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.
10 HS/01 Handle trailing backslash gracefully. (CVE-2019-15846)
16 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917, OVE-20190718-0006)
22 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
23 buffer overrun for (non-chunking) other transports.
25 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
26 TLS1.3, means that a server rejecting a client certificate is not visible
27 to the client until the first read of encrypted data (typically the
28 response to EHLO). Add detection for that case and treat it as a failed
29 TLS connection attempt, so that the normal retry-in-clear can work (if
32 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
33 and/or domain. Found and fixed by Jason Betts.
35 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
36 configuration). If a CNAME target was not a wellformed name pattern, a
39 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
40 the OS reports them interleaved with other addresses.
42 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
43 used both for input and for a verify callout, both encrypted, SMTP
44 responses being sent by the server could be lost. This resulted in
45 dropped connections and sometimes bounces generated by a peer sending
48 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
49 string. Previously it was possible to cause undefined behaviour in a
50 library routine (usually a crash). Found by "zerons".
52 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
55 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
56 API was removed, so update to use the newer ones.
59 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
60 to changes apparently associated with TLS1.3 handling some of the APIs
61 previously used were either nonfunctional or inappropriate. Strings
62 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
63 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
64 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
65 This affects log line X= elements, the $tls_{in,out}_cipher variables,
66 and the use of specific cipher names in the encrypted= ACL condition.
68 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
69 verification result was not updated unless hosts_require_ocsp applied.
71 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
72 queue_list_requires_admin set to false, non-admin users were denied the
75 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
76 directory-of-certs mode. Previously they were advertised despite the
83 JH/01 Remove code calling the customisable local_scan function, unless a new
84 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
86 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
87 non-signal-safe functions being used.
89 JH/03 Bug 2269: When presented with a received message having a stupidly large
90 number of DKIM-Signature headers, disable DKIM verification to avoid
91 a resource-consumption attack. The limit is set at twenty.
93 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
94 report of oldest_pass in ${authres } in consequence, and separate out
95 some descriptions of reasons for verification fail.
97 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
98 files in the spool were present and unlocked. A queue-runner could spot
99 them, resulting in a duplicate delivery. Fix that by doing the unlock
100 after the unlink. Investigation by Tim Stewart. Take the opportunity to
101 add more error-checking on spoolfile handling while that code is being
104 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
105 No known attacks, no CVE, this is defensive hardening.
107 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
108 a queue-runner could start a delivery while other operations were ongoing.
109 Cutthrough delivery was a common victim, resulting in duplicate delivery.
110 Found and investigated by Tim Stewart. Fix by using the open message data
111 file handle rather than opening another, and not locally closing it (which
112 releases a lock) for that case, while creating the temporary .eml format
113 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
115 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
116 $sender_verify_failure/$recipient_verify_failure to "random".
118 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
121 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
122 Previously this would segfault.
124 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
127 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
128 like zero, since the resolver should be doing this for us, But we need one
129 as a CNAME but no MX presence gets the CNAME returned; we need to check
130 that doesn't point to an MX to declare it "no MX returned" rather than
131 "error, loop". A new main option is added so the older capability of
132 following some limited number of chain links is maintained.
134 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
136 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
137 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
138 now-deprecated earlier definitions used only specified the range up to TLS
139 1.2 (in the older-version library docs).
141 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
143 JH/15 Rework TLS client-side context management. Stop using a global, and
144 explicitly pass a context around. This enables future use of TLS for
145 connections to service-daemons (eg. malware scanning) while a client smtp
146 connection is using TLS; with cutthrough connections this is quite likely.
148 JH/16 Fix ARC verification to do AS checks in reverse order.
150 JH/17 Support a "tls" option on the ${readsocket } expansion item.
152 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
153 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
154 Previously the "utf8" would be re-prepended for every additional message.
156 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
157 Previously thery were accepted, resulting in issues when attempting to
158 forward messages to a non-supporting MTA.
160 PP/02 Let -n work with printing macros too, not just options.
162 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
163 one parent address was copied, and bogus data was used at delivery-logging
164 time. Either a crash (after delivery) or bogus log data could result.
165 Discovery and analysis by Tim Stewart.
167 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
168 Previously if the string ended mid-character, we did not insert the
169 promised '?' replacement.
171 PP/04 Documentation: current string operators work on bytes, not codepoints.
173 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
174 should pack well giving a smaller memory footprint so better caching and
175 therefore performance. Group the declarations where this can't be done so
176 that the byte-sized flag variables are not interspersed among pointer
177 variables, giving a better chance of good packing by the compiler.
179 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
180 non-null, to avoid issues with sites running BATV. Previously reports were
181 sent with an empty envelope sender so looked like bounces.
183 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
184 The ignore_error flag wasn't being returned from the filter subprocess so
185 was not set for later routers. Investigation and fix by Matthias Kurz.
187 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
188 and a msg:complete for the whole, when a message is manually removed using
189 -Mrm. Developement by Matthias Kurz, hacked on by JH.
191 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
192 a "Gnu special" function, asprintf() in the DB utility binary builds; I
193 hope that is portable enough.
195 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
196 requiring a known-CA anchor certificate; make it now rely entirely on the
197 TLSA as an anchor. Checking the name on the leaf cert against the name
198 on the A-record for the host is still done for TA (but not for EE mode).
200 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
201 included in delivery lines for non-proxied connections, when compiled with
202 SUPPORT_SOCKS and running with proxy logging enabled.
204 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
205 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
206 move the existing event to fire before the normal logging of message
207 failure so that custom logging is bracketed by normal logging.
209 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
210 msg:fail:internal event. Developement by Matthias Kurz.
212 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
213 far too small for todays use of crypto signatures stored there. Go all
214 the way to the max DNS message size of 64kB, even though this might be
215 overmuch for IOT constrained device use.
217 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
218 copy a string over itself. The library routine is documented as not
219 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
221 JH/32 For main options check_spool_space and check_inode_space, where the
222 platform supports 64b integers, support more than the previous 2^31 kB
223 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
224 the previous G, M, k.
226 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
227 $authenticated_fail_id variable on authentication failure. Previously
230 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
231 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
232 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
233 GNUTLS_SEC_PARAM_MEDIUM.
235 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
236 side. Previously we would continue as if no SNI had been received.
238 JH/36 Harden the handling of string-lists. When a list consisted of a sole
239 "<" character, which should be a list-separator specification, we walked
240 off past the nul-terimation.
242 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
243 causes) even when the retry time is not yet met. Previously they were
244 not, meaning that when (say) an account was over-quota and temp-rejecting,
245 and multiple senders' messages were queued, only one sender would get
246 notified on each configured delay_warning cycle.
248 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
250 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
251 cipher-suites, an error can be left on the stack even for a succeeding
252 accept; this results in impossible error messages when a later operation
255 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
256 return error codes indicating retry. Under TLS1.3 this becomes required.
258 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
259 it only wrote the new authenticators, resulting in a lack of tracking of
260 peer changes of ESMTP extensions until the next cache flush.
262 JH/41 Fix the loop reading a message header line to check for integer overflow,
263 and more-often against header_maxsize. Previously a crafted message could
264 induce a crash of the recive process; now the message is cleanly rejected.
266 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
267 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
273 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
274 When redis_servers is set to a list of > 1 element, and the Redis servers
275 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
276 case of MOVED into a DEFER case instead, thus moving the query onto the
277 next server in the list. For a cluster of N elements, all N servers must
278 be defined in redis_servers.
280 GF/02 Catch and remove uninitialized value warning in exiqsumm
281 Check for existence of @ARGV before looking at $ARGV[0]
283 JH/01 Replace the store_release() internal interface with store_newblock(),
284 which internalises the check required to safely use the old one, plus
285 the allocate and data copy operations duplicated in both (!) of the
286 extant use locations.
288 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
289 modifier. This matches the restriction on the commandline.
291 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
292 Previously only the last row was returned.
294 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
295 we assumed that tags in the header were well-formed, and parsed the
296 element content after inspecting only the first char of the tag.
297 Assumptions at that stage could crash the receive process on malformed
300 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
301 While running the DKIM ACL we operate on the Permanent memory pool so that
302 variables created with "set" persist to the DATA ACL. Also (at any time)
303 DNS lookups that fail create cache records using the Permanent pool. But
304 expansions release any allocations made on the current pool - so a dnsdb
305 lookup expansion done in the DKIM ACL releases the memory used for the
306 DNS negative-cache, and bad things result. Solution is to switch to the
307 Main pool for expansions.
308 While we're in that code, add checks on the DNS cache during store_reset,
309 active in the testsuite.
310 Problem spotted, and debugging aided, by Wolfgang Breyha.
312 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
313 When none of the hosts presented to a transport match an already-open
314 connection, close it and proceed with the list. Previously we would
315 queue the message. Spotted by Lena with Yahoo, probably involving
318 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
319 Previously a spurious "250 OK id=" response was appended to the proper
322 JH/08 The "support for" informational output now, which built with Content
323 Scanning support, has a line for the malware scanner interfaces compiled
324 in. Interface can be individually included or not at build time.
326 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
327 by the template makefile "src/EDITME". The "STREAM" support for an older
328 ClamAV interface method is removed.
330 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
331 rows affected is given instead).
333 JH/11 The runtime Berkeley DB library version is now additionally output by
334 "exim -d -bV". Previously only the compile-time version was shown.
336 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
337 SMTP connection. Previously, when one had more receipients than the
338 first, an abortive onward connection was made. Move to full support for
339 multiple onward connections in sequence, handling cutthrough connection
340 for all multi-message initiating connections.
342 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
343 routers. Previously, a multi-recipient message would fail to match the
344 onward-connection opened for the first recipient, and cause its closure.
346 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
347 a timeout on read on a GnuTLS initiating connection, resulting in the
348 initiating connection being dropped. This mattered most when the callout
349 was marked defer_ok. Fix to keep the two timeout-detection methods
352 JH/15 Relax results from ACL control request to enable cutthrough, in
353 unsupported situations, from error to silently (except under debug)
354 ignoring. This covers use with PRDR, frozen messages, queue-only and
357 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
359 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
360 metadata, resulting in a crash in free().
362 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
363 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
364 Broken also in d185889f4, with init system revamp.
366 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
367 Previously we abruptly closed the connection after reading a malware-
368 found indication; now we go on to read the "scan ok" response line,
371 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
372 ACL. Previously, a crash would result.
374 JH/19 Speed up macro lookups during configuration file read, by skipping non-
375 macro text after a replacement (previously it was only once per line) and
376 by skipping builtin macros when searching for an uppercase lead character.
378 JH/20 DANE support moved from Experimental to mainline. The Makefile control
379 for the build is renamed.
381 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
382 was allocated for every new TLS startup, meaning one per message. Fix
383 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
385 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
386 reported the original. Fix to report (as far as possible) the ACL
387 result replacing the original.
389 JH/23 Fix memory leak during multi-message connections using STARTTLS under
390 OpenSSL. Certificate information is loaded for every new TLS startup,
391 and the resources needed to be freed.
393 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
395 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
398 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
399 DATA response info to the (existing) per-recipient response info for
400 the "C=" log element. It can have useful tracking info from the
401 destination system. Patch from Simon Arlott.
403 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
404 length value. Previously this would segfault.
406 HS/02 Support Avast multiline protoocol, this allows passing flags to
407 newer versions of the scanner.
409 JH/28 Ensure that variables possibly set during message acceptance are marked
410 dead before release of memory in the daemon loop. This stops complaints
411 about them when the debug_store option is enabled. Discovered specifically
412 for sender_rate_period, but applies to a whole set of variables.
413 Do the same for the queue-runner and queue-list loops, for variables set
414 from spool message files. Do the same for the SMTP per-message loop, for
415 certain variables indirectly set in ACL operations.
417 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
418 as a multi-recipient message from a mailinglist manager). The coding had
419 an arbitrary cutoff number of characters while checking for more input;
420 enforced by writing a NUL into the buffer. This corrupted long / fast
421 input. The problem was exposed more widely when more pipelineing of SMTP
422 responses was introduced, and one Exim system was feeding another.
423 The symptom is log complaints of SMTP syntax error (NUL chars) on the
424 receiving system, and refused recipients seen by the sending system
425 (propating to people being dropped from mailing lists).
426 Discovered and pinpointed by David Carter.
428 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
429 replaced by the ${authresults } expansion.
431 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
433 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
434 allows proper process termination in container environments.
436 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
437 Previously the "final dot" had a newline after it; ensure it is CR,LF.
439 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
440 and "err_perm", deprecated since 4.83 when the RFC-defined words
441 "temperror" and "permerror" were introduced.
443 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
444 transport-filters or DKIM-signing. The restriction was lost in the
445 consolidation of verify-callout and delivery SMTP handling.
446 Extend the restriction to also cover ARC-signing.
448 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
449 in defer=pass mode supply a 450 to the initiator. Previously the message
452 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
453 tls_require_ciphers is used as before.
455 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
456 "pass_unscanned". Only tmpfails from the scanner are written to
457 the paniclog, as they may require admin intervention (permission
458 denied, license issues). Other scanner errors (like decompression
459 bombs) do not cause a paniclog entry.
461 JH/36 Fix reinitialisation of DKIM logging variable between messages.
462 Previously it was possible to log spurious information in receive log
465 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
466 triggered odd behaviour from Outlook Express clients.
468 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
471 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
472 since the IETF WG has not yet settled on that versus the original
473 "bare" representation.
475 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
476 Previously the millisecond value corrupted the output.
477 Fix also for syslog_pid=no and log_selector +pid, for which the pid
478 corrupted the output.
484 JH/01 Rework error string handling in TLS interface so that the caller in
485 more cases is responsible for logging. This permits library-sourced
486 string to be attached to addresses during delivery, and collapses
487 pairs of long lines into single ones.
489 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
490 during configuration. Wildcards are allowed and expanded.
492 JH/02 Rework error string handling in DKIM to pass more info back to callers.
493 This permits better logging.
495 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
496 do not close it down and have the child transport start it up again on
497 the passed-on TCP connection. Instead, proxy the child (and any
498 subsequent ones) for TLS via a unix-domain socket channel. Logging is
499 affected: the continued delivery log lines do not have any DNSSEC, TLS
500 Certificate or OCSP information. TLS cipher information is still logged.
502 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
503 identical IP addresses on different listening ports. Will also affect
506 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
507 add noisy ifdef guards to special-case this sillyness.
508 Patch from Bernd Kuhls.
510 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
511 than 255 are no longer allowed.
513 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
514 Disable the session-cache too, which might reduce our load. Since we
515 currrectly use a new context for every connection, both as server and
516 client, there is no benefit for these.
517 GnuTLS appears to not support tickets server-side by default (we don't
518 call gnutls_session_ticket_enable_server()) but client side is enabled
519 by default on recent versions (3.1.3 +) unless the PFS priority string
522 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
523 <https://reproducible-builds.org/specs/source-date-epoch/>.
525 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
526 the check for any unsuccessful recipients did not notice the limit, and
527 erroneously found still-pending ones.
529 JH/08 Pipeline CHUNKING command and data together, on kernels that support
530 MSG_MORE. Only in-clear (not on TLS connections).
532 JH/09 Avoid using a temporary file during transport using dkim. Unless a
533 transport-filter is involved we can buffer the headers in memory for
534 creating the signature, and read the spool data file once for the
535 signature and again for transmission.
537 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
538 4.77 as the kernel support then wasn't solid, having issues in 64bit
539 mode. Now, it's been long enough. Add support for FreeBSD also.
541 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
542 case where the routing stage had gathered several addresses to send to
543 a host before calling the transport for the first, we previously failed
544 to close down TLS in the old transport process before passing the TCP
545 connection to the new process. The new one sent a STARTTLS command
546 which naturally failed, giving a failed delivery and bloating the retry
547 database. Investigation and fix prototype from Wolfgang Breyha.
549 JH/12 Fix check on SMTP command input synchronisation. Previously there were
550 false-negatives in the check that the sender had not preempted a response
551 or prompt from Exim (running as a server), due to that code's lack of
552 awareness of the SMTP input buffering.
554 PP/04 Add commandline_checks_require_admin option.
555 Exim drops privileges sanely, various checks such as -be aren't a
556 security problem, as long as you trust local users with access to their
557 own account. When invoked by services which pass untrusted data to
558 Exim, this might be an issue. Set this option in main configuration
559 AND make fixes to the calling application, such as using `--` to stop
562 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
563 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
564 responses to those, into a single TLS record each way (this usually means
565 a single packet). As a side issue, smtp_enforce_sync now works on TLS
568 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
569 affects you only if you're dancing at the edge of the param size limits.
570 If you are, and this message makes sense to you, then: raise the
571 configured limit or use OpenSSL 1.1. Nothing we can do for older
574 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
575 cmdline element to get the documented default one. Previously it was
578 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
579 are unsuable for non-message-specific reasons - eg. connection timeout,
580 banner-time rejection.
582 JH/16 Fix logging of delivery remote port, when specified by router, under
585 PP/06 Repair manualroute's ability to take options in any order, even if one
586 is the name of a transport.
589 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
591 JH/17 Change the list-building routines interface to use the expanding-string
592 triplet model, for better allocation and copying behaviour.
594 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
595 Previously it was constructed the first time a possibly-matching string
596 was met in the configuration file input during startup; now it is done
599 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
600 compatible one, to avoid the (poorly documented) possibility of a config
601 file in the working directory redirecting the DB files, possibly correpting
602 some existing file. CVE-2017-10140 assigned for BDB.
604 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
605 cache-hot. Previously, although the result was properly cached, the
606 initial verify call returned a defer.
608 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
609 the main verify for receipient in uncached-mode.
611 JH/22 Retire historical build files to an "unsupported" subdir. These are
612 defined as "ones for which we have no current evidence of testing".
614 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
615 if present. Previously it was ignored.
617 JH/24 Start using specified-initialisers in C structure init coding. This is
618 a C99 feature (it's 2017, so now considered safe).
620 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
621 if was a fixed-sized field and bitmask ops via macros; it is now more
624 PP/07 GitHub PR 56: Apply MariaDB build fix.
625 Patch provided by Jaroslav Å karvada.
627 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
628 during Coverity cleanups [4.87 JH/47]
629 Diagnosis and fix provided by Michael Fischer v. Mollard.
631 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
632 the right size to place the terminating semicolon on its own folded
633 line, the header hash was calculated to an incorrect value thanks to
634 the (relaxed) space the fold became.
636 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
637 and confused the parent.
639 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
640 which could crash as a result. This could lead to undeliverable messages.
642 JH/28 Logging: "next input sent too soon" now shows where input was truncated
645 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
646 matters on fast-turnover and PID-randomising systems, which were getting
647 out-of-order delivery.
649 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
650 a possibly-overlapping copy. The symptom was that "Remote host closed
651 connection in response to HELO" was logged instead of the actual 4xx
654 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
655 Previously only that bufferd was discarded, resulting in SYMTP command
658 JH/32 DKIM: when a message has multiple signatures matching an identity given
659 in dkim_verify_signers, run the dkim acl once for each. Previously only
660 one run was done. Bug 2189.
662 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
663 "panic the current process" to "deliberately defer". The panic log is
664 still written with the problem list name; the mail and reject logs now
665 get a temp-reject line for the message that was being handled, saying
666 something like "domains check lookup or other defer". The SMTP 451
667 message is still "Temporary local problem".
669 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
670 A crafted sequence of BDAT commands could result in in-use memory beeing
671 freed. CVE-2017-16943.
673 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
674 from SMTP input. Previously it was always done; now only done for DATA
675 and not BDAT commands. CVE-2017-16944.
677 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
678 to the message (such as an overlong header line). Previously this was
679 not done and we did not exit BDAT mode. Followon from the previous item
680 though a different problem.
686 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
687 than -2003 did; needs libidn2 in addition to libidn.
689 JH/02 The path option on a pipe transport is now expanded before use.
691 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
692 Patch provided by "Björn", documentation fix added too.
694 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
695 missing a wire-to-host endian conversion.
697 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
698 close after a BDAT command line could be taken as a following command,
699 giving a synch failure. Fix by only checking for synch immediately
700 before acknowledging the chunk.
702 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
703 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
704 macro. Patches provided by Josh Soref.
706 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
707 Previously we did not; the RFC seems ambiguous and VRFY is not listed
708 by IANA as a service extension. However, John Klensin suggests that we
711 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
712 the dkim code may be unix-mode line endings rather than smtp wire-format
713 CRLF, so prepend a CR to any bare LF.
715 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
716 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
718 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
719 the first were themselves being wrongly included in the feed into dkim
720 processing; with most chunk sizes in use this resulted in an incorrect
721 body hash calculated value.
723 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
724 DKIM signature block, for verification. Although advised against by
725 standards it is specifically not ruled illegal.
727 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
729 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
730 missing a body hash (the bh= tag).
732 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
733 It seems that HAProxy sends the Proxy Protocol information in clear and
734 only then does a TLS startup, so do the same.
736 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
737 TCP connections (such as for Spamd) unless the daemon successfully set
738 Fast Open mode on its listening sockets. This fixes breakage seen on
739 too-old kernels or those not configured for Fast Open, at the cost of
740 requiring both directions being enabled for TFO, and TFO never being used
741 by non-daemon-related Exim processes.
743 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
744 endings, at least on the first header line. Try to canonify any that get
745 past that check, despite the cost.
747 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
748 now limited to an arbitrary five deep, while parsing addresses with the
749 strip_excess_angle_brackets option enabled.
751 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
752 instead leave the unprompted TLS handshake in socket buffer for the
753 TLS library to consume.
755 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
757 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
759 JH/16 Drop variables when they go out of scope. Memory management drops a whole
760 region in one operation, for speed, and this leaves assigned pointers
761 dangling. Add checks run only under the testsuite which checks all
762 variables at a store-reset and panics on a dangling pointer; add code
763 explicitly nulling out all the variables discovered. Fixes one known
764 bug: a transport crash, where a dangling pointer for $sending_ip_address
765 originally assigned in a verify callout, is re-used.
767 PP/06 Drop '.' from @INC in various Perl scripts.
769 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
771 PP/08 Reduce a number of compilation warnings under clang; building with
772 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
773 should be warning-free.
775 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
777 HS/01 Fix portability problems introduced by PP/08 for platforms where
778 realloc(NULL) is not equivalent to malloc() [SunOS et al].
780 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
781 chunk. This allows us to accept broken chunked messages. We need a more
782 general solution here.
784 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
785 already-broken messages in the queue.
787 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
789 JH/19 Fix reference counting bug in routing-generated-address tracking.
795 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
796 supports it and a size is available (ie. the sending peer gave us one).
798 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
799 years of being deprecated). The replacements are the ACLs
800 acl_smtp_mime and acl_not_smtp_mime.
802 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
803 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
804 or even an in-clear connection were permitted. Now, if the host lookup
805 was dnssec and dane was requested then the host is only used if the
806 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
807 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
808 if one fails this test.
809 This means that a poorly-configured remote DNS will make it incommunicado;
810 but it protects against a DNS-interception attack on it.
812 JH/04 Bug 1810: make continued-use of an open smtp transport connection
813 non-noisy when a race steals the message being considered.
815 JH/05 If main configuration option tls_certificate is unset, generate a
816 self-signed certificate for inbound TLS connections.
818 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
819 in rewrites and routers.
821 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
822 and logged a warning sing 4.83; now they are a configuration file error.
824 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
825 (lacking @domain). Apply the same qualification processing as RCPT.
827 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
829 JH/10 Support ${sha256:} applied to a string (as well as the previous
832 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
833 a cutthrough deliver is pending, as we always want to make a connection.
834 This also avoids re-routing the message when later placing the cutthrough
835 connection after a verify cache hit.
836 Do not update it with the verify result either.
838 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
839 when routing results in more than one destination address.
841 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
842 signing (which inhibits the cutthrough capability). Previously only
843 the presence of an option was tested; now an expansion evaluating as
844 empty is permissible (obviously it should depend only on data available
845 when the cutthrough connection is made).
847 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
848 the relevant preceding SMTP command did not note the pipelining mode.
850 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
851 Previously they were not counted.
853 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
854 as one having no matching records. Previously we deferred the message
855 that needed the lookup.
857 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
858 distinguished as "(=".
860 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
861 for missing MX records. Previously it only worked for missing A records.
863 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
865 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
866 after the data-go-ahead and data-ack. Patch from Jason Betts.
868 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
869 even for a "none" policy. Patch from Tony Meyer.
871 JH/22 Fix continued use of a connection for further deliveries. If a port was
872 specified by a router, it must also match for the delivery to be
875 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
876 When one of the recipients of a message was unsuitable for the connection
877 (has no matching addresses), we lost track of needing to mark it
878 deferred. As a result mail would be lost.
880 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
882 JH/25 Decoding ACL controls is now done using a binary search; the source code
883 takes up less space and should be simpler to maintain. Merge the ACL
884 condition decode tables also, with similar effect.
886 JH/26 Fix problem with one_time used on a redirect router which returned the
887 parent address unchanged. A retry would see the parent address marked as
888 delivered, so not attempt the (identical) child. As a result mail would
891 JH/27 Fix a possible security hole, wherein a process operating with the Exim
892 UID can gain a root shell. Credit to http://www.halfdog.net/ for
893 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
896 JH/28 Enable {spool,log} filesystem space and inode checks as default.
897 Main config options check_{log,spool}_{inodes,space} are now
898 100 inodes, 10MB unless set otherwise in the configuration.
900 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
901 Previously it only applied to the main-section connection policy
904 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
906 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
907 by me. Added RFC7919 DH primes as an alternative.
909 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
910 are not in the system include path.
912 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
913 GnuTLS, when a session startup failed (eg because the client disconnected)
914 Exim did stdio operations after fclose. This was exposed by a recent
915 change which nulled out the file handle after the fclose.
917 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
918 signed directly by the cert-signing cert, rather than an intermediate
919 OCSP-signing cert. This is the model used by LetsEncrypt.
921 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
923 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
924 an incoming connection.
926 HS/02 Bug 1802: Do not half-close the connection after sending a request
929 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
930 fallback to "prime256v1".
932 JH/34 SECURITY: Use proper copy of DATA command in error message.
933 Could leak key material. Remotely exploitable. CVE-2016-9963.
939 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
940 and 3.4.4 - once the server is enabled to respond to an OCSP request
941 it does even when not requested, resulting in a stapling non-aware
942 client dropping the TLS connection.
944 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
945 support variable-length bit vectors. No functional change.
947 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
948 The I= interface field on outgoing lines is now after the H= remote
949 host field, same as incoming lines. There is a separate
950 outgoing_interface log selector which allows you to disable the
953 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
954 If not running log_selector +smtp_connection the mainlog would be held
955 open indefinitely after a "too many connections" event, including to a
956 deleted file after a log rotate. Leave the per net connection logging
957 leaving it open for efficiency as that will be quickly detected by the
958 check on the next write.
960 HS/01 Bug 1671: Fix post transport crash.
961 Processing the wait-<transport> messages could crash the delivery
962 process if the message IDs didn't exist for some reason. When
963 using 'split_spool_directory=yes' the construction of the spool
964 file name failed already, exposing the same netto behaviour.
966 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
967 mime_regex ACL conditions.
969 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
970 to DSN fail messages (bounces): remote IP, remote greeting, remote response
971 to HELO, local diagnostic string.
973 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
974 log line to debug. Even when configured with a tls authenticator many
975 client connections are expected to not authenticate in this way, so
976 an authenticate fail is not an error.
978 HS/02 Add the Exim version string to the process info. This way exiwhat
979 gives some more detail about the running daemon.
981 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
982 matter for fast-change records such as DNSBLs.
984 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
985 retry record, even if constant. There may be multiple transports with
986 different interface settings and the retry behaviour needs to be kept
989 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
991 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
993 JH/10 Bug 840: fix log_defer_output option of pipe transport
995 JH/11 Bug 830: use same host for all RCPTS of a message, even under
996 hosts_randomize. This matters a lot when combined with mua_wrapper.
998 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
999 ${quote_pgsql:<string>} operator.
1001 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1003 JH/14 Change header file name for freeradius-client. Relevant if compiling
1004 with Radius support; from the Gentoo tree and checked under Fedora.
1006 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1008 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1009 option transport_filter, meaning no filtering.
1011 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1013 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1014 defaults to "*" (all hosts). The variable is now available when not built
1015 with TLS, default unset, mainly to enable keeping the testsuite sane.
1016 If a server certificate is not supplied (via tls_certificate) an error is
1017 logged, and clients will find TLS connections fail on startup. Presumably
1018 they will retry in-clear.
1019 Packagers of Exim are strongly encouraged to create a server certificate
1020 at installation time.
1022 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1023 with the $config_file variable.
1025 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1026 in transport context, after the attempt, and per-recipient. The latter type
1027 is per host attempted. The event data is the error message, and the errno
1028 information encodes the lookup type (A vs. MX) used for the (first) host,
1029 and the trailing two digits of the smtp 4xx response.
1031 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1032 to write to mainlog (or rejectlog, paniclog) in the window between file
1033 creation and permissions/ownership being changed. Particularly affects
1034 installations where exicyclog is run as root, rather than exim user;
1035 result is that the running daemon panics and dies.
1037 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1039 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1040 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1041 "pri" and "weight". Note that the previous implicit priority given by the
1042 list order is no longer honoured.
1044 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1045 for DKIM processing.
1047 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1048 by defining SUPPORT_SOCKS.
1050 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1051 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1052 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1053 variables are renamed to proxy_{local,external}_{address,port}.
1055 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1056 for a build by defining SUPPORT_I18N
1058 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1059 of the query string, and make ${quote_redis:} do that quoting.
1061 JH/29 Move Events support from Experimental to mainline, enabled by default
1062 and removable for a build by defining DISABLE_EVENT.
1064 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1066 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1067 cached by the daemon.
1069 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1070 by defining LOOKUP_REDIS. The libhiredis library is required.
1072 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1073 keys are given for lookup.
1075 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1076 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1077 only supported when built with TLS support. The PolarSSL SHA routines
1078 are still used when the TLS library is too old for convenient support.
1080 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1081 openssl_options), for security. OpenSSL forces this from version 1.1.0
1082 server-side so match that on older versions.
1084 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1085 allocation for $value could be released as the expansion processing
1086 concluded, but leaving the global pointer active for it.
1088 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1089 and to use the domains and local_parts ACL conditions.
1091 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1092 incorrectly not doubled on cutthrough transmission, hence seen as a
1093 body-termination at the receiving system - resulting in truncated mails.
1094 Commonly the sender saw a TCP-level error, and retransmitted the message
1095 via the normal store-and-forward channel. This could result in duplicates
1096 received - but deduplicating mailstores were liable to retain only the
1097 initial truncated version.
1099 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1101 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1103 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1104 we're in there, support oversigning also; bug 1309.
1106 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1108 HS/04 Add support for keep_environment and add_environment options.
1110 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1111 either intentional arithmetic overflow during PRNG, or testing config-
1114 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1115 delivery resulted in actual delivery. Cancel cutthrough before DATA
1118 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1119 rejects a recipient: pass the reject to the originator.
1121 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1122 Many were false-positives and ignorable, but it's worth fixing the
1125 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1126 for the new environment-manipulation done at startup. Move the routines
1127 from being local to tls.c to being global via the os.c file.
1129 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1130 an extract embedded as result-arg for a map, the first arg for extract
1131 is unavailable so we cannot tell if this is a numbered or keyed
1132 extraction. Accept either.
1138 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1141 JH/02 The smtp transport option "multi_domain" is now expanded.
1143 JH/03 The smtp transport now requests PRDR by default, if the server offers
1146 JH/04 Certificate name checking on server certificates, when exim is a client,
1147 is now done by default. The transport option tls_verify_cert_hostnames
1148 can be used to disable this per-host. The build option
1149 EXPERIMENTAL_CERTNAMES is withdrawn.
1151 JH/05 The value of the tls_verify_certificates smtp transport and main options
1152 default to the word "system" to access the system default CA bundle.
1153 For GnuTLS, only version 3.0.20 or later.
1155 JH/06 Verification of the server certificate for a TLS connection is now tried
1156 (but not required) by default. The verification status is now logged by
1157 default, for both outbound TLS and client-certificate supplying inbound
1160 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1163 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1164 Status Notification (bounce) messages are now MIME format per RFC 3464.
1165 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1166 under the control of the dsn_advertise_hosts option, and routers may
1167 have a dsn_lasthop option.
1169 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1170 default, modifiable by a malware= option. The list separator for
1171 the options can now be changed in the usual way. Bug 68.
1173 JH/10 The smtp_receive_timeout main option is now expanded before use.
1175 JH/11 The incoming_interface log option now also enables logging of the
1176 local interface on delivery outgoing connections.
1178 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1179 if the interface and destination host and port all match.
1181 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1184 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1185 Patch from Andrew Lewis.
1187 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1188 now supports optional time-restrictions, weighting, and priority
1189 modifiers per server. Patch originally by <rommer@active.by>.
1191 JH/16 The spamd_address main option now supports a mixed list of local
1192 and remote servers. Remote servers can be IPv6 addresses, and
1193 specify a port-range.
1195 JH/17 Bug 68: The spamd_address main option now supports an optional
1196 timeout value per server.
1198 JH/18 Bug 1581: Router and transport options headers_add/remove can
1199 now have the list separator specified.
1201 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1204 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1207 JH/21 Support for the A6 type of dns record is withdrawn.
1209 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1210 rather than the verbs used.
1212 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1213 from 255 to 1024 chars.
1215 JH/24 Verification callouts now attempt to use TLS by default.
1217 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1218 are generic router options now. The defaults didn't change.
1220 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1221 Original patch from Alexander Shikoff, worked over by JH.
1223 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1224 files if ZCAT_COMMAND is not executable.
1226 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1228 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1230 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1231 Normally benign, it bites when the pair was led to by a CNAME;
1232 modern usage is to not canonicalize the domain to a CNAME target
1233 (and we were inconsistent anyway for A-only vs AAAA+A).
1235 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1237 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1238 when evaluating $sender_host_dnssec.
1240 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1241 $sender_helo_dnssec variable.
1243 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1245 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1247 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1249 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1250 documented as working, but never had. Support all but $spam_report.
1252 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1253 added for tls authenticator.
1255 HS/03 Add perl_taintmode main config option
1261 TL/01 When running the test suite, the README says that variables such as
1262 no_msglog_check are global and can be placed anywhere in a specific
1263 test's script, however it was observed that placement needed to be near
1264 the beginning for it to behave that way. Changed the runtest perl
1265 script to read through the entire script once to detect and set these
1266 variables, reset to the beginning of the script, and then run through
1267 the script parsing/test process like normal.
1269 TL/02 The BSD's have an arc4random API. One of the functions to induce
1270 adding randomness was arc4random_stir(), but it has been removed in
1271 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1272 function when detected.
1274 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1275 cause callback expansion.
1277 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1278 syntax errors in an expansion can be treated as a string instead of
1279 logging or causing an error, due to the internal use of bool_lax
1280 instead of bool when processing it.
1282 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1283 server certificates when making smtp deliveries.
1285 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1287 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1289 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1291 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1292 Merged patch from Sebastian Wiedenroth.
1294 JH/05 Fix results-pipe from transport process. Several recipients, combined
1295 with certificate use, exposed issues where response data items split
1296 over buffer boundaries were not parsed properly. This eventually
1297 resulted in duplicates being sent. This issue only became common enough
1298 to notice due to the introduction of connection certificate information,
1299 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1301 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1302 size buffer was used, resulting in syntax errors when an expansion
1305 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1306 version 3.3.6 or later.
1308 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1309 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1310 both become "event_action", the variables become $event_name, $event_data
1311 and $event_defer_errno. There is a new variable $verify_mode, usable in
1312 routers, transports and related events. The tls:cert event is now also
1313 raised for inbound connections, if the main configuration event_action
1316 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1317 early OCSP support, but no stapling (appears to be less than 1.0.0).
1319 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1320 server certificate names available under the smtp transport option
1321 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1324 JH/10 Time-related extraction expansions from certificates now use the main
1325 option "timezone" setting for output formatting, and are consistent
1326 between OpenSSL and GnuTLS compilations. Bug 1541.
1328 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1329 encoded parameter in the incoming message. Bug 1558.
1331 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1332 include certificate info, eximon was claiming there were spoolfile
1335 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1337 JH/14 Log delivery-related information more consistently, using the sequence
1338 "H=<name> [<ip>]" wherever possible.
1340 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1341 are problematic for Debian distribution, omit them from the release
1344 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1346 JH/16 Fix string representation of time values on 64bit time_t architectures.
1349 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1350 output list separator was used.
1355 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1356 checkers that were complaining about end of non-void function with no
1359 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1360 This was a regression introduced in 4.83 by another bugfix.
1362 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1364 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1365 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1371 TF/01 Correctly close the server side of TLS when forking for delivery.
1373 When a message was received over SMTP with TLS, Exim failed to clear up
1374 the incoming connection properly after forking off the child process to
1375 deliver the message. In some situations the subsequent outgoing
1376 delivery connection happened to have the same fd number as the incoming
1377 connection previously had. Exim would try to use TLS and fail, logging
1378 a "Bad file descriptor" error.
1380 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1381 utilities have not been installed.
1383 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1384 temporary space as the ACL may create new global variables.
1386 TL/01 LDAP support uses per connection or global context settings, depending
1387 upon the detected version of the libraries at build time.
1389 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1390 to extract and use the src ip:port in logging and expansions as if it
1391 were a direct connection from the outside internet. PPv2 support was
1392 updated based on HAProxy spec change in May 2014.
1394 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1396 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1397 Properly escape header and check for NULL return.
1399 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1402 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1404 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1405 characters in header names, implemented as a verify condition.
1406 Contributed by Michael Fischer v. Mollard.
1408 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1409 results permerror and temperror. Previous values are deprecated but
1410 still accepted. In a future release, err_perm and err_temp will be
1411 completely removed, which will be a backward incompatibility if the
1412 ACL tests for either of these two old results. Patch contributed by
1413 user bes-internal on the mailing list.
1415 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1417 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1418 selectors, in both main and reject logs.
1420 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1423 JH/07 Add malware type "sock" for talking to simple daemon.
1425 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1427 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1428 routers/transports under cutthrough routing.
1430 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1431 numbers. Touch up "bool" conditional to keep the same definition.
1433 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1435 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1437 JH/12 Expand items in router/transport headers_add or headers_remove lists
1438 individually rather than the list as a whole. Bug 1452.
1440 Required for reasonable handling of multiple headers_ options when
1441 they may be empty; requires that headers_remove items with embedded
1442 colons must have them doubled (or the list-separator changed).
1444 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1445 view the policy declared in the DMARC record. Currently, $dmarc_status
1446 is a combined value of both the record presence and the result of the
1449 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1451 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1452 dnslookup router and the smtp transport (applying to the forward
1455 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1456 of ldap servers used for a specific lookup. Patch provided by Heiko
1459 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1460 New variable $lookup_dnssec_authenticated for observability.
1462 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1463 Patch submitted by Lars Timman.
1465 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1467 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1468 Requires trusted mode and valid format message id, aborts otherwise.
1469 Patch contributed by Heiko Schlichting.
1471 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1472 certextract with support for various fields. Bug 1358.
1474 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1475 is requested by default, modifiable by smtp transport option
1478 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1479 operate on certificate variables to give certificate fingerprints
1480 Also new ${sha256:cert_variable}.
1482 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1484 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1485 Christian Aistleitner.
1487 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1489 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1490 file. Patch from Wolfgang Breyha.
1492 JH/25 Expand the coverage of the delivery $host and $host_address to
1493 client authenticators run in verify callout. Bug 1476.
1495 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1496 align with daemon_smtp_ports. Bug 72.
1498 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1499 support and error reporting did not work properly.
1501 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1502 and is readable. Patch from Andrew Colin Kissa.
1504 TL/14 Enhance documentation of ${run expansion and how it parses the
1505 commandline after expansion, particularly in the case when an
1506 unquoted variable expansion results in an empty value.
1508 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1510 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1511 by renaming away from it. Addresses GH issue 15, reported by
1514 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1516 TL/15 SECURITY: prevent double expansion in math comparison functions
1517 (can expand unsanitized data). Not remotely exploitable.
1524 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1526 PP/02 Make -n do something, by making it not do something.
1527 When combined with -bP, the name of an option is not output.
1529 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1532 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1533 $sender_host_name and config options to manage this, and basic check
1536 PP/05 DSCP support for outbound connections and control modifier for inbound.
1538 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1539 (Only plugin which currently uses this is kerberos4, which nobody should
1540 be using, but we should make it available and other future plugins might
1541 conceivably use it, even though it would break NAT; stuff *should* be
1542 using channel bindings instead).
1544 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1545 name; added for Sendmail compatibility; requires admin caller.
1546 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1547 just ignore it); requires trusted caller.
1548 Also parse but ignore: -Ac -Am -X<logfile>
1551 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1553 TL/02 Add +smtp_confirmation as a default logging option.
1555 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1556 Patch by Magnus Holmgren from 2007-02-20.
1558 TL/04 Bugzilla 1281 - Spec typo.
1559 Bugzilla 1283 - Spec typo.
1560 Bugzilla 1290 - Spec grammar fixes.
1562 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1564 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1566 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1567 mailing list by by Dmitry Isaikin.
1569 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1571 JH/02 Support "G" suffix to numbers in ${if comparisons.
1573 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1575 NM/01 Bugzilla 1197 - Spec typo
1576 Bugzilla 1196 - Spec examples corrections
1578 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1580 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1581 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1584 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1585 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1587 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1588 "acl {{name}{arg}...}", and optional args on acl condition
1591 JH/05 Permit multiple router/transport headers_add/remove lines.
1593 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1595 JH/07 Avoid using a waiting database for a single-message-only transport.
1596 Performance patch from Paul Fisher. Bugzilla 1262.
1598 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1601 JH/09 Add $headers_added variable, with content from use of ACL modifier
1602 add_header (but not yet added to the message). Bugzilla 199.
1604 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1605 Pulled from Bugzilla 817 by Wolfgang Breyha.
1607 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1609 (nb: this is the same fix as in Exim 4.80.1)
1611 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1614 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1617 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1619 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1620 advertise SMTP AUTH mechanism to us, instead of a generic
1621 protocol violation error. Also, make Exim more robust to bad
1622 data from the Dovecot auth socket.
1624 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1626 When a queue runner is handling a message, Exim first routes the
1627 recipient addresses, during which it prunes them based on the retry
1628 hints database. After that it attempts to deliver the message to
1629 any remaining recipients. It then updates the hints database using
1632 So if a recipient address works intermittently, it can get repeatedly
1633 deferred at routing time. The retry hints record remains fresh so the
1634 address never reaches the final cutoff time.
1636 This is a fairly common occurrence when a user is bumping up against
1637 their storage quota. Exim had some logic in its local delivery code
1638 to deal with this. However it did not apply to per-recipient defers
1639 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1641 This change adds a proper retry rule check during routing so that the
1642 final cutoff time is checked against the message's age. We only do
1643 this check if there is an address retry record and there is not a
1644 domain retry record; this implies that previous attempts to handle
1645 the address had the retry_use_local_parts option turned on. We use
1646 this as an approximation for the destination being like a local
1647 delivery, as in LMTP.
1649 I suspect this new check makes the old local delivery cutoff check
1650 redundant, but I have not verified this so I left the code in place.
1652 TF/02 Correct gecos expansion when From: is a prefix of the username.
1654 Test 0254 submits a message to Exim with the header
1658 When I ran the test suite under the user fanf2, Exim expanded
1659 the header to contain my full name, whereas it should have added
1660 a Resent-Sender: header. It erroneously treats any prefix of the
1661 username as equal to the username.
1663 This change corrects that bug.
1665 GF/01 DCC debug and logging tidyup
1666 Error conditions log to paniclog rather than rejectlog.
1667 Debug lines prefixed by "DCC: " to remove any ambiguity.
1669 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1671 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1672 Bug spotted by Jeremy Harris; was flawed since initial commit.
1673 Would have resulted in OCSP responses post-SNI triggering an Exim
1674 NULL dereference and crash.
1676 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1678 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1679 Bug detection, analysis and fix by Samuel Thibault.
1680 Bugzilla 1331, Debian bug #698092.
1682 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1684 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1685 Server implementation by Todd Lyons, client by JH.
1686 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1687 config variable "prdr_enable" controls whether the server
1688 advertises the facility. If the client requests PRDR a new
1689 acl_data_smtp_prdr ACL is called once for each recipient, after
1690 the body content is received and before the acl_smtp_data ACL.
1691 The client is controlled by both of: a hosts_try_prdr option
1692 on the smtp transport, and the server advertisement.
1693 Default client logging of deliveries and rejections involving
1694 PRDR are flagged with the string "PRDR".
1696 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1697 fclose(). Diagnosis by Todd Lyons.
1699 PP/17 Update configure.default to handle IPv6 localhost better.
1700 Patch by Alain Williams (plus minor tweaks).
1703 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1704 This is now consistent with GnuTLS, and is now documented: the
1705 previous undocumented portable approach to treating the option as
1706 unset was to force an expansion failure. That still works, and
1707 an empty string is now equivalent.
1709 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1710 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1711 not performing validation itself.
1713 PP/20 Added force_command boolean option to pipe transport.
1714 Patch from Nick Koston, of cPanel Inc.
1716 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1719 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1721 PP/21 Fix eximon continuous updating with timestamped log-files.
1722 Broken in a format-string cleanup in 4.80, missed when I repaired the
1723 other false fix of the same issue.
1724 Report and fix from Heiko Schlichting.
1727 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1728 Report from Prashanth Katuri.
1730 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1731 It's SecureTransport, so affects any MacOS clients which use the
1732 system-integrated TLS libraries, including email clients.
1734 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1735 using a MIME ACL for non-SMTP local injection.
1736 Report and assistance in diagnosis by Warren Baker.
1738 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1740 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1742 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1743 last id that failed so it may be referenced in subsequent ACL's.
1745 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1748 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1749 ldap library initialization, allowing self-signed CA's to be
1750 used. Also properly sets require_cert option later in code by
1751 using NULL (global ldap config) instead of ldap handle (per
1752 session). Bug diagnosis and testing by alxgomz.
1754 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1755 the src/util/ subdirectory.
1757 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1758 renamed to Transport Post Delivery Action by Jeremy Harris, as
1761 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1762 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1763 redis_servers = needs to be configured which will be used by the redis
1764 lookup. Patch from Warren Baker, of The Packet Hub.
1766 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1768 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1769 hostname or reverse DNS when processing a host list. Used suggestions
1770 from multiple comments on this bug.
1772 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1774 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1775 Missed a few lines, added it to make the runtest require no keyboard
1778 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1779 contains upper case chars. Make router use caseful_local_part.
1781 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1782 support when GnuTLS has been built with p11-kit.
1788 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1790 This, or similar/improved, will also be change PP/11 of 4.82.
1796 PP/01 Handle short writes when writing local log-files.
1797 In practice, only affects FreeBSD (8 onwards).
1798 Bugzilla 1053, with thanks to Dmitry Isaikin.
1800 NM/01 Bugzilla 949 - Documentation tweak
1802 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1805 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1807 PP/02 Implemented gsasl authenticator.
1809 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1811 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1812 `pkg-config foo` for cflags/libs.
1814 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1815 with rest of GSASL and with heimdal_gssapi.
1817 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1818 `pkg-config foo` for cflags/libs for the TLS implementation.
1820 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1821 properties get this fed in as external SSF. A number of robustness
1822 and debugging improvements to the cyrus_sasl authenticator.
1824 PP/08 cyrus_sasl server now expands the server_realm option.
1826 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1827 Patch by Jeremy Harris.
1829 PP/10 Added dbmjz lookup type.
1831 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1833 PP/12 MAIL args handles TAB as well as SP, for better interop with
1834 non-compliant senders.
1835 Analysis and variant patch by Todd Lyons.
1837 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1838 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1839 Patch from Dirk Mueller <dmueller@suse.com>
1841 PP/13 tls_peerdn now print-escaped for spool files.
1842 Observed some $tls_peerdn in wild which contained \n, which resulted
1843 in spool file corruption.
1845 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1846 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1847 or write after TLS renegotiation, which otherwise led to messages
1850 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1851 as a tracking header (ie: a signed header comes before the signature).
1852 Patch from Wolfgang Breyha.
1854 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1855 comma-sep list; embedded commas doubled.
1857 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1859 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1861 Report and patch from Dmitry Banschikov.
1863 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1864 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1865 failures. We appear to now support TLS1.1+ with Exim.
1867 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1868 lets Exim select keys and certificates based upon TLS SNI from client.
1869 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1870 before an outbound SMTP session. New log_selector, +tls_sni.
1872 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1873 NULL dereference. Report and patch from Alun Jones.
1875 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1876 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1877 Not seeing resolver debug output on NetBSD, but suspect this is a
1878 resolver implementation change.
1880 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1881 Left warnings. Added "eximon gdb" invocation mode.
1883 PP/21 Defaulting "accept_8bitmime" to true, not false.
1885 PP/22 Added -bw for inetd wait mode support.
1887 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1888 locate the relevant includes and libraries. Made this the default.
1890 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1891 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1893 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1894 This may cause build issues on older platforms.
1896 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1897 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1898 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1899 Added SNI support via GnuTLS too.
1900 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1902 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1904 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1905 Applied second patch from Janne, implementing suggestion to default
1906 multiple-strings-in-record handling to match SPF spec.
1908 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1910 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1911 read-only, out of scope).
1912 Patch from Wolfgang Breyha, report from Stuart Northfield.
1914 PP/29 Fix three issues highlighted by clang analyser static analysis.
1915 Only crash-plausible issue would require the Cambridge-specific
1916 iplookup router and a misconfiguration.
1917 Report from Marcin Mirosław.
1919 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1921 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1922 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1923 As part of this, removing so much warning spew let me fix some minor
1924 real issues in debug logging.
1926 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1927 assignment on my part. Fixed.
1929 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1930 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1931 Janne Snabb (who went above and beyond: thank you).
1933 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1934 string otherwise requires a connection and a bunch more work and it's
1935 relatively easy to get wrong. Should also expose TLS library linkage
1938 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1939 64-bit ${eval} (JH/03).
1941 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1942 GNU libc to support some of the 64-bit stuff, should not lead to
1943 conflicts. Defined before os.h is pulled in, so if a given platform
1944 needs to override this, it can.
1946 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1947 protection layer was required, which is not implemented.
1948 Bugzilla 1254, patch from Wolfgang Breyha.
1950 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1951 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1952 tls_dhparam take prime identifiers. Also unbreak combination of
1953 OpenSSL+DH_params+TLSSNI.
1955 PP/39 Disable SSLv2 by default in OpenSSL support.
1961 PP/01 Solaris build fix for Oracle's LDAP libraries.
1962 Bugzilla 1109, patch from Stephen Usher.
1964 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1966 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1969 TF/02 Fix a couple more cases where we did not log the error message
1970 when unlink() failed. See also change 4.74-TF/03.
1972 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1973 lock up or crash if it happened to be inside a call to libc when it
1974 got a SIGUSR1 from exiwhat.
1976 The SIGUSR1 handler appends the current process status to the process
1977 log which is later printed by exiwhat. It used to use the general
1978 purpose logging code to do this, but several functions it calls are
1979 not safe for signals.
1981 The new output code in the SIGUSR1 handler is specific to the process
1982 log, and simple enough that it's easy to inspect for signal safety.
1983 Removing some special cases also simplifies the general logging code.
1984 Removing the spurious timestamps from the process log simplifies
1987 TF/04 Improved ratelimit ACL condition.
1989 The /noupdate option has been deprecated in favour of /readonly which
1990 has clearer semantics. The /leaky, /strict, and /readonly update modes
1991 are mutually exclusive. The update mode is no longer included in the
1992 database key; it just determines when the database is updated. (This
1993 means that when you upgrade Exim will forget old rate measurements.)
1995 Exim now checks that the per_* options are used with an update mode that
1996 makes sense for the current ACL. For example, when Exim is processing a
1997 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1998 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1999 must specify per_mail/readonly. If you omit the update mode it defaults to
2000 /leaky where that makes sense (as before) or /readonly where required.
2002 The /noupdate option is now undocumented but still supported for
2003 backwards compatibility. It is equivalent to /readonly except that in
2004 ACLs where /readonly is required you may specify /leaky/noupdate or
2005 /strict/noupdate which are treated the same as /readonly.
2007 A useful new feature is the /count= option. This is a generalization
2008 of the per_byte option, so that you can measure the throughput of other
2009 aggregate values. For example, the per_byte option is now equivalent
2010 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2012 The per_rcpt option has been generalized using the /count= mechanism
2013 (though it's more complicated than the per_byte equivalence). When it is
2014 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2015 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2016 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2017 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2018 Note that using per_rcpt with a non-readonly update mode in more than
2019 one ACL will cause the recipients to be double-counted. (The per_mail
2020 and per_byte options don't have this problem.)
2022 The handling of very low rates has changed slightly. If the computed rate
2023 is less than the event's count (usually one) then this event is the first
2024 after a long gap. In this case the rate is set to the same as this event's
2025 count, so that the first message of a spam run is counted properly.
2027 The major new feature is a mechanism for counting the rate of unique
2028 events. The new per_addr option counts the number of different
2029 recipients that someone has sent messages to in the last time period. It
2030 behaves like per_rcpt if all the recipient addresses are different, but
2031 duplicate recipient addresses do not increase the measured rate. Like
2032 the /count= option this is a general mechanism, so the per_addr option
2033 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2034 example, measure the rate that a client uses different sender addresses
2035 with the options per_mail/unique=$sender_address. There are further
2036 details in the main documentation.
2038 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2040 TF/06 Removed a few PCRE remnants.
2042 TF/07 Automatically extract Exim's version number from tags in the git
2043 repository when doing development or release builds.
2045 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2046 Bugzilla 879. Patch from Paul Fisher.
2048 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2049 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2052 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2054 PP/05 Variable $av_failed, true if the AV scanner deferred.
2055 Bugzilla 1078. Patch from John Horne.
2057 PP/06 Stop make process more reliably on build failure.
2058 Bugzilla 1087. Patch from Heiko Schlittermann.
2060 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2061 Bugzilla 1089. Patch from Heiko Schlittermann.
2063 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2064 Bugzilla 1131. Patch from Holger Weiß.
2066 PP/09 Handle IPv6 addresses with SPF.
2067 Bugzilla 860. Patch from Wolfgang Breyha.
2069 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2071 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2074 PP/11 match_* no longer expand right-hand-side by default.
2075 New compile-time build option, EXPAND_LISTMATCH_RHS.
2076 New expansion conditions, "inlist", "inlisti".
2078 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2080 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2082 PP/14 fix log_write() format string regression from TF/03.
2083 Bugzilla 1152. Patch from Dmitry Isaikin.
2089 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2091 PP/02 Harmonised TLS library version reporting; only show if debugging.
2092 Layout now matches that introduced for other libraries in 4.74 PP/03.
2094 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2096 PP/04 New "dns_use_edns0" global option.
2098 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2101 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2102 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2104 TK/01 Updated PolarSSL code to 0.14.2.
2105 Bugzilla 1097. Patch from Andreas Metzler.
2107 PP/07 Catch divide-by-zero in ${eval:...}.
2108 Fixes bugzilla 1102.
2110 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2113 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2114 format-string attack -- SECURITY: remote arbitrary code execution.
2116 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2117 time unintentionally subject to list matching rules, letting the header
2118 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2119 arbitrary string expansion). This allowed for information disclosure.
2121 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2122 INT_MIN/-1 -- value coerced to INT_MAX.
2128 NM/01 Workaround for PCRE version dependency in version reporting
2131 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2132 This fixes portability to compilers other than gcc, notably
2133 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2135 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2136 makefiles for portability to HP-UX and POSIX correctness.
2138 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2139 Also via indented variable definition in the Makefile.
2140 (Debugging by Oliver Heesakkers).
2142 PP/02 Restore caching of spamd results with expanded spamd_address.
2143 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2145 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2146 Improves build reliability. Fix from: Frank Elsner
2148 NM/02 Fix wide character breakage in the rfc2047 coding
2149 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2151 NM/03 Allow underscore in dnslist lookups
2152 Fixes bug 1026. Patch from Graeme Fowler
2154 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2155 Code patches from Adam Ciarcinski of NetBSD.
2157 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2160 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2161 is logged, to avoid truncation. Patch from John Horne.
2163 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2164 Patch from Jakob Hirsch.
2166 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2167 SQL string expansion failure details.
2168 Patch from Andrey Oktyabrski.
2170 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2171 Patch from Simon Arlott.
2173 PP/09 New lookups functionality failed to compile on old gcc which rejects
2174 extern declarations in function scope.
2175 Patch from Oliver Fleischmann
2177 PP/10 Use sig_atomic_t for flags set from signal handlers.
2178 Check getgroups() return and improve debugging.
2179 Fixed developed for diagnosis in bug 927 (which turned out to be
2182 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2183 Patch from Mark Zealey.
2185 PP/12 Bugzilla 1056: Improved spamd server selection.
2186 Patch from Mark Zealey.
2188 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2189 Based on patch from Heiko Schlittermann.
2191 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2192 Patch from Uwe Doering, sign-off by Michael Haardt.
2194 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2195 variable declaration deep within a block. Bug and patch from
2198 PP/15 lookups-Makefile IRIX compatibility coercion.
2200 PP/16 Make DISABLE_DKIM build knob functional.
2202 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2203 Patch by Simon Arlott
2205 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2206 variable argument macros. Our copy now differs from upstream.
2212 TF/01 Failure to get a lock on a hints database can have serious
2213 consequences so log it to the panic log.
2215 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2216 controlled using the smtp_confirmation log selector.
2218 TF/03 Include the error message when we fail to unlink a spool file.
2220 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2221 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2222 for maintaining out-of-tree patches for some time.
2224 PP/01 Bugzilla 139: Documentation and portability issues.
2225 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2226 Handle per-OS dynamic-module compilation flags.
2228 PP/02 Let /dev/null have normal permissions.
2229 The 4.73 fixes were a little too stringent and complained about the
2230 permissions on /dev/null. Exempt it from some checks.
2231 Reported by Andreas M. Kirchwitz.
2233 PP/03 Report version information for many libraries, including
2234 Exim version information for dynamically loaded libraries. Created
2235 version.h, now support a version extension string for distributors
2236 who patch heavily. Dynamic module ABI change.
2238 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2239 privilege escalation vulnerability whereby the Exim run-time user
2240 can cause root to append content of the attacker's choosing to
2243 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2246 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2247 If dropping privileges for untrusted macros, we disabled normal logging
2248 on the basis that it would fail; for the Exim run-time user, this is not
2249 the case, and it resulted in successful deliveries going unlogged.
2250 Fixed. Reported by Andreas Metzler.
2256 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2257 only prepend for the Resent-* case. Fixes regression introduced in
2258 Exim 4.70 by NM/22 for Bugzilla 607.
2260 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2261 increasing numbers of administrators be bitten by this.
2263 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2265 PP/03 Bugzilla 994: added openssl_options main configuration option.
2267 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2269 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2271 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2273 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2274 without a peer certificate, leading to a segfault because of an
2275 assumption that peers always have certificates. Be a little more
2276 paranoid. Problem reported by Martin Tscholak.
2278 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2279 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2280 NB: ClamAV planning to remove STREAM in "middle of 2010".
2281 CL also introduces -bmalware, various -d+acl logging additions and
2282 more caution in buffer sizes.
2284 PP/09 Implemented reverse_ip expansion operator.
2286 PP/10 Bugzilla 937: provide a "debug" ACL control.
2288 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2290 PP/12 Bugzilla 973: Implement --version.
2292 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2294 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2296 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2298 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2299 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2300 ignore trailing whitespace.
2302 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2304 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2307 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2308 Notification from Dr Andrew Aitchison.
2310 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2311 ExtendedDetectionInfo response format.
2312 Notification from John Horne.
2314 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2317 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2318 XSL and documented dependency on system catalogs, with examples of how
2321 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2324 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2325 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2326 configuration file which is writeable by the Exim user or group.
2328 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2329 of configuration files to cover files specified with the -C option if
2330 they are going to be used with root privileges, not just the default
2333 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2334 option (effectively making it always true).
2336 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2337 files to be used while preserving root privileges.
2339 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2340 that rogue child processes cannot use them.
2342 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2343 run-time user, instead of root.
2345 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2346 Exim run-time user without dropping privileges.
2348 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2349 result string, instead of calling string_vformat() twice with the same
2352 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2353 for other users. Others should always drop root privileges if they use
2354 -C on the command line, even for a whitelisted configure file.
2356 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2358 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2364 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2365 $data_path, and $header_path variables; fixed documentation bugs and
2368 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2369 exipick to access non-standard spools, including the "frozen" queue
2372 NM/01 Bugzilla 965: Support mysql stored procedures.
2373 Patch from Alain Williams
2375 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2377 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2378 Patch from Andreas Metzler
2380 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2381 Patch from Kirill Miazine
2383 NM/05 Bugzilla 671: Added umask to procmail example.
2385 JJ/03 installed exipick 20100323.0, fixing doc bug
2387 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2388 directory. Notification and patch from Dan Rosenberg.
2390 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2392 TK/02 Improve log output when DKIM signing operation fails.
2394 MH/01 Treat the transport option dkim_domain as a colon separated
2395 list, not as a single string, and sign the message with each element,
2396 omitting multiple occurences of the same signer.
2398 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2399 Bugzilla 985, 986. Patch by Simon Arlott
2401 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2402 Patch by Simon Arlott
2404 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2405 MBX locking. Notification from Dan Rosenberg.
2411 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2413 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2415 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2417 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2419 NM/04 Bugzilla 915: Flag broken perl installation during build.
2425 TK/01 Added patch by Johannes Berg that expands the main option
2426 "spamd_address" if it starts with a dollar sign.
2428 TK/02 Write list of recipients to X-Envelope-Sender header when building
2429 the mbox-format spool file for content scanning (suggested by Jakob
2432 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2433 (http://www.dcc-servers.net/) support via dccifd. Activated by
2434 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2436 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2437 by Mark Daniel Reidel <mr@df.eu>.
2439 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2440 When building exim an external PCRE library is now needed -
2441 PCRE is a system library on the majority of modern systems.
2442 See entry on PCRE_LIBS in EDITME file.
2444 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2445 conversation. Added nologin parameter to request.
2446 Patch contributed by Kirill Miazine.
2448 TF/01 Do not log submission mode rewrites if they do not change the address.
2450 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2452 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2453 log files in place. Contributed by Roberto Lima.
2455 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2457 TF/03 Bugzilla 615: When checking the local_parts router precondition
2458 after a local_part_suffix or local_part_prefix option, Exim now
2459 does not use the address's named list lookup cache, since this
2460 contains cached lookups for the whole local part.
2462 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2463 Robert Millan. Documentation is in experimental-spec.txt.
2465 TF/04 Bugzilla 668: Fix parallel build (make -j).
2467 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2469 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2470 Patch provided by Jan Srzednicki.
2472 TF/05 Leading white space used to be stripped from $spam_report which
2473 wrecked the formatting. Now it is preserved.
2475 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2476 that they are available at delivery time.
2478 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2480 TF/08 TLS error reporting now respects the incoming_interface and
2481 incoming_port log selectors.
2483 TF/09 Produce a more useful error message if an SMTP transport's hosts
2484 setting expands to an empty string.
2486 NM/06 Bugzilla 744: EXPN did not work under TLS.
2487 Patch provided by Phil Pennock.
2489 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2490 Patch provided by Richard Godbee.
2492 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2493 acl_smtp_notquit, added index entry.
2495 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2496 Patch provided by Eugene Bujak.
2498 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2499 accept(). Patch provided by Maxim Dounin.
2501 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2502 Patch provided by Phil Pennock.
2504 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2506 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2507 Patch provided by Brad "anomie" Jorsch.
2509 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2510 Patch provided by Dean Brooks.
2512 TK/05 Add native DKIM support (does not depend on external libraries).
2514 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2515 Patch provided by Graeme Fowler.
2517 NM/16 Bugzilla 851: Documentation example syntax fix.
2519 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2521 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2524 NM/19 Bugzilla 745: TLS version reporting.
2525 Patch provided by Phil Pennock.
2527 NM/20 Bugzilla 167: bool: condition support.
2528 Patch provided by Phil Pennock.
2530 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2531 clients. Patch provided by Phil Pennock.
2533 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2534 Patch provided by Brad "anomie" Jorsch.
2536 NM/23 Bugzilla 687: Fix misparses in eximstats.
2537 Patch provided by Heiko Schlittermann.
2539 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2540 Patch provided by Heiko Schlittermann.
2542 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2543 plus update to original patch.
2545 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2547 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2548 Patch provided by David Brownlee.
2550 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2552 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2554 NM/30 Bugzilla 888: TLS documentation bugfixes.
2556 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2558 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2559 Unlike the original bugzilla I have changed all shell scripts in src tree.
2561 NM/33 Bugzilla 898: Transport filter timeout fix.
2562 Patch by Todd Rinaldo.
2564 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2565 Patch by Serge Demonchaux.
2567 NM/35 Bugzilla 39: Base64 decode bug fixes.
2568 Patch by Jakob Hirsch.
2570 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2572 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2574 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2576 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2582 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2583 ALT-N's libdkim that I have put here:
2584 http://duncanthrax.net/exim-experimental/
2586 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2587 were called 'true' and it seems that C99 defines that as a reserved
2588 keyword to be used with 'bool' variable types. That means you could
2589 not include C99-style headers which use bools without triggering
2590 build errors in sieve.c.
2592 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2593 as mailq or other aliases. Changed the --help handling significantly
2594 to do whats expected. exim_usage() emits usage/help information.
2596 SC/01 Added the -bylocaldomain option to eximstats.
2598 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2600 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2602 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2608 PH/01 Another patch from the Sieve maintainer.
2610 PH/02 When an IPv6 address is converted to a string for single-key lookup
2611 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2612 dots are used instead of colons so that keys in lsearch files need not
2613 contain colons. This was done some time before quoting was made available
2614 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2615 (notated using the quote facility) so as to distinguish them from IPv4
2616 keys. This meant that lookups for IP addresses in host lists did not work
2617 for iplsearch lookups.
2619 This has been fixed by arranging for IPv6 addresses to be expressed with
2620 colons if the lookup type is iplsearch. This is not incompatible, because
2621 previously such lookups could never work.
2623 The situation is now rather anomalous, since one *can* have colons in
2624 ordinary lsearch keys. However, making the change in all cases is
2625 incompatible and would probably break a number of configurations.
2627 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2630 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2631 conversion specification without a maximum field width, thereby enabling
2632 a rogue spamd server to cause a buffer overflow. While nobody in their
2633 right mind would setup Exim to query an untrusted spamd server, an
2634 attacker that gains access to a server running spamd could potentially
2635 exploit this vulnerability to run arbitrary code as the Exim user.
2637 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2638 $primary_hostname instead of what libspf2 thinks the hosts name is.
2640 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2641 a directory entry by the name of the lookup key. Previously, if a
2642 symlink pointed to a non-existing file or a file in a directory that
2643 Exim lacked permissions to read, a lookup for a key matching that
2644 symlink would fail. Now it is enough that a matching directory entry
2645 exists, symlink or not. (Bugzilla 503.)
2647 PH/03 The body_linecount and body_zerocount variables are now exported in the
2650 PH/04 Added the $dnslist_matched variable.
2652 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2653 This means they are set thereafter only if the connection becomes
2656 PH/06 Added the client_condition to authenticators so that some can be skipped
2657 by clients under certain conditions.
2659 PH/07 The error message for a badly-placed control=no_multiline_responses left
2660 "_responses" off the end of the name.
2662 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2664 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2665 (without spaces) instead of just copying the configuration text.
2667 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2669 PH/11 Added $max_received_linelength.
2671 PH/12 Added +ignore_defer and +include_defer to host lists.
2673 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2674 way in which PCRE > 7.0 is built.
2676 PH/14 Implemented queue_only_load_latch.
2678 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2679 MAIL command. The effect was to mangle the value on 64-bit systems.
2681 PH/16 Another patch from the Sieve maintainer.
2683 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2685 PH/18 If a system quota error occurred while trying to create the file for
2686 a maildir delivery, the message "Mailbox is full" was not appended to the
2687 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2688 only to a quota excession during the actual writing of the file.
2690 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2691 characters?) which causes problems in log lines. The DN values are now
2692 passed through string_printing() before being added to log lines.
2694 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2695 and InterBase are left for another time.)
2697 PH/21 Added message_body_newlines option.
2699 PH/22 Guard against possible overflow in moan_check_errorcopy().
2701 PH/23 POSIX allows open() to be a macro; guard against that.
2703 PH/24 If the recipient of an error message contained an @ in the local part
2704 (suitably quoted, of course), incorrect values were put in $domain and
2705 $local_part during the evaluation of errors_copy.
2711 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2712 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2715 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2716 issue a MAIL command.
2718 PH/02 In an ACL statement such as
2720 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2722 if a client was not listed at all, or was listed with a value other than
2723 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2724 the condition was not true (as it should be), so access was not denied.
2725 The bug was that the ! inversion was incorrectly passed on to the second
2726 item. This has been fixed.
2728 PH/03 Added additional dnslists conditions == and =& which are different from
2729 = and & when the dns lookup returns more than one IP address.
2731 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2732 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2734 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2735 FSYNC, which compiles an option called disable_fsync that allows for
2736 bypassing fsync(). The documentation is heavily laced with warnings.
2738 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2740 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2741 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2742 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2743 including adding "make clean"; (3) Added -fPIC when compiling the test
2744 dynamically loaded module, to get rid of a warning.
2746 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2747 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2748 = 0s. The bug is otherwise harmless.
2750 PH/07 There was a bug in the dovecot authenticator such that the value of
2751 $auth1 could be overwritten, and so not correctly preserved, after a
2752 successful authentication. This usually meant that the value preserved by
2753 the server_setid option was incorrect.
2755 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2757 PH/09 Installed PCRE release 7.0.
2759 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2760 run for batched SMTP input. It is now run at the start of every message
2761 in the batch. While fixing this I discovered that the process information
2762 (output by running exiwhat) was not always getting set for -bs and -bS
2763 input. This is fixed, and it now also says "batched" for BSMTP.
2765 PH/11 Added control=no_pipelining.
2767 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2768 patch, slightly modified), and move the expansion of helo_data till after
2769 the connection is made in the smtp transport (so it can use these
2772 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2774 PH/14 Added log_selector = +pid.
2776 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2778 PH/16 Add ${if forany and ${if forall.
2780 PH/17 Added dsn_from option to vary the From: line in DSNs.
2782 PH/18 Flush SMTP output before performing a callout, unless control =
2783 no_callout_flush is set.
2785 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2786 was true (the default) a successful delivery failed to delete the retry
2787 item, thus causing premature timeout of the address. The bug is now
2790 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2792 PH/21 Long custom messages for fakedefer and fakereject are now split up
2793 into multiline responses in the same way that messages for "deny" and
2794 other ACL rejections are.
2796 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2797 with slight modification.
2799 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2800 draft, changing the syntax and factoring some duplicate code".
2802 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2803 for deliveries of the second and subsequent messages over the same SMTP
2806 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2807 ${reduce, with only minor "tidies".
2809 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2811 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2812 expansion side effects.
2814 PH/27 When a message times out after an over-quota error from an Exim-imposed
2815 quota, the bounce message says "mailbox is full". This message was not
2816 being given when it was a system quota that was exceeded. It now should
2819 MH/03 Made $recipients available in local_scan(). local_scan() already has
2820 better access to the recipient list through recipients_list[], but
2821 $recipients can be useful in postmaster-provided expansion strings.
2823 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2824 in the case of a MAIL command with additional options following the
2825 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2826 were accidentally chopped off.
2828 PH/29 SMTP synchronization checks are implemented when a command is read -
2829 there is a check that no more input is waiting when there shouldn't be
2830 any. However, for some commands, a delay in an ACL can mean that it is
2831 some time before the response is written. In this time, more input might
2832 arrive, invalidly. So now there are extra checks after an ACL has run for
2833 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2834 pipelining has not been advertised.
2836 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2838 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2839 returned in the SMTP response when smtp_return_error_details was false.
2840 This has been fixed.
2842 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2843 instead of the C I/O that was originally supplied, because problems were
2844 reported on Solaris.
2846 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2847 Exim which did not show up earlier: it was assuming that a call to
2848 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2849 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2850 was a macro that became an assignment, so it seemed to work. This has
2851 changed to a proper function call with a void return, hence the compile
2852 error. Exim's code has been fixed.
2854 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2857 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2859 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2861 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2862 because it turns out to be too big or there is a timeout) there is no
2863 "Completed" line in the log. When some messages of this type were
2864 selected by exigrep, they were listed as "not completed". Others were
2865 picked up by some special patterns. I have improved the selection
2866 criteria to be more general.
2868 PH/38 The host_find_failed option in the manualroute router can now be set
2869 to "ignore", to completely ignore a host whose IP address cannot be
2870 found. If all hosts are ignored, the behaviour is controlled by the new
2871 host_all_ignored option.
2873 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2874 homing or because of multiple MX records with /mx) generated more than
2875 one IP address, and the following item turned out to be the local host,
2876 all the secondary addresses of the first item were incorrectly removed
2877 from the list, along with the local host and any following hosts (which
2878 is what is supposed to happen).
2880 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2881 whoever called Exim into the -H file. In the case of the daemon it was
2882 behaving confusingly. When first started, it used values for whoever
2883 started the daemon, but after a SIGHUP it used the Exim user (because it
2884 calls itself on a restart). I have changed the code so that it now always
2887 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2888 message are rejected with the same error (e.g. no authentication or bad
2889 sender address), and a DATA command is nevertheless sent (as can happen
2890 with PIPELINING or a stupid MUA), the error message that was given to the
2891 RCPT commands is included in the rejection of the DATA command. This is
2892 intended to be helpful for MUAs that show only the final error to their
2895 PH/42 Another patch from the Sieve maintainer.
2897 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2898 Eximstats - Fixed some broken HTML links and added missing column headers
2900 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2901 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2903 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2905 PH/43 Yet another patch from the Sieve maintainer.
2907 PH/44 I found a way to check for a TCP/IP connection going away before sending
2908 the response to the final '.' that terminates a message, but only in the
2909 case where the client has not sent further data following the '.'
2910 (unfortunately, this is allowed). However, in many cases there won't be
2911 any further data because there won't be any more messages to send. A call
2912 to select() can be used: if it shows that the input is "ready", there is
2913 either input waiting, or the socket has been closed. An attempt to read
2914 the next input character can distinguish the two cases. Previously, Exim
2915 would have sent an OK response which the client would never have see.
2916 This could lead to message repetition. This fix should cure that, at
2917 least in a lot of common cases.
2919 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2920 advertised in response to EHLO.
2926 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2927 fixed by 4.65/MH/01 (is this a record?) are fixed:
2929 (i) An empty string was always treated as zero by the numeric comparison
2930 operators. This behaviour has been restored.
2932 (ii) It is documented that the numeric comparison operators always treat
2933 their arguments as decimal numbers. This was broken in that numbers
2934 starting with 0 were being interpreted as octal.
2936 While fixing these problems I realized that there was another issue that
2937 hadn't been noticed. Values of message_size_limit (both the global option
2938 and the transport option) were treated as octal if they started with 0.
2939 The documentation was vague. These values are now always treated as
2940 decimal, and I will make that clear in the documentation.
2946 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2947 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2950 MH/01 Don't check that the operands of numeric comparison operators are
2951 integers when their expansion is in "skipping" mode (fixes bug
2952 introduced by 4.64-PH/07).
2954 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2955 child addresses, Exim now panics and dies. Previously, because the count
2956 is held in a short int, deliveries were likely to be lost. As such a
2957 large number of recipients for a single message is ridiculous
2958 (performance will be very, very poor), I have chosen to impose a limit
2959 rather than extend the field.
2965 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2966 leftover -K file (the existence of which was triggered by #402).
2967 While we were at it, introduced process PID as part of the -K
2968 filename. This should rule out race conditions when creating
2971 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2972 processing considerably. Previous code took too long for large mails,
2973 triggering a timeout which in turn triggers #401.
2975 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2976 in the DK code in transports.c. sendfile() is not really portable,
2977 hence the _LINUX specificness.
2979 TF/01 In the add_headers option to the mail command in an Exim filter,
2980 there was a bug that Exim would claim a syntax error in any
2981 header after the first one which had an odd number of characters
2984 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2985 callout verification, Exim cached a "reject" for the entire domain. This
2986 is correct for most verifications, but it is not correct for a recipient
2987 verification with use_sender or use_postmaster set, because in that case
2988 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2989 case of MAIL FROM:<> rejection from other early rejections (e.g.
2990 rejection of HELO). When verifying a recipient using a non-null MAIL
2991 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2992 Whatever the result of the callout, the value of the domain cache is
2993 left unchanged (for any other kind of callout, getting as far as trying
2994 RCPT means that the domain itself is ok).
2996 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2999 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3000 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3001 session, the daemon ignores ECONNECT errors and logs others; it now
3002 ignores EPIPE as well.
3004 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3005 (quoted-printable decoding).
3007 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3008 later the small subsequent patch to fix an introduced bug.
3010 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3012 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3014 PH/08 An error is now given if message_size_limit is specified negative.
3016 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3017 to be given (somewhat) arbitrary names.
3019 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3022 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3023 miscellaneous code fixes
3025 PH/10 Added the log_reject_target ACL modifier to specify where to log
3028 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3029 hostname. This is wrong, because it relates to the incoming message (and
3030 probably the interface on which it is arriving) and not to the outgoing
3031 callout (which could be using a different interface). This has been
3032 changed to use the value of the helo_data option from the smtp transport
3033 instead - this is what is used when a message is actually being sent. If
3034 there is no remote transport (possible with a router that sets up host
3035 addresses), $smtp_active_hostname is used.
3037 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3038 tweaks were necessary in order to get it to work (see also 21 below):
3039 (a) The code assumed that strncpy() returns a negative number on buffer
3040 overflow, which isn't the case. Replaced with Exim's string_format()
3042 (b) There were several signed/unsigned issues. I just did the minimum
3043 hacking in of casts. There is scope for a larger refactoring.
3044 (c) The code used strcasecmp() which is not a standard C function.
3045 Replaced with Exim's strcmpic() function.
3046 (d) The code set only $1; it now sets $auth1 as well.
3047 (e) A simple test gave the error "authentication client didn't specify
3048 service in request". It would seem that Dovecot has changed its
3049 interface. Fortunately there's a specification; I followed it and
3050 changed what the client sends and it appears to be working now.
3052 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3055 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3056 address A is aliased to B and C, where B exists and C does not. Without
3057 -v the output is "A verified" because verification stops after a
3058 successful redirection if more than one address is generated. However,
3059 with -v the child addresses are also verified. Exim was outputting "A
3060 failed to verify" and then showing the successful verification for C,
3061 with its parentage. It now outputs "B failed to verify", showing B's
3062 parentage before showing the successful verification of C.
3064 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3065 look up a TXT record in a specific list after matching in a combined
3068 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3069 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3070 they consult the DNS. I had assumed they would set it the way they
3071 wanted; and indeed my experiments on Linux seem to show that in some
3072 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3073 To be on the safe side, however, I have now made the interface to
3074 host_find_byname() similar to host_find_bydns(), with an argument
3075 containing the DNS resolver options. The host_find_byname() function now
3076 sets these options at its start, just as host_find_bydns() does. The smtp
3077 transport options dns_qualify_single and dns_search_parents are passed to
3078 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3079 of host_find_byname() use the default settings of RES_DEFNAMES
3080 (qualify_single) but not RES_DNSRCH (search_parents).
3082 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3083 spool_read_header() do less string testing, by means of a preliminary
3084 switch on the second character of optional "-foo" lines. (This is
3085 overdue, caused by the large number of possibilities that now exist.
3086 Originally there were few.) While I was there, I also converted the
3087 str(n)cmp tests so they don't re-test the leading "-" and the first
3088 character, in the hope this might squeeze out yet more improvement.
3090 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3091 flag allowing group syntax was set by the header_syntax check but not
3092 turned off, possible causing trouble later; (2) The flag was not being
3093 set at all for the header_verify test, causing "group"-style headers to
3094 be rejected. I have now set it in this case, and also caused header_
3095 verify to ignore an empty address taken from a group. While doing this, I
3096 came across some other cases where the code for allowing group syntax
3097 while scanning a header line wasn't quite right (mostly, not resetting
3098 the flag correctly in the right place). These bugs could have caused
3099 trouble for malformed header lines. I hope it is now all correct.
3101 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3102 with the "reply" argument non-NULL. The code, however (which originally
3103 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3104 but it didn't always do it. This confused somebody who was copying the
3105 code for some other use. I have removed all the tests.
3107 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3108 feature that was used to support insecure browsers during the U.S. crypto
3109 embargo. It requires special client support, and Exim is probably the
3110 only MTA that supported it -- and would never use it because real RSA is
3111 always available. This code has been removed, because it had the bad
3112 effect of slowing Exim down by computing (never used) parameters for the
3113 RSA_EXPORT functionality.
3115 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3116 authenticator to fail if there's a tab character in the incoming data
3117 (there should never be unless someone is messing about, as it's supposed
3118 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3119 if the connection is using TLS or if the remote IP is the same as the
3120 local IP, and the "valid-client-cert option" if a client certificate has
3123 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3124 authenticators. This can be used for authorization after authentication
3125 succeeds. (In the case of plaintext, it servers for both authentication
3128 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3129 if any retry times were supplied.
3131 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3132 connection, where there is no client IP address to check. In this
3133 situation, the verify now always succeeds.
3135 PH/25 Applied John Jetmore's -Mset patch.
3137 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3139 PH/27 In a string expansion for a processed (not raw) header when multiple
3140 headers of the same name were present, leading whitespace was being
3141 removed from all of them, but trailing whitespace was being removed only
3142 from the last one. Now trailing whitespace is removed from each header
3143 before concatenation. Completely empty headers in a concatenation (as
3144 before) are ignored.
3146 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3147 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3149 PH/29 [Removed. This was a change that I later backed out, and forgot to
3150 correct the ChangeLog entry (that I had efficiently created) before
3151 committing the later change.]
3153 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3154 address errors (4xx response to RCPT) over the same connection as other
3155 messages routed to the same hosts. Such deliveries are always "forced",
3156 so retry times are not inspected. This resulted in far too many retries
3157 for the affected addresses. The effect occurred only when there were more
3158 hosts than the hosts_max_try setting in the smtp transport when it had
3159 the 4xx errors. Those hosts that it had tried were not added to the list
3160 of hosts for which the message was waiting, so if all were tried, there
3161 was no problem. Two fixes have been applied:
3163 (i) If there are any address or message errors in an SMTP delivery, none
3164 of the hosts (tried or untried) are now added to the list of hosts
3165 for which the message is waiting, so the message should not be a
3166 candidate for sending over the same connection that was used for a
3167 successful delivery of some other message. This seems entirely
3168 reasonable: after all the message is NOT "waiting for some host".
3169 This is so "obvious" that I'm not sure why it wasn't done
3170 previously. Hope I haven't missed anything, but it can't do any
3171 harm, as the worst effect is to miss an optimization.
3173 (ii) If, despite (i), such a delivery is accidentally attempted, the
3174 routing retry time is respected, so at least it doesn't keep
3175 hammering the server.
3177 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3178 in ${readsocket because some servers need this prod.
3180 PH/32 Added some extra debug output when updating a wait-xxx database.
3182 PH/33 The hint "could be header name not terminated by colon", which has been
3183 given for certain expansion errors for a long time, was not being given
3184 for the ${if def:h_colon_omitted{... case.
3186 PH/34 The spec says: "With one important exception, whenever a domain list is
3187 being scanned, $domain contains the subject domain." There was at least
3188 one case where this was not true.
3190 PH/35 The error "getsockname() failed: connection reset by peer" was being
3191 written to the panic log as well as the main log, but it isn't really
3192 panic-worthy as it just means the connection died rather early on. I have
3193 removed the panic log writing for the ECONNRESET error when getsockname()
3196 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3197 runs only) independently of the message's sender address. This meant
3198 that, if the 4xx error was in fact related to the sender, a different
3199 message to the same recipient with a different sender could confuse
3200 things. In particular, this can happen when sending to a greylisting
3201 server, but other circumstances could also provoke similar problems.
3202 I have changed the default so that the retry time for these errors is now
3203 based a combination of the sender and recipient addresses. This change
3204 can be overridden by setting address_retry_include_sender=false in the
3207 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3208 remote server are returned as part of bounce messages. This was not
3209 happening for LMTP over a pipe (the lmtp transport), but now it is the
3210 same for both kinds of LMTP.
3212 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3213 in header lines that were in fact CNAMEs. This is no longer the case.
3215 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3216 and queue runs started by the daemon processed all messages. This has
3217 been fixed so that -R and -S can now usefully be given with -q<time>.
3219 PH/40 Import PCRE release 6.7 (fixes some bugs).
3221 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3223 PH/42 Give an error if -q is specified more than once.
3225 PH/43 Renamed the variables $interface_address and $interface_port as
3226 $received_ip_address and $received_port, to make it clear that these
3227 values apply to message reception, and not to the outgoing interface when
3228 a message is delivered. (The old names remain recognized, of course.)
3230 PH/44 There was no timeout on the connect() call when using a Unix domain
3231 socket in the ${readsocket expansion. There now is.
3233 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3234 be meaningful with "accept".
3236 SC/01 Eximstats V1.43
3237 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3239 SC/02 Eximstats V1.44
3240 Use a glob alias rather than an array ref in the generated
3241 parser. This improves both readability and performance.
3243 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3244 Collect SpamAssassin and rejection statistics.
3245 Don't display local sender or destination tables unless
3246 there is data to show.
3247 Added average volumes into the top table text output.
3249 SC/04 Eximstats V1.46
3250 Collect data on the number of addresses (recipients)
3251 as well as the number of messages.
3253 SC/05 Eximstats V1.47
3254 Added 'Message too big' to the list of mail rejection
3255 reasons (thanks to Marco Gaiarin).
3257 SC/06 Eximstats V1.48
3258 Mainlog lines which have GMT offsets and are too short to
3259 have a flag are now skipped.
3261 SC/07 Eximstats V1.49 (Alain Williams)
3262 Added the -emptyok flag.
3264 SC/08 Eximstats V1.50
3265 Fixes for obtaining the IP address from reject messages.
3267 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3268 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3269 whitespace changes from 4.64-PH/27
3271 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3274 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3275 are found, allow negative numbers in numeric criteria)
3277 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3279 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3280 to match changes made in 4.64-PH/43
3282 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3284 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3285 "domain" is actually an IP address, and force a failure. This locks out
3286 those revolvers/nameservers that support "A-for-A" lookups, in
3287 contravention of the specifications.
3289 PH/48 When a host name was looked up from an IP address, and the subsequent
3290 forward lookup of the name timed out, the host name was left in
3291 $sender_host_name, contrary to the specification.
3293 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3294 restricted to single-key lookups, Exim was not diagnosing an error if
3295 * or *@ was used with a query-style lookup.
3297 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3299 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3300 long ago, but noone interested enough thought of it. Let's just say that
3301 the "1.1" means that there are some new functions that weren't there at
3302 some point in the past.
3304 PH/51 Error processing for expansion failure of helo_data from an smtp
3305 transport during callout processing was broken.
3307 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3308 tested/used via the -bh/-bhc/-bs options.
3310 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3311 bug, fixed in subsequent PCRE releases).
3313 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3314 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3316 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3322 SC/01 Use a glob alias rather than an array ref in eximstats generated
3323 parser. This improves both readability and performance.
3325 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3326 Don't display local sender or destination tables in eximstats unless
3327 there is data to show.
3328 Added average volumes into the eximstats top table text output.
3330 SC/03 Collect data on the number of addresses (recipients) as well
3331 as the number of messages in eximstats.
3333 TF/01 Correct an error in the documentation for the redirect router. Exim
3334 does (usually) call initgroups() when daemonizing.
3336 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3337 with consistent privilege compared to when running as a daemon.
3339 TF/03 Note in the spec that $authenticated_id is not set for local
3340 submissions from trusted users.
3342 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3343 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3345 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3346 by adding some example configuration directives to the default
3347 configuration file. A little bit of work is required to uncomment the
3348 directives and define how usernames and passwords are checked, but
3349 there is now a framework to start from.
3351 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3352 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3353 without this. I don't know how relevant this is to other LDAP libraries.
3355 PH/02 Add the verb name to the "unknown ACL verb" error.
3357 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3359 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3361 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3362 directory not expanded when it should be if an expanded home directory
3363 was set for the address (which is overridden by the transport).
3365 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3368 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3369 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3370 because it is too late at that time, and has no effect.
3372 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3373 security issue with \' (bugzilla #107). I could not use the
3374 PQescapeStringConn() function, because it needs a PGconn value as one of
3377 PH/08 When testing addresses using -bt, indicate those final addresses that
3378 are duplicates that would not cause an additional delivery. At least one
3379 person was confused, thinking that -bt output corresponded to deliveries.
3380 (Suppressing duplicates isn't a good idea as you lose the information
3381 about possibly different redirections that led to the duplicates.)
3383 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3384 systems where poll() doesn't work, in particular OS X.
3386 PH/10 Added more information to debugging output for retry time not reached.
3388 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3389 operations in malware.c.
3391 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3394 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3395 syslog_duplication set false, log lines that would normally be written
3396 both the the main log and to the reject log were not written to syslog at
3399 PH/14 In the default configuration, change the use of "message" in ACL warn
3400 statements to "add_header".
3402 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3403 not followed by a command (e.g. "seen endif").
3405 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3406 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3409 PH/17 Added extra conditions to the default value of delay_warning_condition
3413 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3414 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3415 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3418 The Auto-Submitted: and various List- headers are standardised, whereas I
3419 don't think Precedence: ever was.
3421 PH/18 Refactored debugging code in route_finduser() to show more information,
3422 in particular, the error code if getpwnam() issues one.
3424 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3425 This is apparently needed in addition to the PH/07 change above to avoid
3426 any possible encoding problems.
3428 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3429 but not after initializing Perl.
3431 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3432 output them only if debugging. By default they are written stderr,
3433 apparently, which is not desirable.
3435 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3438 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3441 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3443 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3444 authenticated or an ident call has been made. Suppress the default
3445 values for $authenticated_id and $authenticated_sender (but permit -oMai
3446 and -oMas) when testing with -bh.
3448 PH/24 Re-jigged the order of the tests in the default configuration so that the
3449 tests for valid domains and recipients precede the DNS black list and CSA
3450 tests, on the grounds that those ones are more expensive.
3452 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3453 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3454 This bug exists in every version of Exim that I still have, right back to
3457 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3458 However, an attempt to turn on case-sensitivity in a regex key by
3459 including (?-i) didn't work because the subject string was already
3460 lowercased, and the effects were non-intuitive. It turns out that a
3461 one-line patch can be used to allow (?-i) to work as expected.
3467 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3468 other effects) broke the use of negated acl sub-conditions.
3470 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3473 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3474 "Deny" causes Exim to reject the incoming connection with a 554 error.
3475 Unfortunately, if there is a major crisis, such as a disk failure,
3476 tcp-wrappers gives "deny", whereas what one would like would be some
3477 kind of temporary error. A kludge has been added to help with this.
3478 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3479 554 error is used if errno is still zero or contains ENOENT (which occurs
3480 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3483 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3485 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3486 errors. Otherwise a message that provokes a temporary error (when other
3487 messages do not) can cause a whole host to time out.
3489 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3490 addresses were routed directly to files or pipes from a redirect router.
3491 File deliveries just didn't batch; pipe deliveries might have suffered
3494 PH/06 A failure to get a lock for a hints database would erroneously always say
3495 "Failed to get write lock", even when it was really a read lock.
3497 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3498 of 0600. This has been changed to use the value of the lockfile_mode
3499 option (which defaults to 0600).
3501 PH/08 Applied small patch from the Sieve maintainer.
3503 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3504 folder from quota calculations, a direct delivery into this folder messed
3505 up the contents of the maildirsize file. This was because the regex was
3506 used only to exclude .Trash (or whatever) when the size of the mailbox
3507 was calculated. There was no check that a delivery was happening into an
3508 excluded directory. This bug has been fixed by ignoring all quota
3509 processing for deliveries into excluded directories.
3511 PH/10 Added the maildirfolder_create_regex option to appendfile.
3517 PH/01 The code for finding all the local interface addresses on a FreeBSD
3518 system running IPv6 was broken. This may well have applied to all BSD
3519 systems, as well as to others that have similar system calls. The broken
3520 code found IPv4 interfaces correctly, but gave incorrect values for the
3521 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3522 that it would not match correctly against @[] and not recognize the IPv6
3525 PH/02 The ipliteral router was not recognizing addresses of the form user@
3526 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3528 PH/03 Added disable_ipv6.
3530 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3531 decoded form, because it is most often used to construct To: headers
3532 lines in autoreplies, and the decoded form may well be syntactically
3533 invalid. However, $reply_address has leading white space removed, and all
3534 newlines turned into spaces so that the autoreply transport does not
3537 PH/05 If group was specified without a user on a router, and no group or user
3538 was specified on a transport, the group from the router was ignored.
3540 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3541 for visible compile-time settings that can be used to change these
3542 numbers, for those that want even more. Backwards compatibility with old
3543 spool files has been maintained. However, going back to a previous Exim
3544 release will lost any variables that are in spool files.
3546 PH/07 Two small changes when running in the test harness: increase delay when
3547 passing a TCP/IP connection to a new process, in case the original
3548 process has to generate a bounce, and remove special handling of
3549 127.0.0.2 (sic), which is no longer necessary.
3551 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3552 be the same on different OS.
3554 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3557 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3558 whether --show-vars was specified or not
3560 JJ/02 exipick: Added support for new ACL variable spool format introduced
3563 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3564 syntactically invalid From: or Reply-to: line, and a filter used this to
3565 generate an autoreply, and therefore failed to obtain an address for the
3566 autoreply, Exim could try to deliver to a non-existent relative file
3567 name, causing unrelated and misleading errors. What now happens is that
3568 it logs this as a hard delivery error, but does not attempt to create a
3571 PH/11 The exinext utility has a -C option for testing purposes, but although
3572 the given file was scanned by exinext itself; it wasn't being passed on
3573 when Exim was called.
3575 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3576 an end-of-file indication when reading a command response.
3578 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3579 compiled. In many other places in Exim, IPv6 addresses are always
3580 recognized, so I have changed this. It also means that IPv4 domain
3581 literals of the form [IPV4:n.n.n.n] are now always recognized.
3583 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3584 used if the router is not running as root, for example, when verifying at
3585 ACL time, or when using -bh. The debugging output from this situation was
3586 non-existent - all you got was a failure to exec. I have made two
3589 (a) Failures to set uid/gid, the current directory, or a process leader
3590 in a subprocess such as that created by queryprogram now generate
3591 suitable debugging output when -d is set.
3593 (b) The queryprogram router detects when it is not running as root,
3594 outputs suitable debugging information if -d is set, and then runs
3595 the subprocess without attempting to change uid/gid.
3597 PH/15 Minor change to Makefile for building test_host (undocumented testing
3600 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3601 additional section of a DNS packet that returns MX or SRV records.
3602 Instead, it always explicitly searches for A/AAAA records. This avoids
3603 major problems that occur when a DNS server includes only records of one
3604 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3605 fixed another bug: if SRV records were looked up and the corresponding
3606 address records were *not* found in the additional section, the port
3607 values from the SRV records were lost.
3609 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3610 using the correct key (the original address) when searching the retry
3611 rules in order to find which one to use for generating the retry hint.
3613 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3614 adding the default one. Similarly, if it contains a Reply-To: header, the
3615 errors_reply_to option, if set, is not used.
3617 PH/19 When calculating a retry time, Exim used to measure the "time since
3618 failure" by looking at the "first failed" field in the retry record. Now
3619 it does not use this if it is later than than the arrival time of the
3620 message. Instead it uses the arrival time. This makes for better
3621 behaviour in cases where some deliveries succeed, thus re-setting the
3622 "first failed" field. An example is a quota failure for a huge message
3623 when small messages continue to be delivered. Without this change, the
3624 "time since failure" will always be short, possible causing more frequent
3625 delivery attempts for the huge message than are intended.
3626 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3628 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3629 $1, $2, $3) because the numerical variables can be reset during some
3630 expansion items (e.g. "match"), thereby losing the authentication data.
3632 PH/21 Make -bV show the size of off_t variables so that the test suite can
3633 decide whether to run tests for quotas > 2G.
3635 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3636 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3637 filecount value is greater than 2G or if a quota value is greater than 2G
3638 on a system where the size of off_t is not greater than 4, a panic error
3641 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3642 never match. The debug and -bh output now contains an explicit error
3643 message indicating a malformed IPv4 address or mask.
3645 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3646 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3647 PH/23 above applies.
3649 PH/25 Do not write to syslog when running in the test harness. The only
3650 occasion when this arises is a failure to open the main or panic logs
3651 (for which there is an explicit test).
3653 PH/26 Added the /no_tell option to "control=freeze".
3655 PH/27 If a host name lookup failed very early in a connection, for example, if
3656 the IP address matched host_lookup and the reverse lookup yielded a name
3657 that did not have a forward lookup, an error message of the form "no IP
3658 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3659 could be logged. Now it outputs the IP address instead of "NULL".
3661 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3662 allows the sender and the authenticated sender to be set when
3663 submitting a message from within Exim. Since child_open_exim() is
3664 documented for local_scan(), the new function should be too.
3666 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3667 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3668 results in an empty string is now treated as unset.
3670 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3672 PH/31 Added sender_verify_fail logging option.
3674 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3675 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3676 tidied the source and removed it altogether.
3678 PH/33 When a queue run was abandoned because the load average was too high, a
3679 log line was always written; now it is written only if the queue_run log
3680 selector is set. In addition, the log line for abandonment now contains
3681 information about the queue run such as the pid. This is always present
3682 in "start" and "stop" lines but was omitted from the "abandon" line.
3684 PH/34 Omit spaces between a header name and the colon in the error message that
3685 is given when verify = headers_syntax fails (if there are lots of them,
3686 the message gets confusing).
3688 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3689 names, as there are now some PTR records that contain slashes. This check
3690 is only to protect against broken name servers that fall over on strange
3691 characters, so the fact that it applies to all lookups doesn't matter.
3693 PH/36 Now that the new test suite is complete, we can remove some of the
3694 special code in Exim that was needed for the old test suite. For example,
3695 sorting DNS records because real resolvers return them in an arbitrary
3696 order. The new test suite's fake resolver always returns records in the
3699 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3700 bounces) except when queue_only is set, to avoid logging races between
3701 the different processes.
3703 PH/38 Panic-die if .include specifies a non-absolute path.
3705 PH/39 A tweak to the "H" retry rule from its user.
3707 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3708 a label. They prevented compilation on older perls.
3710 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3711 a warning to be raised on newish perls.
3713 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3714 on queue. Changes to match documented behaviour of showing count of
3715 messages matching specified criteria.
3717 PH/40 Changed the default ident timeout from 30s to 5s.
3719 PH/41 Added support for the use of login_cap features, on those BSD systems
3720 that have them, for controlling the resources used by pipe deliveries.
3722 PH/42 The content-scanning code uses fopen() to create files in which to put
3723 message data. Previously it was not paying any attention to the mode of
3724 the files. Exim runs with umask(0) because the rest of the code creates
3725 files with open(), and sets the required mode explicitly. Thus, these
3726 files were ending up world-writeable. This was not a big issue, because,
3727 being within the spool directory, they were not world-accessible. I have
3728 created a function called modefopen, which takes an additional mode
3729 argument. It sets umask(777), creates the file, chmods it to the required
3730 mode, then resets the umask. All the relevant calls to fopen() in the
3731 content scanning code have been changed to use this function.
3733 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3734 to 24 hours. This avoids potential overflow problems when processing G
3735 and H retry rules. I suspect nobody ever tinkers with this value.
3737 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3739 PH/45 When the plaintext authenticator is running as a client, the server's
3740 challenges are checked to ensure they are valid base64 strings. By
3741 default, the authentication attempt is cancelled if an invalid string is
3742 received. Setting client_ignore_invalid_base64 true ignores these errors.
3743 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3744 they are received. Thus, the responses can be made to depend on the
3745 challenges. If an invalid string is ignored, an empty string is placed in
3748 PH/46 Messages that are created by the autoreply transport now contains a
3749 References: header, in accordance with RFCs 2822 and 3834.
3751 PH/47 Added authenticated_sender_force to the smtp transport.
3753 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3755 PH/49 Installed latest patch from the Sieve maintainer.
3757 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3758 was also set, the appendfile transport was unnecessarily scanning a
3759 directory of message files (e.g. for maildir delivery) to find the count
3760 of files (along with the size), even though it did not need this
3761 information. It now does the scan only if it needs to find either the
3762 size of the count of files.
3764 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3766 PH/52 Two bugs concerned with error handling when the smtp transport is
3769 (i) Exim was not creating retry information for temporary errors given
3770 for individual recipients after the DATA command when the smtp transport
3771 was used in LMTP mode. This meant that they could be retried too
3772 frequently, and not timed out correctly.
3774 (ii) Exim was setting the flag that allows error details to be returned
3775 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3776 recipients that were returned after the DATA command.
3778 PH/53 This is related to PH/52, but is more general: for any failing address,
3779 when detailed error information was permitted to be returned to the
3780 sender, but the error was temporary, then after the final timeout, only
3781 "retry timeout exceeded" was returned. Now it returns the full error as
3782 well as "retry timeout exceeded".
3784 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3785 do this, and (what is worse) MTAs that accept it.
3787 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3788 will now be deprecated.
3790 PH/56 New os.c-cygwin from the Cygwin maintainer.
3792 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3793 formats (previously only available in exim formats via -bpr, -bpru,
3794 and -bpra. Now also available in native and exiqgrep formats)
3796 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3797 with very large, slow to parse queues
3799 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3801 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3803 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3804 responses to authentication challenges, though it was showing the
3805 challenges; (ii) I've removed the CR characters from the debug output for
3808 PH/58 Allow for the insertion of a newline as well as a space when a string
3809 is turned into more than one encoded-word during RFC 2047 encoding. The
3810 Sieve code now uses this.
3812 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3813 data_4xx, lost_connection, tls_required.
3815 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3816 message was being sent as an SMTP response.
3818 PH/61 Add -l and -k options to exicyclog.
3820 PH/62 When verifying, if an address was redirected to one new address, so that
3821 verification continued, and the new address failed or deferred after
3822 having set something in $address_data, the value of $address_data was not
3823 passed back to the ACL. This was different to the case when no
3824 redirection occurred. The value is now passed back in both cases.
3826 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3827 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3828 use login_cap.h, so on its own it isn't the distinguishing feature. The
3829 new name refers directly to the setclassresources() function.
3831 PH/65 Added configuration files for NetBSD3.
3833 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3835 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3836 is preferred over IPv4.
3838 PH/68 The bounce_return_message and bounce_return_body options were not being
3839 honoured for bounces generated during the reception of non-SMTP messages.
3840 In particular, this applied to messages rejected by the ACL. This bug has
3841 been fixed. However, if bounce_return_message is true and bounce_return_
3842 body is false, the headers that are returned for a non-SMTP message
3843 include only those that have been read before the error was detected.
3844 (In the case of an ACL rejection, they have all been read.)
3846 PH/69 The HTML version of the specification is now built in a directory called
3847 spec_html instead of spec.html, because the latter looks like a path with
3848 a MIME-type, and this confuses some software.
3850 PH/70 Catch two compiler warnings in sieve.c.
3852 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3853 function verify_get_ident() calls ip_connect() to connect a socket, but
3854 if the "connect()" function timed out, ip_connect() used to close the
3855 socket. However, verify_get_ident() also closes the socket later, and in
3856 between Exim writes to the log, which may get opened at this point. When
3857 the socket was closed in ip_connect(), the log could get the same file
3858 descriptor number as the socket. This naturally causes chaos. The fix is
3859 not to close the socket in ip_connect(); the socket should be closed by
3860 the function that creates it. There was only one place in the code where
3861 this was missing, in the iplookup router, which I don't think anybody now
3862 uses, but I've fixed it anyway.
3864 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3865 well as to direct DNS lookups. Otherwise the handling of names in host
3866 lists is inconsistent and therefore confusing.
3872 PH/01 Two changes to the default runtime configuration:
3874 (1) Move the checks for relay_from_hosts and authenticated clients from
3875 after to before the (commented out) DNS black list checks.
3877 (2) Add control=submission to the relay_from_hosts and authenticated
3878 clients checks, on the grounds that messages accepted by these
3879 statements are most likely to be submissions.
3881 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3883 (1) Generate an error if the third argument for the ${prvs expansion is
3886 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3889 (3) Reset the variables that are obtained from the first argument of
3890 ${prvscheck and used in the second argument before leaving the code,
3891 because their memory is reclaimed, so using them afterwards may do
3894 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3895 one (it's much easier than Tom thought :-).
3897 (5) Because of (4), we can now allow for the use of $prvscheck_result
3898 inside the third argument.
3900 PH/03 For some reason, the default setting of PATH when running a command from
3901 a pipe transport was just "/usr/bin". I have changed it to
3904 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3905 anything to be listed in the output from -bV.
3907 PH/05 When a filter generated an autoreply, the entire To: header line was
3908 quoted in the delivery log line, like this:
3910 => >A.N.Other <ano@some.domain> <original@ddress> ...
3912 This has been changed so that it extracts the operative address. There
3913 may be more than one such address. If so, they are comma-separated, like
3916 => >ano@some.domain,ona@other.domain <original@ddress> ...
3918 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3919 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3920 not being looked up in the rDNS to get a host name, Exim was showing the
3921 IP address twice in Received: lines, even though the IP addresses were
3922 identical. For example:
3924 Received: from [1.2.3.4] (helo=[1.2.3.4])
3926 However, if the real host name was known, it was omitting the HELO data
3927 if it matched the actual IP address. This has been tidied up so that it
3928 doesn't show the same IP address twice.
3930 PH/07 When both +timestamp and +memory debugging was on, the value given by
3931 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3932 called by the debug printing, thereby overwriting the timestamp buffer.
3933 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3935 PH/08 When the original message was included in an autoreply transport, it
3936 always said "this is a copy of the message, including all the headers",
3937 even if body_only or headers_only was set. It now gives an appropriate
3940 PH/09 Applied a patch from the Sieve maintainer which:
3942 o fixes some comments
3943 o adds the (disabled) notify extension core
3944 o adds some debug output for the result of if/elsif tests
3945 o points to the current vacation draft in the documentation
3946 and documents the missing references header update
3950 o fixes a bug in processing the envelope test (when testing
3951 multiple envelope elements, the last element determined the
3954 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3955 Electronic Mail") by including:
3957 Auto-submitted: auto-generated
3959 in the messages that it generates (bounce messages and others, such as
3960 warnings). In the case of bounce messages for non-SMTP messages, there was
3961 also a typo: it was using "Auto_submitted" (underscore instead of
3962 hyphen). Since every message generated by Exim is necessarily in response
3963 to another message, thes have all been changed to:
3965 Auto-Submitted: auto-replied
3967 in accordance with these statements in the RFC:
3969 The auto-replied keyword:
3971 - SHOULD be used on messages sent in direct response to another
3972 message by an automatic process,
3974 - MUST NOT be used on manually-generated messages,
3976 - MAY be used on Delivery Status Notifications (DSNs) and Message
3977 Disposition Notifications (MDNs),
3979 - MUST NOT be used on messages generated by automatic or periodic
3980 processes, except for messages which are automatic responses to
3983 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3984 to the default Received: header definition.
3986 PH/12 Added log selector acl_warn_skipped (default on).
3988 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3989 variables because (a) they are in the wrong storage pool and (b) even if
3990 they were copied, it wouldn't work properly because of the caching.
3992 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3993 checking when decoding. Apparently there are clients that generate
3994 overlong encoded strings. Why am I not surprised?
3996 PH/15 If the first argument of "${if match_address" was not empty, but did not
3997 contain an "@" character, Exim crashed. Now it writes a panic log message
3998 and treats the condition as false.
4000 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4002 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4003 extension "envelope-auth". The code is finished and in agreement with
4004 other implementations, but there is no documentation so far and in fact,
4005 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4006 not changing the active code.
4008 Print executed "if" and "elsif" statements when debugging is used. This
4009 helps a great deal to understand what a filter does.
4011 Document more things not specified clearly in RFC3028. I had all this
4012 sorted out, when out of a sudden new issues came to my mind. Oops."
4014 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4017 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4018 canonical form (as documented). However, after a host name lookup from
4019 the IP address, check_host() was doing a simple string comparison with
4020 addresses acquired from the DNS when checking that the found name did
4021 have the original IP as one of its addresses. Since any found IPv6
4022 addresses are likely to be in abbreviated form, the comparison could
4023 fail. Luckily, there already exists a function for doing the comparison
4024 by converting both addresses to binary, so now that is used instead of
4025 the text comparison.
4027 PH/20 There was another similar case to PH/19, when a complete host name was
4028 given in a host list; looking up its IP address could give an abbreviated
4029 form, whereas the current host's name might or might not be abbreviated.
4030 The same fix has been applied.
4036 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4037 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4040 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4041 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4043 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4045 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4046 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4047 Also, for "Auto-Submitted", treat anything other than "no" as
4048 non-personal, in accordance with RFC 3834. (Previously it treated
4049 anything starting "auto-" as non-personal.)
4051 TF/01 The control=submission/name=... option had a problem with syntax
4052 errors if the name included a slash character. The /name= option
4053 now slurps the rest of the string, so it can include any characters
4054 but it must come last in the list of options (after /sender_retain
4057 PH/05 Some modifications to the interface to the fake nameserver for the new
4065 TK/01 Added the "success_on_redirect" address verification option. See
4066 NewStuff for rationale and an example.
4068 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4070 PH/02 Patch to exigrep to allow it to work on syslog lines.
4072 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4073 fread() to skip over the body file's header line, because in Cygwin the
4074 header line is locked and is inaccessible.
4076 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4077 co-exist for some time) to make it clear that it is the Exim ID that is
4078 referenced, not the Message-ID: header line.
4080 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4081 string_format() function, because snprintf() does not exist on all
4084 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4085 ${stat: expansion item.
4087 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4088 protocol synchronization error", to keep the pedants happy.
4090 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4091 well as for IRIX systems, when gcc is being used. See the host.c source
4094 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4096 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4099 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4100 transport and to the smtp transport in LMTP mode.
4102 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4104 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4105 run a filter in a subprocess. This could lead to confusion in subsequent
4106 lookups in the parent process. There should also be a search_tidyup() at
4107 the end of the subprocess.
4109 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4110 only if the host matched helo_try_verify_hosts, which caused the
4111 verification to occur when the EHLO/HELO command was issued. The ACL just
4112 tested the remembered result. Now, if a previous verification attempt has
4113 not happened, "verify = helo" does it there and then.
4115 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4117 TK/03 Fix log output including CR from clamd.
4119 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4120 exist provoked a memory error which could cause a segfault.
4122 PH/15 Installed PCRE 6.2
4124 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4126 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4127 of the problem. Specifically, suggested +O2 rather than +O1 for the
4130 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4132 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4133 record, the port was not correctly logged when the outgoing_port log
4134 selector was set (it logged the transort's default port).
4136 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4137 fallback_hosts, and "hosts" in the smtp transport.
4139 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4140 host errors such as "Connection refused".
4142 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4143 authentication with radiusclient 0.4.9:
4145 - Error returned from rc_read_config was caught wrongly
4146 - Username/password not passed on to radius server due to wrong length.
4148 The presumption is that some radiusclient API changes for 4.51/PH/17
4149 were not taken care of correctly. The code is still untested by me (my
4150 Linux distribution still has 0.3.2 of radiusclient), but it was
4151 contributed by a Radius user.
4153 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4154 expanding the "port" option of the smtp transport.
4156 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4157 while reading a MIME header. Thanks to Tom Hughes for a patch.
4159 PH/24 Include config.h inside local_scan.h so that configuration settings are
4162 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4163 that in an ACL for RCPT (for example), you can examine exactly what was
4166 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4167 commands, but it was not correctly comparing the address with the actual
4168 client host address. Thus, it would show the EHLO address in Received:
4169 header lines when this was not necessary.
4171 PH/27 Added the % operator to ${eval:}.
4173 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4174 it should be ignoring failures (because with -C, for example, it has lost
4175 privilege). It wasn't ignoring creation failures other than "already
4178 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4181 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4182 because an input line was too long, either on its own, or by virtue of
4183 too many continuations, the temporary file was not being removed, and the
4184 return code was incorrect.
4186 PH/31 Missing "BOOL" in function definition in filtertest.c.
4188 PH/32 Applied Sieve patches from the maintainer.
4190 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4192 PH/33 Added "verify = not_blind".
4194 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4195 Local/Makefile (with some defaults set). These are used in built scripts
4196 such as exicyclog, but they have never been used in the exim_install
4197 script (though there are many overriding facilities there). I have
4198 arranged that the exim_install script now takes note of these two
4201 PH/35 Installed configuration files for Dragonfly.
4203 PH/36 When a locally submitted message by a trusted user did not contain a
4204 From: header, and the sender address was obtained from -f or from an SMTP
4205 MAIL command, and the trusted user did not use -F to supply a sender
4206 name, $originator_name was incorrectly used when constructing a From:
4207 header. Furthermore, $originator_name was used for submission mode
4208 messages from external hosts without From: headers in a similar way,
4209 which is clearly wrong.
4211 PH/37 Added control=suppress_local_fixups.
4213 PH/38 When log_selector = +received_sender was set, and the addition of the
4214 sender made the log line's construction buffer exactly full, or one byte
4215 less than full, an overflow happened when the terminating "\n" was
4218 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4219 when the result of a list match is failure because a DNS lookup failed.
4221 PH/40 RM_COMMAND is now used in the building process.
4223 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4224 the "build-* directories that it finds.
4226 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4227 address in a domain literal was a prefix of an interface address.
4229 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4230 when verifying a sender address, unless rewrite_headers is false.
4232 PH/44 Wrote a long comment about why errors_to addresses are verified as
4233 recipients, not senders.
4235 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4236 the ratelimit ACL was added.
4238 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4240 PH/46 Added extra information about PostgreSQL errors to the error string.
4242 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4243 avoiding the need to install special zones in a real server. This is
4244 backwards compatible; if it can't find the fake resolver, it drops back.
4245 Thus, both old and new test suites can be run.
4247 TF/02 Added util/ratelimit.pl
4249 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4250 clock is set back in time.
4252 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4253 Candler <B.Candler@pobox.com>.
4255 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4256 OK for addresses that are the result of redirections.
4258 PH/48 A number of further additions for the benefit of the new test suite,
4259 including a fake gethostbyname() that interfaces to the fake DNS resolver
4262 TF/06 The fix for widen_domains has also been applied to qualify_single and
4263 search_parents which are the other dnslookup options that can cause
4266 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4269 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4271 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4272 service" if it encountered an MX record with an empty target hostname.
4273 The message is now "an MX or SRV record indicated no SMTP service".
4275 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4276 if the DNS of the sending site is misconfigured. This is quite a
4277 common situation. This change restores the behaviour of treating a
4278 helo verification defer as a failure.
4280 PH/51 If self=fail was set on a router, the bounce message did not include the
4281 actual error message.
4287 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4289 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4290 command itself ended in error, the underlying message about the transport
4291 filter timeout was being overwritten with the pipe command error. Now the
4292 underlying error message should be appended to the second error message.
4294 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4296 PH/02 Reduce the amount of output that "make" produces by default. Full output
4297 can still be requested.
4299 PH/03 The warning log line about a condition test deferring for a "warn" verb
4300 was being output only once per connection, rather than after each
4301 occurrence (because it was using the same function as for successful
4302 "warn" verbs). This seems wrong, so I have changed it.
4304 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4305 it should not have, which might have caused a crash in the right
4306 circumstances, but probably never did.
4308 PH/04 Installed a modified version of Tony Finch's patch to make submission
4309 mode fix the return path as well as the Sender: header line, and to
4310 add a /name= option so that you can make the user's friendly name appear
4313 TF/03 Added the control = fakedefer ACL modifier.
4315 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4316 Mark Lowes for thorough testing.
4318 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4320 TK/03 Merged latest SRS patch from Miles Wilton.
4322 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4323 with the definition in sysexits.h (which is #included earlier).
4324 Fortunately, Exim does not actually use EX_OK. The code used to try to
4325 preserve the sysexits.h value, by assuming that macro definitions were
4326 scanned for macro replacements. I have been disabused of this notion,
4327 so now the code just undefines EX_OK before #including unistd.h.
4329 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4330 in the smtp transport. When a block could not be written in a single
4331 write() function, the timeout was being re-applied to each part-write.
4332 This seems wrong - if the receiver was accepting one byte at a time it
4333 would take for ever. The timeout is now adjusted when this happens. It
4334 doesn't have to be particularly precise.
4336 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4337 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4339 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4340 without a domain if the check to <postmaster@domain> fails.
4342 SC/01 Eximstats: added -xls and the ability to specify output files
4343 (patch written by Frank Heydlauf).
4345 SC/02 Eximstats: use FileHandles for outputting results.
4347 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4349 SC/04 Eximstats: fixed display of large numbers with -nvr option
4351 SC/05 Eximstats: fixed merging of reports with empty tables.
4353 SC/06 Eximstats: added the -include_original_destination flag
4355 SC/07 Eximstats: removed tabs and trailing whitespace.
4357 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4359 TK/06 MBOX spool code: Add real "From " MBOX separator line
4360 so the .eml file is really in mbox format (even though
4361 most programs do not really care). Patch from Alex Miller.
4363 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4364 The latter is generated from $received_to and is only set if the
4365 message has one envelope recipient. SA can use these headers,
4366 obviously out-of-the-box. Patch from Alex Miller.
4368 PH/08 The ${def test on a variable was returning false if the variable's
4369 value was "0", contrary to what the specification has always said!
4370 The result should be true unless the variable is empty.
4372 PH/09 The syntax error of a character other than { following "${if
4373 def:variable_name" (after optional whitespace) was not being diagnosed.
4374 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4375 accidental colon was present, for example, could give incorrect results.
4377 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4378 result is used (not including appendfile, where other changes are about
4381 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4382 This involved changing a lot of size variables from int to off_t. It
4383 should work with maildirs and everything.
4385 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4386 spamd dying while we are connected to it.
4388 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4391 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4392 function for BDB 4.3.
4394 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4396 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4397 the sources for PCRE have been re-arranged and more files are now
4400 PH/15 The code I had for printing potentially long long variables in PH/11
4401 above was not the best (it lost precision). The length of off_t variables
4402 is now inspected at build time, and an appropriate printing format (%ld
4403 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4404 to be "long long int" or "long int". This is needed for the internal
4405 formatting function string_vformat().
4407 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4408 the configuration file to be ":syslog", then the script "guesses" where
4409 the logs files are, rather than using the compiled in default. In our
4410 case the guess is not the same as the compiled default, so the script
4411 suddenly stopped working when I started to use syslog. The patch checks
4412 to see if log_file_path is "". If so, it attempts to read it from exim
4413 with no configuration file to get the compiled in version, before it
4414 falls back to the previous guessing code."
4416 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4417 implementing BATV in an Exim configuration. See NewStuff for the gory
4420 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4421 Makefile that are specific to HP-UX.
4423 PH/18 If the "use_postmaster" option was set for a recipient callout together
4424 with the "random" option, the postmaster address was used as the MAIL
4425 FROM address for the random test, but not for the subsequent recipient
4426 test. It is now used for both.
4428 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4429 patch removes a few documentation additions to RFC 3028, because the
4430 latest draft now contains them. It adds the new en;ascii-case comparator
4431 and a new error check for 8bit text in MIME parts. Comparator and
4432 require names are now matched exactly. I enabled the subaddress
4433 extension, but it is not well tested yet (read: it works for me)."
4435 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4436 rework some of the code of TK/09 above to avoid the hardwired use of
4437 "%lld" and "long long". Replaced the call to snprintf() with a call to
4440 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4441 records point to non-existent hosts", "retry timeout exceeded", and
4442 "retry time not reached for any host after a long failure period".
4444 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4445 experimental DomainKeys support:
4447 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4448 (2) On an error such as an illegally used "control", the wrong name for
4449 the control was given.
4451 These problems did NOT occur unless DomainKeys support was compiled.
4453 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4455 PH/24 Added ${if match_ip condition.
4457 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4458 databases so that it will be absolutely obvious if a crash occurs in the
4459 DB library. This is a regular occurrence (often caused by mis-matched
4462 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4463 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4464 picked up on a user's system that detects such things. There doesn't seem
4465 to be a gcc warning option for this - only an attribute that has to be
4466 put on the function's prototype. It seems that in Fedora Core 4 they have
4467 set this on a number of new functions. No doubt there will be more in due
4470 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4471 specify a transport. However, if an address that was verified by such a
4472 router was the subject of a callout, Exim crashed because it tried to
4473 read the rcpt_include_affixes from the non-existent transport. Now it
4474 just assumes that the setting of that option is false. This bug was
4475 introduced by 4.51/PH/31.
4477 PH/28 Changed -d+all to exclude +memory, because that information is very
4478 rarely of interest, but it makes the output a lot bigger. People tend to
4479 do -d+all out of habit.
4481 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4482 code in os-type was giving problems when libc.so lives in lib64, like on
4485 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4486 aren't the modern standard, and it seems that some systems' include files
4487 don't always have them. Exim was already checking for some of the newer
4488 ones like T_AAAA, and defining it itself. I've added checks for all the
4489 record types that Exim uses.
4491 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4492 not automatically generating a new one, as it is supposed to. This
4493 prevented TLS from working. If the file did exist, but contained invalid
4494 data, a new version was generated, as expected. It was only the case of a
4495 non-existent file that was broken.
4497 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4498 with a change in libdomainkeys > 0.64.
4500 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4501 from DNS. If the selector record carries the flag, it now has
4502 precedence over the domain-wide flag.
4504 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4506 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4507 the use of a port name that isn't defined in /etc/services) occurred, the
4508 message was deferred as in a normal delivery, and thus remained on the
4509 spool, instead of being failed because of the mua_wrapper setting. This
4510 is now fixed, and I tidied up some of the mua_wrapper messages at the
4513 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4514 the messages in a hash of arrays rather than using individual hashes.
4515 This is a bit cleaner and results in dramatic memory savings, albeit
4516 at a slight CPU cost.
4518 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4519 as requested by Marc Sherman.
4521 SC/10 Eximstats: added histograms for user specified patterns as requested
4524 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4526 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4527 fopen() in the content-scanning modules that did not already have it.
4533 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4534 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4536 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4538 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4540 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4541 to test Sieve filters that use "vacation".
4543 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4544 that changes the way the GnuTLS parameters are stored in the cache file.
4545 The new format can be generated externally. For backward compatibility,
4546 if the data in the cache doesn't make sense, Exim assumes it has read an
4547 old-format file, and it generates new data and writes a new file. This
4548 means that you can't go back to an older release without removing the
4551 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4552 work if there are any delivery delays because "one_time" forces the
4553 parent to be marked "delivered", so its unseen clone is never tried
4554 again. For this reason, Exim now forbids the simultaneous setting of
4557 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4558 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4559 entry if you want to know the details. The fix, however, neglected to
4560 consider the case when local delivery batching is involved. The test for
4561 "previously delivered" was not happening when checking to see if an
4562 address could be batched with a previous (undelivered) one; under
4563 certain circumstances this could lead to multiple deliveries to the same
4566 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4567 in its include files, and this causes problems building Exim.
4569 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4570 header_syntax) but Exim was just ignoring anything given after a slash.
4571 In particular, this caused confusion with an attempt to use "verify =
4572 reverse_host_lookup/defer_ok". An error is now given when options are
4573 supplied for verify items that do not have them. (Maybe reverse_host_
4574 lookup should have a defer_ok option, but that's a different point.)
4576 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4577 defined by RFC 821) to 2048, because there were problems with some AUTH
4578 commands, and RFC 1869 says the size should be increased for extended
4579 SMTP commands that take arguments.
4581 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4584 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4585 "unknown" error; now it says that the functionality isn't in the binary.
4587 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4588 an address' error message when a string expansion fails (syntax or
4589 whatever). Otherwise the password may appear in the log. Following change
4590 PH/42 below, there is no longer a chance of it appearing in a bounce
4593 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4595 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4596 of its addresses was ever tried. (Bugzilla bug #2.)
4598 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4599 the result incorrectly in the debug output. (It correctly added a newline
4600 to what was transported.)
4602 TF/01 Added $received_time.
4604 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4605 commented out examples of how to interface to a virus scanner and to
4606 SpamAssassin. Also added commented examples of av_scanner and
4607 spamd_address settings.
4609 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4610 and controls are allowed in which ACLs. There were a couple of minor
4611 errors. Some of the entries in the conditions table (which is a table of
4612 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4613 negation of where the condition IS allowed.
4615 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4617 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4618 header file does not have a version number, so I've had to invent a new
4619 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4620 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4621 radiusclient), but it was contributed by a Radius user.
4623 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4624 files or directories, for OpenSSL.
4626 PH/19 When an Exim process that is running as root has to create an Exim log
4627 file, it does so in a subprocess that runs as exim:exim so as to get the
4628 ownership right at creation (otherwise, other Exim processes might see
4629 the file with the wrong ownership). There was no test for failure of this
4630 fork() call, which would lead to the process getting stuck as it waited
4631 for a non-existent subprocess. Forks do occasionally fail when resources
4632 run out. I reviewed all the other calls to fork(); they all seem to check
4635 PH/20 When checking for unexpected SMTP input at connect time (before writing
4636 the banner), Exim was not dealing correctly with a non-positive return
4637 from the read() function. If the client had disconnected by this time,
4638 the result was a log entry for a synchronization error with an empty
4639 string after "input=" when read() returned zero. If read() returned -1
4640 (an event I could not check), uninitialized data bytes were printed.
4641 There were reports of junk text (parts of files, etc) appearing after
4644 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4646 PH/22 Added support for macro redefinition, and (re)definition in between
4647 driver and ACL definitions.
4649 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4650 forgetting to use the resulting value; it was using the unexpanded value.
4652 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4653 hadn't been configured. The fix is from Juergen Kreileder, who
4654 understands it better than I do:
4656 "Here's what I see happening with three configured cyrus_sasl
4657 authenticators configured (plain, login, cram-md5):
4659 On startup auth_cyrus_sasl_init() gets called for each of these.
4660 This means three calls to sasl_listmech() without a specified mech_list.
4661 => SASL tests which mechs of all available mechs actually work
4662 => three warnings about OTP not working
4663 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4665 With the patch, sasl_listmech() also gets called three times. But now
4666 SASL's mech_list option is set to the server_mech specified in the the
4667 authenticator. Or in other words, the answer from sasl_listmech()
4668 gets limited to just the mech you're testing for (which is different
4670 => the return list contains just 'plain' or 'login', 'cram-md5' or
4671 nothing depending on the value of ob->server_mech.
4673 I've just tested the patch: Authentication still works fine,
4674 unavailable mechs specified in the exim configuration are still
4675 caught, and the auth.log warnings about OTP are gone."
4677 PH/25 When debugging is enabled, the contents of the command line are added
4678 to the debugging output, even when log_selector=+arguments is not
4681 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4682 answer is "GNU", and only if the return is "GNU/something" is the answer
4685 PH/27 $acl_verify_message is now set immediately after the failure of a
4686 verification in an ACL, and so is available in subsequent modifiers. In
4687 particular, the message can be preserved by coding like this:
4689 warn !verify = sender
4690 set acl_m0 = $acl_verify_message
4692 Previously, $acl_verify_message was set only while expanding "message"
4693 and "log_message" when a very denied access.
4695 PH/28 Modified OS/os.c-Linux with
4697 -#ifndef OS_LOAD_AVERAGE
4698 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4700 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4701 nomenclature these days.)
4703 PH/29 Installed patch from the Sieve maintainer that adds the options
4704 sieve_useraddress and sieve_subaddress to the redirect router.
4706 PH/30 In these circumstances:
4707 . Two addresses routed to the same list of hosts;
4708 . First host does not offer TLS;
4709 . First host accepts first address;
4710 . First host gives temporary error to second address;
4711 . Second host offers TLS and a TLS session is established;
4712 . Second host accepts second address.
4713 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4714 and peerdn, if requested) that were in fact used only for the second
4717 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4718 attention to any local part prefix or suffix that was matched by the
4719 router that accepted the address. It now behaves in the same way as it
4720 does for delivery: the affixes are removed from the local part unless
4721 rcpt_include_affixes is set on the transport.
4723 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4724 timeout during the DATA phase of an incoming message.
4726 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4727 applied a patch sanctioned by the Sieve maintainer.
4729 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4730 the uid or gid is negative. A case of a negative gid caused this to be
4731 noticed. The fix allows for either to be negative.
4733 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4734 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4737 PH/36 The change PH/12 above was broken. Fixed it.
4739 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4740 the grounds that routing the same address twice would always produce the
4741 same answer. This might have been true once, but it is certainly no
4742 longer true now. Routing a child address may depend on the previous
4743 routing that produced that child. Some complicated redirection strategies
4744 went wrong when messages had multiple recipients, and made Exim's
4745 behaviour dependent on the order in which the addresses were given.
4747 I have moved the duplicate checking until after the routing is complete.
4748 Exim scans the addresses that are assigned to local and remote
4749 transports, and removes any duplicates. This means that more work will be
4750 done, as duplicates will always all be routed, but duplicates are
4751 presumably rare, so I don't expect this is of any significance.
4753 For deliveries to pipes, files, and autoreplies, the duplicate checking
4754 still happens during the routing process, since they are not going to be
4757 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4758 It corrects a timeout issue with spamd. This is Ian's comment: "The
4759 background is that sometimes spamd either never reads data from a
4760 connection it has accepted, or it never writes response data. The exiscan
4761 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4762 blindly assumes that writes won't block so it may never time out."
4764 PH/39 Allow G after quota size as well as K and M.
4766 PH/40 The value set for $authenticated_id in an authenticator may not contain
4767 binary zeroes or newlines because the value is written to log lines and
4768 to spool files. There was no check on this. Now the value is run through
4769 the string_printing() function so that such characters are converted to
4770 printable escape sequences.
4772 PH/41 $message_linecount is a new variable that contains the total number of
4773 lines in the message. Compare $body_linecount, which is the count for the
4776 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4777 bounce and delay warning messages, except in certain special cases, which
4780 (a) An SMTP error message from a remote host;
4781 (b) A message specified in a :fail: redirection;
4782 (c) A message specified in a "fail" command in a system filter;
4783 (d) A message specified in a FAIL return from the queryprogram router;
4784 (e) A message specified by the cannot_route_message router option.
4786 In these cases only, Exim does include the error details in bounce and
4787 warning messages. There are also a few cases where bland messages such
4788 as "unrouteable address" or "local delivery error" are given.
4790 PH/43 $value is now also set for the "else" part of a ${run expansion.
4792 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4793 being worked on, but at least Exim now implements the latest version to
4796 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4797 process to complete was treated as a delivery failure, a timeout while
4798 writing the message to the pipe was logged, but erroneously treated as a
4799 successful delivery. Such timeouts include transport filter timeouts. For
4800 consistency with the overall process timeout, these timeouts are now
4801 treated as errors, giving rise to delivery failures by default. However,
4802 there is now a new Boolean option for the pipe transport called
4803 timeout_defer, which, if set TRUE, converts the failures into defers for
4804 both kinds of timeout. A transport filter timeout is now identified in
4807 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4808 systems where "make" and "gmake" are different, calling "gmake" at top
4809 level broke things. I've arranged for the value of $(MAKE) to be passed
4810 from the Makefile to this script so that it can call the same version of
4814 A note about Exim versions 4.44 and 4.50
4815 ----------------------------------------
4817 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4818 changes of various kinds. As a consequence, a big documentation update was
4819 needed. This delayed the release for rather longer than seemed good, especially
4820 in the light of a couple of (minor) security issues. Therefore, the changes
4821 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4822 release. So 4.44 and 4.50 are in effect two different branches that both start
4825 I have left the 4.50 change log unchanged; it contains all the changes since
4826 4.43. The change log for 4.44 is below; many of its items are identical to
4827 those for 4.50. This seems to be the most sensible way to preserve the
4828 historical information.
4834 1. Minor wording change to the doc/README.SIEVE file.
4836 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4837 computation of the current number of files was incorrect.
4839 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4840 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4843 4. Give more explanation in the error message when the command for a transport
4844 filter fails to execute.
4846 5. There are several places where Exim runs a non-Exim command in a
4847 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4848 was being done only for the command run by the queryprogram router. It is
4849 now done for all such subprocesses. The other cases are: ${run, transport
4850 filters, and the commands run by the lmtp and pipe transports.
4852 6. Added CONFIGURE_GROUP build-time option.
4854 7. Some older OS have a limit of 256 on the maximum number of file
4855 descriptors. Exim was using setrlimit() to set 1000 as a large value
4856 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4857 systems. I've change it so that if it can't get 1000, it tries for 256.
4859 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4860 was an oversight, and furthermore, ever since the addition of extra
4861 controls (e.g. 4.43/32), the checks on when to allow different forms of
4862 "control" were broken. There should now be diagnostics for all cases when a
4863 control that does not make sense is encountered.
4865 9. Added the /retain_sender option to "control=submission".
4867 10. $recipients is now available in the predata ACL (oversight).
4869 11. Tidy the search cache before the fork to do a delivery from a message
4870 received from the command line. Otherwise the child will trigger a lookup
4871 failure and thereby defer the delivery if it tries to use (for example) a
4872 cached ldap connection that the parent has called unbind on.
4874 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4875 of $address_data from the recipient verification was clobbered by the
4876 sender verification.
4878 13. The value of address_data from a sender verification is now available in
4879 $sender_address_data in subsequent conditions in the ACL statement.
4881 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4883 15. Added a new option "connect=<time>" to callout options, to set a different
4886 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4887 was its contents. (It was OK if the option was not defined at all.)
4889 17. A "Completed" log line is now written for messages that are removed from
4890 the spool by the -Mrm option.
4892 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4893 information about exactly what failed.
4895 19. Added -dd to debug only the daemon process.
4897 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4898 handles timeouts, both on the server side and network timeouts. Renamed the
4899 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4901 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4902 It is now set to "smtps".
4904 22. $host_address is now set to the target address during the checking of
4905 ignore_target_hosts.
4907 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4908 being passed; this would have caused $sender_host_name to have been used if
4909 matching the list had actually called for a host name (not very likely,
4910 since this list is usually IP addresses). A host name is now passed as
4913 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4914 code that specifies a non-restarting handler (typically sigaction() in
4915 modern systems) in an attempt to fix a rare and obscure crash bug.
4917 25. Narrowed the window for a race in the daemon that could cause it to ignore
4918 SIGCHLD signals. This is not a major problem, because they are used only to
4919 wake it up if nothing else does.
4921 26. A malformed maildirsize file could cause Exim to calculate negative values
4922 for the mailbox size or file count. Odd effects could occur as a result.
4923 The maildirsize information is now recalculated if the size or filecount
4926 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4927 support for a long time. Removed HAVE_SYS_VFS_H.
4929 28. Installed the latest version of exipick from John Jetmore.
4931 29. In an address list, if the pattern was not a regular expression, an empty
4932 subject address (from a bounce message) matched only if the pattern was an
4933 empty string. Non-empty patterns were not even tested. This was the wrong
4934 because it is perfectly reasonable to use an empty address as part of a
4935 database query. An empty address is now tested by patterns that are
4936 lookups. However, all the other forms of pattern expect the subject to
4937 contain a local part and a domain, and therefore, for them, an empty
4938 address still always fails if the pattern is not itself empty.
4940 30. Exim went into a mad DNS loop when attempting to do a callout where the
4941 host was specified on an smtp transport, and looking it up yielded more
4942 than one IP address.
4944 31. Re-factored the code for checking spool and log partition space into a
4945 function that finds that data and another that does the check. The former
4946 is then used to implement four new variables: $spool_space, $log_space,
4947 $spool_inodes, and $log_inodes.
4949 32. The RFC2047 encoding function was originally intended for short strings
4950 such as real names; it was not keeping to the 75-character limit for
4951 encoded words that the RFC imposes. It now respects the limit, and
4952 generates multiple encoded words if necessary. To be on the safe side, I
4953 have increased the buffer size for the ${rfc2047: expansion operator from
4956 33. It is now permitted to omit both strings after an "if" condition; if the
4957 condition is true, the result is "true". As before, when the second string
4958 is omitted, a false condition yields an empty string. This makes it less
4959 cumbersome to write custom ACL and router conditions.
4961 34. Failure to deliver a bounce message always caused it to be frozen, even if
4962 there was an errors_to setting on the router. The errors_to setting is now
4965 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4966 canonical form (fully expanded) before being placed in
4967 $sender_host_address.
4969 36. The table in the code that translates DNS record types into text (T_A to
4970 "A" for instance) was missing entries for NS and CNAME. It is just possible
4971 that this could have caused confusion if both these types were looked up
4972 for the same domain, because the text type is used as part of Exim's
4973 per-process caching. But the chance of anyone hitting this buglet seems
4976 37. The dnsdb lookup has been extended in a number of ways.
4978 (1) There is a new type, "zns", which walks up the domain tree until it
4979 finds some nameserver records. It should be used with care.
4981 (2) There is a new type, "mxh", which is like "mx" except that it returns
4982 just the host names, not the priorities.
4984 (3) It is now possible to give a list of domains (or IP addresses) to be
4985 looked up. The behaviour when one of the lookups defers can be
4986 controlled by a keyword.
4988 (4) It is now possible to specify the separator character for use when
4989 multiple records are returned.
4991 38. The dnslists ACL condition has been extended: it is now possible to supply
4992 a list of IP addresses and/or domains to be looked up in a particular DNS
4995 39. Added log_selector=+queue_time_overall.
4997 40. When running the queue in the test harness, wait just a tad after forking a
4998 delivery process, to get repeatability of debugging output.
5000 41. Include certificate and key file names in error message when GnuTLS fails
5001 to set them up, because the GnuTLS error message doesn't include the name
5002 of the failing file when there is a problem reading it.
5004 42. Allow both -bf and -bF in the same test run.
5006 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5008 44. The "Exiscan patch" is now merged into the mainline Exim source.
5010 45. Sometimes the final signoff response after QUIT could fail to get
5011 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5012 < 0 before doing a fflush(). This bug looks as though it goes back to the
5013 introduction of TLS in release 3.20, but "sometimes" must have been rare
5014 because the tests only now provoked it.
5016 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5017 (this can affect the format of dates).
5019 47. exim_tidydb, when checking for the continued existence of a message for
5020 which it has found a message-specific retry record, was not finding
5021 messages that were in split spool directories. Consequently, it was
5022 deleting retry records that should have stayed in existence.
5024 48. Steve fixed some bugs in eximstats.
5026 49. The SPA authentication driver was not abandoning authentication and moving
5027 on to the next authenticator when an expansion was forced to fail,
5028 contradicting the general specification for all authenticators. Instead it
5029 was generating a temporary error. It now behaves as specified.
5031 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5032 (the order specifies the preference for clients). The order is now AES256,
5033 AES128, 3DES, ARCFOUR128.
5035 51. Small patch to Sieve code - explicitly set From: when generating an
5038 52. Exim crashed if a remote delivery caused a very long error message to be
5039 recorded - for instance if somebody sent an entire SpamAssassin report back
5040 as a large number of 550 error lines. This bug was coincidentally fixed by
5041 increasing the size of one of Exim's internal buffers (big_buffer) that
5042 happened as part of the Exiscan merge. However, to be on the safe side, I
5043 have made the code more robust (and fixed the comments that describe what
5046 53. Now that there can be additional text after "Completed" in log lines (if
5047 the queue_time_overall log selector is set), a one-byte patch to exigrep
5048 was needed to allow it to recognize "Completed" as not the last thing in
5051 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5052 patch that reportedly fixes this has been added. I am not expert enough to
5053 create a test for it. This is what the patch creator wrote:
5055 "I found a little strange behaviour of ldap code when working with
5056 Windows 2003 AD Domain, where users was placed in more than one
5057 Organization Units. When I tried to give exim partial DN, the exit code
5058 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5059 But simultaneously result of request was absolutely normal ldap result,
5060 so I produce this patch..."
5062 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5063 REFERENCE, so I have modified the code to exclude the patch when that macro
5066 55. Some experimental protocols are using DNS PTR records for new purposes. The
5067 keys for these records are domain names, not reversed IP addresses. The
5068 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5069 leaves it alone. Component reversal etc. now happens only for IP addresses.
5072 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5074 57. Double the size of the debug message buffer (to 2048) so that more of very
5075 long debug lines gets shown.
5077 58. The exicyclog utility now does better if the number of log files to keep
5078 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5080 59. Two changes related to the smtp_active_hostname option:
5082 (1) $smtp_active_hostname is now available as a variable.
5083 (2) The default for smtp_banner uses $smtp_active_hostname instead
5084 of $primary_hostname.
5086 60. The host_aton() function is supposed to be passed a string that is known
5087 to be a valid IP address. However, in the case of IPv6 addresses, it was
5088 not checking this. This is a hostage to fortune. Exim now panics and dies
5089 if the condition is not met. A case was found where this could be provoked
5090 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5091 components; fortuitously, this particular loophole had already been fixed
5092 by change 4.50/55 above.
5094 If there are any other similar loopholes, the new check in host_aton()
5095 itself should stop them being exploited. The report I received stated that
5096 data on the command line could provoke the exploit when Exim was running as
5097 exim, but did not say which command line option was involved. All I could
5098 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5099 running as the user.
5102 61. There was a buffer overflow vulnerability in the SPA authentication code
5103 (which came originally from the Samba project). I have added a test to the
5104 spa_base64_to_bits() function which I hope fixes it.
5107 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5108 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5110 63. The daemon start-up calls getloadavg() while still root for those OS that
5111 need the first call to be done as root, but it missed one case: when
5112 deliver_queue_load_max is set with deliver_drop_privilege. This is
5113 necessary for the benefit of the queue runner, because there is no re-exec
5114 when deliver_drop_privilege is set.
5116 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5117 This has been fixed.
5119 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5120 was in use, was not putting the data itself into the right store pool;
5121 consequently, it could be overwritten for a subsequent message in the same
5122 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5125 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5127 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5128 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5129 the function were treating the return as a boolean value, which happened to
5130 work because 0=false and not-0=true, but is not correct code.
5132 68. The host_aton() function was not handling scoped IPv6 addresses (those
5133 with, for example, "%eth0" on the end) correctly.
5135 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5136 negated items (that is, ~something) in unsigned ints. Some compilers
5137 apparently mutter when there is no cast.
5139 70. If an address verification called from an ACL failed, and did not produce a
5140 user-specific message (i.e. there was only a "system" message), nothing was
5141 put in $acl_verify_message. In this situation, it now puts the system
5144 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5145 session; change 4.31/43 added the unwanted input to the log line - except
5146 that it did not do this in the start of session case. It now does.
5148 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5149 This is wrong and can cause the other end to generate a synchronization
5150 error if it is another Exim or anything else that does the synchronization
5151 check. A QUIT command is no longer sent after a timeout.
5153 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5154 during host lookups.
5156 74. The defer_ok option of callout verification was not working if it was used
5157 when verifying addresses in header lines, that is, for this case:
5159 verify = header_sender/callout=defer_ok
5161 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5162 those file descriptors could be used for SMTP connections. If anything
5163 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5164 could be sent to the SMTP client, causing chaos. The daemon now opens
5165 stdin, stdout, and stderr to /dev/null when it puts itself into the
5168 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5169 log by default. The user can override this with suitable Perl magic.
5171 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5172 the log message when discard triggers, was not working for the DATA ACL or
5173 for the non-SMTP ACL.
5175 78. Error message wording change in sieve.c.
5177 79. If smtp_accept_max_per_host was set, the number of connections could be
5178 restricted to fewer than expected, because the daemon was trying to set up
5179 a new connection before checking whether the processes handling previous
5180 connections had finished. The check for completed processes is now done
5181 earlier. On busy systems, this bug wouldn't be noticed because something
5182 else would have woken the daemon, and it would have reaped the completed
5185 80. If a message was submitted locally by a user whose login name contained one
5186 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5187 It caused a spool format error. I have fixed the spool reading code. A
5188 related problem was that the "from" clause in the Received: line became
5189 illegal because of the space(s). It is now covered by ${quote_local_part.
5191 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5194 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5196 83. Patch from Sieve maintainer to fix unterminated string problem in
5197 "vacation" handling.
5199 84. Some minor changes to the Linux configuration files to help with other
5200 OS variants using glibc.
5202 85. One more patch for Sieve to update vacation handling to latest spec.
5205 ----------------------------------------------------
5206 See the note above about the 4.44 and 4.50 releases.
5207 ----------------------------------------------------
5213 1. Change 4.43/35 introduced a bug that caused file counts to be
5214 incorrectly computed when quota_filecount was set in an appendfile
5217 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5218 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5221 3. Give more explanation in the error message when the command for a transport
5222 filter fails to execute.
5224 4. There are several places where Exim runs a non-Exim command in a
5225 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5226 was being done only for the command run by the queryprogram router. It is
5227 now done for all such subprocesses. The other cases are: ${run, transport
5228 filters, and the commands run by the lmtp and pipe transports.
5230 5. Some older OS have a limit of 256 on the maximum number of file
5231 descriptors. Exim was using setrlimit() to set 1000 as a large value
5232 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5233 systems. I've change it so that if it can't get 1000, it tries for 256.
5235 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5236 was an oversight, and furthermore, ever since the addition of extra
5237 controls (e.g. 4.43/32), the checks on when to allow different forms of
5238 "control" were broken. There should now be diagnostics for all cases when a
5239 control that does not make sense is encountered.
5241 7. $recipients is now available in the predata ACL (oversight).
5243 8. Tidy the search cache before the fork to do a delivery from a message
5244 received from the command line. Otherwise the child will trigger a lookup
5245 failure and thereby defer the delivery if it tries to use (for example) a
5246 cached ldap connection that the parent has called unbind on.
5248 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5249 of $address_data from the recipient verification was clobbered by the
5250 sender verification.
5252 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5253 was its contents. (It was OK if the option was not defined at all.)
5255 11. A "Completed" log line is now written for messages that are removed from
5256 the spool by the -Mrm option.
5258 12. $host_address is now set to the target address during the checking of
5259 ignore_target_hosts.
5261 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5262 being passed; this would have caused $sender_host_name to have been used if
5263 matching the list had actually called for a host name (not very likely,
5264 since this list is usually IP addresses). A host name is now passed as
5267 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5268 code that specifies a non-restarting handler (typically sigaction() in
5269 modern systems) in an attempt to fix a rare and obscure crash bug.
5271 15. Narrowed the window for a race in the daemon that could cause it to ignore
5272 SIGCHLD signals. This is not a major problem, because they are used only to
5273 wake it up if nothing else does.
5275 16. A malformed maildirsize file could cause Exim to calculate negative values
5276 for the mailbox size or file count. Odd effects could occur as a result.
5277 The maildirsize information is now recalculated if the size or filecount
5280 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5281 support for a long time. Removed HAVE_SYS_VFS_H.
5283 18. Updated exipick to current release from John Jetmore.
5285 19. Allow an empty sender to be matched against a lookup in an address list.
5286 Previously the only cases considered were a regular expression, or an
5289 20. Exim went into a mad DNS lookup loop when doing a callout where the
5290 host was specified on the transport, if the DNS lookup yielded more than
5293 21. The RFC2047 encoding function was originally intended for short strings
5294 such as real names; it was not keeping to the 75-character limit for
5295 encoded words that the RFC imposes. It now respects the limit, and
5296 generates multiple encoded words if necessary. To be on the safe side, I
5297 have increased the buffer size for the ${rfc2047: expansion operator from
5300 22. Failure to deliver a bounce message always caused it to be frozen, even if
5301 there was an errors_to setting on the router. The errors_to setting is now
5304 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5305 canonical form (fully expanded) before being placed in
5306 $sender_host_address.
5308 24. Updated eximstats to version 1.33
5310 25. Include certificate and key file names in error message when GnuTLS fails
5311 to set them up, because the GnuTLS error message doesn't include the name
5312 of the failing file when there is a problem reading it.
5314 26. Expand error message when OpenSSL has problems setting up cert/key files.
5317 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5318 (this can affect the format of dates).
5320 28. exim_tidydb, when checking for the continued existence of a message for
5321 which it has found a message-specific retry record, was not finding
5322 messages that were in split spool directories. Consequently, it was
5323 deleting retry records that should have stayed in existence.
5325 29. eximstats updated to version 1.35
5326 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5327 1.35 - bugfix such that pie charts by volume are generated correctly
5329 30. The SPA authentication driver was not abandoning authentication and moving
5330 on to the next authenticator when an expansion was forced to fail,
5331 contradicting the general specification for all authenticators. Instead it
5332 was generating a temporary error. It now behaves as specified.
5334 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5335 (the order specifies the preference for clients). The order is now AES256,
5336 AES128, 3DES, ARCFOUR128.
5338 31. Small patch to Sieve code - explicitly set From: when generating an
5341 32. Exim crashed if a remote delivery caused a very long error message to be
5342 recorded - for instance if somebody sent an entire SpamAssassin report back
5343 as a large number of 550 error lines. This bug was coincidentally fixed by
5344 increasing the size of one of Exim's internal buffers (big_buffer) that
5345 happened as part of the Exiscan merge. However, to be on the safe side, I
5346 have made the code more robust (and fixed the comments that describe what
5349 33. Some experimental protocols are using DNS PTR records for new purposes. The
5350 keys for these records are domain names, not reversed IP addresses. The
5351 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5352 leaves it alone. Component reversal etc. now happens only for IP addresses.
5355 34. The host_aton() function is supposed to be passed a string that is known
5356 to be a valid IP address. However, in the case of IPv6 addresses, it was
5357 not checking this. This is a hostage to fortune. Exim now panics and dies
5358 if the condition is not met. A case was found where this could be provoked
5359 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5360 components; fortuitously, this particular loophole had already been fixed
5361 by change 4.50/55 or 4.44/33 above.
5363 If there are any other similar loopholes, the new check in host_aton()
5364 itself should stop them being exploited. The report I received stated that
5365 data on the command line could provoke the exploit when Exim was running as
5366 exim, but did not say which command line option was involved. All I could
5367 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5368 running as the user.
5371 35. There was a buffer overflow vulnerability in the SPA authentication code
5372 (which came originally from the Samba project). I have added a test to the
5373 spa_base64_to_bits() function which I hope fixes it.
5376 36. The daemon start-up calls getloadavg() while still root for those OS that
5377 need the first call to be done as root, but it missed one case: when
5378 deliver_queue_load_max is set with deliver_drop_privilege. This is
5379 necessary for the benefit of the queue runner, because there is no re-exec
5380 when deliver_drop_privilege is set.
5382 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5383 was in use, was not putting the data itself into the right store pool;
5384 consequently, it could be overwritten for a subsequent message in the same
5385 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5388 38. Sometimes the final signoff response after QUIT could fail to get
5389 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5390 < 0 before doing a fflush(). This bug looks as though it goes back to the
5391 introduction of TLS in release 3.20, but "sometimes" must have been rare
5392 because the tests only now provoked it.
5398 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5399 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5400 Now it returns an integer. A number of calls were still expecting a T/F
5401 return. Fortuitously, in all cases, the tests worked in OK situations,
5402 which is the norm. However, things would have gone wrong on any write
5403 failures on the smtp file descriptor. This function is used when sending
5404 messages over SMTP and also when doing verify callouts.
5406 2. When Exim is called to do synchronous delivery of a locally submitted
5407 message (the -odf or -odi options), it no longer closes stderr before doing
5410 3. Implemented the mua_wrapper option.
5412 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5414 5. Implemented the functions header_remove(), header_testname(),
5415 header_add_at_position(), and receive_remove_recipient(), and exported them
5418 6. If an ACL "warn" statement specified the addition of headers, Exim already
5419 inserted X-ACL-Warn: at the start if there was no header name. However, it
5420 was not making this test for the second and subsequent header lines if
5421 there were newlines in the string. This meant that an invalid header could
5422 be inserted if Exim was badly configured.
5424 7. Allow an ACL "warn" statement to add header lines at the start or after all
5425 the Received: headers, as well as at the end.
5427 8. Added the rcpt_4xx retry error code.
5429 9. Added postmaster_mailfrom=xxx to callout verification option.
5431 10. Added mailfrom=xxxx to the callout verification option, for verify=
5434 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5435 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5437 12. Inserted some casts to stop certain compilers warning when using pointer
5438 differences as field lengths or precisions in printf-type calls (mostly
5439 affecting debugging statements).
5441 13. Added optional readline() support for -be (dynamically loaded).
5443 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5444 same clock tick as a message's arrival, so that its received time was the
5445 same as the "first fail" time on the retry record, and that message
5446 remained on the queue past the ultimate address timeout, every queue runner
5447 would try a delivery (because it was past the ultimate address timeout) but
5448 after another failure, the ultimate address timeout, which should have then
5449 bounced the address, did not kick in. This was a "< instead of <=" error;
5450 in most cases the first failure would have been in the next clock tick
5451 after the received time, and all would be well.
5453 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5454 being recognized when the domain list was tested by the match_domain
5455 condition in an expansion string.
5457 16. Added the ${str2b64: operator.
5459 17. Exim was always calling setrlimit() to set a large limit for the number of
5460 processes, without checking whether the existing limit was already
5461 adequate. (It did check for the limit on file descriptors.) Furthermore,
5462 errors from getrlimit() and setrlimit() were being ignored. Now they are
5463 logged to the main and panic logs, but Exim does carry on, to try to do its
5464 job under whatever limits there are.
5466 18. Imported PCRE 5.0.
5468 19. Trivial typo in log message " temporarily refused connection" (the leading
5471 20. If the log selector return_path_on_delivery was set and an address was
5472 redirected to /dev/null, the delivery process crashed because it assumed
5473 that a return path would always be set for a "successful" delivery. In this
5474 case, the whole delivery is bypassed as an optimization, and therefore no
5477 21. Internal re-arrangement: the function for sending a challenge and reading
5478 a response while authentication was assuming a zero-terminated challenge
5479 string. It's now changed to take a pointer and a length, to allow for
5480 binary data in such strings.
5482 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5484 23. Exim was not respecting finduser_retries when seeking the login of the
5485 uid under which it was called; it was always trying 10 times. (The default
5486 setting of finduser_retries is zero.) Also, it was sleeping after the final
5487 failure, which is pointless.
5489 24. Implemented tls_on_connect_ports.
5491 25. Implemented acl_smtp_predata.
5493 26. If the domain in control=submission is set empty, Exim assumes that the
5494 authenticated id is a complete email address when it generates From: or
5495 Sender: header lines.
5497 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5498 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5499 chown and chgrp in /bin and hostname in /usr/bin.
5501 28. Exim was keeping the "process log" file open after each use, just as it
5502 does for the main log. This opens the possibility of it remaining open for
5503 long periods when the USR1 signal hits a daemon. Occasional processlog
5504 errors were reported, that could have been caused by this. Anyway, it seems
5505 much more sensible not to leave this file open at all, so that is what now
5508 29. The long-running daemon process does not normally write to the log once it
5509 has entered its main loop, and it closes the log before doing so. This is
5510 so that log files can straightforwardly be renamed and moved. However,
5511 there are a couple of unusual error situations where the daemon does write
5512 log entries, and I had neglected to close the log afterwards.
5514 30. The text of an SMTP error response that was received during a remote
5515 delivery was being truncated at 512 bytes. This is too short for some of
5516 the long messages that one sometimes sees. I've increased the limit to
5519 31. It is now possible to make retry rules that apply only when a message has a
5520 specific sender, in particular, an empty sender.
5522 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5523 it possible to be selective about when SMTP synchronization is enforced.
5525 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5527 32. Implemented hosts_connection_nolog.
5529 33. Added an ACL for QUIT.
5531 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5534 35. Added mailbox_size and mailbox_filecount to appendfile.
5536 36. Added control = no_multiline_responses to ACLs.
5538 37. There was a bug in the logic of the code that waits for the clock to tick
5539 in the case where the clock went backwards by a substantial amount such
5540 that the microsecond fraction of "now" was more than the microsecond
5541 fraction of "then" (but the whole seconds number was less).
5543 38. Added support for the libradius Radius client library this is found on
5544 FreeBSD (previously only the radiusclient library was supported).
5550 1. When certain lookups returned multiple values in the form name=value, the
5551 quoting of the values was not always being done properly. Specifically:
5552 (a) If the value started with a double quote, but contained no whitespace,
5554 (b) If the value contained whitespace other than a space character (i.e.
5555 tabs or newlines or carriage returns) it was not quoted.
5556 This fix has been applied to the mysql and pgsql lookups by writing a
5557 separate quoting function and calling it from the lookup code. The fix
5558 should probably also be applied to nisplus, ibase and oracle lookups, but
5559 since I cannot test any of those, I have not disturbed their existing code.
5561 2. A hit in the callout cache for a specific address caused a log line with no
5562 reason for rejecting RCPT. Now it says "Previous (cached) callout
5563 verification failure".
5565 3. There was an off-by-one bug in the queryprogram router. An over-long
5566 return line was truncated at 256 instead of 255 characters, thereby
5567 overflowing its buffer with the terminating zero. As well as fixing this, I
5568 have increased the buffer size to 1024 (and made a note to document this).
5570 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5571 when Exim is waiting for an SMTP response from a remote server, Exim
5572 restarts its select() call on the socket, thereby resetting its timeout.
5573 This is not a problem when such interrupts are rare. Somebody set up a cron
5574 job to run exiwhat every 2 minutes, which is less than the normal select()
5575 timeout (5 or 10 minutes). This meant that the select() timeout never
5576 kicked in because it was always reset. I have fixed this by comparing the
5577 time when an interrupt arrives with the time at the start of the first call
5578 to select(). If more time than the timeout has elapsed, the interrupt is
5579 treated as a timeout.
5581 5. Some internal re-factoring in preparation for the addition of Sieve
5582 extensions (by MH). In particular, the "personal" test is moved to a
5583 separate function, and given an option for scanning Cc: and Bcc: (which is
5584 not set for Exim filters).
5586 6. When Exim created an email address using the login of the caller as the
5587 local part (e.g. when creating a From: or Sender: header line), it was not
5588 quoting the local part when it contained special characters such as @.
5590 7. Installed new OpenBSD configuration files.
5592 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5593 try to make them clearer.
5595 9. Callout options, other than the timeout value, were being ignored when
5596 verifying sender addresses in header lines. For example, when using
5598 verify = header_sender/callout=no_cache
5600 the cache was (incorrectly) being used.
5602 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5603 only the Cygwin environment.
5605 11. When return_path_on_delivery was set as a log selector, if different remote
5606 addresses in the same message used different return paths and parallel
5607 remote delivery occurred, the wrong values would sometimes be logged.
5608 (Whenever a remote delivery process finished, the return path value from
5609 the most recently started remote delivery process was logged.)
5611 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5612 lines when AUTH and/or TLS are in use. This is the "received protocol"
5613 field. Exim used to use "asmtp" for authenticated SMTP, without any
5614 indication (in the protocol name) for TLS use. Now it follows the RFC and
5615 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5616 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5617 names appear in log lines as well as in Received: header lines.
5619 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5620 extensions, and comparison tests, and to fix some bugs.
5622 14. Changes to the "personal" filter test:
5624 (1) The test was buggy in that it was just doing the equivalent of
5625 "contains" tests on header lines. For example, if a user's address was
5626 anne@some.where, the "personal" test would incorrectly be true for
5628 To: susanne@some.where
5630 This test is now done by extracting each address from the header in turn,
5631 and checking the entire address. Other tests that are part of "personal"
5632 are now done using regular expressions (for example, to check local parts
5633 of addresses in From: header lines).
5635 (2) The list of non-personal local parts in From: addresses has been
5636 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5637 taken from the Sieve specification recommendations.
5639 (3) If the message contains any header line starting with "List-" it is
5640 treated as non-personal.
5642 (4) The test for "circular" in the Subject: header line has been removed
5643 because it now seems ill-conceived.
5645 15. Minor typos in src/EDITME comments corrected.
5647 16. Installed latest exipick from John Jetmore.
5649 17. If headers_add on a router specified a text string that was too long for
5650 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5651 of string_sprintf() is now avoided.
5653 18. $message_body_size was not set (it was always zero) when running the DATA
5654 ACL and the local_scan() function.
5656 19. For the "mail" command in an Exim filter, no default was being set for
5657 the once_repeat time, causing a random time value to be used if "once" was
5658 specified. (If the value happened to be <= 0, no repeat happened.) The
5659 default is now 0s, meaning "never repeat". The "vacation" command was OK
5660 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5661 (I found it when inspecting the code).
5663 20. There is now an overall timeout for performing a callout verification. It
5664 defaults to 4 times the callout timeout, which applies to individual SMTP
5665 commands during the callout. The overall timeout applies when there is more
5666 than one host that can be tried. The timeout is checked before trying the
5667 next host. This prevents very long delays if there are a large number of
5668 hosts and all are timing out (e.g. when the network connections are timing
5669 out). The value of the overall timeout can be changed by specifying an
5670 additional sub-option for "callout", called "maxwait". For example:
5672 verify = sender/callout=5s,maxwait=20s
5674 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5675 to the end before writing, but this should make it even safer).
5677 22. Exim was forgetting that it had advertised PIPELINING for the second and
5678 subsequent messages on an SMTP connection. It was also not resetting its
5679 memory on STARTTLS and an internal HELO.
5681 23. When Exim logs an SMTP synchronization error within a session, it now
5682 records whether PIPELINING has been advertised or not.
5684 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5685 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5686 rather than long int.
5688 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5690 26. Added the never_mail option to autoreply.
5696 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5697 crash if the getsockname() call failed; this can happen if a connection is
5698 closed very soon after it is established. The problem was simply in the
5699 order in which certain operations were done, causing Exim to try to write
5700 to the SMTP stream before it had set up the file descriptor. The bug has
5701 been fixed by making things happen in the correct order.
5707 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5708 before the connection was closed, thus losing the rejection response.
5710 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5711 some early Solaris releases, but causes trouble in current releases where
5712 socklen_t is defined.
5714 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5717 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5720 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5721 that releasing the top of it at the end releases what was used for sub-
5722 expansions (unless the block got too big). However, discard this block if
5723 the first thing is a variable or header, so that we can use its block when
5724 it is dynamic (useful for very large $message_headers, for example).
5726 6. Lookups now cache *every* query, not just the most recent. A new, separate
5727 store pool is used for this. It can be recovered when all lookup caches are
5728 flushed. Lookups now release memory at the end of their result strings.
5729 This has involved some general refactoring of the lookup sources.
5731 7. Some code has been added to the store_xxx() functions to reduce the amount
5732 of flapping under certain conditions.
5734 8. log_incoming_interface used to affect only the <= reception log lines. Now
5735 it causes the local interface and port to be added to several more SMTP log
5736 lines, for example "SMTP connection from", and rejection lines.
5738 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5740 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5742 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5743 could be overwritten at the end of the current message (or the start of a
5744 new message if it was set in a HELO ACL). The value is now preserved for
5745 the duration of the SMTP connection.
5747 12. If a transport had a headers_rewrite setting, and a matching header line
5748 contained an unqualified address, that address was qualified, even if it
5749 did not match any rewriting rules. The underlying bug was that the values
5750 of the flags that permit the existence of unqualified sender and recipient
5751 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5752 non-local messages, and by -bnq for local messages) were not being
5753 preserved with the message after it was received.
5755 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5756 "next input=" as part of the text comprising the host identity instead of
5757 the correct text. The code was using the same buffer for two different
5758 strings. However, depending on which order the printing function evaluated
5759 its arguments, the bug did not always show up. Under Linux, for example, my
5760 test suite worked just fine.
5762 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5763 doesn't work with some older versions of Perl. It has been changed to "my",
5764 which in any case is probably the better facility to use.
5766 15. A really picky compiler found some instances of statements for creating
5767 error messages that either had too many or two few arguments for the format
5770 16. The size of the buffer for calls to the DNS resolver has been increased
5771 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5772 for addresses that have a lot of PTR records. This alleviates a problem; it
5773 does not fully solve it.
5775 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5776 buffer now truncates the list and logs the incident, which is the same
5777 action as happens when Exim is looking up a host name and its aliases.
5778 Previously in this situation something unpredictable would happen;
5779 sometimes it was "internal error: store_reset failed".
5781 18. If a server dropped the connection unexpectedly when an Exim client was
5782 using GnuTLS and trying to read a response, the client delivery process
5783 crashed while trying to generate an error log message.
5785 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5786 string, for example:
5788 warn message = H1: something\nH2: something
5790 the text was added as a single header line from Exim's point of view
5791 though it ended up OK in the delivered message. However, searching for the
5792 second and subsequent header lines using $h_h2: did not work. This has been
5793 fixed. Similarly, if a system filter added multiple headers in this way,
5794 the routers could not see them.
5796 20. Expanded the error message when iplsearch is called with an invalid key to
5797 suggest using net-iplsearch in a host list.
5799 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5800 ACLs are no longer actually imposed (and a message to that effect is
5803 22. If a "gecos" field in a passwd entry contained escaped characters, in
5804 particular, if it contained a \" sequence, Exim got it wrong when building
5805 a From: or a Sender: header from that name. A second bug also caused
5806 incorrect handling when an unquoted " was present following a character
5807 that needed quoting.
5809 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5810 was not being matched caselessly.
5812 24. Arranged for all hyphens in the exim.8 source to be escaped with
5815 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5816 properly. Recipient callout cache records were still being keyed to include
5817 the sender, even when use_sender was set false. This led to far more
5818 callouts that were necessary. The sender is no longer included in the key
5819 when use_sender is false.
5821 26. Added "control = submission" modifier to ACLs.
5823 27. Added the ${base62d: operator to decode base 62 numbers.
5825 28. dnsdb lookups can now access SRV records.
5827 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5828 the configuration file.
5830 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5831 (-v) mode. This makes the output for a verbose queue run more intelligible.
5833 31. Added a use_postmaster feature to recipient callouts.
5835 32. Added the $body_zerocount variable, containing the number of binary zero
5836 bytes in the message body.
5838 33. The time of last modification of the "new" subdirectory is now used as the
5839 "mailbox time last read" when there is a quota error for a maildir
5842 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5844 35. Added +ignore_unknown as a special item in host lists.
5846 36. Code for decoding IPv6 addresses in host lists is now included, even if
5847 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5848 address was recognized as an IP address, but was then not correctly decoded
5849 into binary, causing unexpected and incorrect effects when compared with
5856 1. Very minor rewording of debugging text in manualroute to say "list of
5857 hosts" instead of "hostlist".
5859 2. If verify=header_syntax was set, and a header line with an unqualified
5860 address (no domain) and a large number of spaces between the end of the
5861 name and the colon was received, the reception process suffered a buffer
5862 overflow, and (when I tested it) crashed. This was caused by some obsolete
5863 code that should have been removed. The fix is to remove it!
5865 3. When running in the test harness, delay a bit after writing a bounce
5866 message to get a bit more predictability in the log output.
5868 4. Added a call to search_tidyup() just before forking a reception process. In
5869 theory, someone could use a lookup in the expansion of smtp_accept_max_
5870 per_host which, without the tidyup, could leave open a database connection.
5872 5. Added the variables $recipient_data and $sender_data which get set from a
5873 lookup success in an ACL "recipients" or "senders" condition, or a router
5874 "senders" option, similar to $domain_data and $local_part_data.
5876 6. Moved the writing of debug_print from before to after the "senders" test
5879 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5880 problems for message scanning, either using a data ACL, or using
5881 local_scan() because the Received: header was not generated till after they
5882 were called (in order to set the time as the time of reception completion).
5883 I have revised the way this works. The header is now generated after the
5884 body is received, but before the ACL or local_scan() are called. After they
5885 are run, the timestamp in the header is updated.
5891 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5892 before starting a queue runner without re-exec. This happened only when
5893 deliver_drop_privilege was set or when the Exim user was set to root. The
5894 effect of the bug was that timeouts during subsequent deliveries caused
5895 crashes instead of being properly handled. The handler is now left at its
5896 default (and expected) setting.
5898 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5899 message, again when deliver_drop_privilege is set or Exim is run as root.
5900 The bug described in (1) was not present in this case, but the tidying up
5901 of the other signals was missing. I have made the two cases consistent.
5903 3. The ignore_target_hosts setting on a manualroute router was being ignored
5904 for hosts that were looked up using the /MX notation.
5906 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5909 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5910 operated on the sender address. After changing the $sender_address to <>
5911 for the sender address verify, Exim was re-instated it as the original
5912 (before rewriting) address, but remembering that it had rewritten it, so it
5913 wasn't rewriting it again. This bug also had the effect of breaking the
5914 sender address verification caching when the sender address was rewritten.
5916 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5917 This has been changed so that if the ip literal address matches
5918 ignore_target_hosts, the router declines.
5920 7. Added expansion conditions match_domain, match_address, and match_local_
5921 part (NOT match_host).
5923 8. The placeholder for the Received: header didn't have a length field set.
5925 9. Added code to Exim itself and to exim_lock to test for a specific race
5926 condition that could lead to file corruption when using MBX delivery. The
5927 issue is with the lockfile that is created in /tmp. If this file is removed
5928 after a process has opened it but before that process has acquired a lock,
5929 there is the potential for a second process to recreate the file and also
5930 acquire a lock. This could lead to two Exim processes writing to the file
5931 at the same time. The added code performs the same test as UW imapd; it
5932 checks after acquiring the lock that its file descriptor still refers to
5933 the same named file.
5935 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5936 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5939 11. Added the smtp_active_hostname option. If used, this will typically be made
5940 to depend on the incoming interface address. Because $interface_address is
5941 not set up until the daemon has forked a reception process, error responses
5942 that can happen earlier (such as "too many connections") no longer contain
5945 12. If an expansion in a condition on a "warn" statement fails because a lookup
5946 defers, the "warn" statement is abandoned, and the next ACL statement is
5947 processed. Previously this caused the whole ACL to be aborted.
5949 13. Added the iplsearch lookup type.
5951 14. Added ident_timeout as a log selector.
5953 15. Added tls_certificate_verified as a log selector.
5955 16. Added a global option tls_require_ciphers (compare the smtp transport
5956 option of the same name). This controls incoming TLS connections.
5958 17. I finally figured out how to make tls_require_ciphers do a similar thing
5959 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5960 before starting the TLS session.
5962 18. Tabs are now shown as \t in -bP output.
5964 19. If the log selector return_path_on_delivery was set, Exim crashed when
5965 bouncing a message because it had too many Received: header lines.
5967 20. If two routers both had headers_remove settings, and the first one included
5968 a superfluous trailing colon, the final name in the first list and the
5969 first name in the second list were incorrectly joined into one item (with a
5970 colon in the middle).
5976 1. Added -C and -D options to the exinext utility, mainly to make it easier
5977 to include in the automated testing, but these could be helpful when
5978 multiple configurations are in use.
5980 2. The exinext utility was not formatting the output nicely when there was
5981 an alternate port involved in the retry record key, nor when there was a
5982 message id as well (for retries that were specific to a specific message
5983 and a specific host). It was also confused by IPv6 addresses, because of
5984 the additional colons they contain. I have fixed the IPv4 problem, and
5985 patched it up to do a reasonable job for IPv6.
5987 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5988 delivery, the log line now contains "pipelined" if PIPELINING was used.
5990 4. An SMTP transport process used to panic and die if the bind() call to set
5991 an explicit outgoing interface failed. This has been changed; it is now
5992 treated in the same way as a connect() failure.
5994 5. A reference to $sender_host_name in the part of a conditional expansion
5995 that was being skipped was still causing a DNS lookup. This no longer
5998 6. The def: expansion condition was not recognizing references to header lines
5999 that used bh_ and bheader_.
6001 7. Added the _cache feature to named lists.
6003 8. The code for checking quota_filecount in the appendfile transport was
6004 allowing one more file than it should have been.
6006 9. For compatibility with Sendmail, the command line option
6014 and sets the incoming protocol and host name (for trusted callers). The
6015 host name and its colon can be omitted when only the protocol is to be set.
6016 Note the Exim already has two private options, -pd and -ps, that refer to
6017 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6018 "s", but I don't think that's a major issue.
6020 10. A number of refactoring changes to the code, none of which should affect
6023 (a) The number of logging options was getting close to filling up the
6024 32-bit word that was used as a bit map. I have split them into two classes:
6025 those that are passed in the argument to log_write(), and those that are
6026 only ever tested independently outside of that function. These are now in
6027 separate 32-bit words, so there is plenty of room for expansion again.
6028 There is no change in the user interface or the logging behaviour.
6030 (b) When building, for example, log lines, the code previously used a
6031 macro that called string_cat() twice, in order to add two strings. This is
6032 not really sufficiently general. Furthermore, there was one instance where
6033 it was actually wrong because one of the argument was used twice, and in
6034 one call a function was used. (As it happened, calling the function twice
6035 did not affect the overall behaviour.) The macro has been replaced by a
6036 function that can join an arbitrary number of extra strings onto a growing
6039 (c) The code for expansion conditions now uses a table and a binary chop
6040 instead of a serial search (which was left over from when there were very
6041 few conditions). Also, it now recognizes conditions like "pam" even when
6042 the relevant support is not compiled in: a suitably worded error message is
6043 given if an attempt is made to use such a condition.
6045 11. Added ${time_interval:xxxxx}.
6047 12. A bug was causing one of the ddress fields not to be passed back correctly
6048 from remote delivery subprocesses. The field in question was not being
6049 subsequently used, so this caused to problems in practice.
6051 13. Added new log selectors queue_time and deliver_time.
6053 14. Might have fixed a bug in maildirsizefile handling that threw up
6054 "unexpected character" debug warnings, and recalculated the data
6055 unnecessarily. In any case, I expanded the warning message to give more
6058 15. Added the message "Restricted characters in address" to the statements in
6059 the default ACL that block characters like @ and % in local parts.
6061 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6062 Three changes have been made:
6064 (a) There was a serious bug; a negative response to MAIL caused the whole
6065 recipient domain to be cached as invalid, thereby blocking all messages
6066 to all local parts at the same domain, from all senders. This bug has
6067 been fixed. The domain is no longer cached after a negative response to
6068 MAIL if the sender used is not empty.
6070 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6073 (c) A new callout option, "use_sender" has been added for people who want
6074 the modified behaviour.
6080 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6083 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6084 indeed breaks things for older releases.
6086 3. Added additional logging to the case where there is a problem reading data
6087 from a filter that is running in a subprocess using a pipe, in order to
6088 try to track down a specific problem.
6090 4. Testing facility fudge: when running in the test harness and attempting
6091 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6092 getting "No route to host". Convert this to a timeout.
6094 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6097 6. Some OS don't have socklen_t but use size_t instead. This affects the
6098 fifth argument of getsockopt() amongst other things. This is now
6099 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6100 can be set for individual OS. I have set it for SunOS5, OSF1, and
6101 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6102 some earlier ones do not.
6104 7. Change 4.30/15 was not doing the test caselessly.
6106 8. The standard form for an IPv6 address literal was being rejected by address
6107 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6108 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6109 this, as well as the form without the "IPv6" on the front (but only when
6110 address literals are enabled, of course).
6112 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6114 10. Exim crashed if a message with an empty sender address specified by -f
6115 encountered a router with an errors_to setting. This could be provoked only
6116 by a command such as
6120 where an empty string was supplied; "<>" did not hit this bug.
6122 11. Installed PCRE release 4.5.
6124 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6125 remained set. It is now erased.
6127 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6128 times from message ids (which are base 36 rather than the normal 62).
6130 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6131 were being counted as actual protocol errors, and logged if the log
6132 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6133 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6134 and DATA following a set of rejected RCPTs do not count as protocol errors.
6135 In other words, Exim assumes they were pipelined, though this may not
6136 actually be the case. Of course, in all cases the client gets an
6137 appropriate error code.
6139 15. If a lookup fails in an ACL condition, a message about the failure may
6140 be available; it is used if testing the ACL cannot continue, because most
6141 such messages specify what the cause of the deferral is. However, some
6142 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6143 that caused an old message to be retained and used if a later statement
6144 caused a defer, replacing the real cause of the deferral.
6146 16. If an IP address had so many PTR records that the DNS lookup buffer
6147 was not large enough to hold them, Exim could crash while trying to process
6148 the truncated data. It now detects and logs this case.
6150 17. Further to 4.21/58, another change has been made: if (and only if) the
6151 first line of a message (the first header line) ends with CRLF, a bare LF
6152 in a subsequent header line has a space inserted after it, so as not to
6153 terminate the header.
6155 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6156 unnecessarily, used atoi() instead of strtol(), and didn't check the
6157 termination when getting file sizes from file names by regex.
6159 19. Completely re-implemented the support for maildirsize files, in the light
6160 of a number of problems with the previous contributed implementation
6161 (4.30/29). In particular:
6163 . If the quota is zero, the maildirsize file is maintained, but no quota is
6166 . If the maildir directory does not exist, it is created before any attempt
6167 to write a maildirsize file.
6169 . The quota value in the file is just a cache; if the quota is changed in
6170 the transport, the new value overrides.
6172 . A regular expression is available for excluding directories from the
6175 20. The autoreply transport checks the characters in options that define the
6176 message's headers; it allows continued headers, but it was checking with
6177 isspace() after an embedded newline instead of explicitly looking for a
6180 21. If all the "regular" hosts to which an address was routed had passed their
6181 expiry times, and had not reached their retry times, the address was
6182 bounced, even if fallback hosts were defined. Now Exim should go on to try
6185 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6186 equivalent code in the SMTP transport. Some hosts send humungous responses
6187 to HELO/EHLO, more than 1024 it seems.
6189 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6190 gives compiler warnings in some environments. I've now done it "properly",
6193 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6194 (because of problems with the built-in one) was declared to return uschar *
6195 instead of char *, causing compiler failure.
6197 25. Fixed a file descriptor leak when processing alias/forward files.
6199 26. Fixed a minor format string issue in dbfn.c.
6201 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6203 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6204 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6207 29. When a sender address is verified, it is cached, to save repeating the test
6208 when there is more than one recipient in a message. However, when the
6209 verification involves a callout, it is possible for different callout
6210 options to be set for different recipients. It is too complicated to keep
6211 track of this in the cache, so now Exim always runs a verification when a
6212 callout is required, relying on the callout cache for the optimization.
6213 The overhead is duplication of the address routing, but this should not be
6216 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6217 to be verified with callout=postmaster, and the main callout worked but the
6218 postmaster check failed, the verification correctly failed. However, if a
6219 subsequent RCPT command asked for sender verification *without* the
6220 postmaster check, incorrect caching caused this verification also to fail,
6223 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6224 it was not caching the DNS options (qualify_single, search_parents) that
6225 were used when the lookup failed. A subsequent lookup with different
6226 options therefore always gave the same answer, though there were cases
6227 where it should not have. (Example: a "domains = !$mx_any" option on a
6228 dnslookup router: the "domains" option is always processed without any
6229 widening, but the router might have qualify_single set.) Now Exim uses the
6230 cached value only when the same options are set.
6232 32. Added John Jetmore's "exipick" utility to the distribution.
6234 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6235 than a timeout (e.g. a certificate is required, and is not provided), an
6236 Exim server now closes the connection immediately. Previously it waited for
6237 the client to close - but if the client is SSL, it seems that they each
6238 wait for each other, leading to a delay before one of them times out.
6240 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6241 maintained 0.8.x compatibility because I don't think many are using it, and
6242 it is clearly obsolete.
6244 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6247 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6248 client certificate was expired. A simple patch fixes this, though I don't
6249 understand the full logic of why the verify callback is called multiple
6252 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6253 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6254 which causes problems with some clients (such as the Certicom SSL Plus
6255 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6256 disables the coutermeasure allowing Eudora to connect."
6258 38. Exim was not checking that a write() to a log file succeeded. This could
6259 lead to Bad Things if a log got too big, in particular if it hit a file
6260 size limit. Exim now panics and dies if it cannot write to a log file, just
6261 as it does if it cannot open a log file.
6263 39. Modified OS/Makefile-Linux so that it now contains
6265 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6267 The two -D definitions ensure that Exim is compiled with large file
6268 support, which makes it possible to handle log files that are bigger than
6271 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6272 instance) a domain was checked against a named list that involved a lookup,
6273 causing $domain_data to be set, then another domain was checked against the
6274 same list, then the first domain was re-checked, the value of $domain_data
6275 after the final check could be wrong. In particular, if the second check
6276 failed, it could be set empty. This bug probably also applied to
6279 41. The strip_trailing_dot option was not being applied to the address given
6280 with the -f command-line option.
6282 42. The code for reading a message's header from the spool was incrementing
6283 $received_count, but never initializing it. This meant that the value was
6284 incorrect (doubled) while delivering a message in the same process in which
6285 it was received. In the most common configuration of Exim, this never
6286 happens - a fresh exec is done - but it can happen when
6287 deliver_drop_privilege is set.
6289 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6290 it now includes up to 150 characters of the unexpected data in the log
6293 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6294 and building data strings. The size of both of these buffers was 10 000
6295 bytes - far larger than anybody would *ever* want, thought I. Needless to
6296 say, somebody hit the limit. I have increased the maximum line length to
6297 20 000 and the maximum data length of concatenated lines to 100 000. I have
6298 also fixed two bugs, because there was no checking on these buffers. Tsk,
6299 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6300 buffer is too small.
6302 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6303 lsearch lookups. Now it does.
6305 46. When parsing a route_list item in a manualroute router, a fixed-length
6306 buffer was used for the list of hosts. I made this 1024 bytes long,
6307 thinking that nobody would ever have a list of hosts that long. Wrong.
6308 Somebody had a whole pile of complicated expansion conditions, and the
6309 string was silently truncated, leading to an expansion error. It turns out
6310 that it is easier to change to an unlimited length (owing to other changes
6311 that have happened since this code was originally written) than to build
6312 structure for giving a limitation error. The length of the item that
6313 expands into the list of hosts is now unlimited.
6315 47. The lsearch lookup could not handle data where the length of text line was
6316 more than 4095 characters. Such lines were truncated, leading to shortened
6317 data being returned. It should now handle lines of any length.
6319 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6320 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6323 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6324 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6325 no longer happens in comment lines. A list of replacements is now placed
6326 at the head of all of the source files, except those whose only change is
6327 to replace PERL_COMMAND in the very first #! line.
6329 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6330 messages on the queue, with a bottom-up merge sort, using code contributed
6331 by Michael Haardt. This should make operations like -bp somewhat faster on
6332 large queues. It won't affect queue runners, except when queue_run_in_order
6335 51. Installed eximstats 1.31 in the distribution.
6337 52. Added support for SRV lookups to the dnslookup router.
6339 53. If an ACL referred to $message_body or $message_body_end, the value was not
6340 reset for any messages that followed in the same SMTP session.
6342 54. The store-handling optimization for building very long strings was not
6343 differentiating between the different store pools. I don't think this
6344 actually made any difference in practice, but I've tidied it.
6346 55. While running the routers to verify a sender address, $sender_address
6347 was still set to the sender address. This is wrong, because when routing to
6348 send a bounce to the sender, it would be empty. Therefore, I have changed
6349 it so that, while verifying a sender address, $sender_address is set to <>.
6350 (There is no change to what happens when verifying a recipient address.)
6352 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6353 target A or AAAA records (if not already returned) without resetting the
6354 qualify_single or search_parents options of the DNS resolver. These are
6355 inappropriate in this case because the targets of MX and SRV records must
6356 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6357 target that, when qualified, matched something in the local domain. These
6358 two options are now turned off when doing these lookups.
6360 57. It seems that at least some releases of Reiserfs (which does not have the
6361 concept of a fixed number of inodes) returns zero and not -1 for the
6362 number of available inodes. This interacted badly with check_spool_inodes,
6363 which assumed that -1 was the "no such thing" setting. What I have done is
6364 to check that the total number of inodes is greater than zero before doing
6365 the test of how many are available.
6367 58. When a "warn" ACL statement has a log_message modifier, the message is
6368 remembered, and not repeated. This is to avoid a lot of repetition when a
6369 message has many recipients that cause the same warning to be written.
6370 However, Exim was preserving the list of already written lines for an
6371 entire SMTP session, which doesn't seem right. The memory is now reset if a
6372 new message is started.
6374 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6375 debugging output unless log_rewrite was also set.
6377 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6378 of (int)(handle) when we know that handle contains (void *)(-1).
6380 61. The Exim daemon panic-logs an error return when it closes the incoming
6381 connection. However "connection reset by peer" seems to be common, and
6382 isn't really an error worthy of noting specially, so that particular error
6385 62. When Exim is trying to find all the local interfaces, it used to panic and
6386 die if the ioctl to get the interface flags failed. However, it seems that
6387 on at least one OS (Solaris 9) it is possible to have an interface that is
6388 included in the list of interfaces, but for which you get a failure error
6389 for this call. This happens when the interface is not "plumbed" into a
6390 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6391 failure of the "get flags" call assumes that the interface is down.
6393 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6394 makes life easier for people who are doing arithmetic on fields extracted
6395 from dates, where you often get leading zeros that should not be
6396 interpreted as octal.
6398 64. Added qualify_domain to the redirect router, to override the global
6401 65. If a pathologically long header line contained very many addresses (the
6402 report of this problem mentioned 10 000) and each of them was rewritten,
6403 Exim could use up a very large amount of memory. (It kept on making new
6404 copies of the header line as it rewrote, and never released the old ones.)
6405 At the expense of a bit more processing, the header rewriting function has
6406 been changed so that it no longer eats memory in this way.
6408 66. The generation of the Received: header has been moved from the time that a
6409 message starts to be received, to the time that it finishes. The timestamp
6410 in the Received: header should now be very close to that of the <= log
6411 line. There are two side-effects of this change:
6413 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6414 logged header lines no longer include the local Received: line, because
6415 it has not yet been created. The same applies to a copy of the message
6416 that is returned to a non-SMTP sender when a message is rejected.
6418 (b) When a filter file is tested using -bf, no additional Received: header
6419 is added to the test message. After some thought, I decided that this
6422 This change does not affect the value of $received_for. It is still set
6423 after address rewriting, but before local_scan() is called.
6425 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6427 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6428 gave an unhelpful panic error message, and a defer error. I have managed to
6429 change this behaviour so that it now rejects any supplied certificate,
6430 which seems right, as the list of acceptable certificates is empty.
6432 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6433 gave an unhelpful defer error. I have not managed to make this reject any
6434 supplied certificates, but the error message it gives is "no certificate
6435 supplied", which is not helpful.
6437 70. exigrep's output now also includes lines that are not associated with any
6438 message, but which match the given pattern. Implemented by a patch from
6439 Martin Sluka, which also tidied up the Perl a bit.
6441 71. Recipient callout verification, like sender verification, was using <> in
6442 the MAIL FROM command. This isn't really the right thing, since the actual
6443 sender may affect whether the remote host accepts the recipient or not. I
6444 have changed it to use the actual sender in the callout; this means that
6445 the cache record is now keyed on a recipient/sender pair, not just the
6446 recipient address. There doesn't seem to be a real danger of callout loops,
6447 since a callout by the remote host to check the sender would use <>.
6448 [SEE ABOVE: changed after hitting problems.]
6450 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6451 temporary errors. However, in the case of such a code being given after
6452 the end of a data transmission (i.e. after ".") Exim was failing to write
6453 a retry record for the message. (Yes, there was some broken host that was
6454 actually sending 8xx at this point.)
6456 73. An unknown lookup type in a host list could cause Exim to panic-die when
6457 the list was checked. (An example that provoked this was putting <; in the
6458 middle of a list instead of at the start.) If this happened during a DATA
6459 ACL check, a -D file could be left lying around. This kind of configuration
6460 error no longer causes Exim to die; instead it causes a defer error. The
6461 incident is still logged to the main and panic logs.
6463 74. Buglet left over from Exim 3 conversion. The message "too many messages
6464 in one connection" was written to the rejectlog but not the mainlog, except
6465 when address rewriting (yes!) was being logged.
6467 75. Added write_rejectlog option.
6469 76. When a system filter was run not as root (that is, when system_filter_user
6470 was set), the values of the $n variables were not being returned to the
6471 main process; thus, they were not subsequently available in the $sn
6474 77. Added +return_path_on_delivery log selector.
6476 78. A connection timeout was being treated differently from recipients deferred
6477 when testing hosts_max_try with a message that was older than the host's
6478 retry timeout. (The host should not be counted, thus allowing all hosts to
6479 be tried at least once before bouncing.) This may have been the cause of an
6480 occasionally reported bug whereby a message would remain on the queue
6481 longer than the retry timeout, but would be bounced if a delivery was
6482 forced. I say "may" because I never totally pinned down the problem;
6483 setting up timeout/retry tests is difficult. See also the next item.
6485 79. The ultimate address timeout was not being applied to errors that involved
6486 a combination of host plus message (for example, a timeout on a MAIL
6487 command). When an address resolved to a number of possible hosts, and they
6488 were not all tried for each delivery (e.g. because of hosts_max_try), a
6489 message could remain on the queue longer than the retry timeout.
6491 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6494 81. Fixed an obscure SMTP outgoing bug which required at least the following
6495 conditions: (a) there was another message waiting for the same server;
6496 (b) the server returned 5xx to all RCPT commands in the first message so
6497 that the message was not completed; (c) the server dropped the connection
6498 or gave a negative response to the RSET that Exim sends to abort the
6499 transaction. The observed case was a dropped connection after DATA that had
6500 been sent in pipelining mode. That is, the server had advertised PIPELINING
6501 but was not implementing it correctly. The effect of the bug was incorrect
6502 behaviour, such as trying another host, and this could lead to a crash.
6508 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6509 and daemon.c were passed as pointers to ints; they should have been
6510 pointers to socklen_t variables (which are typically unsigned ints).
6512 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6515 3. Fixed a really odd bug that affected only the testing scheme; patching a
6516 certain fixed string in the binary changed the value of another string that
6517 happened to be identical to the end of the original first string.
6519 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6520 name", it returns that address as the IP address. On some operating
6521 systems (e.g. Solaris), it also passes back the IP address string as the
6522 "host name". However, on others (e.g. Linux), it passes back an empty
6523 string. Exim wasn't checking for this, and was changing the host name to an
6524 empty string, assuming it had been canonicalized.
6526 5. Although rare, it is permitted to have more than one PTR record for a given
6527 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6528 all the names associated with an address, because they do in Solaris.
6529 However, it seems that they do not in Linux for data that comes from the
6530 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6531 I found this out when I moved to a new Linux workstation and tried to run
6532 the Exim test suite.
6534 To get round this problem I have changed the code so that it now does its
6535 own call to the DNS to look up PTR records when searching for a host name.
6536 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6537 addresses that are only in /etc/hosts are still found.
6539 This behaviour is, however, controlled by an option called host_lookup_
6540 order, which defaults to "bydns:byaddr". If people want to use the other
6541 order, or indeed, just use one or the other means of lookup, they can
6542 specify it in this variable.
6544 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6545 some operating systems, this comes back from gethostbyaddr() as an empty
6546 string, and this is what Exim used to test for. However, it seems that in
6547 other systems, "." is yielded. Exim now tests for this case too.
6549 7. The values of check_spool_space and check_log_space are now held internally
6550 as a number of kilobytes instead of an absolute number of bytes. If a
6551 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6552 kilobyte. This means that much larger values can be stored.
6554 8. Exim monitor: an attempt to get the action menu when not actually pointing
6555 at a message produces an empty menu entitled "No message selected". This
6556 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6557 no entries in it ("Shell widget menu has zero width and/or height"). So I
6558 have added a single, blank menu entry in this case.
6560 9. Added ${quote_local_part.
6562 10. MIME decoding is now applied to the contents of Subject: header lines when
6565 11. Now that a reference to $sender_host_address automatically causes a reverse
6566 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6567 host lookup before query-style lookups in lists that might use this
6568 variable. This has therefore been abolished, and the "net-" prefix is no
6569 longer necessary for query-style lookups.
6571 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6572 have been a typo for LFLAGS, so it has been changed.
6574 13. The install script calls Exim with "-C /dev/null" in order to find the
6575 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6576 to be output. However, since Exim outputs its version number before the
6577 error, it didn't break the script. It just looked ugly. I fixed this by
6578 always allowing "-C /dev/null" if the caller is root.
6580 14. Ignore overlarge ACL variable number when reading spool file - insurance
6581 against a later release with more variables having written the file.
6583 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6584 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6585 this, as well as the form without the "IPv6" on the front.
6587 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6588 OS/Makefile-Darwin file.
6590 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6591 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6593 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6594 variable. However, some OS use ints for inodes. I've added cast to long int
6595 to get rid of the compiler warning.
6597 19. I had forgotten to lock out "/../" in configuration file names when
6598 ALT_CONFIG_PREFIX was set.
6600 20. Routers used for verification do not need to specify transports. However,
6601 if such a router generated a host list, and callout was configured, Exim
6602 crashed, because it could not find a port number from the (non-existent)
6603 transport. It now assumes port 25 in this circumstance.
6605 21. Added the -t option to exigrep.
6607 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6608 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6609 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6610 them is set, LOOKUP_LSEARCH is forced.
6612 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6614 24. Added sender and host information to the "rejected by local_scan()" log
6615 line; previously there was no indication of these.
6617 25. Added .include_if_exists.
6619 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6620 Linux. It turns out that not all file systems support this. Apparently some
6621 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6622 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6623 not supported on the file descriptor, is now ignored when Exim is trying to
6624 sync a directory. This applies only to Linux.
6626 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6628 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6629 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6632 29. Added support for maildirsize files from supplied patch (modified a bit).
6634 30. The use of :fail: followed by an empty string could lead Exim to respond to
6635 sender verification failures with (e.g.):
6637 550 Verification failed for <xxx>
6638 550 Sender verify failed
6640 where the first response line was missing the '-' that indicates it is not
6641 the final line of the response.
6643 31. The loop for finding the name of the user that called Exim had a hardwired
6644 limit of 10; it now uses the value of finduser_retries, which is used for
6645 all other user lookups.
6647 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6650 33. Exim was neglecting to zero errno before one call of strtol() when
6651 expanding a string and expecting an integer value. On some systems this
6652 resulted in spurious "integer overflow" errors. Also, it was casting the
6653 result into an int without checking.
6655 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6656 did not work. The code looks as if it has *never* worked, though it appears
6657 to have been documented since at least release 1.62. I have made it work.
6659 35. The "timeout_DNS" error in retry rules, also documented since at least
6660 1.62, also never worked. As it isn't clear exactly what this means, and
6661 clearly it isn't a major issue, I have abolished the feature by treating it
6662 as "timeout", and writing a warning to the main and panic logs.
6664 36. The display of retry rules for -brt wasn't always showing the error code
6667 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6668 timeout_connect_A, timeout_connect_MX.
6670 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6671 to the empty sender.
6673 39. The daemon was not analysing the content of -oX till after it had closed
6674 stderr and disconnected from the controlling terminal. This meant that any
6675 syntax errors were only noted on the panic log, and the return code from
6676 the command was 0. By re-arranging the code a little, I've made the
6677 decoding happen first, so such errors now appear on stderr, and the return
6678 code is 1. However, the actual setting up of the sockets still happens in
6679 the disconnected process, so errors there are still only recorded on the
6682 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6683 connections (as happens on some IP stacks) was logged at start up time as
6684 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6685 it from "IPv6 and IPv4", which means that two separate sockets are being
6688 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6689 says whether AF_INET or AF_INET6 was passed as an argument.
6691 42. Exiwhat output was messed up when time zones were included in log
6694 43. Exiwhat now gives more information about the daemon's listening ports,
6695 and whether -tls-on-connect was used.
6697 44. The "port" option of the smtp transport is now expanded.
6699 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6700 silently ignored. Now an error message is written to the main and panic
6703 46. There's a new ACL modifier called "logwrite" which writes to a log file
6704 as soon as it is encountered.
6706 47. Added $local_user_uid and $local_user_gid at routing time.
6708 48. Exim crashed when trying to verify a sender address that was being
6711 49. Exim was recognizing only a space character after ".include". It now also
6712 recognizes a tab character.
6714 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6715 extracting the relevant information from the specification. The man page no
6716 longer contains scrambled data for the -d option, and I've added a section
6717 at the front about calling Exim under different names.
6719 51. Added "extra_headers" argument to the "mail" command in filter files.
6721 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6724 53. Installed eximstats 1.29.
6726 54. Added transport_filter_timeout as a generic transport option.
6728 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6729 Cc: header lines. This was required by RFC 822, but it not required by RFC
6732 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6733 incoming messages that did not have them. Now it does so only if the
6734 message originates locally, that is, if there is no associated remote host
6735 address. When Resent- header lines are present, this applies to the Resent-
6736 lines rather than the non-Resent- lines.
6738 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6739 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6741 58. Messages for configuration errors now include the name of the main
6742 configuration file - useful now that there may be more than one file in a
6743 list (.included file names were always shown).
6745 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6746 for those rare installations that do not start the daemon as root or run it
6747 setuid root. I've cut out the call to initgroups() if the daemon is not
6750 60. The Exim user and group can now be bound into the binary as text strings
6751 that are looked up at the start of Exim's processing.
6753 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6755 62. Added $mailstore_basename variable.
6757 63. Installed patch to sieve.c from Michael Haardt.
6759 64. When Exim failed to open the panic log after failing to open the main log,
6760 the original message it was trying to log was written to stderr and debug
6761 output, but if they were not available (the usual case in production), it
6762 was lost. Now it is written to syslog before the two lines that record the
6763 failures to open the logs.
6765 65. Users' Exim filters run in subprocesses under the user's uid. It is
6766 possible for a "deliver" command or an alias in a "personal" command to
6767 provoke an address rewrite. If logging of address rewriting is configured,
6768 this fails because the process is not running as root or exim. There may be
6769 a better way of dealing with this, but for the moment (because 4.30 needs
6770 to be released), I have disabled address rewrite logging when running a
6771 filter in a non-root, non-exim process.
6777 1. The buildconfig auxiliary program wasn't quoting the value set for
6778 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6779 not defined. This bug was masked in 4.22 by the effect that was fixed in
6782 2. Some messages that were rejected after a message id was allocated were
6783 shown as "incomplete" by exigrep. It no longer does this for messages that
6784 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6786 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6787 have allow_domain_literals set, the ID did not get logged in the <= line.
6788 Domain literals are now always recognized in Message-ID: header lines.
6790 4. The first argument for a ${extract expansion item is the key name or field
6791 number. Leading and trailing spaces in this item were not being ignored,
6792 causing some misleading effects.
6794 5. When deliver_drop_privilege was set, single queue runner processes started
6795 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6796 same command in the process it spins off) were not dropping privilege.
6798 6. When the daemon running as "exim" started a queue runner, it always
6799 re-executed Exim in the spun-off process. This is a waste of effort when
6800 deliver_drop_privilege is set. The new process now just calls the
6801 queue-runner function directly.
6807 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6810 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6811 ignored. Though the use of -oP was forcing the writing of a pid file, it
6812 was always written to the default place.
6814 3. If the message "no IP address found for host xxxx" is generated during
6815 incoming verification, it is now followed by identification of the incoming
6816 connection (so you can more easily find what provoked it).
6818 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6820 5. Added some features to "harden" Exim a bit more against certain attacks:
6822 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6823 be put in Local/Makefile. This is like the never_users runtime option,
6824 but it cannot be overridden. The default setting is "root".
6826 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6827 prefix string with which any file named in a -C command line option
6830 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6831 is retained for -C and -D only if the caller of Exim is root. Without
6832 it, the exim user may also use -C and -D and retain privilege.
6834 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6835 command line option is disabled.
6837 6. Macro names set by the -D option must start with an upper case letter, just
6838 like macro names defined in the configuration file.
6840 7. Added "dereference=" facility to LDAP.
6842 8. Two instances of the typo "uknown" in the source files are fixed.
6844 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6845 the Configure-Makefile script screwed up while processing it.
6847 10. Incorporated PCRE 4.4.
6849 11. The SMTP synchronization check was not operating right at the start of an
6850 SMTP session. For example, it could not catch a HELO sent before the client
6851 waited for the greeting. There is now a check for outstanding input at the
6852 point when the greeting is written. Because of the duplex, asynchronous
6853 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6854 way, but not yet received, when the check is performed.
6856 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6857 on TCP/IP sockets, because this apparently causes some broken clients to
6860 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6861 unchanged) from the Cygwin maintainer.
6863 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6864 was supported instead of testing for mbx. Effectively a typo.
6866 15. The spa authenticator server code was not checking that the input it
6867 received was valid base64.
6869 16. The debug output line for the "set" modifier in ACLs was not showing the
6870 name of the variable that was being set.
6872 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6874 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6875 lookup on its own. Something else was needed to trigger the lookup. For
6876 example, a match in host_lookup or the need for a host name in a host list.
6877 Now, if $sender_host_name is referenced and the host name has not yet been
6878 looked up, a lookup is performed. If the lookup fails, the variable remains
6879 empty, and $host_lookup_failed is set to "1".
6881 19. Added "eqi" as a case-independent comparison operator.
6883 20. The saslauthd authentication condition could segfault if neither service
6884 nor realm was specified.
6886 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6887 error message that was logged was misleading, and incoming SMTP
6888 connections were dropped. The message is now more accurate, and temporary
6889 errors are given to SMTP connections.
6891 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6892 (except RSET) with a 421 error, until QUIT is received. However, it was
6893 failing to send a response to QUIT.
6895 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6896 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6897 ACL is now run after the helo_try_verify_hosts code.
6899 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6900 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6901 case-independent, so other case variants are also recognized). Apparently
6902 some systems use these upper case variants.
6904 25. If more than two messages were waiting for the same host, and a transport
6905 filter was specified for the transport, Exim sent two messages over the
6906 same TCP/IP connection, and then failed with "socket operation on non-
6907 socket" when it tried to send the third.
6909 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6911 27. The extern definition of crypt16() in expand.c was not being excluded when
6912 the OS had its own crypt16() function.
6914 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6915 as a preferred synonym for return_size_limit, both as an option and as an
6918 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6920 30. Changed the default configuration ACL to relax the local part checking rule
6921 for addresses that are not in any local domains. For these addresses,
6922 slashes and pipe symbols are allowed within local parts, but the sequence
6923 /../ is explicitly forbidden.
6925 31. SPA server authentication was not clearing the challenge buffer before
6928 32. log_message in a "warn" ACL statement was writing to the reject log as
6929 well as to the main log, which contradicts the documentation and doesn't
6930 seem right (because no rejection is happening). So I have stopped it.
6932 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6933 However, I am unable to do any testing of this.
6935 34. Fixed an infelicity in the appendfile transport. When checking directories
6936 for a mailbox, to see if any needed to be created, it was accidentally
6937 using path names with one or more superfluous leading slashes; tracing
6938 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6940 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6941 contents are added to the log line that is written for every discarded
6942 recipient. (Previously a log_message setting was ignored.)
6944 36. The ${quote: operator now quotes the string if it is empty.
6946 37. The install script runs exim in order to find its version number. If for
6947 some reason other than non-existence or emptiness, which it checks, it
6948 could not run './exim', it was installing it with an empty version number,
6949 i.e. as "exim-". This error state is now caught, and the installation is
6952 38. An argument was missing from the function that creates an error message
6953 when Exim fails to connect to the socket for saslauthd authentication.
6954 This could cause Exim to crash, or give a corrupted message.
6956 39. Added isip, isip4, and isip6 to ${if conditions.
6958 40. The ACL variables $acl_xx are now saved with the message, and can be
6959 accessed later in routers, transports, and filters.
6961 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6962 strings in the file are not string-expanded.
6964 42. If a MAIL command specified a SIZE value that was too large to fit into an
6965 int variable, the check against message_size_limit failed. Such values are
6966 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6967 day this will have to be increased, but I don't think I want to be around
6968 when emails are that large.
6975 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6976 iconv() is not standard in FreeBSD.
6978 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6979 IPv6 enabled. The observed symptom was a segmentation fault on return from
6980 the function os_common_find_running_interfaces() in src/os.c.
6982 3. In the check_special_case() function in daemon.c I had used "errno" as an
6983 argument name, which causes warnings on some systems. This was basically a
6984 typo, since it was named "eno" in the comments!
6986 4. The code that waits for the clock to tick (at a resolution of some fraction
6987 of a second) so as to ensure message-id uniqueness was always waiting for
6988 at least one whole tick, when it could have waited for less. [This is
6989 almost certainly not relevant at current processor speeds, where it is
6990 unlikely to ever wait at all. But we try to future-proof.]
6992 5. The function that sleeps for a time interval that includes fractions of a
6993 second contained a race. It did not block SIGALRM between setting the
6994 timer, and suspending (a couple of lines later). If the interval was short
6995 and the sigsuspend() was delayed until after it had expired, the suspension
6996 never ended. On busy systems this could lead to processes getting stuck for
6999 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7000 process, before it forks any delivery processes. The open lookup caching
7001 mechanism meant that the open file or database connection was passed into
7002 the delivery process. The problem was that delivery processes always tidy
7003 up cached lookup data. This could cause a problem for the next delivery
7004 process started by the queue runner, because the external queue runner
7005 process does not know about the closure. So the next delivery process
7006 still has data in the lookup cache. In the case of a file lookup, there was
7007 no problem because closing a file descriptor in a subprocess doesn't affect
7008 the parent. However, if the lookup was caching a connection to a database,
7009 the connection was closed, and the second delivery process was likely to
7010 see errors such as "PGSQL: query failed: server closed the connection
7011 unexpectedly". The problem has been fixed by closing all cached lookups
7012 in a queue runner before running a delivery process.
7014 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7015 seem to have the "const" qualifier which it has on other OS. I've
7018 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7019 *of the same type* (client or server) with the same public name that an
7020 error should be diagnosed.
7022 9. When Exim looked up a host name for an IP address, but failed to find the
7023 original IP address when looking up the host name (a safety check), it
7024 output the message "<ip address> does not match any IP for NULL", which was
7025 confusing, to say the least. The bug was that the host name should have
7026 appeared instead of "NULL".
7028 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7029 user that is built into the binary, and the -C or -D options is used, root
7030 privilege is dropped before the configuration file is read. In addition,
7031 logging is switched to stderr instead of the normal log files. If the
7032 configuration then re-defines the Exim user, the unprivileged environment
7033 is probably not what is expected, so Exim logs a panic warning message (but
7036 However, if deliver_drop_privilege is set, the unprivileged state may well
7037 be exactly what is intended, so the warning has been cut out in that case,
7038 and Exim is allowed to try to write to its normal log files.
7044 1. smtp_return_error_details was not giving details for temporary sender
7045 or receiver verification errors.
7047 2. Diagnose a configuration error if two authenticators have the same public
7050 3. Exim used not to create the message log file for a message until the first
7051 delivery attempt. This could be confusing when incoming messages were held
7052 for policy or load reasons. The message log file is now created at the time
7053 the message is received, and an initial "Received" line is written to it.
7055 4. The automatically generated man page for command line options had a minor
7056 bug that caused no ill effects; however, a more serious problem was that
7057 the procedure for building the man page automatically didn't always
7058 operate. Consequently, release 4.20 contains an out-of-date version. This
7059 shouldn't happen again.
7061 5. When building Exim with embedded Perl support, the script that builds the
7062 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7063 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7065 6. The freeze_tell option was not being used for messages that were frozen on
7066 arrival, either by an ACL or by local_scan().
7068 7. Added the smtp_incomplete_transaction log selector.
7070 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7071 was accepting AUTH without a new EHLO.
7073 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7074 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7077 10. Logging of TCP/IP connections (when configured) now happens in the main
7078 daemon process instead of the child process, so that the TCP/IP connection
7079 count is more accurate (but it can never be perfect).
7081 11. The use of "drop" in a nested ACL was not being handled correctly in the
7082 outer ACL. Now, if condition failure induced by the nested "drop" causes
7083 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7084 or "require"), the connection is dropped.
7086 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7087 that yield "discard" can now be used with an "accept" or a "discard" verb,
7088 but an error is generated for any others (because I can't see a useful way
7089 to define what should happen).
7091 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7092 are now processed in the same way as lines in the Exim configuration file.
7093 In particular, continuation lines are supported.
7095 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7097 15. Added -ti meaning -t -i.
7099 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7100 domains, and warn by logging if others are found.
7102 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7103 returned from GIFCONF when Exim is trying to find the list of interfaces on
7104 a host. The code in os.c has been modified to copy each ifreq to an aligned
7105 structure in all cases.
7107 Also, in some cases, the returned ifreq's were being copied to a 'struct
7108 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7109 means the last couple of bytes of an IPv6 address could be chopped if the
7110 ifreq contained only a normal sockaddr (14 bytes storage).
7112 18. Named domain lists were not supported in the hosts_treat_as_local option.
7113 An entry such as +xxxx was not recognized, and was treated as a literal
7116 19. Ensure that header lines added by a DATA ACL are included in the reject log
7117 if the ACL subsequently rejects the message.
7119 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7120 MD5 (which is deprecated).
7122 21. When testing a filter file using -bf, Exim was writing a message when it
7123 took the sender from a "From " line in the message, but it was not doing so
7124 when it took $return_path from a Return-Path: header line. It now does.
7126 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7127 with a valid header line field name (a series of printing characters
7128 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7130 23. Changed "disc" in the source to "disk" to conform to the documentation and
7131 the book and for uniformity.
7133 24. Ignore Sendmail's -Ooption=value command line item.
7135 25. When execve() failed while trying to run a command in a pipe transport,
7136 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7137 could be confused with a return value of 69 from the command itself. This
7138 has been changed to 127, the value the shell returns if it is asked to run
7139 a non-existent command. The wording for the related log line suggests a
7140 non-existent command as the problem.
7142 26. If received_header_text expands to an empty string, do not add a Received:
7143 header line to the message. (Well, it adds a token one on the spool, but
7144 marks it "old" so that it doesn't get used or transmitted.)
7146 27. Installed eximstats 1.28 (addition of -nt option).
7148 28. There was no check for failure on the call to getsockname() in the daemon
7149 code. This can fail if there is a shortage of resources on the system, with
7150 ENOMEM, for example. A temporary error is now given on failure.
7152 29. Contrary to the C standard, it seems that in some environments, the
7153 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7154 program. Exim now does this explicitly; it affects the formatting of
7155 timestamps using strftime().
7157 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7158 complaints. I've now initialized all the variables, to avoid this.
7160 32. Header lines added by a system filter were not being "seen" during
7161 transport-time rewrites.
7163 33. The info_callback() function passed to OpenSSL is set up with type void
7164 (*)(SSL *, int, int), as described somewhere. However, when calling the
7165 function (actually a macro) that sets it up, the type void(*)() is
7166 expected. I've put in a cast to prevent warnings from picky compilers.
7168 34. If a DNS black list lookup found a CNAME record, but there were no A
7169 records associated with the domain it pointed at, Exim crashed.
7171 35. If a DNS black list lookup returned more than one A record, Exim ignored
7172 all but the first. It now scans all returned addresses if a particular IP
7173 value is being sought. In this situation, the contents of the
7174 $dnslist_value variable are a list of all the addresses, separated by a
7177 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7178 to accept a host name and all its aliases if the forward lookup for any of
7179 them yielded the IP address of the incoming connection. Now it accepts only
7180 those names whose forward lookup yields the correct IP address. Any other
7181 names are discarded. This closes a loophole whereby a rogue DNS
7182 administrator could create reverse DNS records to break through a
7183 wildcarded host restriction in an ACL.
7185 37. If a user filter or a system filter that ran in a subprocess used any of
7186 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7187 the wrong values were passed to the pipe command ($thisaddress had the
7188 value of $0, $0 had the value of $1, etc). This bug was introduced by
7189 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7191 38. Improved the line breaking for long SMTP error messages from ACLs.
7192 Previously, if there was no break point between 40 and 75 characters, Exim
7193 left the rest of the message alone. Two changes have been made: (a) I've
7194 reduced the minimum length to 35 characters; (b) if it can't find a break
7195 point between 35 and 75 characters, it looks ahead and uses the first one
7196 that it finds. This may give the occasional overlong line, but at least the
7197 remaining text gets split now.
7199 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7200 file descriptors might be low, and that setting 1000 would always raise it.
7201 It turns out that in some environments, the limit is already over 1000 and
7202 that lowering it causes trouble. So now Exim takes care not to decrease it.
7204 40. When delivering a message, the value of $return_path is set to $sender_
7205 address at the start of routing (routers may change the value). By an
7206 oversight, this default was not being set up when an address was tested by
7207 -bt or -bv, which affected the outcome if any router or filter referred to
7210 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7211 log only when "message" or "log_message" are set. However, if one of the
7212 conditions was an address verification, or a call to a nested ACL, the
7213 messages generated by the underlying test were being passed through. This
7214 no longer happens. The underlying message is available in $acl_verify_
7215 message for both "message" and "log_message" expansions, so it can be
7216 passed through if needed.
7218 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7219 new expansion $bh_ to give the encoded byte string without charset
7220 translation. Translation happens only if iconv() is available; HAVE_ICONV
7221 indicates this at build time. HEADERS_CHARSET gives the charset to
7222 translate to; headers_charset can change it in the configuration, and
7223 "headers charset" can change it in an individual filter file.
7225 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7226 that creates RFC 2047 encoded "words" labels them as that charset instead
7227 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7228 expansion operator; (ii) when Exim creates a From: line for a local
7229 message; (iii) when a header line is rewritten to include a "phrase" part.
7231 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7232 buggy, causing it to skip the first lines of messages whose message ID
7233 ended in 'D'. This would not have bitten before Exim release 4.14, because
7234 message IDs were unlikely to end in 'D' before then. The effect was to have
7235 incorrect size information for certain domains.
7237 45. #include "config.h" was missing at the start of the crypt16.c module. This
7238 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7241 46. If there was a timeout during a "random" callout check, Exim treated it as
7242 a failure of the random address, and carried on sending RSET and the real
7243 address. If the delay was just some slowness somewhere, the response to the
7244 original RCPT would be taken as a response to RSET and so on, causing
7245 mayhem of various kinds.
7247 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7248 when I implemented it. It didn't allow for the fact that some option values
7249 may legitimately be negative (e.g. size_addition), and it didn't even do
7250 the right test for positive values.
7252 48. Domain names in DNS records are case-independent. Exim always looks them up
7253 in lower case. Some resolvers return domain names in exactly the case they
7254 appear in the zone file, that is, they may contain uppercase letters. Not
7255 all resolvers do this - some return always lower case. Exim was treating a
7256 change of case by a resolver as a change of domain, similar to a widening
7257 of a domain abbreviation. This triggered its re-routing code and so it was
7258 trying to route what was effectively the same domain again. This normally
7259 caused routing to fail (because the router wouldn't handle the domain
7260 twice). Now Exim checks for this case specially, and just changes the
7261 casing of the domain that it ultimately uses when it transmits the message
7264 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7267 50. If a filter generated a file delivery with a non-absolute name (possible if
7268 no home directory exists for the router), the forbid_file option was not
7271 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7272 the existing equality matching.
7274 52. Exim was using ints instead of ino_t variables in some places where it was
7275 dealing with inode numbers.
7277 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7278 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7279 called TMPDIR, and if it finds it is different, it changes its value.
7281 54. The smtp_printf() function is now made available to local_scan() so
7282 additional output lines can be written before returning. There is also an
7283 smtp_fflush() function to enable the detection of a dropped connection.
7284 The variables smtp_input and smtp_batched_input are exported to
7287 55. Changed the default runtime configuration: the message "Unknown user"
7288 has been removed from the ACL, and instead placed on the localuser router,
7289 using the cannot_route_message feature. This means that any verification
7290 failures that generate their own messages won't get overridden. Similarly,
7291 the "Unrouteable address" message that was in the ACL for unverifiable
7292 relay addresses has also been removed.
7294 56. Added hosts_avoid_esmtp to the smtp transport.
7296 57. The exicyclog script was not checking for the esoteric option
7297 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7298 will work only if exicyclog is run under the appropriate euid.
7300 58. Following a discussion on the list, the rules by which Exim recognises line
7301 endings on incoming messages have been changed. The -dropcr and drop_cr
7302 options are now no-ops, retained only for backwards compatibility. The
7303 following line terminators are recognized: LF CRLF CR. However, special
7304 processing applies to CR:
7306 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7307 nor a local message in the state where . is a terminator.
7309 (ii) If a bare CR is encountered in a header line, an extra space is added
7310 after the line terminator so as not to end the header. The reasoning
7311 behind this is that bare CRs in header lines are most likely either
7312 to be mistakes, or people trying to play silly games.
7314 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7315 was being incorrectly given as 18 bytes larger than it should have been.
7316 This is a VOB (very old bug).
7318 60. This may never have affected anything current, but just in case it has:
7319 When the local host is found other than at the start of a list of hosts,
7320 the local host, those with the same MX, and any that follow, are discarded.
7321 When the list in question was part of a longer list of hosts, the following
7322 hosts (not currently being processed) were also being discarded. This no
7323 longer happens. I'm not sure if this situation could ever has previously
7326 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7329 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7330 header. This is something that is recommended in a new Internet Draft, and
7331 is something that is documented as being done by Sendmail. There are two
7332 possible values. For messages generated by the autoreply transport, Exim
7335 Auto-Submitted: auto-replied
7337 whereas for all other generated messages (e.g. bounces) it adds
7339 Auto-Submitted: auto-generated
7341 63. The "personal" condition in filters now includes a test for the
7342 Auto-Submitted: header. If it contains the string "auto-" the message it
7343 not considered personal.
7345 64. Added rcpt_include_affixes as a generic transport option.
7347 65. Added queue_only_override (default true).
7349 66. Added the syslog_duplication option.
7351 67. If what should have been the first header line of a message consisted of
7352 a space followed by a colon, Exim was mis-interpreting it as a header line.
7353 It isn't of course - it is syntactically invalid and should therefore be
7354 treated as the start of the message body. The misbehaviour could have
7355 caused a number of strange effects, including loss of data in subsequent
7356 header lines, and spool format errors.
7358 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7359 client host had authenticated. This control can now be exercised by an ACL
7360 for more flexibility.
7362 69. By default, callouts do not happen when testing with -bh. There is now a
7363 variant, -bhc, which does actually run the callout code, including
7364 consulting and updating the callout cache.
7366 70. Added support for saslauthd authentication, courtesy of Alexander
7369 71. If statvfs() failed on the spool or log directories while checking their
7370 size for availability, Exim confusingly gave the error "space shortage".
7371 Furthermore, in debugging mode it crashed with a floating point exception.
7372 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7373 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7374 serious problem, Exim now writes to the main and panic logs when this
7375 happens, with details of the failure. It then refuses to accept the
7376 incoming message, giving the message "spool directory problem" or "log
7377 directory problem" with a 421 code for SMTP messages.
7379 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7380 0, 1, and 2 exist, because some OS complain for execs without them (see
7381 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7382 descriptors. However, the code omitted to check that the open succeeded,
7383 causing mysterious errors if for some reason the permissions on /dev/null
7384 got screwed. Now Exim writes a message to the main and panic logs, and
7385 bombs out if it can't open /dev/null.
7387 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7388 interact so that it is all more flexible. It is supposed to remain
7389 backwards compatible. Also added extra_local_interfaces.
7391 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7392 to bomb out with an assertion failure - to the client this appears as a
7393 connection drop. This problem occurs in the part of the code that was taken
7394 from the Samba project. Fortunately, the assertion is in a very simple
7395 function, so I have fixed this by reproducing the function inline in the
7396 one place where it is called, and arranging for authentication to fail
7397 instead of killing the process with assert().
7399 75. The SPA client code was not working when the server requested OEM rather
7400 than Unicode encoding.
7402 76. Added code to make require_files with a specific uid setting more usable in
7403 the case where statting the file as root fails - usually a non-root-mounted
7404 NFS file system. When this happens and the failure is EACCES, Exim now
7405 forks a subprocess and does the per-uid checking as the relevant uid.
7407 77. Added process_log_path.
7409 78. If log_file_path was not explicitly set, a setting of check_log_space or
7410 check_log_inodes was ignored.
7412 79. If a space check for the spool or log partitions fails, the incident is now
7413 logged. Of course, in the latter case the data may get lost...
7415 80. Added the %p formatting code to string_format() so that it can be used to
7416 print addresses in debug_print(). Adjusted all the address printing in the
7417 debugging in store.c to use %p rather than %d.
7419 81. There was a concern that a line of code in smtp_in.c could overflow a
7420 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7421 initially expressed, the concern was not well-founded, because trailing
7422 spaces are removed early. However, if the trailing spaces were followed by
7423 a NULL, they did not get removed, so the overflow was possible. Two fixes
7426 (a) I re-wrote the offending code in a cleaner fashion.
7427 (b) If an incoming SMTP command contains a NULL character, it is rejected
7430 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7431 runs initgroups(), so that if the Exim user is in any additional groups,
7432 they will be used during message reception.
7438 The change log for 4.20 and earlier releases has been archived.