 actual error message.
3723 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3725 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3726 command itself ended in error, the underlying message about the transport
3727 filter timeout was being overwritten with the pipe command error. Now the
3728 underlying error message should be appended to the second error message.
3730 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3732 PH/02 Reduce the amount of output that "make" produces by default. Full output
3733 can still be requested.
3735 PH/03 The warning log line about a condition test deferring for a "warn" verb
3736 was being output only once per connection, rather than after each
3737 occurrence (because it was using the same function as for successful
3738 "warn" verbs). This seems wrong, so I have changed it.
3740 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3741 it should not have, which might have caused a crash in the right
3742 circumstances, but probably never did.
3744 PH/04 Installed a modified version of Tony Finch's patch to make submission
3745 mode fix the return path as well as the Sender: header line, and to
3746 add a /name= option so that you can make the user's friendly name appear
3749 TF/03 Added the control = fakedefer ACL modifier.
3751 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3752 Mark Lowes for thorough testing.
3754 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3756 TK/03 Merged latest SRS patch from Miles Wilton.
3758 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3759 with the definition in sysexits.h (which is #included earlier).
3760 Fortunately, Exim does not actually use EX_OK. The code used to try to
3761 preserve the sysexits.h value, by assuming that macro definitions were
3762 scanned for macro replacements. I have been disabused of this notion,
3763 so now the code just undefines EX_OK before #including unistd.h.
3765 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3766 in the smtp transport. When a block could not be written in a single
3767 write() function, the timeout was being re-applied to each part-write.
3768 This seems wrong - if the receiver was accepting one byte at a time it
3769 would take for ever. The timeout is now adjusted when this happens. It
3770 doesn't have to be particularly precise.
3772 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3773 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3775 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3776 without a domain if the check to <postmaster@domain> fails.
3778 SC/01 Eximstats: added -xls and the ability to specify output files
3779 (patch written by Frank Heydlauf).
3781 SC/02 Eximstats: use FileHandles for outputting results.
3783 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3785 SC/04 Eximstats: fixed display of large numbers with -nvr option
3787 SC/05 Eximstats: fixed merging of reports with empty tables.
3789 SC/06 Eximstats: added the -include_original_destination flag
3791 SC/07 Eximstats: removed tabs and trailing whitespace.
3793 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3795 TK/06 MBOX spool code: Add real "From " MBOX separator line
3796 so the .eml file is really in mbox format (even though
3797 most programs do not really care). Patch from Alex Miller.
3799 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3800 The latter is generated from $received_to and is only set if the
3801 message has one envelope recipient. SA can use these headers,
3802 obviously out-of-the-box. Patch from Alex Miller.
3804 PH/08 The ${def test on a variable was returning false if the variable's
3805 value was "0", contrary to what the specification has always said!
3806 The result should be true unless the variable is empty.
3808 PH/09 The syntax error of a character other than { following "${if
3809 def:variable_name" (after optional whitespace) was not being diagnosed.
3810 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3811 accidental colon was present, for example, could give incorrect results.
3813 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3814 result is used (not including appendfile, where other changes are about
3817 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3818 This involved changing a lot of size variables from int to off_t. It
3819 should work with maildirs and everything.
3821 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3822 spamd dying while we are connected to it.
3824 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3827 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3828 function for BDB 4.3.
3830 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3832 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3833 the sources for PCRE have been re-arranged and more files are now
3836 PH/15 The code I had for printing potentially long long variables in PH/11
3837 above was not the best (it lost precision). The length of off_t variables
3838 is now inspected at build time, and an appropriate printing format (%ld
3839 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3840 to be "long long int" or "long int". This is needed for the internal
3841 formatting function string_vformat().
3843 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3844 the configuration file to be ":syslog", then the script "guesses" where
3845 the logs files are, rather than using the compiled in default. In our
3846 case the guess is not the same as the compiled default, so the script
3847 suddenly stopped working when I started to use syslog. The patch checks
3848 to see if log_file_path is "". If so, it attempts to read it from exim
3849 with no configuration file to get the compiled in version, before it
3850 falls back to the previous guessing code."
3852 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3853 implementing BATV in an Exim configuration. See NewStuff for the gory
3856 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3857 Makefile that are specific to HP-UX.
3859 PH/18 If the "use_postmaster" option was set for a recipient callout together
3860 with the "random" option, the postmaster address was used as the MAIL
3861 FROM address for the random test, but not for the subsequent recipient
3862 test. It is now used for both.
3864 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3865 patch removes a few documentation additions to RFC 3028, because the
3866 latest draft now contains them. It adds the new en;ascii-case comparator
3867 and a new error check for 8bit text in MIME parts. Comparator and
3868 require names are now matched exactly. I enabled the subaddress
3869 extension, but it is not well tested yet (read: it works for me)."
3871 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3872 rework some of the code of TK/09 above to avoid the hardwired use of
3873 "%lld" and "long long". Replaced the call to snprintf() with a call to
3876 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3877 records point to non-existent hosts", "retry timeout exceeded", and
3878 "retry time not reached for any host after a long failure period".
3880 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3881 experimental DomainKeys support:
3883 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3884 (2) On an error such as an illegally used "control", the wrong name for
3885 the control was given.
3887 These problems did NOT occur unless DomainKeys support was compiled.
3889 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3891 PH/24 Added ${if match_ip condition.
3893 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3894 databases so that it will be absolutely obvious if a crash occurs in the
3895 DB library. This is a regular occurrence (often caused by mis-matched
3898 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3899 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3900 picked up on a user's system that detects such things. There doesn't seem
3901 to be a gcc warning option for this - only an attribute that has to be
3902 put on the function's prototype. It seems that in Fedora Core 4 they have
3903 set this on a number of new functions. No doubt there will be more in due
3906 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3907 specify a transport. However, if an address that was verified by such a
3908 router was the subject of a callout, Exim crashed because it tried to
3909 read the rcpt_include_affixes from the non-existent transport. Now it
3910 just assumes that the setting of that option is false. This bug was
3911 introduced by 4.51/PH/31.
3913 PH/28 Changed -d+all to exclude +memory, because that information is very
3914 rarely of interest, but it makes the output a lot bigger. People tend to
3915 do -d+all out of habit.
3917 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3918 code in os-type was giving problems when libc.so lives in lib64, like on
3921 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3922 aren't the modern standard, and it seems that some systems' include files
3923 don't always have them. Exim was already checking for some of the newer
3924 ones like T_AAAA, and defining it itself. I've added checks for all the
3925 record types that Exim uses.
3927 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3928 not automatically generating a new one, as it is supposed to. This
3929 prevented TLS from working. If the file did exist, but contained invalid
3930 data, a new version was generated, as expected. It was only the case of a
3931 non-existent file that was broken.
3933 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3934 with a change in libdomainkeys > 0.64.
3936 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3937 from DNS. If the selector record carries the flag, it now has
3938 precedence over the domain-wide flag.
3940 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3942 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3943 the use of a port name that isn't defined in /etc/services) occurred, the
3944 message was deferred as in a normal delivery, and thus remained on the
3945 spool, instead of being failed because of the mua_wrapper setting. This
3946 is now fixed, and I tidied up some of the mua_wrapper messages at the
3949 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3950 the messages in a hash of arrays rather than using individual hashes.
3951 This is a bit cleaner and results in dramatic memory savings, albeit
3952 at a slight CPU cost.
3954 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3955 as requested by Marc Sherman.
3957 SC/10 Eximstats: added histograms for user specified patterns as requested
3960 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3962 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3963 fopen() in the content-scanning modules that did not already have it.
3969 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3970 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3972 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3974 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3976 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3977 to test Sieve filters that use "vacation".
3979 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3980 that changes the way the GnuTLS parameters are stored in the cache file.
3981 The new format can be generated externally. For backward compatibility,
3982 if the data in the cache doesn't make sense, Exim assumes it has read an
3983 old-format file, and it generates new data and writes a new file. This
3984 means that you can't go back to an older release without removing the
3987 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3988 work if there are any delivery delays because "one_time" forces the
3989 parent to be marked "delivered", so its unseen clone is never tried
3990 again. For this reason, Exim now forbids the simultaneous setting of
3993 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3994 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3995 entry if you want to know the details. The fix, however, neglected to
3996 consider the case when local delivery batching is involved. The test for
3997 "previously delivered" was not happening when checking to see if an
3998 address could be batched with a previous (undelivered) one; under
3999 certain circumstances this could lead to multiple deliveries to the same
4002 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4003 in its include files, and this causes problems building Exim.
4005 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4006 header_syntax) but Exim was just ignoring anything given after a slash.
4007 In particular, this caused confusion with an attempt to use "verify =
4008 reverse_host_lookup/defer_ok". An error is now given when options are
4009 supplied for verify items that do not have them. (Maybe reverse_host_
4010 lookup should have a defer_ok option, but that's a different point.)
4012 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4013 defined by RFC 821) to 2048, because there were problems with some AUTH
4014 commands, and RFC 1869 says the size should be increased for extended
4015 SMTP commands that take arguments.
4017 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4020 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4021 "unknown" error; now it says that the functionality isn't in the binary.
4023 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4024 an address' error message when a string expansion fails (syntax or
4025 whatever). Otherwise the password may appear in the log. Following change
4026 PH/42 below, there is no longer a chance of it appearing in a bounce
4029 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4031 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4032 of its addresses was ever tried. (Bugzilla bug #2.)
4034 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4035 the result incorrectly in the debug output. (It correctly added a newline
4036 to what was transported.)
4038 TF/01 Added $received_time.
4040 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4041 commented out examples of how to interface to a virus scanner and to
4042 SpamAssassin. Also added commented examples of av_scanner and
4043 spamd_address settings.
4045 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4046 and controls are allowed in which ACLs. There were a couple of minor
4047 errors. Some of the entries in the conditions table (which is a table of
4048 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4049 negation of where the condition IS allowed.
4051 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4053 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4054 header file does not have a version number, so I've had to invent a new
4055 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4056 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4057 radiusclient), but it was contributed by a Radius user.
4059 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4060 files or directories, for OpenSSL.
4062 PH/19 When an Exim process that is running as root has to create an Exim log
4063 file, it does so in a subprocess that runs as exim:exim so as to get the
4064 ownership right at creation (otherwise, other Exim processes might see
4065 the file with the wrong ownership). There was no test for failure of this
4066 fork() call, which would lead to the process getting stuck as it waited
4067 for a non-existent subprocess. Forks do occasionally fail when resources
4068 run out. I reviewed all the other calls to fork(); they all seem to check
4071 PH/20 When checking for unexpected SMTP input at connect time (before writing
4072 the banner), Exim was not dealing correctly with a non-positive return
4073 from the read() function. If the client had disconnected by this time,
4074 the result was a log entry for a synchronization error with an empty
4075 string after "input=" when read() returned zero. If read() returned -1
4076 (an event I could not check), uninitialized data bytes were printed.
4077 There were reports of junk text (parts of files, etc) appearing after
4080 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4082 PH/22 Added support for macro redefinition, and (re)definition in between
4083 driver and ACL definitions.
4085 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4086 forgetting to use the resulting value; it was using the unexpanded value.
4088 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4089 hadn't been configured. The fix is from Juergen Kreileder, who
4090 understands it better than I do:
4092 "Here's what I see happening with three configured cyrus_sasl
4093 authenticators configured (plain, login, cram-md5):
4095 On startup auth_cyrus_sasl_init() gets called for each of these.
4096 This means three calls to sasl_listmech() without a specified mech_list.
4097 => SASL tests which mechs of all available mechs actually work
4098 => three warnings about OTP not working
4099 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4101 With the patch, sasl_listmech() also gets called three times. But now
4102 SASL's mech_list option is set to the server_mech specified in the the
4103 authenticator. Or in other words, the answer from sasl_listmech()
4104 gets limited to just the mech you're testing for (which is different
4106 => the return list contains just 'plain' or 'login', 'cram-md5' or
4107 nothing depending on the value of ob->server_mech.
4109 I've just tested the patch: Authentication still works fine,
4110 unavailable mechs specified in the exim configuration are still
4111 caught, and the auth.log warnings about OTP are gone."
4113 PH/25 When debugging is enabled, the contents of the command line are added
4114 to the debugging output, even when log_selector=+arguments is not
4117 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4118 answer is "GNU", and only if the return is "GNU/something" is the answer
4121 PH/27 $acl_verify_message is now set immediately after the failure of a
4122 verification in an ACL, and so is available in subsequent modifiers. In
4123 particular, the message can be preserved by coding like this:
4125 warn !verify = sender
4126 set acl_m0 = $acl_verify_message
4128 Previously, $acl_verify_message was set only while expanding "message"
4129 and "log_message" when a very denied access.
4131 PH/28 Modified OS/os.c-Linux with
4133 -#ifndef OS_LOAD_AVERAGE
4134 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4136 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4137 nomenclature these days.)
4139 PH/29 Installed patch from the Sieve maintainer that adds the options
4140 sieve_useraddress and sieve_subaddress to the redirect router.
4142 PH/30 In these circumstances:
4143 . Two addresses routed to the same list of hosts;
4144 . First host does not offer TLS;
4145 . First host accepts first address;
4146 . First host gives temporary error to second address;
4147 . Second host offers TLS and a TLS session is established;
4148 . Second host accepts second address.
4149 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4150 and peerdn, if requested) that were in fact used only for the second
4153 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4154 attention to any local part prefix or suffix that was matched by the
4155 router that accepted the address. It now behaves in the same way as it
4156 does for delivery: the affixes are removed from the local part unless
4157 rcpt_include_affixes is set on the transport.
4159 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4160 timeout during the DATA phase of an incoming message.
4162 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4163 applied a patch sanctioned by the Sieve maintainer.
4165 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4166 the uid or gid is negative. A case of a negative gid caused this to be
4167 noticed. The fix allows for either to be negative.
4169 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4170 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4173 PH/36 The change PH/12 above was broken. Fixed it.
4175 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4176 the grounds that routing the same address twice would always produce the
4177 same answer. This might have been true once, but it is certainly no
4178 longer true now. Routing a child address may depend on the previous
4179 routing that produced that child. Some complicated redirection strategies
4180 went wrong when messages had multiple recipients, and made Exim's
4181 behaviour dependent on the order in which the addresses were given.
4183 I have moved the duplicate checking until after the routing is complete.
4184 Exim scans the addresses that are assigned to local and remote
4185 transports, and removes any duplicates. This means that more work will be
4186 done, as duplicates will always all be routed, but duplicates are
4187 presumably rare, so I don't expect this is of any significance.
4189 For deliveries to pipes, files, and autoreplies, the duplicate checking
4190 still happens during the routing process, since they are not going to be
4193 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4194 It corrects a timeout issue with spamd. This is Ian's comment: "The
4195 background is that sometimes spamd either never reads data from a
4196 connection it has accepted, or it never writes response data. The exiscan
4197 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4198 blindly assumes that writes won't block so it may never time out."
4200 PH/39 Allow G after quota size as well as K and M.
4202 PH/40 The value set for $authenticated_id in an authenticator may not contain
4203 binary zeroes or newlines because the value is written to log lines and
4204 to spool files. There was no check on this. Now the value is run through
4205 the string_printing() function so that such characters are converted to
4206 printable escape sequences.
4208 PH/41 $message_linecount is a new variable that contains the total number of
4209 lines in the message. Compare $body_linecount, which is the count for the
4212 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4213 bounce and delay warning messages, except in certain special cases, which
4216 (a) An SMTP error message from a remote host;
4217 (b) A message specified in a :fail: redirection;
4218 (c) A message specified in a "fail" command in a system filter;
4219 (d) A message specified in a FAIL return from the queryprogram router;
4220 (e) A message specified by the cannot_route_message router option.
4222 In these cases only, Exim does include the error details in bounce and
4223 warning messages. There are also a few cases where bland messages such
4224 as "unrouteable address" or "local delivery error" are given.
4226 PH/43 $value is now also set for the "else" part of a ${run expansion.
4228 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4229 being worked on, but at least Exim now implements the latest version to
4232 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4233 process to complete was treated as a delivery failure, a timeout while
4234 writing the message to the pipe was logged, but erroneously treated as a
4235 successful delivery. Such timeouts include transport filter timeouts. For
4236 consistency with the overall process timeout, these timeouts are now
4237 treated as errors, giving rise to delivery failures by default. However,
4238 there is now a new Boolean option for the pipe transport called
4239 timeout_defer, which, if set TRUE, converts the failures into defers for
4240 both kinds of timeout. A transport filter timeout is now identified in
4243 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4244 systems where "make" and "gmake" are different, calling "gmake" at top
4245 level broke things. I've arranged for the value of $(MAKE) to be passed
4246 from the Makefile to this script so that it can call the same version of
4250 A note about Exim versions 4.44 and 4.50
4251 ----------------------------------------
4253 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4254 changes of various kinds. As a consequence, a big documentation update was
4255 needed. This delayed the release for rather longer than seemed good, especially
4256 in the light of a couple of (minor) security issues. Therefore, the changes
4257 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4258 release. So 4.44 and 4.50 are in effect two different branches that both start
4261 I have left the 4.50 change log unchanged; it contains all the changes since
4262 4.43. The change log for 4.44 is below; many of its items are identical to
4263 those for 4.50. This seems to be the most sensible way to preserve the
4264 historical information.
4270 1. Minor wording change to the doc/README.SIEVE file.
4272 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4273 computation of the current number of files was incorrect.
4275 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4276 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4279 4. Give more explanation in the error message when the command for a transport
4280 filter fails to execute.
4282 5. There are several places where Exim runs a non-Exim command in a
4283 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4284 was being done only for the command run by the queryprogram router. It is
4285 now done for all such subprocesses. The other cases are: ${run, transport
4286 filters, and the commands run by the lmtp and pipe transports.
4288 6. Added CONFIGURE_GROUP build-time option.
4290 7. Some older OS have a limit of 256 on the maximum number of file
4291 descriptors. Exim was using setrlimit() to set 1000 as a large value
4292 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4293 systems. I've change it so that if it can't get 1000, it tries for 256.
4295 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4296 was an oversight, and furthermore, ever since the addition of extra
4297 controls (e.g. 4.43/32), the checks on when to allow different forms of
4298 "control" were broken. There should now be diagnostics for all cases when a
4299 control that does not make sense is encountered.
4301 9. Added the /retain_sender option to "control=submission".
4303 10. $recipients is now available in the predata ACL (oversight).
4305 11. Tidy the search cache before the fork to do a delivery from a message
4306 received from the command line. Otherwise the child will trigger a lookup
4307 failure and thereby defer the delivery if it tries to use (for example) a
4308 cached ldap connection that the parent has called unbind on.
4310 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4311 of $address_data from the recipient verification was clobbered by the
4312 sender verification.
4314 13. The value of address_data from a sender verification is now available in
4315 $sender_address_data in subsequent conditions in the ACL statement.
4317 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4319 15. Added a new option "connect=<time>" to callout options, to set a different
4322 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4323 was its contents. (It was OK if the option was not defined at all.)
4325 17. A "Completed" log line is now written for messages that are removed from
4326 the spool by the -Mrm option.
4328 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4329 information about exactly what failed.
4331 19. Added -dd to debug only the daemon process.
4333 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4334 handles timeouts, both on the server side and network timeouts. Renamed the
4335 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4337 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4338 It is now set to "smtps".
4340 22. $host_address is now set to the target address during the checking of
4341 ignore_target_hosts.
4343 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4344 being passed; this would have caused $sender_host_name to have been used if
4345 matching the list had actually called for a host name (not very likely,
4346 since this list is usually IP addresses). A host name is now passed as
4349 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4350 code that specifies a non-restarting handler (typically sigaction() in
4351 modern systems) in an attempt to fix a rare and obscure crash bug.
4353 25. Narrowed the window for a race in the daemon that could cause it to ignore
4354 SIGCHLD signals. This is not a major problem, because they are used only to
4355 wake it up if nothing else does.
4357 26. A malformed maildirsize file could cause Exim to calculate negative values
4358 for the mailbox size or file count. Odd effects could occur as a result.
4359 The maildirsize information is now recalculated if the size or filecount
4362 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4363 support for a long time. Removed HAVE_SYS_VFS_H.
4365 28. Installed the latest version of exipick from John Jetmore.
4367 29. In an address list, if the pattern was not a regular expression, an empty
4368 subject address (from a bounce message) matched only if the pattern was an
4369 empty string. Non-empty patterns were not even tested. This was the wrong
4370 because it is perfectly reasonable to use an empty address as part of a
4371 database query. An empty address is now tested by patterns that are
4372 lookups. However, all the other forms of pattern expect the subject to
4373 contain a local part and a domain, and therefore, for them, an empty
4374 address still always fails if the pattern is not itself empty.
4376 30. Exim went into a mad DNS loop when attempting to do a callout where the
4377 host was specified on an smtp transport, and looking it up yielded more
4378 than one IP address.
4380 31. Re-factored the code for checking spool and log partition space into a
4381 function that finds that data and another that does the check. The former
4382 is then used to implement four new variables: $spool_space, $log_space,
4383 $spool_inodes, and $log_inodes.
4385 32. The RFC2047 encoding function was originally intended for short strings
4386 such as real names; it was not keeping to the 75-character limit for
4387 encoded words that the RFC imposes. It now respects the limit, and
4388 generates multiple encoded words if necessary. To be on the safe side, I
4389 have increased the buffer size for the ${rfc2047: expansion operator from
4392 33. It is now permitted to omit both strings after an "if" condition; if the
4393 condition is true, the result is "true". As before, when the second string
4394 is omitted, a false condition yields an empty string. This makes it less
4395 cumbersome to write custom ACL and router conditions.
4397 34. Failure to deliver a bounce message always caused it to be frozen, even if
4398 there was an errors_to setting on the router. The errors_to setting is now
4401 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4402 canonical form (fully expanded) before being placed in
4403 $sender_host_address.
4405 36. The table in the code that translates DNS record types into text (T_A to
4406 "A" for instance) was missing entries for NS and CNAME. It is just possible
4407 that this could have caused confusion if both these types were looked up
4408 for the same domain, because the text type is used as part of Exim's
4409 per-process caching. But the chance of anyone hitting this buglet seems
4412 37. The dnsdb lookup has been extended in a number of ways.
4414 (1) There is a new type, "zns", which walks up the domain tree until it
4415 finds some nameserver records. It should be used with care.
4417 (2) There is a new type, "mxh", which is like "mx" except that it returns
4418 just the host names, not the priorities.
4420 (3) It is now possible to give a list of domains (or IP addresses) to be
4421 looked up. The behaviour when one of the lookups defers can be
4422 controlled by a keyword.
4424 (4) It is now possible to specify the separator character for use when
4425 multiple records are returned.
4427 38. The dnslists ACL condition has been extended: it is now possible to supply
4428 a list of IP addresses and/or domains to be looked up in a particular DNS
4431 39. Added log_selector=+queue_time_overall.
4433 40. When running the queue in the test harness, wait just a tad after forking a
4434 delivery process, to get repeatability of debugging output.
4436 41. Include certificate and key file names in error message when GnuTLS fails
4437 to set them up, because the GnuTLS error message doesn't include the name
4438 of the failing file when there is a problem reading it.
4440 42. Allow both -bf and -bF in the same test run.
4442 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4444 44. The "Exiscan patch" is now merged into the mainline Exim source.
4446 45. Sometimes the final signoff response after QUIT could fail to get
4447 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4448 < 0 before doing a fflush(). This bug looks as though it goes back to the
4449 introduction of TLS in release 3.20, but "sometimes" must have been rare
4450 because the tests only now provoked it.
4452 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4453 (this can affect the format of dates).
4455 47. exim_tidydb, when checking for the continued existence of a message for
4456 which it has found a message-specific retry record, was not finding
4457 messages that were in split spool directories. Consequently, it was
4458 deleting retry records that should have stayed in existence.
4460 48. Steve fixed some bugs in eximstats.
4462 49. The SPA authentication driver was not abandoning authentication and moving
4463 on to the next authenticator when an expansion was forced to fail,
4464 contradicting the general specification for all authenticators. Instead it
4465 was generating a temporary error. It now behaves as specified.
4467 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4468 (the order specifies the preference for clients). The order is now AES256,
4469 AES128, 3DES, ARCFOUR128.
4471 51. Small patch to Sieve code - explicitly set From: when generating an
4474 52. Exim crashed if a remote delivery caused a very long error message to be
4475 recorded - for instance if somebody sent an entire SpamAssassin report back
4476 as a large number of 550 error lines. This bug was coincidentally fixed by
4477 increasing the size of one of Exim's internal buffers (big_buffer) that
4478 happened as part of the Exiscan merge. However, to be on the safe side, I
4479 have made the code more robust (and fixed the comments that describe what
4482 53. Now that there can be additional text after "Completed" in log lines (if
4483 the queue_time_overall log selector is set), a one-byte patch to exigrep
4484 was needed to allow it to recognize "Completed" as not the last thing in
4487 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4488 patch that reportedly fixes this has been added. I am not expert enough to
4489 create a test for it. This is what the patch creator wrote:
4491 "I found a little strange behaviour of ldap code when working with
4492 Windows 2003 AD Domain, where users was placed in more than one
4493 Organization Units. When I tried to give exim partial DN, the exit code
4494 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4495 But simultaneously result of request was absolutely normal ldap result,
4496 so I produce this patch..."
4498 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4499 REFERENCE, so I have modified the code to exclude the patch when that macro
4502 55. Some experimental protocols are using DNS PTR records for new purposes. The
4503 keys for these records are domain names, not reversed IP addresses. The
4504 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4505 leaves it alone. Component reversal etc. now happens only for IP addresses.
4508 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4510 57. Double the size of the debug message buffer (to 2048) so that more of very
4511 long debug lines gets shown.
4513 58. The exicyclog utility now does better if the number of log files to keep
4514 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4516 59. Two changes related to the smtp_active_hostname option:
4518 (1) $smtp_active_hostname is now available as a variable.
4519 (2) The default for smtp_banner uses $smtp_active_hostname instead
4520 of $primary_hostname.
4522 60. The host_aton() function is supposed to be passed a string that is known
4523 to be a valid IP address. However, in the case of IPv6 addresses, it was
4524 not checking this. This is a hostage to fortune. Exim now panics and dies
4525 if the condition is not met. A case was found where this could be provoked
4526 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4527 components; fortuitously, this particular loophole had already been fixed
4528 by change 4.50/55 above.
4530 If there are any other similar loopholes, the new check in host_aton()
4531 itself should stop them being exploited. The report I received stated that
4532 data on the command line could provoke the exploit when Exim was running as
4533 exim, but did not say which command line option was involved. All I could
4534 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4535 running as the user.
4538 61. There was a buffer overflow vulnerability in the SPA authentication code
4539 (which came originally from the Samba project). I have added a test to the
4540 spa_base64_to_bits() function which I hope fixes it.
4543 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4544 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4546 63. The daemon start-up calls getloadavg() while still root for those OS that
4547 need the first call to be done as root, but it missed one case: when
4548 deliver_queue_load_max is set with deliver_drop_privilege. This is
4549 necessary for the benefit of the queue runner, because there is no re-exec
4550 when deliver_drop_privilege is set.
4552 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4553 This has been fixed.
4555 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4556 was in use, was not putting the data itself into the right store pool;
4557 consequently, it could be overwritten for a subsequent message in the same
4558 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4561 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4563 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4564 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4565 the function were treating the return as a boolean value, which happened to
4566 work because 0=false and not-0=true, but is not correct code.
4568 68. The host_aton() function was not handling scoped IPv6 addresses (those
4569 with, for example, "%eth0" on the end) correctly.
4571 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4572 negated items (that is, ~something) in unsigned ints. Some compilers
4573 apparently mutter when there is no cast.
4575 70. If an address verification called from an ACL failed, and did not produce a
4576 user-specific message (i.e. there was only a "system" message), nothing was
4577 put in $acl_verify_message. In this situation, it now puts the system
4580 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4581 session; change 4.31/43 added the unwanted input to the log line - except
4582 that it did not do this in the start of session case. It now does.
4584 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4585 This is wrong and can cause the other end to generate a synchronization
4586 error if it is another Exim or anything else that does the synchronization
4587 check. A QUIT command is no longer sent after a timeout.
4589 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4590 during host lookups.
4592 74. The defer_ok option of callout verification was not working if it was used
4593 when verifying addresses in header lines, that is, for this case:
4595 verify = header_sender/callout=defer_ok
4597 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4598 those file descriptors could be used for SMTP connections. If anything
4599 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4600 could be sent to the SMTP client, causing chaos. The daemon now opens
4601 stdin, stdout, and stderr to /dev/null when it puts itself into the
4604 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4605 log by default. The user can override this with suitable Perl magic.
4607 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4608 the log message when discard triggers, was not working for the DATA ACL or
4609 for the non-SMTP ACL.
4611 78. Error message wording change in sieve.c.
4613 79. If smtp_accept_max_per_host was set, the number of connections could be
4614 restricted to fewer than expected, because the daemon was trying to set up
4615 a new connection before checking whether the processes handling previous
4616 connections had finished. The check for completed processes is now done
4617 earlier. On busy systems, this bug wouldn't be noticed because something
4618 else would have woken the daemon, and it would have reaped the completed
4621 80. If a message was submitted locally by a user whose login name contained one
4622 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4623 It caused a spool format error. I have fixed the spool reading code. A
4624 related problem was that the "from" clause in the Received: line became
4625 illegal because of the space(s). It is now covered by ${quote_local_part.
4627 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4630 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4632 83. Patch from Sieve maintainer to fix unterminated string problem in
4633 "vacation" handling.
4635 84. Some minor changes to the Linux configuration files to help with other
4636 OS variants using glibc.
4638 85. One more patch for Sieve to update vacation handling to latest spec.
4641 ----------------------------------------------------
4642 See the note above about the 4.44 and 4.50 releases.
4643 ----------------------------------------------------
4649 1. Change 4.43/35 introduced a bug that caused file counts to be
4650 incorrectly computed when quota_filecount was set in an appendfile
4653 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4654 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4657 3. Give more explanation in the error message when the command for a transport
4658 filter fails to execute.
4660 4. There are several places where Exim runs a non-Exim command in a
4661 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4662 was being done only for the command run by the queryprogram router. It is
4663 now done for all such subprocesses. The other cases are: ${run, transport
4664 filters, and the commands run by the lmtp and pipe transports.
4666 5. Some older OS have a limit of 256 on the maximum number of file
4667 descriptors. Exim was using setrlimit() to set 1000 as a large value
4668 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4669 systems. I've change it so that if it can't get 1000, it tries for 256.
4671 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4672 was an oversight, and furthermore, ever since the addition of extra
4673 controls (e.g. 4.43/32), the checks on when to allow different forms of
4674 "control" were broken. There should now be diagnostics for all cases when a
4675 control that does not make sense is encountered.
4677 7. $recipients is now available in the predata ACL (oversight).
4679 8. Tidy the search cache before the fork to do a delivery from a message
4680 received from the command line. Otherwise the child will trigger a lookup
4681 failure and thereby defer the delivery if it tries to use (for example) a
4682 cached ldap connection that the parent has called unbind on.
4684 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4685 of $address_data from the recipient verification was clobbered by the
4686 sender verification.
4688 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4689 was its contents. (It was OK if the option was not defined at all.)
4691 11. A "Completed" log line is now written for messages that are removed from
4692 the spool by the -Mrm option.
4694 12. $host_address is now set to the target address during the checking of
4695 ignore_target_hosts.
4697 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4698 being passed; this would have caused $sender_host_name to have been used if
4699 matching the list had actually called for a host name (not very likely,
4700 since this list is usually IP addresses). A host name is now passed as
4703 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4704 code that specifies a non-restarting handler (typically sigaction() in
4705 modern systems) in an attempt to fix a rare and obscure crash bug.
4707 15. Narrowed the window for a race in the daemon that could cause it to ignore
4708 SIGCHLD signals. This is not a major problem, because they are used only to
4709 wake it up if nothing else does.
4711 16. A malformed maildirsize file could cause Exim to calculate negative values
4712 for the mailbox size or file count. Odd effects could occur as a result.
4713 The maildirsize information is now recalculated if the size or filecount
4716 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4717 support for a long time. Removed HAVE_SYS_VFS_H.
4719 18. Updated exipick to current release from John Jetmore.
4721 19. Allow an empty sender to be matched against a lookup in an address list.
4722 Previously the only cases considered were a regular expression, or an
4725 20. Exim went into a mad DNS lookup loop when doing a callout where the
4726 host was specified on the transport, if the DNS lookup yielded more than
4729 21. The RFC2047 encoding function was originally intended for short strings
4730 such as real names; it was not keeping to the 75-character limit for
4731 encoded words that the RFC imposes. It now respects the limit, and
4732 generates multiple encoded words if necessary. To be on the safe side, I
4733 have increased the buffer size for the ${rfc2047: expansion operator from
4736 22. Failure to deliver a bounce message always caused it to be frozen, even if
4737 there was an errors_to setting on the router. The errors_to setting is now
4740 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4741 canonical form (fully expanded) before being placed in
4742 $sender_host_address.
4744 24. Updated eximstats to version 1.33
4746 25. Include certificate and key file names in error message when GnuTLS fails
4747 to set them up, because the GnuTLS error message doesn't include the name
4748 of the failing file when there is a problem reading it.
4750 26. Expand error message when OpenSSL has problems setting up cert/key files.
4753 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4754 (this can affect the format of dates).
4756 28. exim_tidydb, when checking for the continued existence of a message for
4757 which it has found a message-specific retry record, was not finding
4758 messages that were in split spool directories. Consequently, it was
4759 deleting retry records that should have stayed in existence.
4761 29. eximstats updated to version 1.35
4762 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4763 1.35 - bugfix such that pie charts by volume are generated correctly
4765 30. The SPA authentication driver was not abandoning authentication and moving
4766 on to the next authenticator when an expansion was forced to fail,
4767 contradicting the general specification for all authenticators. Instead it
4768 was generating a temporary error. It now behaves as specified.
4770 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4771 (the order specifies the preference for clients). The order is now AES256,
4772 AES128, 3DES, ARCFOUR128.
4774 31. Small patch to Sieve code - explicitly set From: when generating an
4777 32. Exim crashed if a remote delivery caused a very long error message to be
4778 recorded - for instance if somebody sent an entire SpamAssassin report back
4779 as a large number of 550 error lines. This bug was coincidentally fixed by
4780 increasing the size of one of Exim's internal buffers (big_buffer) that
4781 happened as part of the Exiscan merge. However, to be on the safe side, I
4782 have made the code more robust (and fixed the comments that describe what
4785 33. Some experimental protocols are using DNS PTR records for new purposes. The
4786 keys for these records are domain names, not reversed IP addresses. The
4787 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4788 leaves it alone. Component reversal etc. now happens only for IP addresses.
4791 34. The host_aton() function is supposed to be passed a string that is known
4792 to be a valid IP address. However, in the case of IPv6 addresses, it was
4793 not checking this. This is a hostage to fortune. Exim now panics and dies
4794 if the condition is not met. A case was found where this could be provoked
4795 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4796 components; fortuitously, this particular loophole had already been fixed
4797 by change 4.50/55 or 4.44/33 above.
4799 If there are any other similar loopholes, the new check in host_aton()
4800 itself should stop them being exploited. The report I received stated that
4801 data on the command line could provoke the exploit when Exim was running as
4802 exim, but did not say which command line option was involved. All I could
4803 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4804 running as the user.
4807 35. There was a buffer overflow vulnerability in the SPA authentication code
4808 (which came originally from the Samba project). I have added a test to the
4809 spa_base64_to_bits() function which I hope fixes it.
4812 36. The daemon start-up calls getloadavg() while still root for those OS that
4813 need the first call to be done as root, but it missed one case: when
4814 deliver_queue_load_max is set with deliver_drop_privilege. This is
4815 necessary for the benefit of the queue runner, because there is no re-exec
4816 when deliver_drop_privilege is set.
4818 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4819 was in use, was not putting the data itself into the right store pool;
4820 consequently, it could be overwritten for a subsequent message in the same
4821 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4824 38. Sometimes the final signoff response after QUIT could fail to get
4825 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4826 < 0 before doing a fflush(). This bug looks as though it goes back to the
4827 introduction of TLS in release 3.20, but "sometimes" must have been rare
4828 because the tests only now provoked it.
4834 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4835 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4836 Now it returns an integer. A number of calls were still expecting a T/F
4837 return. Fortuitously, in all cases, the tests worked in OK situations,
4838 which is the norm. However, things would have gone wrong on any write
4839 failures on the smtp file descriptor. This function is used when sending
4840 messages over SMTP and also when doing verify callouts.
4842 2. When Exim is called to do synchronous delivery of a locally submitted
4843 message (the -odf or -odi options), it no longer closes stderr before doing
4846 3. Implemented the mua_wrapper option.
4848 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4850 5. Implemented the functions header_remove(), header_testname(),
4851 header_add_at_position(), and receive_remove_recipient(), and exported them
4854 6. If an ACL "warn" statement specified the addition of headers, Exim already
4855 inserted X-ACL-Warn: at the start if there was no header name. However, it
4856 was not making this test for the second and subsequent header lines if
4857 there were newlines in the string. This meant that an invalid header could
4858 be inserted if Exim was badly configured.
4860 7. Allow an ACL "warn" statement to add header lines at the start or after all
4861 the Received: headers, as well as at the end.
4863 8. Added the rcpt_4xx retry error code.
4865 9. Added postmaster_mailfrom=xxx to callout verification option.
4867 10. Added mailfrom=xxxx to the callout verification option, for verify=
4870 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4871 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4873 12. Inserted some casts to stop certain compilers warning when using pointer
4874 differences as field lengths or precisions in printf-type calls (mostly
4875 affecting debugging statements).
4877 13. Added optional readline() support for -be (dynamically loaded).
4879 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4880 same clock tick as a message's arrival, so that its received time was the
4881 same as the "first fail" time on the retry record, and that message
4882 remained on the queue past the ultimate address timeout, every queue runner
4883 would try a delivery (because it was past the ultimate address timeout) but
4884 after another failure, the ultimate address timeout, which should have then
4885 bounced the address, did not kick in. This was a "< instead of <=" error;
4886 in most cases the first failure would have been in the next clock tick
4887 after the received time, and all would be well.
4889 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4890 being recognized when the domain list was tested by the match_domain
4891 condition in an expansion string.
4893 16. Added the ${str2b64: operator.
4895 17. Exim was always calling setrlimit() to set a large limit for the number of
4896 processes, without checking whether the existing limit was already
4897 adequate. (It did check for the limit on file descriptors.) Furthermore,
4898 errors from getrlimit() and setrlimit() were being ignored. Now they are
4899 logged to the main and panic logs, but Exim does carry on, to try to do its
4900 job under whatever limits there are.
4902 18. Imported PCRE 5.0.
4904 19. Trivial typo in log message " temporarily refused connection" (the leading
4907 20. If the log selector return_path_on_delivery was set and an address was
4908 redirected to /dev/null, the delivery process crashed because it assumed
4909 that a return path would always be set for a "successful" delivery. In this
4910 case, the whole delivery is bypassed as an optimization, and therefore no
4913 21. Internal re-arrangement: the function for sending a challenge and reading
4914 a response while authentication was assuming a zero-terminated challenge
4915 string. It's now changed to take a pointer and a length, to allow for
4916 binary data in such strings.
4918 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4920 23. Exim was not respecting finduser_retries when seeking the login of the
4921 uid under which it was called; it was always trying 10 times. (The default
4922 setting of finduser_retries is zero.) Also, it was sleeping after the final
4923 failure, which is pointless.
4925 24. Implemented tls_on_connect_ports.
4927 25. Implemented acl_smtp_predata.
4929 26. If the domain in control=submission is set empty, Exim assumes that the
4930 authenticated id is a complete email address when it generates From: or
4931 Sender: header lines.
4933 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4934 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4935 chown and chgrp in /bin and hostname in /usr/bin.
4937 28. Exim was keeping the "process log" file open after each use, just as it
4938 does for the main log. This opens the possibility of it remaining open for
4939 long periods when the USR1 signal hits a daemon. Occasional processlog
4940 errors were reported, that could have been caused by this. Anyway, it seems
4941 much more sensible not to leave this file open at all, so that is what now
4944 29. The long-running daemon process does not normally write to the log once it
4945 has entered its main loop, and it closes the log before doing so. This is
4946 so that log files can straightforwardly be renamed and moved. However,
4947 there are a couple of unusual error situations where the daemon does write
4948 log entries, and I had neglected to close the log afterwards.
4950 30. The text of an SMTP error response that was received during a remote
4951 delivery was being truncated at 512 bytes. This is too short for some of
4952 the long messages that one sometimes sees. I've increased the limit to
4955 31. It is now possible to make retry rules that apply only when a message has a
4956 specific sender, in particular, an empty sender.
4958 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4959 it possible to be selective about when SMTP synchronization is enforced.
4961 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4963 32. Implemented hosts_connection_nolog.
4965 33. Added an ACL for QUIT.
4967 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4970 35. Added mailbox_size and mailbox_filecount to appendfile.
4972 36. Added control = no_multiline_responses to ACLs.
4974 37. There was a bug in the logic of the code that waits for the clock to tick
4975 in the case where the clock went backwards by a substantial amount such
4976 that the microsecond fraction of "now" was more than the microsecond
4977 fraction of "then" (but the whole seconds number was less).
4979 38. Added support for the libradius Radius client library this is found on
4980 FreeBSD (previously only the radiusclient library was supported).
4986 1. When certain lookups returned multiple values in the form name=value, the
4987 quoting of the values was not always being done properly. Specifically:
4988 (a) If the value started with a double quote, but contained no whitespace,
4990 (b) If the value contained whitespace other than a space character (i.e.
4991 tabs or newlines or carriage returns) it was not quoted.
4992 This fix has been applied to the mysql and pgsql lookups by writing a
4993 separate quoting function and calling it from the lookup code. The fix
4994 should probably also be applied to nisplus, ibase and oracle lookups, but
4995 since I cannot test any of those, I have not disturbed their existing code.
4997 2. A hit in the callout cache for a specific address caused a log line with no
4998 reason for rejecting RCPT. Now it says "Previous (cached) callout
4999 verification failure".
5001 3. There was an off-by-one bug in the queryprogram router. An over-long
5002 return line was truncated at 256 instead of 255 characters, thereby
5003 overflowing its buffer with the terminating zero. As well as fixing this, I
5004 have increased the buffer size to 1024 (and made a note to document this).
5006 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5007 when Exim is waiting for an SMTP response from a remote server, Exim
5008 restarts its select() call on the socket, thereby resetting its timeout.
5009 This is not a problem when such interrupts are rare. Somebody set up a cron
5010 job to run exiwhat every 2 minutes, which is less than the normal select()
5011 timeout (5 or 10 minutes). This meant that the select() timeout never
5012 kicked in because it was always reset. I have fixed this by comparing the
5013 time when an interrupt arrives with the time at the start of the first call
5014 to select(). If more time than the timeout has elapsed, the interrupt is
5015 treated as a timeout.
5017 5. Some internal re-factoring in preparation for the addition of Sieve
5018 extensions (by MH). In particular, the "personal" test is moved to a
5019 separate function, and given an option for scanning Cc: and Bcc: (which is
5020 not set for Exim filters).
5022 6. When Exim created an email address using the login of the caller as the
5023 local part (e.g. when creating a From: or Sender: header line), it was not
5024 quoting the local part when it contained special characters such as @.
5026 7. Installed new OpenBSD configuration files.
5028 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5029 try to make them clearer.
5031 9. Callout options, other than the timeout value, were being ignored when
5032 verifying sender addresses in header lines. For example, when using
5034 verify = header_sender/callout=no_cache
5036 the cache was (incorrectly) being used.
5038 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5039 only the Cygwin environment.
5041 11. When return_path_on_delivery was set as a log selector, if different remote
5042 addresses in the same message used different return paths and parallel
5043 remote delivery occurred, the wrong values would sometimes be logged.
5044 (Whenever a remote delivery process finished, the return path value from
5045 the most recently started remote delivery process was logged.)
5047 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5048 lines when AUTH and/or TLS are in use. This is the "received protocol"
5049 field. Exim used to use "asmtp" for authenticated SMTP, without any
5050 indication (in the protocol name) for TLS use. Now it follows the RFC and
5051 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5052 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5053 names appear in log lines as well as in Received: header lines.
5055 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5056 extensions, and comparison tests, and to fix some bugs.
5058 14. Changes to the "personal" filter test:
5060 (1) The test was buggy in that it was just doing the equivalent of
5061 "contains" tests on header lines. For example, if a user's address was
5062 anne@some.where, the "personal" test would incorrectly be true for
5064 To: susanne@some.where
5066 This test is now done by extracting each address from the header in turn,
5067 and checking the entire address. Other tests that are part of "personal"
5068 are now done using regular expressions (for example, to check local parts
5069 of addresses in From: header lines).
5071 (2) The list of non-personal local parts in From: addresses has been
5072 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5073 taken from the Sieve specification recommendations.
5075 (3) If the message contains any header line starting with "List-" it is
5076 treated as non-personal.
5078 (4) The test for "circular" in the Subject: header line has been removed
5079 because it now seems ill-conceived.
5081 15. Minor typos in src/EDITME comments corrected.
5083 16. Installed latest exipick from John Jetmore.
5085 17. If headers_add on a router specified a text string that was too long for
5086 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5087 of string_sprintf() is now avoided.
5089 18. $message_body_size was not set (it was always zero) when running the DATA
5090 ACL and the local_scan() function.
5092 19. For the "mail" command in an Exim filter, no default was being set for
5093 the once_repeat time, causing a random time value to be used if "once" was
5094 specified. (If the value happened to be <= 0, no repeat happened.) The
5095 default is now 0s, meaning "never repeat". The "vacation" command was OK
5096 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5097 (I found it when inspecting the code).
5099 20. There is now an overall timeout for performing a callout verification. It
5100 defaults to 4 times the callout timeout, which applies to individual SMTP
5101 commands during the callout. The overall timeout applies when there is more
5102 than one host that can be tried. The timeout is checked before trying the
5103 next host. This prevents very long delays if there are a large number of
5104 hosts and all are timing out (e.g. when the network connections are timing
5105 out). The value of the overall timeout can be changed by specifying an
5106 additional sub-option for "callout", called "maxwait". For example:
5108 verify = sender/callout=5s,maxwait=20s
5110 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5111 to the end before writing, but this should make it even safer).
5113 22. Exim was forgetting that it had advertised PIPELINING for the second and
5114 subsequent messages on an SMTP connection. It was also not resetting its
5115 memory on STARTTLS and an internal HELO.
5117 23. When Exim logs an SMTP synchronization error within a session, it now
5118 records whether PIPELINING has been advertised or not.
5120 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5121 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5122 rather than long int.
5124 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5126 26. Added the never_mail option to autoreply.
5132 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5133 crash if the getsockname() call failed; this can happen if a connection is
5134 closed very soon after it is established. The problem was simply in the
5135 order in which certain operations were done, causing Exim to try to write
5136 to the SMTP stream before it had set up the file descriptor. The bug has
5137 been fixed by making things happen in the correct order.
5143 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5144 before the connection was closed, thus losing the rejection response.
5146 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5147 some early Solaris releases, but causes trouble in current releases where
5148 socklen_t is defined.
5150 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5153 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5156 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5157 that releasing the top of it at the end releases what was used for sub-
5158 expansions (unless the block got too big). However, discard this block if
5159 the first thing is a variable or header, so that we can use its block when
5160 it is dynamic (useful for very large $message_headers, for example).
5162 6. Lookups now cache *every* query, not just the most recent. A new, separate
5163 store pool is used for this. It can be recovered when all lookup caches are
5164 flushed. Lookups now release memory at the end of their result strings.
5165 This has involved some general refactoring of the lookup sources.
5167 7. Some code has been added to the store_xxx() functions to reduce the amount
5168 of flapping under certain conditions.
5170 8. log_incoming_interface used to affect only the <= reception log lines. Now
5171 it causes the local interface and port to be added to several more SMTP log
5172 lines, for example "SMTP connection from", and rejection lines.
5174 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5176 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5178 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5179 could be overwritten at the end of the current message (or the start of a
5180 new message if it was set in a HELO ACL). The value is now preserved for
5181 the duration of the SMTP connection.
5183 12. If a transport had a headers_rewrite setting, and a matching header line
5184 contained an unqualified address, that address was qualified, even if it
5185 did not match any rewriting rules. The underlying bug was that the values
5186 of the flags that permit the existence of unqualified sender and recipient
5187 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5188 non-local messages, and by -bnq for local messages) were not being
5189 preserved with the message after it was received.
5191 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5192 "next input=" as part of the text comprising the host identity instead of
5193 the correct text. The code was using the same buffer for two different
5194 strings. However, depending on which order the printing function evaluated
5195 its arguments, the bug did not always show up. Under Linux, for example, my
5196 test suite worked just fine.
5198 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5199 doesn't work with some older versions of Perl. It has been changed to "my",
5200 which in any case is probably the better facility to use.
5202 15. A really picky compiler found some instances of statements for creating
5203 error messages that either had too many or two few arguments for the format
5206 16. The size of the buffer for calls to the DNS resolver has been increased
5207 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5208 for addresses that have a lot of PTR records. This alleviates a problem; it
5209 does not fully solve it.
5211 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5212 buffer now truncates the list and logs the incident, which is the same
5213 action as happens when Exim is looking up a host name and its aliases.
5214 Previously in this situation something unpredictable would happen;
5215 sometimes it was "internal error: store_reset failed".
5217 18. If a server dropped the connection unexpectedly when an Exim client was
5218 using GnuTLS and trying to read a response, the client delivery process
5219 crashed while trying to generate an error log message.
5221 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5222 string, for example:
5224 warn message = H1: something\nH2: something
5226 the text was added as a single header line from Exim's point of view
5227 though it ended up OK in the delivered message. However, searching for the
5228 second and subsequent header lines using $h_h2: did not work. This has been
5229 fixed. Similarly, if a system filter added multiple headers in this way,
5230 the routers could not see them.
5232 20. Expanded the error message when iplsearch is called with an invalid key to
5233 suggest using net-iplsearch in a host list.
5235 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5236 ACLs are no longer actually imposed (and a message to that effect is
5239 22. If a "gecos" field in a passwd entry contained escaped characters, in
5240 particular, if it contained a \" sequence, Exim got it wrong when building
5241 a From: or a Sender: header from that name. A second bug also caused
5242 incorrect handling when an unquoted " was present following a character
5243 that needed quoting.
5245 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5246 was not being matched caselessly.
5248 24. Arranged for all hyphens in the exim.8 source to be escaped with
5251 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5252 properly. Recipient callout cache records were still being keyed to include
5253 the sender, even when use_sender was set false. This led to far more
5254 callouts that were necessary. The sender is no longer included in the key
5255 when use_sender is false.
5257 26. Added "control = submission" modifier to ACLs.
5259 27. Added the ${base62d: operator to decode base 62 numbers.
5261 28. dnsdb lookups can now access SRV records.
5263 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5264 the configuration file.
5266 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5267 (-v) mode. This makes the output for a verbose queue run more intelligible.
5269 31. Added a use_postmaster feature to recipient callouts.
5271 32. Added the $body_zerocount variable, containing the number of binary zero
5272 bytes in the message body.
5274 33. The time of last modification of the "new" subdirectory is now used as the
5275 "mailbox time last read" when there is a quota error for a maildir
5278 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5280 35. Added +ignore_unknown as a special item in host lists.
5282 36. Code for decoding IPv6 addresses in host lists is now included, even if
5283 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5284 address was recognized as an IP address, but was then not correctly decoded
5285 into binary, causing unexpected and incorrect effects when compared with
5292 1. Very minor rewording of debugging text in manualroute to say "list of
5293 hosts" instead of "hostlist".
5295 2. If verify=header_syntax was set, and a header line with an unqualified
5296 address (no domain) and a large number of spaces between the end of the
5297 name and the colon was received, the reception process suffered a buffer
5298 overflow, and (when I tested it) crashed. This was caused by some obsolete
5299 code that should have been removed. The fix is to remove it!
5301 3. When running in the test harness, delay a bit after writing a bounce
5302 message to get a bit more predictability in the log output.
5304 4. Added a call to search_tidyup() just before forking a reception process. In
5305 theory, someone could use a lookup in the expansion of smtp_accept_max_
5306 per_host which, without the tidyup, could leave open a database connection.
5308 5. Added the variables $recipient_data and $sender_data which get set from a
5309 lookup success in an ACL "recipients" or "senders" condition, or a router
5310 "senders" option, similar to $domain_data and $local_part_data.
5312 6. Moved the writing of debug_print from before to after the "senders" test
5315 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5316 problems for message scanning, either using a data ACL, or using
5317 local_scan() because the Received: header was not generated till after they
5318 were called (in order to set the time as the time of reception completion).
5319 I have revised the way this works. The header is now generated after the
5320 body is received, but before the ACL or local_scan() are called. After they
5321 are run, the timestamp in the header is updated.
5327 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5328 before starting a queue runner without re-exec. This happened only when
5329 deliver_drop_privilege was set or when the Exim user was set to root. The
5330 effect of the bug was that timeouts during subsequent deliveries caused
5331 crashes instead of being properly handled. The handler is now left at its
5332 default (and expected) setting.
5334 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5335 message, again when deliver_drop_privilege is set or Exim is run as root.
5336 The bug described in (1) was not present in this case, but the tidying up
5337 of the other signals was missing. I have made the two cases consistent.
5339 3. The ignore_target_hosts setting on a manualroute router was being ignored
5340 for hosts that were looked up using the /MX notation.
5342 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5345 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5346 operated on the sender address. After changing the $sender_address to <>
5347 for the sender address verify, Exim was re-instated it as the original
5348 (before rewriting) address, but remembering that it had rewritten it, so it
5349 wasn't rewriting it again. This bug also had the effect of breaking the
5350 sender address verification caching when the sender address was rewritten.
5352 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5353 This has been changed so that if the ip literal address matches
5354 ignore_target_hosts, the router declines.
5356 7. Added expansion conditions match_domain, match_address, and match_local_
5357 part (NOT match_host).
5359 8. The placeholder for the Received: header didn't have a length field set.
5361 9. Added code to Exim itself and to exim_lock to test for a specific race
5362 condition that could lead to file corruption when using MBX delivery. The
5363 issue is with the lockfile that is created in /tmp. If this file is removed
5364 after a process has opened it but before that process has acquired a lock,
5365 there is the potential for a second process to recreate the file and also
5366 acquire a lock. This could lead to two Exim processes writing to the file
5367 at the same time. The added code performs the same test as UW imapd; it
5368 checks after acquiring the lock that its file descriptor still refers to
5369 the same named file.
5371 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5372 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5375 11. Added the smtp_active_hostname option. If used, this will typically be made
5376 to depend on the incoming interface address. Because $interface_address is
5377 not set up until the daemon has forked a reception process, error responses
5378 that can happen earlier (such as "too many connections") no longer contain
5381 12. If an expansion in a condition on a "warn" statement fails because a lookup
5382 defers, the "warn" statement is abandoned, and the next ACL statement is
5383 processed. Previously this caused the whole ACL to be aborted.
5385 13. Added the iplsearch lookup type.
5387 14. Added ident_timeout as a log selector.
5389 15. Added tls_certificate_verified as a log selector.
5391 16. Added a global option tls_require_ciphers (compare the smtp transport
5392 option of the same name). This controls incoming TLS connections.
5394 17. I finally figured out how to make tls_require_ciphers do a similar thing
5395 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5396 before starting the TLS session.
5398 18. Tabs are now shown as \t in -bP output.
5400 19. If the log selector return_path_on_delivery was set, Exim crashed when
5401 bouncing a message because it had too many Received: header lines.
5403 20. If two routers both had headers_remove settings, and the first one included
5404 a superfluous trailing colon, the final name in the first list and the
5405 first name in the second list were incorrectly joined into one item (with a
5406 colon in the middle).
5412 1. Added -C and -D options to the exinext utility, mainly to make it easier
5413 to include in the automated testing, but these could be helpful when
5414 multiple configurations are in use.
5416 2. The exinext utility was not formatting the output nicely when there was
5417 an alternate port involved in the retry record key, nor when there was a
5418 message id as well (for retries that were specific to a specific message
5419 and a specific host). It was also confused by IPv6 addresses, because of
5420 the additional colons they contain. I have fixed the IPv4 problem, and
5421 patched it up to do a reasonable job for IPv6.
5423 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5424 delivery, the log line now contains "pipelined" if PIPELINING was used.
5426 4. An SMTP transport process used to panic and die if the bind() call to set
5427 an explicit outgoing interface failed. This has been changed; it is now
5428 treated in the same way as a connect() failure.
5430 5. A reference to $sender_host_name in the part of a conditional expansion
5431 that was being skipped was still causing a DNS lookup. This no longer
5434 6. The def: expansion condition was not recognizing references to header lines
5435 that used bh_ and bheader_.
5437 7. Added the _cache feature to named lists.
5439 8. The code for checking quota_filecount in the appendfile transport was
5440 allowing one more file than it should have been.
5442 9. For compatibility with Sendmail, the command line option
5450 and sets the incoming protocol and host name (for trusted callers). The
5451 host name and its colon can be omitted when only the protocol is to be set.
5452 Note the Exim already has two private options, -pd and -ps, that refer to
5453 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5454 "s", but I don't think that's a major issue.
5456 10. A number of refactoring changes to the code, none of which should affect
5459 (a) The number of logging options was getting close to filling up the
5460 32-bit word that was used as a bit map. I have split them into two classes:
5461 those that are passed in the argument to log_write(), and those that are
5462 only ever tested independently outside of that function. These are now in
5463 separate 32-bit words, so there is plenty of room for expansion again.
5464 There is no change in the user interface or the logging behaviour.
5466 (b) When building, for example, log lines, the code previously used a
5467 macro that called string_cat() twice, in order to add two strings. This is
5468 not really sufficiently general. Furthermore, there was one instance where
5469 it was actually wrong because one of the argument was used twice, and in
5470 one call a function was used. (As it happened, calling the function twice
5471 did not affect the overall behaviour.) The macro has been replaced by a
5472 function that can join an arbitrary number of extra strings onto a growing
5475 (c) The code for expansion conditions now uses a table and a binary chop
5476 instead of a serial search (which was left over from when there were very
5477 few conditions). Also, it now recognizes conditions like "pam" even when
5478 the relevant support is not compiled in: a suitably worded error message is
5479 given if an attempt is made to use such a condition.
5481 11. Added ${time_interval:xxxxx}.
5483 12. A bug was causing one of the ddress fields not to be passed back correctly
5484 from remote delivery subprocesses. The field in question was not being
5485 subsequently used, so this caused to problems in practice.
5487 13. Added new log selectors queue_time and deliver_time.
5489 14. Might have fixed a bug in maildirsizefile handling that threw up
5490 "unexpected character" debug warnings, and recalculated the data
5491 unnecessarily. In any case, I expanded the warning message to give more
5494 15. Added the message "Restricted characters in address" to the statements in
5495 the default ACL that block characters like @ and % in local parts.
5497 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5498 Three changes have been made:
5500 (a) There was a serious bug; a negative response to MAIL caused the whole
5501 recipient domain to be cached as invalid, thereby blocking all messages
5502 to all local parts at the same domain, from all senders. This bug has
5503 been fixed. The domain is no longer cached after a negative response to
5504 MAIL if the sender used is not empty.
5506 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5509 (c) A new callout option, "use_sender" has been added for people who want
5510 the modified behaviour.
5516 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5519 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5520 indeed breaks things for older releases.
5522 3. Added additional logging to the case where there is a problem reading data
5523 from a filter that is running in a subprocess using a pipe, in order to
5524 try to track down a specific problem.
5526 4. Testing facility fudge: when running in the test harness and attempting
5527 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5528 getting "No route to host". Convert this to a timeout.
5530 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5533 6. Some OS don't have socklen_t but use size_t instead. This affects the
5534 fifth argument of getsockopt() amongst other things. This is now
5535 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5536 can be set for individual OS. I have set it for SunOS5, OSF1, and
5537 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5538 some earlier ones do not.
5540 7. Change 4.30/15 was not doing the test caselessly.
5542 8. The standard form for an IPv6 address literal was being rejected by address
5543 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5544 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5545 this, as well as the form without the "IPv6" on the front (but only when
5546 address literals are enabled, of course).
5548 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5550 10. Exim crashed if a message with an empty sender address specified by -f
5551 encountered a router with an errors_to setting. This could be provoked only
5552 by a command such as
5556 where an empty string was supplied; "<>" did not hit this bug.
5558 11. Installed PCRE release 4.5.
5560 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5561 remained set. It is now erased.
5563 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5564 times from message ids (which are base 36 rather than the normal 62).
5566 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5567 were being counted as actual protocol errors, and logged if the log
5568 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5569 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5570 and DATA following a set of rejected RCPTs do not count as protocol errors.
5571 In other words, Exim assumes they were pipelined, though this may not
5572 actually be the case. Of course, in all cases the client gets an
5573 appropriate error code.
5575 15. If a lookup fails in an ACL condition, a message about the failure may
5576 be available; it is used if testing the ACL cannot continue, because most
5577 such messages specify what the cause of the deferral is. However, some
5578 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5579 that caused an old message to be retained and used if a later statement
5580 caused a defer, replacing the real cause of the deferral.
5582 16. If an IP address had so many PTR records that the DNS lookup buffer
5583 was not large enough to hold them, Exim could crash while trying to process
5584 the truncated data. It now detects and logs this case.
5586 17. Further to 4.21/58, another change has been made: if (and only if) the
5587 first line of a message (the first header line) ends with CRLF, a bare LF
5588 in a subsequent header line has a space inserted after it, so as not to
5589 terminate the header.
5591 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5592 unnecessarily, used atoi() instead of strtol(), and didn't check the
5593 termination when getting file sizes from file names by regex.
5595 19. Completely re-implemented the support for maildirsize files, in the light
5596 of a number of problems with the previous contributed implementation
5597 (4.30/29). In particular:
5599 . If the quota is zero, the maildirsize file is maintained, but no quota is
5602 . If the maildir directory does not exist, it is created before any attempt
5603 to write a maildirsize file.
5605 . The quota value in the file is just a cache; if the quota is changed in
5606 the transport, the new value overrides.
5608 . A regular expression is available for excluding directories from the
5611 20. The autoreply transport checks the characters in options that define the
5612 message's headers; it allows continued headers, but it was checking with
5613 isspace() after an embedded newline instead of explicitly looking for a
5616 21. If all the "regular" hosts to which an address was routed had passed their
5617 expiry times, and had not reached their retry times, the address was
5618 bounced, even if fallback hosts were defined. Now Exim should go on to try
5621 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5622 equivalent code in the SMTP transport. Some hosts send humungous responses
5623 to HELO/EHLO, more than 1024 it seems.
5625 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5626 gives compiler warnings in some environments. I've now done it "properly",
5629 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5630 (because of problems with the built-in one) was declared to return uschar *
5631 instead of char *, causing compiler failure.
5633 25. Fixed a file descriptor leak when processing alias/forward files.
5635 26. Fixed a minor format string issue in dbfn.c.
5637 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5639 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5640 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5643 29. When a sender address is verified, it is cached, to save repeating the test
5644 when there is more than one recipient in a message. However, when the
5645 verification involves a callout, it is possible for different callout
5646 options to be set for different recipients. It is too complicated to keep
5647 track of this in the cache, so now Exim always runs a verification when a
5648 callout is required, relying on the callout cache for the optimization.
5649 The overhead is duplication of the address routing, but this should not be
5652 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5653 to be verified with callout=postmaster, and the main callout worked but the
5654 postmaster check failed, the verification correctly failed. However, if a
5655 subsequent RCPT command asked for sender verification *without* the
5656 postmaster check, incorrect caching caused this verification also to fail,
5659 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5660 it was not caching the DNS options (qualify_single, search_parents) that
5661 were used when the lookup failed. A subsequent lookup with different
5662 options therefore always gave the same answer, though there were cases
5663 where it should not have. (Example: a "domains = !$mx_any" option on a
5664 dnslookup router: the "domains" option is always processed without any
5665 widening, but the router might have qualify_single set.) Now Exim uses the
5666 cached value only when the same options are set.
5668 32. Added John Jetmore's "exipick" utility to the distribution.
5670 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5671 than a timeout (e.g. a certificate is required, and is not provided), an
5672 Exim server now closes the connection immediately. Previously it waited for
5673 the client to close - but if the client is SSL, it seems that they each
5674 wait for each other, leading to a delay before one of them times out.
5676 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5677 maintained 0.8.x compatibility because I don't think many are using it, and
5678 it is clearly obsolete.
5680 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5683 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5684 client certificate was expired. A simple patch fixes this, though I don't
5685 understand the full logic of why the verify callback is called multiple
5688 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5689 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5690 which causes problems with some clients (such as the Certicom SSL Plus
5691 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5692 disables the coutermeasure allowing Eudora to connect."
5694 38. Exim was not checking that a write() to a log file succeeded. This could
5695 lead to Bad Things if a log got too big, in particular if it hit a file
5696 size limit. Exim now panics and dies if it cannot write to a log file, just
5697 as it does if it cannot open a log file.
5699 39. Modified OS/Makefile-Linux so that it now contains
5701 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5703 The two -D definitions ensure that Exim is compiled with large file
5704 support, which makes it possible to handle log files that are bigger than
5707 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5708 instance) a domain was checked against a named list that involved a lookup,
5709 causing $domain_data to be set, then another domain was checked against the
5710 same list, then the first domain was re-checked, the value of $domain_data
5711 after the final check could be wrong. In particular, if the second check
5712 failed, it could be set empty. This bug probably also applied to
5715 41. The strip_trailing_dot option was not being applied to the address given
5716 with the -f command-line option.
5718 42. The code for reading a message's header from the spool was incrementing
5719 $received_count, but never initializing it. This meant that the value was
5720 incorrect (doubled) while delivering a message in the same process in which
5721 it was received. In the most common configuration of Exim, this never
5722 happens - a fresh exec is done - but it can happen when
5723 deliver_drop_privilege is set.
5725 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5726 it now includes up to 150 characters of the unexpected data in the log
5729 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5730 and building data strings. The size of both of these buffers was 10 000
5731 bytes - far larger than anybody would *ever* want, thought I. Needless to
5732 say, somebody hit the limit. I have increased the maximum line length to
5733 20 000 and the maximum data length of concatenated lines to 100 000. I have
5734 also fixed two bugs, because there was no checking on these buffers. Tsk,
5735 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5736 buffer is too small.
5738 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5739 lsearch lookups. Now it does.
5741 46. When parsing a route_list item in a manualroute router, a fixed-length
5742 buffer was used for the list of hosts. I made this 1024 bytes long,
5743 thinking that nobody would ever have a list of hosts that long. Wrong.
5744 Somebody had a whole pile of complicated expansion conditions, and the
5745 string was silently truncated, leading to an expansion error. It turns out
5746 that it is easier to change to an unlimited length (owing to other changes
5747 that have happened since this code was originally written) than to build
5748 structure for giving a limitation error. The length of the item that
5749 expands into the list of hosts is now unlimited.
5751 47. The lsearch lookup could not handle data where the length of text line was
5752 more than 4095 characters. Such lines were truncated, leading to shortened
5753 data being returned. It should now handle lines of any length.
5755 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5756 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5759 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5760 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5761 no longer happens in comment lines. A list of replacements is now placed
5762 at the head of all of the source files, except those whose only change is
5763 to replace PERL_COMMAND in the very first #! line.
5765 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5766 messages on the queue, with a bottom-up merge sort, using code contributed
5767 by Michael Haardt. This should make operations like -bp somewhat faster on
5768 large queues. It won't affect queue runners, except when queue_run_in_order
5771 51. Installed eximstats 1.31 in the distribution.
5773 52. Added support for SRV lookups to the dnslookup router.
5775 53. If an ACL referred to $message_body or $message_body_end, the value was not
5776 reset for any messages that followed in the same SMTP session.
5778 54. The store-handling optimization for building very long strings was not
5779 differentiating between the different store pools. I don't think this
5780 actually made any difference in practice, but I've tidied it.
5782 55. While running the routers to verify a sender address, $sender_address
5783 was still set to the sender address. This is wrong, because when routing to
5784 send a bounce to the sender, it would be empty. Therefore, I have changed
5785 it so that, while verifying a sender address, $sender_address is set to <>.
5786 (There is no change to what happens when verifying a recipient address.)
5788 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5789 target A or AAAA records (if not already returned) without resetting the
5790 qualify_single or search_parents options of the DNS resolver. These are
5791 inappropriate in this case because the targets of MX and SRV records must
5792 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5793 target that, when qualified, matched something in the local domain. These
5794 two options are now turned off when doing these lookups.
5796 57. It seems that at least some releases of Reiserfs (which does not have the
5797 concept of a fixed number of inodes) returns zero and not -1 for the
5798 number of available inodes. This interacted badly with check_spool_inodes,
5799 which assumed that -1 was the "no such thing" setting. What I have done is
5800 to check that the total number of inodes is greater than zero before doing
5801 the test of how many are available.
5803 58. When a "warn" ACL statement has a log_message modifier, the message is
5804 remembered, and not repeated. This is to avoid a lot of repetition when a
5805 message has many recipients that cause the same warning to be written.
5806 However, Exim was preserving the list of already written lines for an
5807 entire SMTP session, which doesn't seem right. The memory is now reset if a
5808 new message is started.
5810 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5811 debugging output unless log_rewrite was also set.
5813 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5814 of (int)(handle) when we know that handle contains (void *)(-1).
5816 61. The Exim daemon panic-logs an error return when it closes the incoming
5817 connection. However "connection reset by peer" seems to be common, and
5818 isn't really an error worthy of noting specially, so that particular error
5821 62. When Exim is trying to find all the local interfaces, it used to panic and
5822 die if the ioctl to get the interface flags failed. However, it seems that
5823 on at least one OS (Solaris 9) it is possible to have an interface that is
5824 included in the list of interfaces, but for which you get a failure error
5825 for this call. This happens when the interface is not "plumbed" into a
5826 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5827 failure of the "get flags" call assumes that the interface is down.
5829 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5830 makes life easier for people who are doing arithmetic on fields extracted
5831 from dates, where you often get leading zeros that should not be
5832 interpreted as octal.
5834 64. Added qualify_domain to the redirect router, to override the global
5837 65. If a pathologically long header line contained very many addresses (the
5838 report of this problem mentioned 10 000) and each of them was rewritten,
5839 Exim could use up a very large amount of memory. (It kept on making new
5840 copies of the header line as it rewrote, and never released the old ones.)
5841 At the expense of a bit more processing, the header rewriting function has
5842 been changed so that it no longer eats memory in this way.
5844 66. The generation of the Received: header has been moved from the time that a
5845 message starts to be received, to the time that it finishes. The timestamp
5846 in the Received: header should now be very close to that of the <= log
5847 line. There are two side-effects of this change:
5849 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5850 logged header lines no longer include the local Received: line, because
5851 it has not yet been created. The same applies to a copy of the message
5852 that is returned to a non-SMTP sender when a message is rejected.
5854 (b) When a filter file is tested using -bf, no additional Received: header
5855 is added to the test message. After some thought, I decided that this
5858 This change does not affect the value of $received_for. It is still set
5859 after address rewriting, but before local_scan() is called.
5861 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5863 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5864 gave an unhelpful panic error message, and a defer error. I have managed to
5865 change this behaviour so that it now rejects any supplied certificate,
5866 which seems right, as the list of acceptable certificates is empty.
5868 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5869 gave an unhelpful defer error. I have not managed to make this reject any
5870 supplied certificates, but the error message it gives is "no certificate
5871 supplied", which is not helpful.
5873 70. exigrep's output now also includes lines that are not associated with any
5874 message, but which match the given pattern. Implemented by a patch from
5875 Martin Sluka, which also tidied up the Perl a bit.
5877 71. Recipient callout verification, like sender verification, was using <> in
5878 the MAIL FROM command. This isn't really the right thing, since the actual
5879 sender may affect whether the remote host accepts the recipient or not. I
5880 have changed it to use the actual sender in the callout; this means that
5881 the cache record is now keyed on a recipient/sender pair, not just the
5882 recipient address. There doesn't seem to be a real danger of callout loops,
5883 since a callout by the remote host to check the sender would use <>.
5884 [SEE ABOVE: changed after hitting problems.]
5886 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5887 temporary errors. However, in the case of such a code being given after
5888 the end of a data transmission (i.e. after ".") Exim was failing to write
5889 a retry record for the message. (Yes, there was some broken host that was
5890 actually sending 8xx at this point.)
5892 73. An unknown lookup type in a host list could cause Exim to panic-die when
5893 the list was checked. (An example that provoked this was putting <; in the
5894 middle of a list instead of at the start.) If this happened during a DATA
5895 ACL check, a -D file could be left lying around. This kind of configuration
5896 error no longer causes Exim to die; instead it causes a defer error. The
5897 incident is still logged to the main and panic logs.
5899 74. Buglet left over from Exim 3 conversion. The message "too many messages
5900 in one connection" was written to the rejectlog but not the mainlog, except
5901 when address rewriting (yes!) was being logged.
5903 75. Added write_rejectlog option.
5905 76. When a system filter was run not as root (that is, when system_filter_user
5906 was set), the values of the $n variables were not being returned to the
5907 main process; thus, they were not subsequently available in the $sn
5910 77. Added +return_path_on_delivery log selector.
5912 78. A connection timeout was being treated differently from recipients deferred
5913 when testing hosts_max_try with a message that was older than the host's
5914 retry timeout. (The host should not be counted, thus allowing all hosts to
5915 be tried at least once before bouncing.) This may have been the cause of an
5916 occasionally reported bug whereby a message would remain on the queue
5917 longer than the retry timeout, but would be bounced if a delivery was
5918 forced. I say "may" because I never totally pinned down the problem;
5919 setting up timeout/retry tests is difficult. See also the next item.
5921 79. The ultimate address timeout was not being applied to errors that involved
5922 a combination of host plus message (for example, a timeout on a MAIL
5923 command). When an address resolved to a number of possible hosts, and they
5924 were not all tried for each delivery (e.g. because of hosts_max_try), a
5925 message could remain on the queue longer than the retry timeout.
5927 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5930 81. Fixed an obscure SMTP outgoing bug which required at least the following
5931 conditions: (a) there was another message waiting for the same server;
5932 (b) the server returned 5xx to all RCPT commands in the first message so
5933 that the message was not completed; (c) the server dropped the connection
5934 or gave a negative response to the RSET that Exim sends to abort the
5935 transaction. The observed case was a dropped connection after DATA that had
5936 been sent in pipelining mode. That is, the server had advertised PIPELINING
5937 but was not implementing it correctly. The effect of the bug was incorrect
5938 behaviour, such as trying another host, and this could lead to a crash.
5944 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5945 and daemon.c were passed as pointers to ints; they should have been
5946 pointers to socklen_t variables (which are typically unsigned ints).
5948 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5951 3. Fixed a really odd bug that affected only the testing scheme; patching a
5952 certain fixed string in the binary changed the value of another string that
5953 happened to be identical to the end of the original first string.
5955 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5956 name", it returns that address as the IP address. On some operating
5957 systems (e.g. Solaris), it also passes back the IP address string as the
5958 "host name". However, on others (e.g. Linux), it passes back an empty
5959 string. Exim wasn't checking for this, and was changing the host name to an
5960 empty string, assuming it had been canonicalized.
5962 5. Although rare, it is permitted to have more than one PTR record for a given
5963 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5964 all the names associated with an address, because they do in Solaris.
5965 However, it seems that they do not in Linux for data that comes from the
5966 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5967 I found this out when I moved to a new Linux workstation and tried to run
5968 the Exim test suite.
5970 To get round this problem I have changed the code so that it now does its
5971 own call to the DNS to look up PTR records when searching for a host name.
5972 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5973 addresses that are only in /etc/hosts are still found.
5975 This behaviour is, however, controlled by an option called host_lookup_
5976 order, which defaults to "bydns:byaddr". If people want to use the other
5977 order, or indeed, just use one or the other means of lookup, they can
5978 specify it in this variable.
5980 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5981 some operating systems, this comes back from gethostbyaddr() as an empty
5982 string, and this is what Exim used to test for. However, it seems that in
5983 other systems, "." is yielded. Exim now tests for this case too.
5985 7. The values of check_spool_space and check_log_space are now held internally
5986 as a number of kilobytes instead of an absolute number of bytes. If a
5987 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5988 kilobyte. This means that much larger values can be stored.
5990 8. Exim monitor: an attempt to get the action menu when not actually pointing
5991 at a message produces an empty menu entitled "No message selected". This
5992 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5993 no entries in it ("Shell widget menu has zero width and/or height"). So I
5994 have added a single, blank menu entry in this case.
5996 9. Added ${quote_local_part.
5998 10. MIME decoding is now applied to the contents of Subject: header lines when
6001 11. Now that a reference to $sender_host_address automatically causes a reverse
6002 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6003 host lookup before query-style lookups in lists that might use this
6004 variable. This has therefore been abolished, and the "net-" prefix is no
6005 longer necessary for query-style lookups.
6007 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6008 have been a typo for LFLAGS, so it has been changed.
6010 13. The install script calls Exim with "-C /dev/null" in order to find the
6011 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6012 to be output. However, since Exim outputs its version number before the
6013 error, it didn't break the script. It just looked ugly. I fixed this by
6014 always allowing "-C /dev/null" if the caller is root.
6016 14. Ignore overlarge ACL variable number when reading spool file - insurance
6017 against a later release with more variables having written the file.
6019 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6020 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6021 this, as well as the form without the "IPv6" on the front.
6023 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6024 OS/Makefile-Darwin file.
6026 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6027 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6029 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6030 variable. However, some OS use ints for inodes. I've added cast to long int
6031 to get rid of the compiler warning.
6033 19. I had forgotten to lock out "/../" in configuration file names when
6034 ALT_CONFIG_PREFIX was set.
6036 20. Routers used for verification do not need to specify transports. However,
6037 if such a router generated a host list, and callout was configured, Exim
6038 crashed, because it could not find a port number from the (non-existent)
6039 transport. It now assumes port 25 in this circumstance.
6041 21. Added the -t option to exigrep.
6043 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6044 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6045 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6046 them is set, LOOKUP_LSEARCH is forced.
6048 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6050 24. Added sender and host information to the "rejected by local_scan()" log
6051 line; previously there was no indication of these.
6053 25. Added .include_if_exists.
6055 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6056 Linux. It turns out that not all file systems support this. Apparently some
6057 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6058 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6059 not supported on the file descriptor, is now ignored when Exim is trying to
6060 sync a directory. This applies only to Linux.
6062 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6064 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6065 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6068 29. Added support for maildirsize files from supplied patch (modified a bit).
6070 30. The use of :fail: followed by an empty string could lead Exim to respond to
6071 sender verification failures with (e.g.):
6073 550 Verification failed for <xxx>
6074 550 Sender verify failed
6076 where the first response line was missing the '-' that indicates it is not
6077 the final line of the response.
6079 31. The loop for finding the name of the user that called Exim had a hardwired
6080 limit of 10; it now uses the value of finduser_retries, which is used for
6081 all other user lookups.
6083 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6086 33. Exim was neglecting to zero errno before one call of strtol() when
6087 expanding a string and expecting an integer value. On some systems this
6088 resulted in spurious "integer overflow" errors. Also, it was casting the
6089 result into an int without checking.
6091 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6092 did not work. The code looks as if it has *never* worked, though it appears
6093 to have been documented since at least release 1.62. I have made it work.
6095 35. The "timeout_DNS" error in retry rules, also documented since at least
6096 1.62, also never worked. As it isn't clear exactly what this means, and
6097 clearly it isn't a major issue, I have abolished the feature by treating it
6098 as "timeout", and writing a warning to the main and panic logs.
6100 36. The display of retry rules for -brt wasn't always showing the error code
6103 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6104 timeout_connect_A, timeout_connect_MX.
6106 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6107 to the empty sender.
6109 39. The daemon was not analysing the content of -oX till after it had closed
6110 stderr and disconnected from the controlling terminal. This meant that any
6111 syntax errors were only noted on the panic log, and the return code from
6112 the command was 0. By re-arranging the code a little, I've made the
6113 decoding happen first, so such errors now appear on stderr, and the return
6114 code is 1. However, the actual setting up of the sockets still happens in
6115 the disconnected process, so errors there are still only recorded on the
6118 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6119 connections (as happens on some IP stacks) was logged at start up time as
6120 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6121 it from "IPv6 and IPv4", which means that two separate sockets are being
6124 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6125 says whether AF_INET or AF_INET6 was passed as an argument.
6127 42. Exiwhat output was messed up when time zones were included in log
6130 43. Exiwhat now gives more information about the daemon's listening ports,
6131 and whether -tls-on-connect was used.
6133 44. The "port" option of the smtp transport is now expanded.
6135 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6136 silently ignored. Now an error message is written to the main and panic
6139 46. There's a new ACL modifier called "logwrite" which writes to a log file
6140 as soon as it is encountered.
6142 47. Added $local_user_uid and $local_user_gid at routing time.
6144 48. Exim crashed when trying to verify a sender address that was being
6147 49. Exim was recognizing only a space character after ".include". It now also
6148 recognizes a tab character.
6150 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6151 extracting the relevant information from the specification. The man page no
6152 longer contains scrambled data for the -d option, and I've added a section
6153 at the front about calling Exim under different names.
6155 51. Added "extra_headers" argument to the "mail" command in filter files.
6157 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6160 53. Installed eximstats 1.29.
6162 54. Added transport_filter_timeout as a generic transport option.
6164 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6165 Cc: header lines. This was required by RFC 822, but it not required by RFC
6168 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6169 incoming messages that did not have them. Now it does so only if the
6170 message originates locally, that is, if there is no associated remote host
6171 address. When Resent- header lines are present, this applies to the Resent-
6172 lines rather than the non-Resent- lines.
6174 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6175 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6177 58. Messages for configuration errors now include the name of the main
6178 configuration file - useful now that there may be more than one file in a
6179 list (.included file names were always shown).
6181 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6182 for those rare installations that do not start the daemon as root or run it
6183 setuid root. I've cut out the call to initgroups() if the daemon is not
6186 60. The Exim user and group can now be bound into the binary as text strings
6187 that are looked up at the start of Exim's processing.
6189 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6191 62. Added $mailstore_basename variable.
6193 63. Installed patch to sieve.c from Michael Haardt.
6195 64. When Exim failed to open the panic log after failing to open the main log,
6196 the original message it was trying to log was written to stderr and debug
6197 output, but if they were not available (the usual case in production), it
6198 was lost. Now it is written to syslog before the two lines that record the
6199 failures to open the logs.
6201 65. Users' Exim filters run in subprocesses under the user's uid. It is
6202 possible for a "deliver" command or an alias in a "personal" command to
6203 provoke an address rewrite. If logging of address rewriting is configured,
6204 this fails because the process is not running as root or exim. There may be
6205 a better way of dealing with this, but for the moment (because 4.30 needs
6206 to be released), I have disabled address rewrite logging when running a
6207 filter in a non-root, non-exim process.
6213 1. The buildconfig auxiliary program wasn't quoting the value set for
6214 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6215 not defined. This bug was masked in 4.22 by the effect that was fixed in
6218 2. Some messages that were rejected after a message id was allocated were
6219 shown as "incomplete" by exigrep. It no longer does this for messages that
6220 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6222 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6223 have allow_domain_literals set, the ID did not get logged in the <= line.
6224 Domain literals are now always recognized in Message-ID: header lines.
6226 4. The first argument for a ${extract expansion item is the key name or field
6227 number. Leading and trailing spaces in this item were not being ignored,
6228 causing some misleading effects.
6230 5. When deliver_drop_privilege was set, single queue runner processes started
6231 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6232 same command in the process it spins off) were not dropping privilege.
6234 6. When the daemon running as "exim" started a queue runner, it always
6235 re-executed Exim in the spun-off process. This is a waste of effort when
6236 deliver_drop_privilege is set. The new process now just calls the
6237 queue-runner function directly.
6243 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6246 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6247 ignored. Though the use of -oP was forcing the writing of a pid file, it
6248 was always written to the default place.
6250 3. If the message "no IP address found for host xxxx" is generated during
6251 incoming verification, it is now followed by identification of the incoming
6252 connection (so you can more easily find what provoked it).
6254 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6256 5. Added some features to "harden" Exim a bit more against certain attacks:
6258 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6259 be put in Local/Makefile. This is like the never_users runtime option,
6260 but it cannot be overridden. The default setting is "root".
6262 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6263 prefix string with which any file named in a -C command line option
6266 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6267 is retained for -C and -D only if the caller of Exim is root. Without
6268 it, the exim user may also use -C and -D and retain privilege.
6270 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6271 command line option is disabled.
6273 6. Macro names set by the -D option must start with an upper case letter, just
6274 like macro names defined in the configuration file.
6276 7. Added "dereference=" facility to LDAP.
6278 8. Two instances of the typo "uknown" in the source files are fixed.
6280 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6281 the Configure-Makefile script screwed up while processing it.
6283 10. Incorporated PCRE 4.4.
6285 11. The SMTP synchronization check was not operating right at the start of an
6286 SMTP session. For example, it could not catch a HELO sent before the client
6287 waited for the greeting. There is now a check for outstanding input at the
6288 point when the greeting is written. Because of the duplex, asynchronous
6289 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6290 way, but not yet received, when the check is performed.
6292 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6293 on TCP/IP sockets, because this apparently causes some broken clients to
6296 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6297 unchanged) from the Cygwin maintainer.
6299 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6300 was supported instead of testing for mbx. Effectively a typo.
6302 15. The spa authenticator server code was not checking that the input it
6303 received was valid base64.
6305 16. The debug output line for the "set" modifier in ACLs was not showing the
6306 name of the variable that was being set.
6308 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6310 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6311 lookup on its own. Something else was needed to trigger the lookup. For
6312 example, a match in host_lookup or the need for a host name in a host list.
6313 Now, if $sender_host_name is referenced and the host name has not yet been
6314 looked up, a lookup is performed. If the lookup fails, the variable remains
6315 empty, and $host_lookup_failed is set to "1".
6317 19. Added "eqi" as a case-independent comparison operator.
6319 20. The saslauthd authentication condition could segfault if neither service
6320 nor realm was specified.
6322 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6323 error message that was logged was misleading, and incoming SMTP
6324 connections were dropped. The message is now more accurate, and temporary
6325 errors are given to SMTP connections.
6327 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6328 (except RSET) with a 421 error, until QUIT is received. However, it was
6329 failing to send a response to QUIT.
6331 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6332 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6333 ACL is now run after the helo_try_verify_hosts code.
6335 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6336 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6337 case-independent, so other case variants are also recognized). Apparently
6338 some systems use these upper case variants.
6340 25. If more than two messages were waiting for the same host, and a transport
6341 filter was specified for the transport, Exim sent two messages over the
6342 same TCP/IP connection, and then failed with "socket operation on non-
6343 socket" when it tried to send the third.
6345 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6347 27. The extern definition of crypt16() in expand.c was not being excluded when
6348 the OS had its own crypt16() function.
6350 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6351 as a preferred synonym for return_size_limit, both as an option and as an
6354 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6356 30. Changed the default configuration ACL to relax the local part checking rule
6357 for addresses that are not in any local domains. For these addresses,
6358 slashes and pipe symbols are allowed within local parts, but the sequence
6359 /../ is explicitly forbidden.
6361 31. SPA server authentication was not clearing the challenge buffer before
6364 32. log_message in a "warn" ACL statement was writing to the reject log as
6365 well as to the main log, which contradicts the documentation and doesn't
6366 seem right (because no rejection is happening). So I have stopped it.
6368 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6369 However, I am unable to do any testing of this.
6371 34. Fixed an infelicity in the appendfile transport. When checking directories
6372 for a mailbox, to see if any needed to be created, it was accidentally
6373 using path names with one or more superfluous leading slashes; tracing
6374 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6376 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6377 contents are added to the log line that is written for every discarded
6378 recipient. (Previously a log_message setting was ignored.)
6380 36. The ${quote: operator now quotes the string if it is empty.
6382 37. The install script runs exim in order to find its version number. If for
6383 some reason other than non-existence or emptiness, which it checks, it
6384 could not run './exim', it was installing it with an empty version number,
6385 i.e. as "exim-". This error state is now caught, and the installation is
6388 38. An argument was missing from the function that creates an error message
6389 when Exim fails to connect to the socket for saslauthd authentication.
6390 This could cause Exim to crash, or give a corrupted message.
6392 39. Added isip, isip4, and isip6 to ${if conditions.
6394 40. The ACL variables $acl_xx are now saved with the message, and can be
6395 accessed later in routers, transports, and filters.
6397 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6398 strings in the file are not string-expanded.
6400 42. If a MAIL command specified a SIZE value that was too large to fit into an
6401 int variable, the check against message_size_limit failed. Such values are
6402 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6403 day this will have to be increased, but I don't think I want to be around
6404 when emails are that large.
6411 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6412 iconv() is not standard in FreeBSD.
6414 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6415 IPv6 enabled. The observed symptom was a segmentation fault on return from
6416 the function os_common_find_running_interfaces() in src/os.c.
6418 3. In the check_special_case() function in daemon.c I had used "errno" as an
6419 argument name, which causes warnings on some systems. This was basically a
6420 typo, since it was named "eno" in the comments!
6422 4. The code that waits for the clock to tick (at a resolution of some fraction
6423 of a second) so as to ensure message-id uniqueness was always waiting for
6424 at least one whole tick, when it could have waited for less. [This is
6425 almost certainly not relevant at current processor speeds, where it is
6426 unlikely to ever wait at all. But we try to future-proof.]
6428 5. The function that sleeps for a time interval that includes fractions of a
6429 second contained a race. It did not block SIGALRM between setting the
6430 timer, and suspending (a couple of lines later). If the interval was short
6431 and the sigsuspend() was delayed until after it had expired, the suspension
6432 never ended. On busy systems this could lead to processes getting stuck for
6435 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6436 process, before it forks any delivery processes. The open lookup caching
6437 mechanism meant that the open file or database connection was passed into
6438 the delivery process. The problem was that delivery processes always tidy
6439 up cached lookup data. This could cause a problem for the next delivery
6440 process started by the queue runner, because the external queue runner
6441 process does not know about the closure. So the next delivery process
6442 still has data in the lookup cache. In the case of a file lookup, there was
6443 no problem because closing a file descriptor in a subprocess doesn't affect
6444 the parent. However, if the lookup was caching a connection to a database,
6445 the connection was closed, and the second delivery process was likely to
6446 see errors such as "PGSQL: query failed: server closed the connection
6447 unexpectedly". The problem has been fixed by closing all cached lookups
6448 in a queue runner before running a delivery process.
6450 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6451 seem to have the "const" qualifier which it has on other OS. I've
6454 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6455 *of the same type* (client or server) with the same public name that an
6456 error should be diagnosed.
6458 9. When Exim looked up a host name for an IP address, but failed to find the
6459 original IP address when looking up the host name (a safety check), it
6460 output the message "<ip address> does not match any IP for NULL", which was
6461 confusing, to say the least. The bug was that the host name should have
6462 appeared instead of "NULL".
6464 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6465 user that is built into the binary, and the -C or -D options is used, root
6466 privilege is dropped before the configuration file is read. In addition,
6467 logging is switched to stderr instead of the normal log files. If the
6468 configuration then re-defines the Exim user, the unprivileged environment
6469 is probably not what is expected, so Exim logs a panic warning message (but
6472 However, if deliver_drop_privilege is set, the unprivileged state may well
6473 be exactly what is intended, so the warning has been cut out in that case,
6474 and Exim is allowed to try to write to its normal log files.
6480 1. smtp_return_error_details was not giving details for temporary sender
6481 or receiver verification errors.
6483 2. Diagnose a configuration error if two authenticators have the same public
6486 3. Exim used not to create the message log file for a message until the first
6487 delivery attempt. This could be confusing when incoming messages were held
6488 for policy or load reasons. The message log file is now created at the time
6489 the message is received, and an initial "Received" line is written to it.
6491 4. The automatically generated man page for command line options had a minor
6492 bug that caused no ill effects; however, a more serious problem was that
6493 the procedure for building the man page automatically didn't always
6494 operate. Consequently, release 4.20 contains an out-of-date version. This
6495 shouldn't happen again.
6497 5. When building Exim with embedded Perl support, the script that builds the
6498 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6499 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6501 6. The freeze_tell option was not being used for messages that were frozen on
6502 arrival, either by an ACL or by local_scan().
6504 7. Added the smtp_incomplete_transaction log selector.
6506 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6507 was accepting AUTH without a new EHLO.
6509 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6510 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6513 10. Logging of TCP/IP connections (when configured) now happens in the main
6514 daemon process instead of the child process, so that the TCP/IP connection
6515 count is more accurate (but it can never be perfect).
6517 11. The use of "drop" in a nested ACL was not being handled correctly in the
6518 outer ACL. Now, if condition failure induced by the nested "drop" causes
6519 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6520 or "require"), the connection is dropped.
6522 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6523 that yield "discard" can now be used with an "accept" or a "discard" verb,
6524 but an error is generated for any others (because I can't see a useful way
6525 to define what should happen).
6527 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6528 are now processed in the same way as lines in the Exim configuration file.
6529 In particular, continuation lines are supported.
6531 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6533 15. Added -ti meaning -t -i.
6535 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6536 domains, and warn by logging if others are found.
6538 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6539 returned from GIFCONF when Exim is trying to find the list of interfaces on
6540 a host. The code in os.c has been modified to copy each ifreq to an aligned
6541 structure in all cases.
6543 Also, in some cases, the returned ifreq's were being copied to a 'struct
6544 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6545 means the last couple of bytes of an IPv6 address could be chopped if the
6546 ifreq contained only a normal sockaddr (14 bytes storage).
6548 18. Named domain lists were not supported in the hosts_treat_as_local option.
6549 An entry such as +xxxx was not recognized, and was treated as a literal
6552 19. Ensure that header lines added by a DATA ACL are included in the reject log
6553 if the ACL subsequently rejects the message.
6555 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6556 MD5 (which is deprecated).
6558 21. When testing a filter file using -bf, Exim was writing a message when it
6559 took the sender from a "From " line in the message, but it was not doing so
6560 when it took $return_path from a Return-Path: header line. It now does.
6562 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6563 with a valid header line field name (a series of printing characters
6564 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6566 23. Changed "disc" in the source to "disk" to conform to the documentation and
6567 the book and for uniformity.
6569 24. Ignore Sendmail's -Ooption=value command line item.
6571 25. When execve() failed while trying to run a command in a pipe transport,
6572 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6573 could be confused with a return value of 69 from the command itself. This
6574 has been changed to 127, the value the shell returns if it is asked to run
6575 a non-existent command. The wording for the related log line suggests a
6576 non-existent command as the problem.
6578 26. If received_header_text expands to an empty string, do not add a Received:
6579 header line to the message. (Well, it adds a token one on the spool, but
6580 marks it "old" so that it doesn't get used or transmitted.)
6582 27. Installed eximstats 1.28 (addition of -nt option).
6584 28. There was no check for failure on the call to getsockname() in the daemon
6585 code. This can fail if there is a shortage of resources on the system, with
6586 ENOMEM, for example. A temporary error is now given on failure.
6588 29. Contrary to the C standard, it seems that in some environments, the
6589 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6590 program. Exim now does this explicitly; it affects the formatting of
6591 timestamps using strftime().
6593 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6594 complaints. I've now initialized all the variables, to avoid this.
6596 32. Header lines added by a system filter were not being "seen" during
6597 transport-time rewrites.
6599 33. The info_callback() function passed to OpenSSL is set up with type void
6600 (*)(SSL *, int, int), as described somewhere. However, when calling the
6601 function (actually a macro) that sets it up, the type void(*)() is
6602 expected. I've put in a cast to prevent warnings from picky compilers.
6604 34. If a DNS black list lookup found a CNAME record, but there were no A
6605 records associated with the domain it pointed at, Exim crashed.
6607 35. If a DNS black list lookup returned more than one A record, Exim ignored
6608 all but the first. It now scans all returned addresses if a particular IP
6609 value is being sought. In this situation, the contents of the
6610 $dnslist_value variable are a list of all the addresses, separated by a
6613 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6614 to accept a host name and all its aliases if the forward lookup for any of
6615 them yielded the IP address of the incoming connection. Now it accepts only
6616 those names whose forward lookup yields the correct IP address. Any other
6617 names are discarded. This closes a loophole whereby a rogue DNS
6618 administrator could create reverse DNS records to break through a
6619 wildcarded host restriction in an ACL.
6621 37. If a user filter or a system filter that ran in a subprocess used any of
6622 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6623 the wrong values were passed to the pipe command ($thisaddress had the
6624 value of $0, $0 had the value of $1, etc). This bug was introduced by
6625 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6627 38. Improved the line breaking for long SMTP error messages from ACLs.
6628 Previously, if there was no break point between 40 and 75 characters, Exim
6629 left the rest of the message alone. Two changes have been made: (a) I've
6630 reduced the minimum length to 35 characters; (b) if it can't find a break
6631 point between 35 and 75 characters, it looks ahead and uses the first one
6632 that it finds. This may give the occasional overlong line, but at least the
6633 remaining text gets split now.
6635 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6636 file descriptors might be low, and that setting 1000 would always raise it.
6637 It turns out that in some environments, the limit is already over 1000 and
6638 that lowering it causes trouble. So now Exim takes care not to decrease it.
6640 40. When delivering a message, the value of $return_path is set to $sender_
6641 address at the start of routing (routers may change the value). By an
6642 oversight, this default was not being set up when an address was tested by
6643 -bt or -bv, which affected the outcome if any router or filter referred to
6646 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6647 log only when "message" or "log_message" are set. However, if one of the
6648 conditions was an address verification, or a call to a nested ACL, the
6649 messages generated by the underlying test were being passed through. This
6650 no longer happens. The underlying message is available in $acl_verify_
6651 message for both "message" and "log_message" expansions, so it can be
6652 passed through if needed.
6654 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6655 new expansion $bh_ to give the encoded byte string without charset
6656 translation. Translation happens only if iconv() is available; HAVE_ICONV
6657 indicates this at build time. HEADERS_CHARSET gives the charset to
6658 translate to; headers_charset can change it in the configuration, and
6659 "headers charset" can change it in an individual filter file.
6661 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6662 that creates RFC 2047 encoded "words" labels them as that charset instead
6663 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6664 expansion operator; (ii) when Exim creates a From: line for a local
6665 message; (iii) when a header line is rewritten to include a "phrase" part.
6667 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6668 buggy, causing it to skip the first lines of messages whose message ID
6669 ended in 'D'. This would not have bitten before Exim release 4.14, because
6670 message IDs were unlikely to end in 'D' before then. The effect was to have
6671 incorrect size information for certain domains.
6673 45. #include "config.h" was missing at the start of the crypt16.c module. This
6674 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6677 46. If there was a timeout during a "random" callout check, Exim treated it as
6678 a failure of the random address, and carried on sending RSET and the real
6679 address. If the delay was just some slowness somewhere, the response to the
6680 original RCPT would be taken as a response to RSET and so on, causing
6681 mayhem of various kinds.
6683 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6684 when I implemented it. It didn't allow for the fact that some option values
6685 may legitimately be negative (e.g. size_addition), and it didn't even do
6686 the right test for positive values.
6688 48. Domain names in DNS records are case-independent. Exim always looks them up
6689 in lower case. Some resolvers return domain names in exactly the case they
6690 appear in the zone file, that is, they may contain uppercase letters. Not
6691 all resolvers do this - some return always lower case. Exim was treating a
6692 change of case by a resolver as a change of domain, similar to a widening
6693 of a domain abbreviation. This triggered its re-routing code and so it was
6694 trying to route what was effectively the same domain again. This normally
6695 caused routing to fail (because the router wouldn't handle the domain
6696 twice). Now Exim checks for this case specially, and just changes the
6697 casing of the domain that it ultimately uses when it transmits the message
6700 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6703 50. If a filter generated a file delivery with a non-absolute name (possible if
6704 no home directory exists for the router), the forbid_file option was not
6707 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6708 the existing equality matching.
6710 52. Exim was using ints instead of ino_t variables in some places where it was
6711 dealing with inode numbers.
6713 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6714 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6715 called TMPDIR, and if it finds it is different, it changes its value.
6717 54. The smtp_printf() function is now made available to local_scan() so
6718 additional output lines can be written before returning. There is also an
6719 smtp_fflush() function to enable the detection of a dropped connection.
6720 The variables smtp_input and smtp_batched_input are exported to
6723 55. Changed the default runtime configuration: the message "Unknown user"
6724 has been removed from the ACL, and instead placed on the localuser router,
6725 using the cannot_route_message feature. This means that any verification
6726 failures that generate their own messages won't get overridden. Similarly,
6727 the "Unrouteable address" message that was in the ACL for unverifiable
6728 relay addresses has also been removed.
6730 56. Added hosts_avoid_esmtp to the smtp transport.
6732 57. The exicyclog script was not checking for the esoteric option
6733 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6734 will work only if exicyclog is run under the appropriate euid.
6736 58. Following a discussion on the list, the rules by which Exim recognises line
6737 endings on incoming messages have been changed. The -dropcr and drop_cr
6738 options are now no-ops, retained only for backwards compatibility. The
6739 following line terminators are recognized: LF CRLF CR. However, special
6740 processing applies to CR:
6742 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6743 nor a local message in the state where . is a terminator.
6745 (ii) If a bare CR is encountered in a header line, an extra space is added
6746 after the line terminator so as not to end the header. The reasoning
6747 behind this is that bare CRs in header lines are most likely either
6748 to be mistakes, or people trying to play silly games.
6750 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6751 was being incorrectly given as 18 bytes larger than it should have been.
6752 This is a VOB (very old bug).
6754 60. This may never have affected anything current, but just in case it has:
6755 When the local host is found other than at the start of a list of hosts,
6756 the local host, those with the same MX, and any that follow, are discarded.
6757 When the list in question was part of a longer list of hosts, the following
6758 hosts (not currently being processed) were also being discarded. This no
6759 longer happens. I'm not sure if this situation could ever has previously
6762 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6765 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6766 header. This is something that is recommended in a new Internet Draft, and
6767 is something that is documented as being done by Sendmail. There are two
6768 possible values. For messages generated by the autoreply transport, Exim
6771 Auto-Submitted: auto-replied
6773 whereas for all other generated messages (e.g. bounces) it adds
6775 Auto-Submitted: auto-generated
6777 63. The "personal" condition in filters now includes a test for the
6778 Auto-Submitted: header. If it contains the string "auto-" the message it
6779 not considered personal.
6781 64. Added rcpt_include_affixes as a generic transport option.
6783 65. Added queue_only_override (default true).
6785 66. Added the syslog_duplication option.
6787 67. If what should have been the first header line of a message consisted of
6788 a space followed by a colon, Exim was mis-interpreting it as a header line.
6789 It isn't of course - it is syntactically invalid and should therefore be
6790 treated as the start of the message body. The misbehaviour could have
6791 caused a number of strange effects, including loss of data in subsequent
6792 header lines, and spool format errors.
6794 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6795 client host had authenticated. This control can now be exercised by an ACL
6796 for more flexibility.
6798 69. By default, callouts do not happen when testing with -bh. There is now a
6799 variant, -bhc, which does actually run the callout code, including
6800 consulting and updating the callout cache.
6802 70. Added support for saslauthd authentication, courtesy of Alexander
6805 71. If statvfs() failed on the spool or log directories while checking their
6806 size for availability, Exim confusingly gave the error "space shortage".
6807 Furthermore, in debugging mode it crashed with a floating point exception.
6808 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6809 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6810 serious problem, Exim now writes to the main and panic logs when this
6811 happens, with details of the failure. It then refuses to accept the
6812 incoming message, giving the message "spool directory problem" or "log
6813 directory problem" with a 421 code for SMTP messages.
6815 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6816 0, 1, and 2 exist, because some OS complain for execs without them (see
6817 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6818 descriptors. However, the code omitted to check that the open succeeded,
6819 causing mysterious errors if for some reason the permissions on /dev/null
6820 got screwed. Now Exim writes a message to the main and panic logs, and
6821 bombs out if it can't open /dev/null.
6823 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6824 interact so that it is all more flexible. It is supposed to remain
6825 backwards compatible. Also added extra_local_interfaces.
6827 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6828 to bomb out with an assertion failure - to the client this appears as a
6829 connection drop. This problem occurs in the part of the code that was taken
6830 from the Samba project. Fortunately, the assertion is in a very simple
6831 function, so I have fixed this by reproducing the function inline in the
6832 one place where it is called, and arranging for authentication to fail
6833 instead of killing the process with assert().
6835 75. The SPA client code was not working when the server requested OEM rather
6836 than Unicode encoding.
6838 76. Added code to make require_files with a specific uid setting more usable in
6839 the case where statting the file as root fails - usually a non-root-mounted
6840 NFS file system. When this happens and the failure is EACCES, Exim now
6841 forks a subprocess and does the per-uid checking as the relevant uid.
6843 77. Added process_log_path.
6845 78. If log_file_path was not explicitly set, a setting of check_log_space or
6846 check_log_inodes was ignored.
6848 79. If a space check for the spool or log partitions fails, the incident is now
6849 logged. Of course, in the latter case the data may get lost...
6851 80. Added the %p formatting code to string_format() so that it can be used to
6852 print addresses in debug_print(). Adjusted all the address printing in the
6853 debugging in store.c to use %p rather than %d.
6855 81. There was a concern that a line of code in smtp_in.c could overflow a
6856 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6857 initially expressed, the concern was not well-founded, because trailing
6858 spaces are removed early. However, if the trailing spaces were followed by
6859 a NULL, they did not get removed, so the overflow was possible. Two fixes
6862 (a) I re-wrote the offending code in a cleaner fashion.
6863 (b) If an incoming SMTP command contains a NULL character, it is rejected
6866 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6867 runs initgroups(), so that if the Exim user is in any additional groups,
6868 they will be used during message reception.
6874 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob