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
79 JH/01 Remove code calling the customisable local_scan function, unless a new
80 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
82 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
83 non-signal-safe functions being used.
85 JH/03 Bug 2269: When presented with a received message having a stupidly large
86 number of DKIM-Signature headers, disable DKIM verification to avoid
87 a resource-consumption attack. The limit is set at twenty.
89 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
90 report of oldest_pass in ${authres } in consequence, and separate out
91 some descriptions of reasons for verification fail.
93 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
94 files in the spool were present and unlocked. A queue-runner could spot
95 them, resulting in a duplicate delivery. Fix that by doing the unlock
96 after the unlink. Investigation by Tim Stewart. Take the opportunity to
97 add more error-checking on spoolfile handling while that code is being
100 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
101 No known attacks, no CVE, this is defensive hardening.
103 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
104 a queue-runner could start a delivery while other operations were ongoing.
105 Cutthrough delivery was a common victim, resulting in duplicate delivery.
106 Found and investigated by Tim Stewart. Fix by using the open message data
107 file handle rather than opening another, and not locally closing it (which
108 releases a lock) for that case, while creating the temporary .eml format
109 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
111 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
112 $sender_verify_failure/$recipient_verify_failure to "random".
114 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
117 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
118 Previously this would segfault.
120 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
123 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
124 like zero, since the resolver should be doing this for us, But we need one
125 as a CNAME but no MX presence gets the CNAME returned; we need to check
126 that doesn't point to an MX to declare it "no MX returned" rather than
127 "error, loop". A new main option is added so the older capability of
128 following some limited number of chain links is maintained.
130 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
132 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
133 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
134 now-deprecated earlier definitions used only specified the range up to TLS
135 1.2 (in the older-version library docs).
137 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
139 JH/15 Rework TLS client-side context management. Stop using a global, and
140 explicitly pass a context around. This enables future use of TLS for
141 connections to service-daemons (eg. malware scanning) while a client smtp
142 connection is using TLS; with cutthrough connections this is quite likely.
144 JH/16 Fix ARC verification to do AS checks in reverse order.
146 JH/17 Support a "tls" option on the ${readsocket } expansion item.
148 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
149 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
150 Previously the "utf8" would be re-prepended for every additional message.
152 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
153 Previously thery were accepted, resulting in issues when attempting to
154 forward messages to a non-supporting MTA.
156 PP/02 Let -n work with printing macros too, not just options.
158 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
159 one parent address was copied, and bogus data was used at delivery-logging
160 time. Either a crash (after delivery) or bogus log data could result.
161 Discovery and analysis by Tim Stewart.
163 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
164 Previously if the string ended mid-character, we did not insert the
165 promised '?' replacement.
167 PP/04 Documentation: current string operators work on bytes, not codepoints.
169 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
170 should pack well giving a smaller memory footprint so better caching and
171 therefore performance. Group the declarations where this can't be done so
172 that the byte-sized flag variables are not interspersed among pointer
173 variables, giving a better chance of good packing by the compiler.
175 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
176 non-null, to avoid issues with sites running BATV. Previously reports were
177 sent with an empty envelope sender so looked like bounces.
179 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
180 The ignore_error flag wasn't being returned from the filter subprocess so
181 was not set for later routers. Investigation and fix by Matthias Kurz.
183 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
184 and a msg:complete for the whole, when a message is manually removed using
185 -Mrm. Developement by Matthias Kurz, hacked on by JH.
187 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
188 a "Gnu special" function, asprintf() in the DB utility binary builds; I
189 hope that is portable enough.
191 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
192 requiring a known-CA anchor certificate; make it now rely entirely on the
193 TLSA as an anchor. Checking the name on the leaf cert against the name
194 on the A-record for the host is still done for TA (but not for EE mode).
196 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
197 included in delivery lines for non-proxied connections, when compiled with
198 SUPPORT_SOCKS and running with proxy logging enabled.
200 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
201 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
202 move the existing event to fire before the normal logging of message
203 failure so that custom logging is bracketed by normal logging.
205 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
206 msg:fail:internal event. Developement by Matthias Kurz.
208 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
209 far too small for todays use of crypto signatures stored there. Go all
210 the way to the max DNS message size of 64kB, even though this might be
211 overmuch for IOT constrained device use.
213 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
214 copy a string over itself. The library routine is documented as not
215 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
217 JH/32 For main options check_spool_space and check_inode_space, where the
218 platform supports 64b integers, support more than the previous 2^31 kB
219 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
220 the previous G, M, k.
222 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
223 $authenticated_fail_id variable on authentication failure. Previously
226 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
227 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
228 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
229 GNUTLS_SEC_PARAM_MEDIUM.
231 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
232 side. Previously we would continue as if no SNI had been received.
234 JH/36 Harden the handling of string-lists. When a list consisted of a sole
235 "<" character, which should be a list-separator specification, we walked
236 off past the nul-terimation.
238 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
239 causes) even when the retry time is not yet met. Previously they were
240 not, meaning that when (say) an account was over-quota and temp-rejecting,
241 and multiple senders' messages were queued, only one sender would get
242 notified on each configured delay_warning cycle.
244 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
246 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
247 cipher-suites, an error can be left on the stack even for a succeeding
248 accept; this results in impossible error messages when a later operation
251 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
252 return error codes indicating retry. Under TLS1.3 this becomes required.
254 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
255 it only wrote the new authenticators, resulting in a lack of tracking of
256 peer changes of ESMTP extensions until the next cache flush.
258 JH/41 Fix the loop reading a message header line to check for integer overflow,
259 and more-often against header_maxsize. Previously a crafted message could
260 induce a crash of the recive process; now the message is cleanly rejected.
262 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
263 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
269 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
270 When redis_servers is set to a list of > 1 element, and the Redis servers
271 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
272 case of MOVED into a DEFER case instead, thus moving the query onto the
273 next server in the list. For a cluster of N elements, all N servers must
274 be defined in redis_servers.
276 GF/02 Catch and remove uninitialized value warning in exiqsumm
277 Check for existence of @ARGV before looking at $ARGV[0]
279 JH/01 Replace the store_release() internal interface with store_newblock(),
280 which internalises the check required to safely use the old one, plus
281 the allocate and data copy operations duplicated in both (!) of the
282 extant use locations.
284 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
285 modifier. This matches the restriction on the commandline.
287 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
288 Previously only the last row was returned.
290 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
291 we assumed that tags in the header were well-formed, and parsed the
292 element content after inspecting only the first char of the tag.
293 Assumptions at that stage could crash the receive process on malformed
296 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
297 While running the DKIM ACL we operate on the Permanent memory pool so that
298 variables created with "set" persist to the DATA ACL. Also (at any time)
299 DNS lookups that fail create cache records using the Permanent pool. But
300 expansions release any allocations made on the current pool - so a dnsdb
301 lookup expansion done in the DKIM ACL releases the memory used for the
302 DNS negative-cache, and bad things result. Solution is to switch to the
303 Main pool for expansions.
304 While we're in that code, add checks on the DNS cache during store_reset,
305 active in the testsuite.
306 Problem spotted, and debugging aided, by Wolfgang Breyha.
308 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
309 When none of the hosts presented to a transport match an already-open
310 connection, close it and proceed with the list. Previously we would
311 queue the message. Spotted by Lena with Yahoo, probably involving
314 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
315 Previously a spurious "250 OK id=" response was appended to the proper
318 JH/08 The "support for" informational output now, which built with Content
319 Scanning support, has a line for the malware scanner interfaces compiled
320 in. Interface can be individually included or not at build time.
322 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
323 by the template makefile "src/EDITME". The "STREAM" support for an older
324 ClamAV interface method is removed.
326 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
327 rows affected is given instead).
329 JH/11 The runtime Berkeley DB library version is now additionally output by
330 "exim -d -bV". Previously only the compile-time version was shown.
332 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
333 SMTP connection. Previously, when one had more receipients than the
334 first, an abortive onward connection was made. Move to full support for
335 multiple onward connections in sequence, handling cutthrough connection
336 for all multi-message initiating connections.
338 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
339 routers. Previously, a multi-recipient message would fail to match the
340 onward-connection opened for the first recipient, and cause its closure.
342 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
343 a timeout on read on a GnuTLS initiating connection, resulting in the
344 initiating connection being dropped. This mattered most when the callout
345 was marked defer_ok. Fix to keep the two timeout-detection methods
348 JH/15 Relax results from ACL control request to enable cutthrough, in
349 unsupported situations, from error to silently (except under debug)
350 ignoring. This covers use with PRDR, frozen messages, queue-only and
353 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
355 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
356 metadata, resulting in a crash in free().
358 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
359 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
360 Broken also in d185889f4, with init system revamp.
362 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
363 Previously we abruptly closed the connection after reading a malware-
364 found indication; now we go on to read the "scan ok" response line,
367 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
368 ACL. Previously, a crash would result.
370 JH/19 Speed up macro lookups during configuration file read, by skipping non-
371 macro text after a replacement (previously it was only once per line) and
372 by skipping builtin macros when searching for an uppercase lead character.
374 JH/20 DANE support moved from Experimental to mainline. The Makefile control
375 for the build is renamed.
377 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
378 was allocated for every new TLS startup, meaning one per message. Fix
379 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
381 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
382 reported the original. Fix to report (as far as possible) the ACL
383 result replacing the original.
385 JH/23 Fix memory leak during multi-message connections using STARTTLS under
386 OpenSSL. Certificate information is loaded for every new TLS startup,
387 and the resources needed to be freed.
389 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
391 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
394 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
395 DATA response info to the (existing) per-recipient response info for
396 the "C=" log element. It can have useful tracking info from the
397 destination system. Patch from Simon Arlott.
399 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
400 length value. Previously this would segfault.
402 HS/02 Support Avast multiline protoocol, this allows passing flags to
403 newer versions of the scanner.
405 JH/28 Ensure that variables possibly set during message acceptance are marked
406 dead before release of memory in the daemon loop. This stops complaints
407 about them when the debug_store option is enabled. Discovered specifically
408 for sender_rate_period, but applies to a whole set of variables.
409 Do the same for the queue-runner and queue-list loops, for variables set
410 from spool message files. Do the same for the SMTP per-message loop, for
411 certain variables indirectly set in ACL operations.
413 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
414 as a multi-recipient message from a mailinglist manager). The coding had
415 an arbitrary cutoff number of characters while checking for more input;
416 enforced by writing a NUL into the buffer. This corrupted long / fast
417 input. The problem was exposed more widely when more pipelineing of SMTP
418 responses was introduced, and one Exim system was feeding another.
419 The symptom is log complaints of SMTP syntax error (NUL chars) on the
420 receiving system, and refused recipients seen by the sending system
421 (propating to people being dropped from mailing lists).
422 Discovered and pinpointed by David Carter.
424 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
425 replaced by the ${authresults } expansion.
427 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
429 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
430 allows proper process termination in container environments.
432 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
433 Previously the "final dot" had a newline after it; ensure it is CR,LF.
435 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
436 and "err_perm", deprecated since 4.83 when the RFC-defined words
437 "temperror" and "permerror" were introduced.
439 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
440 transport-filters or DKIM-signing. The restriction was lost in the
441 consolidation of verify-callout and delivery SMTP handling.
442 Extend the restriction to also cover ARC-signing.
444 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
445 in defer=pass mode supply a 450 to the initiator. Previously the message
448 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
449 tls_require_ciphers is used as before.
451 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
452 "pass_unscanned". Only tmpfails from the scanner are written to
453 the paniclog, as they may require admin intervention (permission
454 denied, license issues). Other scanner errors (like decompression
455 bombs) do not cause a paniclog entry.
457 JH/36 Fix reinitialisation of DKIM logging variable between messages.
458 Previously it was possible to log spurious information in receive log
461 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
462 triggered odd behaviour from Outlook Express clients.
464 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
467 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
468 since the IETF WG has not yet settled on that versus the original
469 "bare" representation.
471 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
472 Previously the millisecond value corrupted the output.
473 Fix also for syslog_pid=no and log_selector +pid, for which the pid
474 corrupted the output.
480 JH/01 Rework error string handling in TLS interface so that the caller in
481 more cases is responsible for logging. This permits library-sourced
482 string to be attached to addresses during delivery, and collapses
483 pairs of long lines into single ones.
485 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
486 during configuration. Wildcards are allowed and expanded.
488 JH/02 Rework error string handling in DKIM to pass more info back to callers.
489 This permits better logging.
491 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
492 do not close it down and have the child transport start it up again on
493 the passed-on TCP connection. Instead, proxy the child (and any
494 subsequent ones) for TLS via a unix-domain socket channel. Logging is
495 affected: the continued delivery log lines do not have any DNSSEC, TLS
496 Certificate or OCSP information. TLS cipher information is still logged.
498 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
499 identical IP addresses on different listening ports. Will also affect
502 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
503 add noisy ifdef guards to special-case this sillyness.
504 Patch from Bernd Kuhls.
506 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
507 than 255 are no longer allowed.
509 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
510 Disable the session-cache too, which might reduce our load. Since we
511 currrectly use a new context for every connection, both as server and
512 client, there is no benefit for these.
513 GnuTLS appears to not support tickets server-side by default (we don't
514 call gnutls_session_ticket_enable_server()) but client side is enabled
515 by default on recent versions (3.1.3 +) unless the PFS priority string
518 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
519 <https://reproducible-builds.org/specs/source-date-epoch/>.
521 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
522 the check for any unsuccessful recipients did not notice the limit, and
523 erroneously found still-pending ones.
525 JH/08 Pipeline CHUNKING command and data together, on kernels that support
526 MSG_MORE. Only in-clear (not on TLS connections).
528 JH/09 Avoid using a temporary file during transport using dkim. Unless a
529 transport-filter is involved we can buffer the headers in memory for
530 creating the signature, and read the spool data file once for the
531 signature and again for transmission.
533 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
534 4.77 as the kernel support then wasn't solid, having issues in 64bit
535 mode. Now, it's been long enough. Add support for FreeBSD also.
537 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
538 case where the routing stage had gathered several addresses to send to
539 a host before calling the transport for the first, we previously failed
540 to close down TLS in the old transport process before passing the TCP
541 connection to the new process. The new one sent a STARTTLS command
542 which naturally failed, giving a failed delivery and bloating the retry
543 database. Investigation and fix prototype from Wolfgang Breyha.
545 JH/12 Fix check on SMTP command input synchronisation. Previously there were
546 false-negatives in the check that the sender had not preempted a response
547 or prompt from Exim (running as a server), due to that code's lack of
548 awareness of the SMTP input buffering.
550 PP/04 Add commandline_checks_require_admin option.
551 Exim drops privileges sanely, various checks such as -be aren't a
552 security problem, as long as you trust local users with access to their
553 own account. When invoked by services which pass untrusted data to
554 Exim, this might be an issue. Set this option in main configuration
555 AND make fixes to the calling application, such as using `--` to stop
558 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
559 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
560 responses to those, into a single TLS record each way (this usually means
561 a single packet). As a side issue, smtp_enforce_sync now works on TLS
564 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
565 affects you only if you're dancing at the edge of the param size limits.
566 If you are, and this message makes sense to you, then: raise the
567 configured limit or use OpenSSL 1.1. Nothing we can do for older
570 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
571 cmdline element to get the documented default one. Previously it was
574 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
575 are unsuable for non-message-specific reasons - eg. connection timeout,
576 banner-time rejection.
578 JH/16 Fix logging of delivery remote port, when specified by router, under
581 PP/06 Repair manualroute's ability to take options in any order, even if one
582 is the name of a transport.
585 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
587 JH/17 Change the list-building routines interface to use the expanding-string
588 triplet model, for better allocation and copying behaviour.
590 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
591 Previously it was constructed the first time a possibly-matching string
592 was met in the configuration file input during startup; now it is done
595 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
596 compatible one, to avoid the (poorly documented) possibility of a config
597 file in the working directory redirecting the DB files, possibly correpting
598 some existing file. CVE-2017-10140 assigned for BDB.
600 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
601 cache-hot. Previously, although the result was properly cached, the
602 initial verify call returned a defer.
604 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
605 the main verify for receipient in uncached-mode.
607 JH/22 Retire historical build files to an "unsupported" subdir. These are
608 defined as "ones for which we have no current evidence of testing".
610 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
611 if present. Previously it was ignored.
613 JH/24 Start using specified-initialisers in C structure init coding. This is
614 a C99 feature (it's 2017, so now considered safe).
616 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
617 if was a fixed-sized field and bitmask ops via macros; it is now more
620 PP/07 GitHub PR 56: Apply MariaDB build fix.
621 Patch provided by Jaroslav Å karvada.
623 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
624 during Coverity cleanups [4.87 JH/47]
625 Diagnosis and fix provided by Michael Fischer v. Mollard.
627 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
628 the right size to place the terminating semicolon on its own folded
629 line, the header hash was calculated to an incorrect value thanks to
630 the (relaxed) space the fold became.
632 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
633 and confused the parent.
635 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
636 which could crash as a result. This could lead to undeliverable messages.
638 JH/28 Logging: "next input sent too soon" now shows where input was truncated
641 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
642 matters on fast-turnover and PID-randomising systems, which were getting
643 out-of-order delivery.
645 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
646 a possibly-overlapping copy. The symptom was that "Remote host closed
647 connection in response to HELO" was logged instead of the actual 4xx
650 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
651 Previously only that bufferd was discarded, resulting in SYMTP command
654 JH/32 DKIM: when a message has multiple signatures matching an identity given
655 in dkim_verify_signers, run the dkim acl once for each. Previously only
656 one run was done. Bug 2189.
658 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
659 "panic the current process" to "deliberately defer". The panic log is
660 still written with the problem list name; the mail and reject logs now
661 get a temp-reject line for the message that was being handled, saying
662 something like "domains check lookup or other defer". The SMTP 451
663 message is still "Temporary local problem".
665 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
666 A crafted sequence of BDAT commands could result in in-use memory beeing
667 freed. CVE-2017-16943.
669 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
670 from SMTP input. Previously it was always done; now only done for DATA
671 and not BDAT commands. CVE-2017-16944.
673 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
674 to the message (such as an overlong header line). Previously this was
675 not done and we did not exit BDAT mode. Followon from the previous item
676 though a different problem.
682 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
683 than -2003 did; needs libidn2 in addition to libidn.
685 JH/02 The path option on a pipe transport is now expanded before use.
687 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
688 Patch provided by "Björn", documentation fix added too.
690 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
691 missing a wire-to-host endian conversion.
693 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
694 close after a BDAT command line could be taken as a following command,
695 giving a synch failure. Fix by only checking for synch immediately
696 before acknowledging the chunk.
698 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
699 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
700 macro. Patches provided by Josh Soref.
702 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
703 Previously we did not; the RFC seems ambiguous and VRFY is not listed
704 by IANA as a service extension. However, John Klensin suggests that we
707 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
708 the dkim code may be unix-mode line endings rather than smtp wire-format
709 CRLF, so prepend a CR to any bare LF.
711 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
712 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
714 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
715 the first were themselves being wrongly included in the feed into dkim
716 processing; with most chunk sizes in use this resulted in an incorrect
717 body hash calculated value.
719 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
720 DKIM signature block, for verification. Although advised against by
721 standards it is specifically not ruled illegal.
723 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
725 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
726 missing a body hash (the bh= tag).
728 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
729 It seems that HAProxy sends the Proxy Protocol information in clear and
730 only then does a TLS startup, so do the same.
732 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
733 TCP connections (such as for Spamd) unless the daemon successfully set
734 Fast Open mode on its listening sockets. This fixes breakage seen on
735 too-old kernels or those not configured for Fast Open, at the cost of
736 requiring both directions being enabled for TFO, and TFO never being used
737 by non-daemon-related Exim processes.
739 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
740 endings, at least on the first header line. Try to canonify any that get
741 past that check, despite the cost.
743 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
744 now limited to an arbitrary five deep, while parsing addresses with the
745 strip_excess_angle_brackets option enabled.
747 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
748 instead leave the unprompted TLS handshake in socket buffer for the
749 TLS library to consume.
751 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
753 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
755 JH/16 Drop variables when they go out of scope. Memory management drops a whole
756 region in one operation, for speed, and this leaves assigned pointers
757 dangling. Add checks run only under the testsuite which checks all
758 variables at a store-reset and panics on a dangling pointer; add code
759 explicitly nulling out all the variables discovered. Fixes one known
760 bug: a transport crash, where a dangling pointer for $sending_ip_address
761 originally assigned in a verify callout, is re-used.
763 PP/06 Drop '.' from @INC in various Perl scripts.
765 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
767 PP/08 Reduce a number of compilation warnings under clang; building with
768 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
769 should be warning-free.
771 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
773 HS/01 Fix portability problems introduced by PP/08 for platforms where
774 realloc(NULL) is not equivalent to malloc() [SunOS et al].
776 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
777 chunk. This allows us to accept broken chunked messages. We need a more
778 general solution here.
780 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
781 already-broken messages in the queue.
783 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
785 JH/19 Fix reference counting bug in routing-generated-address tracking.
791 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
792 supports it and a size is available (ie. the sending peer gave us one).
794 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
795 years of being deprecated). The replacements are the ACLs
796 acl_smtp_mime and acl_not_smtp_mime.
798 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
799 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
800 or even an in-clear connection were permitted. Now, if the host lookup
801 was dnssec and dane was requested then the host is only used if the
802 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
803 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
804 if one fails this test.
805 This means that a poorly-configured remote DNS will make it incommunicado;
806 but it protects against a DNS-interception attack on it.
808 JH/04 Bug 1810: make continued-use of an open smtp transport connection
809 non-noisy when a race steals the message being considered.
811 JH/05 If main configuration option tls_certificate is unset, generate a
812 self-signed certificate for inbound TLS connections.
814 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
815 in rewrites and routers.
817 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
818 and logged a warning sing 4.83; now they are a configuration file error.
820 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
821 (lacking @domain). Apply the same qualification processing as RCPT.
823 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
825 JH/10 Support ${sha256:} applied to a string (as well as the previous
828 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
829 a cutthrough deliver is pending, as we always want to make a connection.
830 This also avoids re-routing the message when later placing the cutthrough
831 connection after a verify cache hit.
832 Do not update it with the verify result either.
834 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
835 when routing results in more than one destination address.
837 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
838 signing (which inhibits the cutthrough capability). Previously only
839 the presence of an option was tested; now an expansion evaluating as
840 empty is permissible (obviously it should depend only on data available
841 when the cutthrough connection is made).
843 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
844 the relevant preceding SMTP command did not note the pipelining mode.
846 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
847 Previously they were not counted.
849 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
850 as one having no matching records. Previously we deferred the message
851 that needed the lookup.
853 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
854 distinguished as "(=".
856 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
857 for missing MX records. Previously it only worked for missing A records.
859 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
861 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
862 after the data-go-ahead and data-ack. Patch from Jason Betts.
864 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
865 even for a "none" policy. Patch from Tony Meyer.
867 JH/22 Fix continued use of a connection for further deliveries. If a port was
868 specified by a router, it must also match for the delivery to be
871 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
872 When one of the recipients of a message was unsuitable for the connection
873 (has no matching addresses), we lost track of needing to mark it
874 deferred. As a result mail would be lost.
876 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
878 JH/25 Decoding ACL controls is now done using a binary search; the source code
879 takes up less space and should be simpler to maintain. Merge the ACL
880 condition decode tables also, with similar effect.
882 JH/26 Fix problem with one_time used on a redirect router which returned the
883 parent address unchanged. A retry would see the parent address marked as
884 delivered, so not attempt the (identical) child. As a result mail would
887 JH/27 Fix a possible security hole, wherein a process operating with the Exim
888 UID can gain a root shell. Credit to http://www.halfdog.net/ for
889 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
892 JH/28 Enable {spool,log} filesystem space and inode checks as default.
893 Main config options check_{log,spool}_{inodes,space} are now
894 100 inodes, 10MB unless set otherwise in the configuration.
896 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
897 Previously it only applied to the main-section connection policy
900 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
902 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
903 by me. Added RFC7919 DH primes as an alternative.
905 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
906 are not in the system include path.
908 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
909 GnuTLS, when a session startup failed (eg because the client disconnected)
910 Exim did stdio operations after fclose. This was exposed by a recent
911 change which nulled out the file handle after the fclose.
913 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
914 signed directly by the cert-signing cert, rather than an intermediate
915 OCSP-signing cert. This is the model used by LetsEncrypt.
917 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
919 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
920 an incoming connection.
922 HS/02 Bug 1802: Do not half-close the connection after sending a request
925 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
926 fallback to "prime256v1".
928 JH/34 SECURITY: Use proper copy of DATA command in error message.
929 Could leak key material. Remotely exploitable. CVE-2016-9963.
935 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
936 and 3.4.4 - once the server is enabled to respond to an OCSP request
937 it does even when not requested, resulting in a stapling non-aware
938 client dropping the TLS connection.
940 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
941 support variable-length bit vectors. No functional change.
943 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
944 The I= interface field on outgoing lines is now after the H= remote
945 host field, same as incoming lines. There is a separate
946 outgoing_interface log selector which allows you to disable the
949 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
950 If not running log_selector +smtp_connection the mainlog would be held
951 open indefinitely after a "too many connections" event, including to a
952 deleted file after a log rotate. Leave the per net connection logging
953 leaving it open for efficiency as that will be quickly detected by the
954 check on the next write.
956 HS/01 Bug 1671: Fix post transport crash.
957 Processing the wait-<transport> messages could crash the delivery
958 process if the message IDs didn't exist for some reason. When
959 using 'split_spool_directory=yes' the construction of the spool
960 file name failed already, exposing the same netto behaviour.
962 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
963 mime_regex ACL conditions.
965 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
966 to DSN fail messages (bounces): remote IP, remote greeting, remote response
967 to HELO, local diagnostic string.
969 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
970 log line to debug. Even when configured with a tls authenticator many
971 client connections are expected to not authenticate in this way, so
972 an authenticate fail is not an error.
974 HS/02 Add the Exim version string to the process info. This way exiwhat
975 gives some more detail about the running daemon.
977 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
978 matter for fast-change records such as DNSBLs.
980 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
981 retry record, even if constant. There may be multiple transports with
982 different interface settings and the retry behaviour needs to be kept
985 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
987 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
989 JH/10 Bug 840: fix log_defer_output option of pipe transport
991 JH/11 Bug 830: use same host for all RCPTS of a message, even under
992 hosts_randomize. This matters a lot when combined with mua_wrapper.
994 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
995 ${quote_pgsql:<string>} operator.
997 JH/13 Bug 1708: avoid misaligned access in cached lookup.
999 JH/14 Change header file name for freeradius-client. Relevant if compiling
1000 with Radius support; from the Gentoo tree and checked under Fedora.
1002 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1004 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1005 option transport_filter, meaning no filtering.
1007 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1009 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1010 defaults to "*" (all hosts). The variable is now available when not built
1011 with TLS, default unset, mainly to enable keeping the testsuite sane.
1012 If a server certificate is not supplied (via tls_certificate) an error is
1013 logged, and clients will find TLS connections fail on startup. Presumably
1014 they will retry in-clear.
1015 Packagers of Exim are strongly encouraged to create a server certificate
1016 at installation time.
1018 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1019 with the $config_file variable.
1021 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1022 in transport context, after the attempt, and per-recipient. The latter type
1023 is per host attempted. The event data is the error message, and the errno
1024 information encodes the lookup type (A vs. MX) used for the (first) host,
1025 and the trailing two digits of the smtp 4xx response.
1027 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1028 to write to mainlog (or rejectlog, paniclog) in the window between file
1029 creation and permissions/ownership being changed. Particularly affects
1030 installations where exicyclog is run as root, rather than exim user;
1031 result is that the running daemon panics and dies.
1033 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1035 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1036 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1037 "pri" and "weight". Note that the previous implicit priority given by the
1038 list order is no longer honoured.
1040 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1041 for DKIM processing.
1043 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1044 by defining SUPPORT_SOCKS.
1046 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1047 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1048 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1049 variables are renamed to proxy_{local,external}_{address,port}.
1051 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1052 for a build by defining SUPPORT_I18N
1054 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1055 of the query string, and make ${quote_redis:} do that quoting.
1057 JH/29 Move Events support from Experimental to mainline, enabled by default
1058 and removable for a build by defining DISABLE_EVENT.
1060 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1062 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1063 cached by the daemon.
1065 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1066 by defining LOOKUP_REDIS. The libhiredis library is required.
1068 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1069 keys are given for lookup.
1071 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1072 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1073 only supported when built with TLS support. The PolarSSL SHA routines
1074 are still used when the TLS library is too old for convenient support.
1076 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1077 openssl_options), for security. OpenSSL forces this from version 1.1.0
1078 server-side so match that on older versions.
1080 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1081 allocation for $value could be released as the expansion processing
1082 concluded, but leaving the global pointer active for it.
1084 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1085 and to use the domains and local_parts ACL conditions.
1087 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1088 incorrectly not doubled on cutthrough transmission, hence seen as a
1089 body-termination at the receiving system - resulting in truncated mails.
1090 Commonly the sender saw a TCP-level error, and retransmitted the message
1091 via the normal store-and-forward channel. This could result in duplicates
1092 received - but deduplicating mailstores were liable to retain only the
1093 initial truncated version.
1095 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1097 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1099 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1100 we're in there, support oversigning also; bug 1309.
1102 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1104 HS/04 Add support for keep_environment and add_environment options.
1106 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1107 either intentional arithmetic overflow during PRNG, or testing config-
1110 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1111 delivery resulted in actual delivery. Cancel cutthrough before DATA
1114 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1115 rejects a recipient: pass the reject to the originator.
1117 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1118 Many were false-positives and ignorable, but it's worth fixing the
1121 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1122 for the new environment-manipulation done at startup. Move the routines
1123 from being local to tls.c to being global via the os.c file.
1125 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1126 an extract embedded as result-arg for a map, the first arg for extract
1127 is unavailable so we cannot tell if this is a numbered or keyed
1128 extraction. Accept either.
1134 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1137 JH/02 The smtp transport option "multi_domain" is now expanded.
1139 JH/03 The smtp transport now requests PRDR by default, if the server offers
1142 JH/04 Certificate name checking on server certificates, when exim is a client,
1143 is now done by default. The transport option tls_verify_cert_hostnames
1144 can be used to disable this per-host. The build option
1145 EXPERIMENTAL_CERTNAMES is withdrawn.
1147 JH/05 The value of the tls_verify_certificates smtp transport and main options
1148 default to the word "system" to access the system default CA bundle.
1149 For GnuTLS, only version 3.0.20 or later.
1151 JH/06 Verification of the server certificate for a TLS connection is now tried
1152 (but not required) by default. The verification status is now logged by
1153 default, for both outbound TLS and client-certificate supplying inbound
1156 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1159 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1160 Status Notification (bounce) messages are now MIME format per RFC 3464.
1161 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1162 under the control of the dsn_advertise_hosts option, and routers may
1163 have a dsn_lasthop option.
1165 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1166 default, modifiable by a malware= option. The list separator for
1167 the options can now be changed in the usual way. Bug 68.
1169 JH/10 The smtp_receive_timeout main option is now expanded before use.
1171 JH/11 The incoming_interface log option now also enables logging of the
1172 local interface on delivery outgoing connections.
1174 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1175 if the interface and destination host and port all match.
1177 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1180 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1181 Patch from Andrew Lewis.
1183 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1184 now supports optional time-restrictions, weighting, and priority
1185 modifiers per server. Patch originally by <rommer@active.by>.
1187 JH/16 The spamd_address main option now supports a mixed list of local
1188 and remote servers. Remote servers can be IPv6 addresses, and
1189 specify a port-range.
1191 JH/17 Bug 68: The spamd_address main option now supports an optional
1192 timeout value per server.
1194 JH/18 Bug 1581: Router and transport options headers_add/remove can
1195 now have the list separator specified.
1197 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1200 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1203 JH/21 Support for the A6 type of dns record is withdrawn.
1205 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1206 rather than the verbs used.
1208 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1209 from 255 to 1024 chars.
1211 JH/24 Verification callouts now attempt to use TLS by default.
1213 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1214 are generic router options now. The defaults didn't change.
1216 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1217 Original patch from Alexander Shikoff, worked over by JH.
1219 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1220 files if ZCAT_COMMAND is not executable.
1222 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1224 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1226 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1227 Normally benign, it bites when the pair was led to by a CNAME;
1228 modern usage is to not canonicalize the domain to a CNAME target
1229 (and we were inconsistent anyway for A-only vs AAAA+A).
1231 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1233 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1234 when evaluating $sender_host_dnssec.
1236 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1237 $sender_helo_dnssec variable.
1239 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1241 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1243 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1245 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1246 documented as working, but never had. Support all but $spam_report.
1248 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1249 added for tls authenticator.
1251 HS/03 Add perl_taintmode main config option
1257 TL/01 When running the test suite, the README says that variables such as
1258 no_msglog_check are global and can be placed anywhere in a specific
1259 test's script, however it was observed that placement needed to be near
1260 the beginning for it to behave that way. Changed the runtest perl
1261 script to read through the entire script once to detect and set these
1262 variables, reset to the beginning of the script, and then run through
1263 the script parsing/test process like normal.
1265 TL/02 The BSD's have an arc4random API. One of the functions to induce
1266 adding randomness was arc4random_stir(), but it has been removed in
1267 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1268 function when detected.
1270 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1271 cause callback expansion.
1273 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1274 syntax errors in an expansion can be treated as a string instead of
1275 logging or causing an error, due to the internal use of bool_lax
1276 instead of bool when processing it.
1278 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1279 server certificates when making smtp deliveries.
1281 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1283 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1285 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1287 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1288 Merged patch from Sebastian Wiedenroth.
1290 JH/05 Fix results-pipe from transport process. Several recipients, combined
1291 with certificate use, exposed issues where response data items split
1292 over buffer boundaries were not parsed properly. This eventually
1293 resulted in duplicates being sent. This issue only became common enough
1294 to notice due to the introduction of connection certificate information,
1295 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1297 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1298 size buffer was used, resulting in syntax errors when an expansion
1301 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1302 version 3.3.6 or later.
1304 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1305 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1306 both become "event_action", the variables become $event_name, $event_data
1307 and $event_defer_errno. There is a new variable $verify_mode, usable in
1308 routers, transports and related events. The tls:cert event is now also
1309 raised for inbound connections, if the main configuration event_action
1312 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1313 early OCSP support, but no stapling (appears to be less than 1.0.0).
1315 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1316 server certificate names available under the smtp transport option
1317 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1320 JH/10 Time-related extraction expansions from certificates now use the main
1321 option "timezone" setting for output formatting, and are consistent
1322 between OpenSSL and GnuTLS compilations. Bug 1541.
1324 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1325 encoded parameter in the incoming message. Bug 1558.
1327 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1328 include certificate info, eximon was claiming there were spoolfile
1331 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1333 JH/14 Log delivery-related information more consistently, using the sequence
1334 "H=<name> [<ip>]" wherever possible.
1336 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1337 are problematic for Debian distribution, omit them from the release
1340 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1342 JH/16 Fix string representation of time values on 64bit time_t architectures.
1345 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1346 output list separator was used.
1351 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1352 checkers that were complaining about end of non-void function with no
1355 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1356 This was a regression introduced in 4.83 by another bugfix.
1358 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1360 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1361 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1367 TF/01 Correctly close the server side of TLS when forking for delivery.
1369 When a message was received over SMTP with TLS, Exim failed to clear up
1370 the incoming connection properly after forking off the child process to
1371 deliver the message. In some situations the subsequent outgoing
1372 delivery connection happened to have the same fd number as the incoming
1373 connection previously had. Exim would try to use TLS and fail, logging
1374 a "Bad file descriptor" error.
1376 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1377 utilities have not been installed.
1379 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1380 temporary space as the ACL may create new global variables.
1382 TL/01 LDAP support uses per connection or global context settings, depending
1383 upon the detected version of the libraries at build time.
1385 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1386 to extract and use the src ip:port in logging and expansions as if it
1387 were a direct connection from the outside internet. PPv2 support was
1388 updated based on HAProxy spec change in May 2014.
1390 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1392 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1393 Properly escape header and check for NULL return.
1395 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1398 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1400 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1401 characters in header names, implemented as a verify condition.
1402 Contributed by Michael Fischer v. Mollard.
1404 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1405 results permerror and temperror. Previous values are deprecated but
1406 still accepted. In a future release, err_perm and err_temp will be
1407 completely removed, which will be a backward incompatibility if the
1408 ACL tests for either of these two old results. Patch contributed by
1409 user bes-internal on the mailing list.
1411 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1413 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1414 selectors, in both main and reject logs.
1416 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1419 JH/07 Add malware type "sock" for talking to simple daemon.
1421 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1423 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1424 routers/transports under cutthrough routing.
1426 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1427 numbers. Touch up "bool" conditional to keep the same definition.
1429 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1431 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1433 JH/12 Expand items in router/transport headers_add or headers_remove lists
1434 individually rather than the list as a whole. Bug 1452.
1436 Required for reasonable handling of multiple headers_ options when
1437 they may be empty; requires that headers_remove items with embedded
1438 colons must have them doubled (or the list-separator changed).
1440 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1441 view the policy declared in the DMARC record. Currently, $dmarc_status
1442 is a combined value of both the record presence and the result of the
1445 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1447 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1448 dnslookup router and the smtp transport (applying to the forward
1451 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1452 of ldap servers used for a specific lookup. Patch provided by Heiko
1455 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1456 New variable $lookup_dnssec_authenticated for observability.
1458 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1459 Patch submitted by Lars Timman.
1461 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1463 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1464 Requires trusted mode and valid format message id, aborts otherwise.
1465 Patch contributed by Heiko Schlichting.
1467 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1468 certextract with support for various fields. Bug 1358.
1470 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1471 is requested by default, modifiable by smtp transport option
1474 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1475 operate on certificate variables to give certificate fingerprints
1476 Also new ${sha256:cert_variable}.
1478 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1480 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1481 Christian Aistleitner.
1483 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1485 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1486 file. Patch from Wolfgang Breyha.
1488 JH/25 Expand the coverage of the delivery $host and $host_address to
1489 client authenticators run in verify callout. Bug 1476.
1491 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1492 align with daemon_smtp_ports. Bug 72.
1494 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1495 support and error reporting did not work properly.
1497 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1498 and is readable. Patch from Andrew Colin Kissa.
1500 TL/14 Enhance documentation of ${run expansion and how it parses the
1501 commandline after expansion, particularly in the case when an
1502 unquoted variable expansion results in an empty value.
1504 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1506 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1507 by renaming away from it. Addresses GH issue 15, reported by
1510 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1512 TL/15 SECURITY: prevent double expansion in math comparison functions
1513 (can expand unsanitized data). Not remotely exploitable.
1520 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1522 PP/02 Make -n do something, by making it not do something.
1523 When combined with -bP, the name of an option is not output.
1525 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1528 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1529 $sender_host_name and config options to manage this, and basic check
1532 PP/05 DSCP support for outbound connections and control modifier for inbound.
1534 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1535 (Only plugin which currently uses this is kerberos4, which nobody should
1536 be using, but we should make it available and other future plugins might
1537 conceivably use it, even though it would break NAT; stuff *should* be
1538 using channel bindings instead).
1540 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1541 name; added for Sendmail compatibility; requires admin caller.
1542 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1543 just ignore it); requires trusted caller.
1544 Also parse but ignore: -Ac -Am -X<logfile>
1547 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1549 TL/02 Add +smtp_confirmation as a default logging option.
1551 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1552 Patch by Magnus Holmgren from 2007-02-20.
1554 TL/04 Bugzilla 1281 - Spec typo.
1555 Bugzilla 1283 - Spec typo.
1556 Bugzilla 1290 - Spec grammar fixes.
1558 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1560 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1562 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1563 mailing list by by Dmitry Isaikin.
1565 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1567 JH/02 Support "G" suffix to numbers in ${if comparisons.
1569 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1571 NM/01 Bugzilla 1197 - Spec typo
1572 Bugzilla 1196 - Spec examples corrections
1574 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1576 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1577 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1580 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1581 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1583 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1584 "acl {{name}{arg}...}", and optional args on acl condition
1587 JH/05 Permit multiple router/transport headers_add/remove lines.
1589 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1591 JH/07 Avoid using a waiting database for a single-message-only transport.
1592 Performance patch from Paul Fisher. Bugzilla 1262.
1594 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1597 JH/09 Add $headers_added variable, with content from use of ACL modifier
1598 add_header (but not yet added to the message). Bugzilla 199.
1600 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1601 Pulled from Bugzilla 817 by Wolfgang Breyha.
1603 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1605 (nb: this is the same fix as in Exim 4.80.1)
1607 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1610 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1613 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1615 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1616 advertise SMTP AUTH mechanism to us, instead of a generic
1617 protocol violation error. Also, make Exim more robust to bad
1618 data from the Dovecot auth socket.
1620 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1622 When a queue runner is handling a message, Exim first routes the
1623 recipient addresses, during which it prunes them based on the retry
1624 hints database. After that it attempts to deliver the message to
1625 any remaining recipients. It then updates the hints database using
1628 So if a recipient address works intermittently, it can get repeatedly
1629 deferred at routing time. The retry hints record remains fresh so the
1630 address never reaches the final cutoff time.
1632 This is a fairly common occurrence when a user is bumping up against
1633 their storage quota. Exim had some logic in its local delivery code
1634 to deal with this. However it did not apply to per-recipient defers
1635 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1637 This change adds a proper retry rule check during routing so that the
1638 final cutoff time is checked against the message's age. We only do
1639 this check if there is an address retry record and there is not a
1640 domain retry record; this implies that previous attempts to handle
1641 the address had the retry_use_local_parts option turned on. We use
1642 this as an approximation for the destination being like a local
1643 delivery, as in LMTP.
1645 I suspect this new check makes the old local delivery cutoff check
1646 redundant, but I have not verified this so I left the code in place.
1648 TF/02 Correct gecos expansion when From: is a prefix of the username.
1650 Test 0254 submits a message to Exim with the header
1654 When I ran the test suite under the user fanf2, Exim expanded
1655 the header to contain my full name, whereas it should have added
1656 a Resent-Sender: header. It erroneously treats any prefix of the
1657 username as equal to the username.
1659 This change corrects that bug.
1661 GF/01 DCC debug and logging tidyup
1662 Error conditions log to paniclog rather than rejectlog.
1663 Debug lines prefixed by "DCC: " to remove any ambiguity.
1665 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1667 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1668 Bug spotted by Jeremy Harris; was flawed since initial commit.
1669 Would have resulted in OCSP responses post-SNI triggering an Exim
1670 NULL dereference and crash.
1672 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1674 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1675 Bug detection, analysis and fix by Samuel Thibault.
1676 Bugzilla 1331, Debian bug #698092.
1678 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1680 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1681 Server implementation by Todd Lyons, client by JH.
1682 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1683 config variable "prdr_enable" controls whether the server
1684 advertises the facility. If the client requests PRDR a new
1685 acl_data_smtp_prdr ACL is called once for each recipient, after
1686 the body content is received and before the acl_smtp_data ACL.
1687 The client is controlled by both of: a hosts_try_prdr option
1688 on the smtp transport, and the server advertisement.
1689 Default client logging of deliveries and rejections involving
1690 PRDR are flagged with the string "PRDR".
1692 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1693 fclose(). Diagnosis by Todd Lyons.
1695 PP/17 Update configure.default to handle IPv6 localhost better.
1696 Patch by Alain Williams (plus minor tweaks).
1699 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1700 This is now consistent with GnuTLS, and is now documented: the
1701 previous undocumented portable approach to treating the option as
1702 unset was to force an expansion failure. That still works, and
1703 an empty string is now equivalent.
1705 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1706 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1707 not performing validation itself.
1709 PP/20 Added force_command boolean option to pipe transport.
1710 Patch from Nick Koston, of cPanel Inc.
1712 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1715 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1717 PP/21 Fix eximon continuous updating with timestamped log-files.
1718 Broken in a format-string cleanup in 4.80, missed when I repaired the
1719 other false fix of the same issue.
1720 Report and fix from Heiko Schlichting.
1723 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1724 Report from Prashanth Katuri.
1726 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1727 It's SecureTransport, so affects any MacOS clients which use the
1728 system-integrated TLS libraries, including email clients.
1730 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1731 using a MIME ACL for non-SMTP local injection.
1732 Report and assistance in diagnosis by Warren Baker.
1734 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1736 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1738 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1739 last id that failed so it may be referenced in subsequent ACL's.
1741 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1744 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1745 ldap library initialization, allowing self-signed CA's to be
1746 used. Also properly sets require_cert option later in code by
1747 using NULL (global ldap config) instead of ldap handle (per
1748 session). Bug diagnosis and testing by alxgomz.
1750 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1751 the src/util/ subdirectory.
1753 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1754 renamed to Transport Post Delivery Action by Jeremy Harris, as
1757 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1758 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1759 redis_servers = needs to be configured which will be used by the redis
1760 lookup. Patch from Warren Baker, of The Packet Hub.
1762 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1764 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1765 hostname or reverse DNS when processing a host list. Used suggestions
1766 from multiple comments on this bug.
1768 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1770 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1771 Missed a few lines, added it to make the runtest require no keyboard
1774 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1775 contains upper case chars. Make router use caseful_local_part.
1777 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1778 support when GnuTLS has been built with p11-kit.
1784 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1786 This, or similar/improved, will also be change PP/11 of 4.82.
1792 PP/01 Handle short writes when writing local log-files.
1793 In practice, only affects FreeBSD (8 onwards).
1794 Bugzilla 1053, with thanks to Dmitry Isaikin.
1796 NM/01 Bugzilla 949 - Documentation tweak
1798 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1801 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1803 PP/02 Implemented gsasl authenticator.
1805 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1807 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1808 `pkg-config foo` for cflags/libs.
1810 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1811 with rest of GSASL and with heimdal_gssapi.
1813 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1814 `pkg-config foo` for cflags/libs for the TLS implementation.
1816 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1817 properties get this fed in as external SSF. A number of robustness
1818 and debugging improvements to the cyrus_sasl authenticator.
1820 PP/08 cyrus_sasl server now expands the server_realm option.
1822 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1823 Patch by Jeremy Harris.
1825 PP/10 Added dbmjz lookup type.
1827 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1829 PP/12 MAIL args handles TAB as well as SP, for better interop with
1830 non-compliant senders.
1831 Analysis and variant patch by Todd Lyons.
1833 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1834 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1835 Patch from Dirk Mueller <dmueller@suse.com>
1837 PP/13 tls_peerdn now print-escaped for spool files.
1838 Observed some $tls_peerdn in wild which contained \n, which resulted
1839 in spool file corruption.
1841 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1842 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1843 or write after TLS renegotiation, which otherwise led to messages
1846 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1847 as a tracking header (ie: a signed header comes before the signature).
1848 Patch from Wolfgang Breyha.
1850 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1851 comma-sep list; embedded commas doubled.
1853 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1855 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1857 Report and patch from Dmitry Banschikov.
1859 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1860 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1861 failures. We appear to now support TLS1.1+ with Exim.
1863 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1864 lets Exim select keys and certificates based upon TLS SNI from client.
1865 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1866 before an outbound SMTP session. New log_selector, +tls_sni.
1868 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1869 NULL dereference. Report and patch from Alun Jones.
1871 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1872 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1873 Not seeing resolver debug output on NetBSD, but suspect this is a
1874 resolver implementation change.
1876 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1877 Left warnings. Added "eximon gdb" invocation mode.
1879 PP/21 Defaulting "accept_8bitmime" to true, not false.
1881 PP/22 Added -bw for inetd wait mode support.
1883 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1884 locate the relevant includes and libraries. Made this the default.
1886 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1887 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1889 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1890 This may cause build issues on older platforms.
1892 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1893 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1894 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1895 Added SNI support via GnuTLS too.
1896 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1898 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1900 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1901 Applied second patch from Janne, implementing suggestion to default
1902 multiple-strings-in-record handling to match SPF spec.
1904 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1906 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1907 read-only, out of scope).
1908 Patch from Wolfgang Breyha, report from Stuart Northfield.
1910 PP/29 Fix three issues highlighted by clang analyser static analysis.
1911 Only crash-plausible issue would require the Cambridge-specific
1912 iplookup router and a misconfiguration.
1913 Report from Marcin Mirosław.
1915 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1917 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1918 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1919 As part of this, removing so much warning spew let me fix some minor
1920 real issues in debug logging.
1922 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1923 assignment on my part. Fixed.
1925 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1926 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1927 Janne Snabb (who went above and beyond: thank you).
1929 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1930 string otherwise requires a connection and a bunch more work and it's
1931 relatively easy to get wrong. Should also expose TLS library linkage
1934 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1935 64-bit ${eval} (JH/03).
1937 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1938 GNU libc to support some of the 64-bit stuff, should not lead to
1939 conflicts. Defined before os.h is pulled in, so if a given platform
1940 needs to override this, it can.
1942 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1943 protection layer was required, which is not implemented.
1944 Bugzilla 1254, patch from Wolfgang Breyha.
1946 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1947 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1948 tls_dhparam take prime identifiers. Also unbreak combination of
1949 OpenSSL+DH_params+TLSSNI.
1951 PP/39 Disable SSLv2 by default in OpenSSL support.
1957 PP/01 Solaris build fix for Oracle's LDAP libraries.
1958 Bugzilla 1109, patch from Stephen Usher.
1960 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1962 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1965 TF/02 Fix a couple more cases where we did not log the error message
1966 when unlink() failed. See also change 4.74-TF/03.
1968 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1969 lock up or crash if it happened to be inside a call to libc when it
1970 got a SIGUSR1 from exiwhat.
1972 The SIGUSR1 handler appends the current process status to the process
1973 log which is later printed by exiwhat. It used to use the general
1974 purpose logging code to do this, but several functions it calls are
1975 not safe for signals.
1977 The new output code in the SIGUSR1 handler is specific to the process
1978 log, and simple enough that it's easy to inspect for signal safety.
1979 Removing some special cases also simplifies the general logging code.
1980 Removing the spurious timestamps from the process log simplifies
1983 TF/04 Improved ratelimit ACL condition.
1985 The /noupdate option has been deprecated in favour of /readonly which
1986 has clearer semantics. The /leaky, /strict, and /readonly update modes
1987 are mutually exclusive. The update mode is no longer included in the
1988 database key; it just determines when the database is updated. (This
1989 means that when you upgrade Exim will forget old rate measurements.)
1991 Exim now checks that the per_* options are used with an update mode that
1992 makes sense for the current ACL. For example, when Exim is processing a
1993 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1994 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1995 must specify per_mail/readonly. If you omit the update mode it defaults to
1996 /leaky where that makes sense (as before) or /readonly where required.
1998 The /noupdate option is now undocumented but still supported for
1999 backwards compatibility. It is equivalent to /readonly except that in
2000 ACLs where /readonly is required you may specify /leaky/noupdate or
2001 /strict/noupdate which are treated the same as /readonly.
2003 A useful new feature is the /count= option. This is a generalization
2004 of the per_byte option, so that you can measure the throughput of other
2005 aggregate values. For example, the per_byte option is now equivalent
2006 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2008 The per_rcpt option has been generalized using the /count= mechanism
2009 (though it's more complicated than the per_byte equivalence). When it is
2010 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2011 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2012 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2013 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2014 Note that using per_rcpt with a non-readonly update mode in more than
2015 one ACL will cause the recipients to be double-counted. (The per_mail
2016 and per_byte options don't have this problem.)
2018 The handling of very low rates has changed slightly. If the computed rate
2019 is less than the event's count (usually one) then this event is the first
2020 after a long gap. In this case the rate is set to the same as this event's
2021 count, so that the first message of a spam run is counted properly.
2023 The major new feature is a mechanism for counting the rate of unique
2024 events. The new per_addr option counts the number of different
2025 recipients that someone has sent messages to in the last time period. It
2026 behaves like per_rcpt if all the recipient addresses are different, but
2027 duplicate recipient addresses do not increase the measured rate. Like
2028 the /count= option this is a general mechanism, so the per_addr option
2029 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2030 example, measure the rate that a client uses different sender addresses
2031 with the options per_mail/unique=$sender_address. There are further
2032 details in the main documentation.
2034 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2036 TF/06 Removed a few PCRE remnants.
2038 TF/07 Automatically extract Exim's version number from tags in the git
2039 repository when doing development or release builds.
2041 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2042 Bugzilla 879. Patch from Paul Fisher.
2044 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2045 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2048 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2050 PP/05 Variable $av_failed, true if the AV scanner deferred.
2051 Bugzilla 1078. Patch from John Horne.
2053 PP/06 Stop make process more reliably on build failure.
2054 Bugzilla 1087. Patch from Heiko Schlittermann.
2056 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2057 Bugzilla 1089. Patch from Heiko Schlittermann.
2059 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2060 Bugzilla 1131. Patch from Holger Weiß.
2062 PP/09 Handle IPv6 addresses with SPF.
2063 Bugzilla 860. Patch from Wolfgang Breyha.
2065 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2067 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2070 PP/11 match_* no longer expand right-hand-side by default.
2071 New compile-time build option, EXPAND_LISTMATCH_RHS.
2072 New expansion conditions, "inlist", "inlisti".
2074 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2076 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2078 PP/14 fix log_write() format string regression from TF/03.
2079 Bugzilla 1152. Patch from Dmitry Isaikin.
2085 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2087 PP/02 Harmonised TLS library version reporting; only show if debugging.
2088 Layout now matches that introduced for other libraries in 4.74 PP/03.
2090 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2092 PP/04 New "dns_use_edns0" global option.
2094 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2097 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2098 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2100 TK/01 Updated PolarSSL code to 0.14.2.
2101 Bugzilla 1097. Patch from Andreas Metzler.
2103 PP/07 Catch divide-by-zero in ${eval:...}.
2104 Fixes bugzilla 1102.
2106 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2109 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2110 format-string attack -- SECURITY: remote arbitrary code execution.
2112 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2113 time unintentionally subject to list matching rules, letting the header
2114 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2115 arbitrary string expansion). This allowed for information disclosure.
2117 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2118 INT_MIN/-1 -- value coerced to INT_MAX.
2124 NM/01 Workaround for PCRE version dependency in version reporting
2127 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2128 This fixes portability to compilers other than gcc, notably
2129 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2131 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2132 makefiles for portability to HP-UX and POSIX correctness.
2134 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2135 Also via indented variable definition in the Makefile.
2136 (Debugging by Oliver Heesakkers).
2138 PP/02 Restore caching of spamd results with expanded spamd_address.
2139 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2141 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2142 Improves build reliability. Fix from: Frank Elsner
2144 NM/02 Fix wide character breakage in the rfc2047 coding
2145 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2147 NM/03 Allow underscore in dnslist lookups
2148 Fixes bug 1026. Patch from Graeme Fowler
2150 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2151 Code patches from Adam Ciarcinski of NetBSD.
2153 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2156 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2157 is logged, to avoid truncation. Patch from John Horne.
2159 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2160 Patch from Jakob Hirsch.
2162 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2163 SQL string expansion failure details.
2164 Patch from Andrey Oktyabrski.
2166 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2167 Patch from Simon Arlott.
2169 PP/09 New lookups functionality failed to compile on old gcc which rejects
2170 extern declarations in function scope.
2171 Patch from Oliver Fleischmann
2173 PP/10 Use sig_atomic_t for flags set from signal handlers.
2174 Check getgroups() return and improve debugging.
2175 Fixed developed for diagnosis in bug 927 (which turned out to be
2178 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2179 Patch from Mark Zealey.
2181 PP/12 Bugzilla 1056: Improved spamd server selection.
2182 Patch from Mark Zealey.
2184 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2185 Based on patch from Heiko Schlittermann.
2187 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2188 Patch from Uwe Doering, sign-off by Michael Haardt.
2190 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2191 variable declaration deep within a block. Bug and patch from
2194 PP/15 lookups-Makefile IRIX compatibility coercion.
2196 PP/16 Make DISABLE_DKIM build knob functional.
2198 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2199 Patch by Simon Arlott
2201 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2202 variable argument macros. Our copy now differs from upstream.
2208 TF/01 Failure to get a lock on a hints database can have serious
2209 consequences so log it to the panic log.
2211 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2212 controlled using the smtp_confirmation log selector.
2214 TF/03 Include the error message when we fail to unlink a spool file.
2216 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2217 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2218 for maintaining out-of-tree patches for some time.
2220 PP/01 Bugzilla 139: Documentation and portability issues.
2221 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2222 Handle per-OS dynamic-module compilation flags.
2224 PP/02 Let /dev/null have normal permissions.
2225 The 4.73 fixes were a little too stringent and complained about the
2226 permissions on /dev/null. Exempt it from some checks.
2227 Reported by Andreas M. Kirchwitz.
2229 PP/03 Report version information for many libraries, including
2230 Exim version information for dynamically loaded libraries. Created
2231 version.h, now support a version extension string for distributors
2232 who patch heavily. Dynamic module ABI change.
2234 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2235 privilege escalation vulnerability whereby the Exim run-time user
2236 can cause root to append content of the attacker's choosing to
2239 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2242 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2243 If dropping privileges for untrusted macros, we disabled normal logging
2244 on the basis that it would fail; for the Exim run-time user, this is not
2245 the case, and it resulted in successful deliveries going unlogged.
2246 Fixed. Reported by Andreas Metzler.
2252 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2253 only prepend for the Resent-* case. Fixes regression introduced in
2254 Exim 4.70 by NM/22 for Bugzilla 607.
2256 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2257 increasing numbers of administrators be bitten by this.
2259 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2261 PP/03 Bugzilla 994: added openssl_options main configuration option.
2263 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2265 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2267 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2269 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2270 without a peer certificate, leading to a segfault because of an
2271 assumption that peers always have certificates. Be a little more
2272 paranoid. Problem reported by Martin Tscholak.
2274 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2275 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2276 NB: ClamAV planning to remove STREAM in "middle of 2010".
2277 CL also introduces -bmalware, various -d+acl logging additions and
2278 more caution in buffer sizes.
2280 PP/09 Implemented reverse_ip expansion operator.
2282 PP/10 Bugzilla 937: provide a "debug" ACL control.
2284 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2286 PP/12 Bugzilla 973: Implement --version.
2288 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2290 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2292 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2294 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2295 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2296 ignore trailing whitespace.
2298 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2300 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2303 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2304 Notification from Dr Andrew Aitchison.
2306 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2307 ExtendedDetectionInfo response format.
2308 Notification from John Horne.
2310 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2313 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2314 XSL and documented dependency on system catalogs, with examples of how
2317 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2320 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2321 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2322 configuration file which is writeable by the Exim user or group.
2324 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2325 of configuration files to cover files specified with the -C option if
2326 they are going to be used with root privileges, not just the default
2329 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2330 option (effectively making it always true).
2332 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2333 files to be used while preserving root privileges.
2335 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2336 that rogue child processes cannot use them.
2338 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2339 run-time user, instead of root.
2341 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2342 Exim run-time user without dropping privileges.
2344 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2345 result string, instead of calling string_vformat() twice with the same
2348 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2349 for other users. Others should always drop root privileges if they use
2350 -C on the command line, even for a whitelisted configure file.
2352 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2354 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2360 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2361 $data_path, and $header_path variables; fixed documentation bugs and
2364 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2365 exipick to access non-standard spools, including the "frozen" queue
2368 NM/01 Bugzilla 965: Support mysql stored procedures.
2369 Patch from Alain Williams
2371 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2373 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2374 Patch from Andreas Metzler
2376 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2377 Patch from Kirill Miazine
2379 NM/05 Bugzilla 671: Added umask to procmail example.
2381 JJ/03 installed exipick 20100323.0, fixing doc bug
2383 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2384 directory. Notification and patch from Dan Rosenberg.
2386 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2388 TK/02 Improve log output when DKIM signing operation fails.
2390 MH/01 Treat the transport option dkim_domain as a colon separated
2391 list, not as a single string, and sign the message with each element,
2392 omitting multiple occurences of the same signer.
2394 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2395 Bugzilla 985, 986. Patch by Simon Arlott
2397 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2398 Patch by Simon Arlott
2400 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2401 MBX locking. Notification from Dan Rosenberg.
2407 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2409 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2411 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2413 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2415 NM/04 Bugzilla 915: Flag broken perl installation during build.
2421 TK/01 Added patch by Johannes Berg that expands the main option
2422 "spamd_address" if it starts with a dollar sign.
2424 TK/02 Write list of recipients to X-Envelope-Sender header when building
2425 the mbox-format spool file for content scanning (suggested by Jakob
2428 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2429 (http://www.dcc-servers.net/) support via dccifd. Activated by
2430 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2432 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2433 by Mark Daniel Reidel <mr@df.eu>.
2435 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2436 When building exim an external PCRE library is now needed -
2437 PCRE is a system library on the majority of modern systems.
2438 See entry on PCRE_LIBS in EDITME file.
2440 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2441 conversation. Added nologin parameter to request.
2442 Patch contributed by Kirill Miazine.
2444 TF/01 Do not log submission mode rewrites if they do not change the address.
2446 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2448 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2449 log files in place. Contributed by Roberto Lima.
2451 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2453 TF/03 Bugzilla 615: When checking the local_parts router precondition
2454 after a local_part_suffix or local_part_prefix option, Exim now
2455 does not use the address's named list lookup cache, since this
2456 contains cached lookups for the whole local part.
2458 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2459 Robert Millan. Documentation is in experimental-spec.txt.
2461 TF/04 Bugzilla 668: Fix parallel build (make -j).
2463 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2465 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2466 Patch provided by Jan Srzednicki.
2468 TF/05 Leading white space used to be stripped from $spam_report which
2469 wrecked the formatting. Now it is preserved.
2471 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2472 that they are available at delivery time.
2474 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2476 TF/08 TLS error reporting now respects the incoming_interface and
2477 incoming_port log selectors.
2479 TF/09 Produce a more useful error message if an SMTP transport's hosts
2480 setting expands to an empty string.
2482 NM/06 Bugzilla 744: EXPN did not work under TLS.
2483 Patch provided by Phil Pennock.
2485 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2486 Patch provided by Richard Godbee.
2488 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2489 acl_smtp_notquit, added index entry.
2491 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2492 Patch provided by Eugene Bujak.
2494 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2495 accept(). Patch provided by Maxim Dounin.
2497 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2498 Patch provided by Phil Pennock.
2500 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2502 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2503 Patch provided by Brad "anomie" Jorsch.
2505 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2506 Patch provided by Dean Brooks.
2508 TK/05 Add native DKIM support (does not depend on external libraries).
2510 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2511 Patch provided by Graeme Fowler.
2513 NM/16 Bugzilla 851: Documentation example syntax fix.
2515 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2517 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2520 NM/19 Bugzilla 745: TLS version reporting.
2521 Patch provided by Phil Pennock.
2523 NM/20 Bugzilla 167: bool: condition support.
2524 Patch provided by Phil Pennock.
2526 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2527 clients. Patch provided by Phil Pennock.
2529 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2530 Patch provided by Brad "anomie" Jorsch.
2532 NM/23 Bugzilla 687: Fix misparses in eximstats.
2533 Patch provided by Heiko Schlittermann.
2535 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2536 Patch provided by Heiko Schlittermann.
2538 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2539 plus update to original patch.
2541 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2543 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2544 Patch provided by David Brownlee.
2546 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2548 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2550 NM/30 Bugzilla 888: TLS documentation bugfixes.
2552 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2554 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2555 Unlike the original bugzilla I have changed all shell scripts in src tree.
2557 NM/33 Bugzilla 898: Transport filter timeout fix.
2558 Patch by Todd Rinaldo.
2560 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2561 Patch by Serge Demonchaux.
2563 NM/35 Bugzilla 39: Base64 decode bug fixes.
2564 Patch by Jakob Hirsch.
2566 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2568 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2570 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2572 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2578 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2579 ALT-N's libdkim that I have put here:
2580 http://duncanthrax.net/exim-experimental/
2582 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2583 were called 'true' and it seems that C99 defines that as a reserved
2584 keyword to be used with 'bool' variable types. That means you could
2585 not include C99-style headers which use bools without triggering
2586 build errors in sieve.c.
2588 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2589 as mailq or other aliases. Changed the --help handling significantly
2590 to do whats expected. exim_usage() emits usage/help information.
2592 SC/01 Added the -bylocaldomain option to eximstats.
2594 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2596 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2598 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2604 PH/01 Another patch from the Sieve maintainer.
2606 PH/02 When an IPv6 address is converted to a string for single-key lookup
2607 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2608 dots are used instead of colons so that keys in lsearch files need not
2609 contain colons. This was done some time before quoting was made available
2610 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2611 (notated using the quote facility) so as to distinguish them from IPv4
2612 keys. This meant that lookups for IP addresses in host lists did not work
2613 for iplsearch lookups.
2615 This has been fixed by arranging for IPv6 addresses to be expressed with
2616 colons if the lookup type is iplsearch. This is not incompatible, because
2617 previously such lookups could never work.
2619 The situation is now rather anomalous, since one *can* have colons in
2620 ordinary lsearch keys. However, making the change in all cases is
2621 incompatible and would probably break a number of configurations.
2623 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2626 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2627 conversion specification without a maximum field width, thereby enabling
2628 a rogue spamd server to cause a buffer overflow. While nobody in their
2629 right mind would setup Exim to query an untrusted spamd server, an
2630 attacker that gains access to a server running spamd could potentially
2631 exploit this vulnerability to run arbitrary code as the Exim user.
2633 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2634 $primary_hostname instead of what libspf2 thinks the hosts name is.
2636 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2637 a directory entry by the name of the lookup key. Previously, if a
2638 symlink pointed to a non-existing file or a file in a directory that
2639 Exim lacked permissions to read, a lookup for a key matching that
2640 symlink would fail. Now it is enough that a matching directory entry
2641 exists, symlink or not. (Bugzilla 503.)
2643 PH/03 The body_linecount and body_zerocount variables are now exported in the
2646 PH/04 Added the $dnslist_matched variable.
2648 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2649 This means they are set thereafter only if the connection becomes
2652 PH/06 Added the client_condition to authenticators so that some can be skipped
2653 by clients under certain conditions.
2655 PH/07 The error message for a badly-placed control=no_multiline_responses left
2656 "_responses" off the end of the name.
2658 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2660 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2661 (without spaces) instead of just copying the configuration text.
2663 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2665 PH/11 Added $max_received_linelength.
2667 PH/12 Added +ignore_defer and +include_defer to host lists.
2669 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2670 way in which PCRE > 7.0 is built.
2672 PH/14 Implemented queue_only_load_latch.
2674 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2675 MAIL command. The effect was to mangle the value on 64-bit systems.
2677 PH/16 Another patch from the Sieve maintainer.
2679 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2681 PH/18 If a system quota error occurred while trying to create the file for
2682 a maildir delivery, the message "Mailbox is full" was not appended to the
2683 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2684 only to a quota excession during the actual writing of the file.
2686 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2687 characters?) which causes problems in log lines. The DN values are now
2688 passed through string_printing() before being added to log lines.
2690 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2691 and InterBase are left for another time.)
2693 PH/21 Added message_body_newlines option.
2695 PH/22 Guard against possible overflow in moan_check_errorcopy().
2697 PH/23 POSIX allows open() to be a macro; guard against that.
2699 PH/24 If the recipient of an error message contained an @ in the local part
2700 (suitably quoted, of course), incorrect values were put in $domain and
2701 $local_part during the evaluation of errors_copy.
2707 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2708 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2711 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2712 issue a MAIL command.
2714 PH/02 In an ACL statement such as
2716 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2718 if a client was not listed at all, or was listed with a value other than
2719 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2720 the condition was not true (as it should be), so access was not denied.
2721 The bug was that the ! inversion was incorrectly passed on to the second
2722 item. This has been fixed.
2724 PH/03 Added additional dnslists conditions == and =& which are different from
2725 = and & when the dns lookup returns more than one IP address.
2727 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2728 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2730 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2731 FSYNC, which compiles an option called disable_fsync that allows for
2732 bypassing fsync(). The documentation is heavily laced with warnings.
2734 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2736 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2737 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2738 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2739 including adding "make clean"; (3) Added -fPIC when compiling the test
2740 dynamically loaded module, to get rid of a warning.
2742 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2743 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2744 = 0s. The bug is otherwise harmless.
2746 PH/07 There was a bug in the dovecot authenticator such that the value of
2747 $auth1 could be overwritten, and so not correctly preserved, after a
2748 successful authentication. This usually meant that the value preserved by
2749 the server_setid option was incorrect.
2751 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2753 PH/09 Installed PCRE release 7.0.
2755 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2756 run for batched SMTP input. It is now run at the start of every message
2757 in the batch. While fixing this I discovered that the process information
2758 (output by running exiwhat) was not always getting set for -bs and -bS
2759 input. This is fixed, and it now also says "batched" for BSMTP.
2761 PH/11 Added control=no_pipelining.
2763 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2764 patch, slightly modified), and move the expansion of helo_data till after
2765 the connection is made in the smtp transport (so it can use these
2768 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2770 PH/14 Added log_selector = +pid.
2772 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2774 PH/16 Add ${if forany and ${if forall.
2776 PH/17 Added dsn_from option to vary the From: line in DSNs.
2778 PH/18 Flush SMTP output before performing a callout, unless control =
2779 no_callout_flush is set.
2781 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2782 was true (the default) a successful delivery failed to delete the retry
2783 item, thus causing premature timeout of the address. The bug is now
2786 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2788 PH/21 Long custom messages for fakedefer and fakereject are now split up
2789 into multiline responses in the same way that messages for "deny" and
2790 other ACL rejections are.
2792 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2793 with slight modification.
2795 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2796 draft, changing the syntax and factoring some duplicate code".
2798 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2799 for deliveries of the second and subsequent messages over the same SMTP
2802 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2803 ${reduce, with only minor "tidies".
2805 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2807 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2808 expansion side effects.
2810 PH/27 When a message times out after an over-quota error from an Exim-imposed
2811 quota, the bounce message says "mailbox is full". This message was not
2812 being given when it was a system quota that was exceeded. It now should
2815 MH/03 Made $recipients available in local_scan(). local_scan() already has
2816 better access to the recipient list through recipients_list[], but
2817 $recipients can be useful in postmaster-provided expansion strings.
2819 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2820 in the case of a MAIL command with additional options following the
2821 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2822 were accidentally chopped off.
2824 PH/29 SMTP synchronization checks are implemented when a command is read -
2825 there is a check that no more input is waiting when there shouldn't be
2826 any. However, for some commands, a delay in an ACL can mean that it is
2827 some time before the response is written. In this time, more input might
2828 arrive, invalidly. So now there are extra checks after an ACL has run for
2829 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2830 pipelining has not been advertised.
2832 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2834 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2835 returned in the SMTP response when smtp_return_error_details was false.
2836 This has been fixed.
2838 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2839 instead of the C I/O that was originally supplied, because problems were
2840 reported on Solaris.
2842 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2843 Exim which did not show up earlier: it was assuming that a call to
2844 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2845 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2846 was a macro that became an assignment, so it seemed to work. This has
2847 changed to a proper function call with a void return, hence the compile
2848 error. Exim's code has been fixed.
2850 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2853 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2855 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2857 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2858 because it turns out to be too big or there is a timeout) there is no
2859 "Completed" line in the log. When some messages of this type were
2860 selected by exigrep, they were listed as "not completed". Others were
2861 picked up by some special patterns. I have improved the selection
2862 criteria to be more general.
2864 PH/38 The host_find_failed option in the manualroute router can now be set
2865 to "ignore", to completely ignore a host whose IP address cannot be
2866 found. If all hosts are ignored, the behaviour is controlled by the new
2867 host_all_ignored option.
2869 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2870 homing or because of multiple MX records with /mx) generated more than
2871 one IP address, and the following item turned out to be the local host,
2872 all the secondary addresses of the first item were incorrectly removed
2873 from the list, along with the local host and any following hosts (which
2874 is what is supposed to happen).
2876 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2877 whoever called Exim into the -H file. In the case of the daemon it was
2878 behaving confusingly. When first started, it used values for whoever
2879 started the daemon, but after a SIGHUP it used the Exim user (because it
2880 calls itself on a restart). I have changed the code so that it now always
2883 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2884 message are rejected with the same error (e.g. no authentication or bad
2885 sender address), and a DATA command is nevertheless sent (as can happen
2886 with PIPELINING or a stupid MUA), the error message that was given to the
2887 RCPT commands is included in the rejection of the DATA command. This is
2888 intended to be helpful for MUAs that show only the final error to their
2891 PH/42 Another patch from the Sieve maintainer.
2893 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2894 Eximstats - Fixed some broken HTML links and added missing column headers
2896 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2897 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2899 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2901 PH/43 Yet another patch from the Sieve maintainer.
2903 PH/44 I found a way to check for a TCP/IP connection going away before sending
2904 the response to the final '.' that terminates a message, but only in the
2905 case where the client has not sent further data following the '.'
2906 (unfortunately, this is allowed). However, in many cases there won't be
2907 any further data because there won't be any more messages to send. A call
2908 to select() can be used: if it shows that the input is "ready", there is
2909 either input waiting, or the socket has been closed. An attempt to read
2910 the next input character can distinguish the two cases. Previously, Exim
2911 would have sent an OK response which the client would never have see.
2912 This could lead to message repetition. This fix should cure that, at
2913 least in a lot of common cases.
2915 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2916 advertised in response to EHLO.
2922 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2923 fixed by 4.65/MH/01 (is this a record?) are fixed:
2925 (i) An empty string was always treated as zero by the numeric comparison
2926 operators. This behaviour has been restored.
2928 (ii) It is documented that the numeric comparison operators always treat
2929 their arguments as decimal numbers. This was broken in that numbers
2930 starting with 0 were being interpreted as octal.
2932 While fixing these problems I realized that there was another issue that
2933 hadn't been noticed. Values of message_size_limit (both the global option
2934 and the transport option) were treated as octal if they started with 0.
2935 The documentation was vague. These values are now always treated as
2936 decimal, and I will make that clear in the documentation.
2942 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2943 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2946 MH/01 Don't check that the operands of numeric comparison operators are
2947 integers when their expansion is in "skipping" mode (fixes bug
2948 introduced by 4.64-PH/07).
2950 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2951 child addresses, Exim now panics and dies. Previously, because the count
2952 is held in a short int, deliveries were likely to be lost. As such a
2953 large number of recipients for a single message is ridiculous
2954 (performance will be very, very poor), I have chosen to impose a limit
2955 rather than extend the field.
2961 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2962 leftover -K file (the existence of which was triggered by #402).
2963 While we were at it, introduced process PID as part of the -K
2964 filename. This should rule out race conditions when creating
2967 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2968 processing considerably. Previous code took too long for large mails,
2969 triggering a timeout which in turn triggers #401.
2971 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2972 in the DK code in transports.c. sendfile() is not really portable,
2973 hence the _LINUX specificness.
2975 TF/01 In the add_headers option to the mail command in an Exim filter,
2976 there was a bug that Exim would claim a syntax error in any
2977 header after the first one which had an odd number of characters
2980 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2981 callout verification, Exim cached a "reject" for the entire domain. This
2982 is correct for most verifications, but it is not correct for a recipient
2983 verification with use_sender or use_postmaster set, because in that case
2984 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2985 case of MAIL FROM:<> rejection from other early rejections (e.g.
2986 rejection of HELO). When verifying a recipient using a non-null MAIL
2987 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2988 Whatever the result of the callout, the value of the domain cache is
2989 left unchanged (for any other kind of callout, getting as far as trying
2990 RCPT means that the domain itself is ok).
2992 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2995 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2996 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2997 session, the daemon ignores ECONNECT errors and logs others; it now
2998 ignores EPIPE as well.
3000 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3001 (quoted-printable decoding).
3003 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3004 later the small subsequent patch to fix an introduced bug.
3006 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3008 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3010 PH/08 An error is now given if message_size_limit is specified negative.
3012 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3013 to be given (somewhat) arbitrary names.
3015 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3018 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3019 miscellaneous code fixes
3021 PH/10 Added the log_reject_target ACL modifier to specify where to log
3024 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3025 hostname. This is wrong, because it relates to the incoming message (and
3026 probably the interface on which it is arriving) and not to the outgoing
3027 callout (which could be using a different interface). This has been
3028 changed to use the value of the helo_data option from the smtp transport
3029 instead - this is what is used when a message is actually being sent. If
3030 there is no remote transport (possible with a router that sets up host
3031 addresses), $smtp_active_hostname is used.
3033 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3034 tweaks were necessary in order to get it to work (see also 21 below):
3035 (a) The code assumed that strncpy() returns a negative number on buffer
3036 overflow, which isn't the case. Replaced with Exim's string_format()
3038 (b) There were several signed/unsigned issues. I just did the minimum
3039 hacking in of casts. There is scope for a larger refactoring.
3040 (c) The code used strcasecmp() which is not a standard C function.
3041 Replaced with Exim's strcmpic() function.
3042 (d) The code set only $1; it now sets $auth1 as well.
3043 (e) A simple test gave the error "authentication client didn't specify
3044 service in request". It would seem that Dovecot has changed its
3045 interface. Fortunately there's a specification; I followed it and
3046 changed what the client sends and it appears to be working now.
3048 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3051 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3052 address A is aliased to B and C, where B exists and C does not. Without
3053 -v the output is "A verified" because verification stops after a
3054 successful redirection if more than one address is generated. However,
3055 with -v the child addresses are also verified. Exim was outputting "A
3056 failed to verify" and then showing the successful verification for C,
3057 with its parentage. It now outputs "B failed to verify", showing B's
3058 parentage before showing the successful verification of C.
3060 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3061 look up a TXT record in a specific list after matching in a combined
3064 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3065 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3066 they consult the DNS. I had assumed they would set it the way they
3067 wanted; and indeed my experiments on Linux seem to show that in some
3068 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3069 To be on the safe side, however, I have now made the interface to
3070 host_find_byname() similar to host_find_bydns(), with an argument
3071 containing the DNS resolver options. The host_find_byname() function now
3072 sets these options at its start, just as host_find_bydns() does. The smtp
3073 transport options dns_qualify_single and dns_search_parents are passed to
3074 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3075 of host_find_byname() use the default settings of RES_DEFNAMES
3076 (qualify_single) but not RES_DNSRCH (search_parents).
3078 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3079 spool_read_header() do less string testing, by means of a preliminary
3080 switch on the second character of optional "-foo" lines. (This is
3081 overdue, caused by the large number of possibilities that now exist.
3082 Originally there were few.) While I was there, I also converted the
3083 str(n)cmp tests so they don't re-test the leading "-" and the first
3084 character, in the hope this might squeeze out yet more improvement.
3086 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3087 flag allowing group syntax was set by the header_syntax check but not
3088 turned off, possible causing trouble later; (2) The flag was not being
3089 set at all for the header_verify test, causing "group"-style headers to
3090 be rejected. I have now set it in this case, and also caused header_
3091 verify to ignore an empty address taken from a group. While doing this, I
3092 came across some other cases where the code for allowing group syntax
3093 while scanning a header line wasn't quite right (mostly, not resetting
3094 the flag correctly in the right place). These bugs could have caused
3095 trouble for malformed header lines. I hope it is now all correct.
3097 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3098 with the "reply" argument non-NULL. The code, however (which originally
3099 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3100 but it didn't always do it. This confused somebody who was copying the
3101 code for some other use. I have removed all the tests.
3103 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3104 feature that was used to support insecure browsers during the U.S. crypto
3105 embargo. It requires special client support, and Exim is probably the
3106 only MTA that supported it -- and would never use it because real RSA is
3107 always available. This code has been removed, because it had the bad
3108 effect of slowing Exim down by computing (never used) parameters for the
3109 RSA_EXPORT functionality.
3111 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3112 authenticator to fail if there's a tab character in the incoming data
3113 (there should never be unless someone is messing about, as it's supposed
3114 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3115 if the connection is using TLS or if the remote IP is the same as the
3116 local IP, and the "valid-client-cert option" if a client certificate has
3119 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3120 authenticators. This can be used for authorization after authentication
3121 succeeds. (In the case of plaintext, it servers for both authentication
3124 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3125 if any retry times were supplied.
3127 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3128 connection, where there is no client IP address to check. In this
3129 situation, the verify now always succeeds.
3131 PH/25 Applied John Jetmore's -Mset patch.
3133 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3135 PH/27 In a string expansion for a processed (not raw) header when multiple
3136 headers of the same name were present, leading whitespace was being
3137 removed from all of them, but trailing whitespace was being removed only
3138 from the last one. Now trailing whitespace is removed from each header
3139 before concatenation. Completely empty headers in a concatenation (as
3140 before) are ignored.
3142 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3143 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3145 PH/29 [Removed. This was a change that I later backed out, and forgot to
3146 correct the ChangeLog entry (that I had efficiently created) before
3147 committing the later change.]
3149 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3150 address errors (4xx response to RCPT) over the same connection as other
3151 messages routed to the same hosts. Such deliveries are always "forced",
3152 so retry times are not inspected. This resulted in far too many retries
3153 for the affected addresses. The effect occurred only when there were more
3154 hosts than the hosts_max_try setting in the smtp transport when it had
3155 the 4xx errors. Those hosts that it had tried were not added to the list
3156 of hosts for which the message was waiting, so if all were tried, there
3157 was no problem. Two fixes have been applied:
3159 (i) If there are any address or message errors in an SMTP delivery, none
3160 of the hosts (tried or untried) are now added to the list of hosts
3161 for which the message is waiting, so the message should not be a
3162 candidate for sending over the same connection that was used for a
3163 successful delivery of some other message. This seems entirely
3164 reasonable: after all the message is NOT "waiting for some host".
3165 This is so "obvious" that I'm not sure why it wasn't done
3166 previously. Hope I haven't missed anything, but it can't do any
3167 harm, as the worst effect is to miss an optimization.
3169 (ii) If, despite (i), such a delivery is accidentally attempted, the
3170 routing retry time is respected, so at least it doesn't keep
3171 hammering the server.
3173 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3174 in ${readsocket because some servers need this prod.
3176 PH/32 Added some extra debug output when updating a wait-xxx database.
3178 PH/33 The hint "could be header name not terminated by colon", which has been
3179 given for certain expansion errors for a long time, was not being given
3180 for the ${if def:h_colon_omitted{... case.
3182 PH/34 The spec says: "With one important exception, whenever a domain list is
3183 being scanned, $domain contains the subject domain." There was at least
3184 one case where this was not true.
3186 PH/35 The error "getsockname() failed: connection reset by peer" was being
3187 written to the panic log as well as the main log, but it isn't really
3188 panic-worthy as it just means the connection died rather early on. I have
3189 removed the panic log writing for the ECONNRESET error when getsockname()
3192 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3193 runs only) independently of the message's sender address. This meant
3194 that, if the 4xx error was in fact related to the sender, a different
3195 message to the same recipient with a different sender could confuse
3196 things. In particular, this can happen when sending to a greylisting
3197 server, but other circumstances could also provoke similar problems.
3198 I have changed the default so that the retry time for these errors is now
3199 based a combination of the sender and recipient addresses. This change
3200 can be overridden by setting address_retry_include_sender=false in the
3203 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3204 remote server are returned as part of bounce messages. This was not
3205 happening for LMTP over a pipe (the lmtp transport), but now it is the
3206 same for both kinds of LMTP.
3208 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3209 in header lines that were in fact CNAMEs. This is no longer the case.
3211 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3212 and queue runs started by the daemon processed all messages. This has
3213 been fixed so that -R and -S can now usefully be given with -q<time>.
3215 PH/40 Import PCRE release 6.7 (fixes some bugs).
3217 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3219 PH/42 Give an error if -q is specified more than once.
3221 PH/43 Renamed the variables $interface_address and $interface_port as
3222 $received_ip_address and $received_port, to make it clear that these
3223 values apply to message reception, and not to the outgoing interface when
3224 a message is delivered. (The old names remain recognized, of course.)
3226 PH/44 There was no timeout on the connect() call when using a Unix domain
3227 socket in the ${readsocket expansion. There now is.
3229 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3230 be meaningful with "accept".
3232 SC/01 Eximstats V1.43
3233 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3235 SC/02 Eximstats V1.44
3236 Use a glob alias rather than an array ref in the generated
3237 parser. This improves both readability and performance.
3239 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3240 Collect SpamAssassin and rejection statistics.
3241 Don't display local sender or destination tables unless
3242 there is data to show.
3243 Added average volumes into the top table text output.
3245 SC/04 Eximstats V1.46
3246 Collect data on the number of addresses (recipients)
3247 as well as the number of messages.
3249 SC/05 Eximstats V1.47
3250 Added 'Message too big' to the list of mail rejection
3251 reasons (thanks to Marco Gaiarin).
3253 SC/06 Eximstats V1.48
3254 Mainlog lines which have GMT offsets and are too short to
3255 have a flag are now skipped.
3257 SC/07 Eximstats V1.49 (Alain Williams)
3258 Added the -emptyok flag.
3260 SC/08 Eximstats V1.50
3261 Fixes for obtaining the IP address from reject messages.
3263 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3264 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3265 whitespace changes from 4.64-PH/27
3267 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3270 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3271 are found, allow negative numbers in numeric criteria)
3273 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3275 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3276 to match changes made in 4.64-PH/43
3278 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3280 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3281 "domain" is actually an IP address, and force a failure. This locks out
3282 those revolvers/nameservers that support "A-for-A" lookups, in
3283 contravention of the specifications.
3285 PH/48 When a host name was looked up from an IP address, and the subsequent
3286 forward lookup of the name timed out, the host name was left in
3287 $sender_host_name, contrary to the specification.
3289 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3290 restricted to single-key lookups, Exim was not diagnosing an error if
3291 * or *@ was used with a query-style lookup.
3293 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3295 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3296 long ago, but noone interested enough thought of it. Let's just say that
3297 the "1.1" means that there are some new functions that weren't there at
3298 some point in the past.
3300 PH/51 Error processing for expansion failure of helo_data from an smtp
3301 transport during callout processing was broken.
3303 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3304 tested/used via the -bh/-bhc/-bs options.
3306 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3307 bug, fixed in subsequent PCRE releases).
3309 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3310 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3312 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3318 SC/01 Use a glob alias rather than an array ref in eximstats generated
3319 parser. This improves both readability and performance.
3321 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3322 Don't display local sender or destination tables in eximstats unless
3323 there is data to show.
3324 Added average volumes into the eximstats top table text output.
3326 SC/03 Collect data on the number of addresses (recipients) as well
3327 as the number of messages in eximstats.
3329 TF/01 Correct an error in the documentation for the redirect router. Exim
3330 does (usually) call initgroups() when daemonizing.
3332 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3333 with consistent privilege compared to when running as a daemon.
3335 TF/03 Note in the spec that $authenticated_id is not set for local
3336 submissions from trusted users.
3338 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3339 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3341 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3342 by adding some example configuration directives to the default
3343 configuration file. A little bit of work is required to uncomment the
3344 directives and define how usernames and passwords are checked, but
3345 there is now a framework to start from.
3347 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3348 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3349 without this. I don't know how relevant this is to other LDAP libraries.
3351 PH/02 Add the verb name to the "unknown ACL verb" error.
3353 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3355 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3357 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3358 directory not expanded when it should be if an expanded home directory
3359 was set for the address (which is overridden by the transport).
3361 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3364 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3365 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3366 because it is too late at that time, and has no effect.
3368 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3369 security issue with \' (bugzilla #107). I could not use the
3370 PQescapeStringConn() function, because it needs a PGconn value as one of
3373 PH/08 When testing addresses using -bt, indicate those final addresses that
3374 are duplicates that would not cause an additional delivery. At least one
3375 person was confused, thinking that -bt output corresponded to deliveries.
3376 (Suppressing duplicates isn't a good idea as you lose the information
3377 about possibly different redirections that led to the duplicates.)
3379 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3380 systems where poll() doesn't work, in particular OS X.
3382 PH/10 Added more information to debugging output for retry time not reached.
3384 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3385 operations in malware.c.
3387 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3390 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3391 syslog_duplication set false, log lines that would normally be written
3392 both the the main log and to the reject log were not written to syslog at
3395 PH/14 In the default configuration, change the use of "message" in ACL warn
3396 statements to "add_header".
3398 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3399 not followed by a command (e.g. "seen endif").
3401 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3402 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3405 PH/17 Added extra conditions to the default value of delay_warning_condition
3409 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3410 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3411 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3414 The Auto-Submitted: and various List- headers are standardised, whereas I
3415 don't think Precedence: ever was.
3417 PH/18 Refactored debugging code in route_finduser() to show more information,
3418 in particular, the error code if getpwnam() issues one.
3420 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3421 This is apparently needed in addition to the PH/07 change above to avoid
3422 any possible encoding problems.
3424 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3425 but not after initializing Perl.
3427 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3428 output them only if debugging. By default they are written stderr,
3429 apparently, which is not desirable.
3431 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3434 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3437 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3439 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3440 authenticated or an ident call has been made. Suppress the default
3441 values for $authenticated_id and $authenticated_sender (but permit -oMai
3442 and -oMas) when testing with -bh.
3444 PH/24 Re-jigged the order of the tests in the default configuration so that the
3445 tests for valid domains and recipients precede the DNS black list and CSA
3446 tests, on the grounds that those ones are more expensive.
3448 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3449 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3450 This bug exists in every version of Exim that I still have, right back to
3453 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3454 However, an attempt to turn on case-sensitivity in a regex key by
3455 including (?-i) didn't work because the subject string was already
3456 lowercased, and the effects were non-intuitive. It turns out that a
3457 one-line patch can be used to allow (?-i) to work as expected.
3463 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3464 other effects) broke the use of negated acl sub-conditions.
3466 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3469 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3470 "Deny" causes Exim to reject the incoming connection with a 554 error.
3471 Unfortunately, if there is a major crisis, such as a disk failure,
3472 tcp-wrappers gives "deny", whereas what one would like would be some
3473 kind of temporary error. A kludge has been added to help with this.
3474 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3475 554 error is used if errno is still zero or contains ENOENT (which occurs
3476 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3479 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3481 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3482 errors. Otherwise a message that provokes a temporary error (when other
3483 messages do not) can cause a whole host to time out.
3485 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3486 addresses were routed directly to files or pipes from a redirect router.
3487 File deliveries just didn't batch; pipe deliveries might have suffered
3490 PH/06 A failure to get a lock for a hints database would erroneously always say
3491 "Failed to get write lock", even when it was really a read lock.
3493 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3494 of 0600. This has been changed to use the value of the lockfile_mode
3495 option (which defaults to 0600).
3497 PH/08 Applied small patch from the Sieve maintainer.
3499 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3500 folder from quota calculations, a direct delivery into this folder messed
3501 up the contents of the maildirsize file. This was because the regex was
3502 used only to exclude .Trash (or whatever) when the size of the mailbox
3503 was calculated. There was no check that a delivery was happening into an
3504 excluded directory. This bug has been fixed by ignoring all quota
3505 processing for deliveries into excluded directories.
3507 PH/10 Added the maildirfolder_create_regex option to appendfile.
3513 PH/01 The code for finding all the local interface addresses on a FreeBSD
3514 system running IPv6 was broken. This may well have applied to all BSD
3515 systems, as well as to others that have similar system calls. The broken
3516 code found IPv4 interfaces correctly, but gave incorrect values for the
3517 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3518 that it would not match correctly against @[] and not recognize the IPv6
3521 PH/02 The ipliteral router was not recognizing addresses of the form user@
3522 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3524 PH/03 Added disable_ipv6.
3526 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3527 decoded form, because it is most often used to construct To: headers
3528 lines in autoreplies, and the decoded form may well be syntactically
3529 invalid. However, $reply_address has leading white space removed, and all
3530 newlines turned into spaces so that the autoreply transport does not
3533 PH/05 If group was specified without a user on a router, and no group or user
3534 was specified on a transport, the group from the router was ignored.
3536 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3537 for visible compile-time settings that can be used to change these
3538 numbers, for those that want even more. Backwards compatibility with old
3539 spool files has been maintained. However, going back to a previous Exim
3540 release will lost any variables that are in spool files.
3542 PH/07 Two small changes when running in the test harness: increase delay when
3543 passing a TCP/IP connection to a new process, in case the original
3544 process has to generate a bounce, and remove special handling of
3545 127.0.0.2 (sic), which is no longer necessary.
3547 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3548 be the same on different OS.
3550 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3553 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3554 whether --show-vars was specified or not
3556 JJ/02 exipick: Added support for new ACL variable spool format introduced
3559 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3560 syntactically invalid From: or Reply-to: line, and a filter used this to
3561 generate an autoreply, and therefore failed to obtain an address for the
3562 autoreply, Exim could try to deliver to a non-existent relative file
3563 name, causing unrelated and misleading errors. What now happens is that
3564 it logs this as a hard delivery error, but does not attempt to create a
3567 PH/11 The exinext utility has a -C option for testing purposes, but although
3568 the given file was scanned by exinext itself; it wasn't being passed on
3569 when Exim was called.
3571 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3572 an end-of-file indication when reading a command response.
3574 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3575 compiled. In many other places in Exim, IPv6 addresses are always
3576 recognized, so I have changed this. It also means that IPv4 domain
3577 literals of the form [IPV4:n.n.n.n] are now always recognized.
3579 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3580 used if the router is not running as root, for example, when verifying at
3581 ACL time, or when using -bh. The debugging output from this situation was
3582 non-existent - all you got was a failure to exec. I have made two
3585 (a) Failures to set uid/gid, the current directory, or a process leader
3586 in a subprocess such as that created by queryprogram now generate
3587 suitable debugging output when -d is set.
3589 (b) The queryprogram router detects when it is not running as root,
3590 outputs suitable debugging information if -d is set, and then runs
3591 the subprocess without attempting to change uid/gid.
3593 PH/15 Minor change to Makefile for building test_host (undocumented testing
3596 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3597 additional section of a DNS packet that returns MX or SRV records.
3598 Instead, it always explicitly searches for A/AAAA records. This avoids
3599 major problems that occur when a DNS server includes only records of one
3600 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3601 fixed another bug: if SRV records were looked up and the corresponding
3602 address records were *not* found in the additional section, the port
3603 values from the SRV records were lost.
3605 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3606 using the correct key (the original address) when searching the retry
3607 rules in order to find which one to use for generating the retry hint.
3609 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3610 adding the default one. Similarly, if it contains a Reply-To: header, the
3611 errors_reply_to option, if set, is not used.
3613 PH/19 When calculating a retry time, Exim used to measure the "time since
3614 failure" by looking at the "first failed" field in the retry record. Now
3615 it does not use this if it is later than than the arrival time of the
3616 message. Instead it uses the arrival time. This makes for better
3617 behaviour in cases where some deliveries succeed, thus re-setting the
3618 "first failed" field. An example is a quota failure for a huge message
3619 when small messages continue to be delivered. Without this change, the
3620 "time since failure" will always be short, possible causing more frequent
3621 delivery attempts for the huge message than are intended.
3622 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3624 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3625 $1, $2, $3) because the numerical variables can be reset during some
3626 expansion items (e.g. "match"), thereby losing the authentication data.
3628 PH/21 Make -bV show the size of off_t variables so that the test suite can
3629 decide whether to run tests for quotas > 2G.
3631 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3632 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3633 filecount value is greater than 2G or if a quota value is greater than 2G
3634 on a system where the size of off_t is not greater than 4, a panic error
3637 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3638 never match. The debug and -bh output now contains an explicit error
3639 message indicating a malformed IPv4 address or mask.
3641 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3642 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3643 PH/23 above applies.
3645 PH/25 Do not write to syslog when running in the test harness. The only
3646 occasion when this arises is a failure to open the main or panic logs
3647 (for which there is an explicit test).
3649 PH/26 Added the /no_tell option to "control=freeze".
3651 PH/27 If a host name lookup failed very early in a connection, for example, if
3652 the IP address matched host_lookup and the reverse lookup yielded a name
3653 that did not have a forward lookup, an error message of the form "no IP
3654 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3655 could be logged. Now it outputs the IP address instead of "NULL".
3657 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3658 allows the sender and the authenticated sender to be set when
3659 submitting a message from within Exim. Since child_open_exim() is
3660 documented for local_scan(), the new function should be too.
3662 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3663 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3664 results in an empty string is now treated as unset.
3666 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3668 PH/31 Added sender_verify_fail logging option.
3670 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3671 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3672 tidied the source and removed it altogether.
3674 PH/33 When a queue run was abandoned because the load average was too high, a
3675 log line was always written; now it is written only if the queue_run log
3676 selector is set. In addition, the log line for abandonment now contains
3677 information about the queue run such as the pid. This is always present
3678 in "start" and "stop" lines but was omitted from the "abandon" line.
3680 PH/34 Omit spaces between a header name and the colon in the error message that
3681 is given when verify = headers_syntax fails (if there are lots of them,
3682 the message gets confusing).
3684 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3685 names, as there are now some PTR records that contain slashes. This check
3686 is only to protect against broken name servers that fall over on strange
3687 characters, so the fact that it applies to all lookups doesn't matter.
3689 PH/36 Now that the new test suite is complete, we can remove some of the
3690 special code in Exim that was needed for the old test suite. For example,
3691 sorting DNS records because real resolvers return them in an arbitrary
3692 order. The new test suite's fake resolver always returns records in the
3695 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3696 bounces) except when queue_only is set, to avoid logging races between
3697 the different processes.
3699 PH/38 Panic-die if .include specifies a non-absolute path.
3701 PH/39 A tweak to the "H" retry rule from its user.
3703 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3704 a label. They prevented compilation on older perls.
3706 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3707 a warning to be raised on newish perls.
3709 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3710 on queue. Changes to match documented behaviour of showing count of
3711 messages matching specified criteria.
3713 PH/40 Changed the default ident timeout from 30s to 5s.
3715 PH/41 Added support for the use of login_cap features, on those BSD systems
3716 that have them, for controlling the resources used by pipe deliveries.
3718 PH/42 The content-scanning code uses fopen() to create files in which to put
3719 message data. Previously it was not paying any attention to the mode of
3720 the files. Exim runs with umask(0) because the rest of the code creates
3721 files with open(), and sets the required mode explicitly. Thus, these
3722 files were ending up world-writeable. This was not a big issue, because,
3723 being within the spool directory, they were not world-accessible. I have
3724 created a function called modefopen, which takes an additional mode
3725 argument. It sets umask(777), creates the file, chmods it to the required
3726 mode, then resets the umask. All the relevant calls to fopen() in the
3727 content scanning code have been changed to use this function.
3729 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3730 to 24 hours. This avoids potential overflow problems when processing G
3731 and H retry rules. I suspect nobody ever tinkers with this value.
3733 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3735 PH/45 When the plaintext authenticator is running as a client, the server's
3736 challenges are checked to ensure they are valid base64 strings. By
3737 default, the authentication attempt is cancelled if an invalid string is
3738 received. Setting client_ignore_invalid_base64 true ignores these errors.
3739 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3740 they are received. Thus, the responses can be made to depend on the
3741 challenges. If an invalid string is ignored, an empty string is placed in
3744 PH/46 Messages that are created by the autoreply transport now contains a
3745 References: header, in accordance with RFCs 2822 and 3834.
3747 PH/47 Added authenticated_sender_force to the smtp transport.
3749 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3751 PH/49 Installed latest patch from the Sieve maintainer.
3753 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3754 was also set, the appendfile transport was unnecessarily scanning a
3755 directory of message files (e.g. for maildir delivery) to find the count
3756 of files (along with the size), even though it did not need this
3757 information. It now does the scan only if it needs to find either the
3758 size of the count of files.
3760 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3762 PH/52 Two bugs concerned with error handling when the smtp transport is
3765 (i) Exim was not creating retry information for temporary errors given
3766 for individual recipients after the DATA command when the smtp transport
3767 was used in LMTP mode. This meant that they could be retried too
3768 frequently, and not timed out correctly.
3770 (ii) Exim was setting the flag that allows error details to be returned
3771 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3772 recipients that were returned after the DATA command.
3774 PH/53 This is related to PH/52, but is more general: for any failing address,
3775 when detailed error information was permitted to be returned to the
3776 sender, but the error was temporary, then after the final timeout, only
3777 "retry timeout exceeded" was returned. Now it returns the full error as
3778 well as "retry timeout exceeded".
3780 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3781 do this, and (what is worse) MTAs that accept it.
3783 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3784 will now be deprecated.
3786 PH/56 New os.c-cygwin from the Cygwin maintainer.
3788 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3789 formats (previously only available in exim formats via -bpr, -bpru,
3790 and -bpra. Now also available in native and exiqgrep formats)
3792 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3793 with very large, slow to parse queues
3795 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3797 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3799 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3800 responses to authentication challenges, though it was showing the
3801 challenges; (ii) I've removed the CR characters from the debug output for
3804 PH/58 Allow for the insertion of a newline as well as a space when a string
3805 is turned into more than one encoded-word during RFC 2047 encoding. The
3806 Sieve code now uses this.
3808 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3809 data_4xx, lost_connection, tls_required.
3811 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3812 message was being sent as an SMTP response.
3814 PH/61 Add -l and -k options to exicyclog.
3816 PH/62 When verifying, if an address was redirected to one new address, so that
3817 verification continued, and the new address failed or deferred after
3818 having set something in $address_data, the value of $address_data was not
3819 passed back to the ACL. This was different to the case when no
3820 redirection occurred. The value is now passed back in both cases.
3822 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3823 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3824 use login_cap.h, so on its own it isn't the distinguishing feature. The
3825 new name refers directly to the setclassresources() function.
3827 PH/65 Added configuration files for NetBSD3.
3829 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3831 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3832 is preferred over IPv4.
3834 PH/68 The bounce_return_message and bounce_return_body options were not being
3835 honoured for bounces generated during the reception of non-SMTP messages.
3836 In particular, this applied to messages rejected by the ACL. This bug has
3837 been fixed. However, if bounce_return_message is true and bounce_return_
3838 body is false, the headers that are returned for a non-SMTP message
3839 include only those that have been read before the error was detected.
3840 (In the case of an ACL rejection, they have all been read.)
3842 PH/69 The HTML version of the specification is now built in a directory called
3843 spec_html instead of spec.html, because the latter looks like a path with
3844 a MIME-type, and this confuses some software.
3846 PH/70 Catch two compiler warnings in sieve.c.
3848 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3849 function verify_get_ident() calls ip_connect() to connect a socket, but
3850 if the "connect()" function timed out, ip_connect() used to close the
3851 socket. However, verify_get_ident() also closes the socket later, and in
3852 between Exim writes to the log, which may get opened at this point. When
3853 the socket was closed in ip_connect(), the log could get the same file
3854 descriptor number as the socket. This naturally causes chaos. The fix is
3855 not to close the socket in ip_connect(); the socket should be closed by
3856 the function that creates it. There was only one place in the code where
3857 this was missing, in the iplookup router, which I don't think anybody now
3858 uses, but I've fixed it anyway.
3860 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3861 well as to direct DNS lookups. Otherwise the handling of names in host
3862 lists is inconsistent and therefore confusing.
3868 PH/01 Two changes to the default runtime configuration:
3870 (1) Move the checks for relay_from_hosts and authenticated clients from
3871 after to before the (commented out) DNS black list checks.
3873 (2) Add control=submission to the relay_from_hosts and authenticated
3874 clients checks, on the grounds that messages accepted by these
3875 statements are most likely to be submissions.
3877 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3879 (1) Generate an error if the third argument for the ${prvs expansion is
3882 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3885 (3) Reset the variables that are obtained from the first argument of
3886 ${prvscheck and used in the second argument before leaving the code,
3887 because their memory is reclaimed, so using them afterwards may do
3890 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3891 one (it's much easier than Tom thought :-).
3893 (5) Because of (4), we can now allow for the use of $prvscheck_result
3894 inside the third argument.
3896 PH/03 For some reason, the default setting of PATH when running a command from
3897 a pipe transport was just "/usr/bin". I have changed it to
3900 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3901 anything to be listed in the output from -bV.
3903 PH/05 When a filter generated an autoreply, the entire To: header line was
3904 quoted in the delivery log line, like this:
3906 => >A.N.Other <ano@some.domain> <original@ddress> ...
3908 This has been changed so that it extracts the operative address. There
3909 may be more than one such address. If so, they are comma-separated, like
3912 => >ano@some.domain,ona@other.domain <original@ddress> ...
3914 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3915 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3916 not being looked up in the rDNS to get a host name, Exim was showing the
3917 IP address twice in Received: lines, even though the IP addresses were
3918 identical. For example:
3920 Received: from [1.2.3.4] (helo=[1.2.3.4])
3922 However, if the real host name was known, it was omitting the HELO data
3923 if it matched the actual IP address. This has been tidied up so that it
3924 doesn't show the same IP address twice.
3926 PH/07 When both +timestamp and +memory debugging was on, the value given by
3927 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3928 called by the debug printing, thereby overwriting the timestamp buffer.
3929 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3931 PH/08 When the original message was included in an autoreply transport, it
3932 always said "this is a copy of the message, including all the headers",
3933 even if body_only or headers_only was set. It now gives an appropriate
3936 PH/09 Applied a patch from the Sieve maintainer which:
3938 o fixes some comments
3939 o adds the (disabled) notify extension core
3940 o adds some debug output for the result of if/elsif tests
3941 o points to the current vacation draft in the documentation
3942 and documents the missing references header update
3946 o fixes a bug in processing the envelope test (when testing
3947 multiple envelope elements, the last element determined the
3950 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3951 Electronic Mail") by including:
3953 Auto-submitted: auto-generated
3955 in the messages that it generates (bounce messages and others, such as
3956 warnings). In the case of bounce messages for non-SMTP messages, there was
3957 also a typo: it was using "Auto_submitted" (underscore instead of
3958 hyphen). Since every message generated by Exim is necessarily in response
3959 to another message, thes have all been changed to:
3961 Auto-Submitted: auto-replied
3963 in accordance with these statements in the RFC:
3965 The auto-replied keyword:
3967 - SHOULD be used on messages sent in direct response to another
3968 message by an automatic process,
3970 - MUST NOT be used on manually-generated messages,
3972 - MAY be used on Delivery Status Notifications (DSNs) and Message
3973 Disposition Notifications (MDNs),
3975 - MUST NOT be used on messages generated by automatic or periodic
3976 processes, except for messages which are automatic responses to
3979 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3980 to the default Received: header definition.
3982 PH/12 Added log selector acl_warn_skipped (default on).
3984 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3985 variables because (a) they are in the wrong storage pool and (b) even if
3986 they were copied, it wouldn't work properly because of the caching.
3988 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3989 checking when decoding. Apparently there are clients that generate
3990 overlong encoded strings. Why am I not surprised?
3992 PH/15 If the first argument of "${if match_address" was not empty, but did not
3993 contain an "@" character, Exim crashed. Now it writes a panic log message
3994 and treats the condition as false.
3996 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3998 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3999 extension "envelope-auth". The code is finished and in agreement with
4000 other implementations, but there is no documentation so far and in fact,
4001 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4002 not changing the active code.
4004 Print executed "if" and "elsif" statements when debugging is used. This
4005 helps a great deal to understand what a filter does.
4007 Document more things not specified clearly in RFC3028. I had all this
4008 sorted out, when out of a sudden new issues came to my mind. Oops."
4010 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4013 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4014 canonical form (as documented). However, after a host name lookup from
4015 the IP address, check_host() was doing a simple string comparison with
4016 addresses acquired from the DNS when checking that the found name did
4017 have the original IP as one of its addresses. Since any found IPv6
4018 addresses are likely to be in abbreviated form, the comparison could
4019 fail. Luckily, there already exists a function for doing the comparison
4020 by converting both addresses to binary, so now that is used instead of
4021 the text comparison.
4023 PH/20 There was another similar case to PH/19, when a complete host name was
4024 given in a host list; looking up its IP address could give an abbreviated
4025 form, whereas the current host's name might or might not be abbreviated.
4026 The same fix has been applied.
4032 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4033 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4036 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4037 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4039 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4041 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4042 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4043 Also, for "Auto-Submitted", treat anything other than "no" as
4044 non-personal, in accordance with RFC 3834. (Previously it treated
4045 anything starting "auto-" as non-personal.)
4047 TF/01 The control=submission/name=... option had a problem with syntax
4048 errors if the name included a slash character. The /name= option
4049 now slurps the rest of the string, so it can include any characters
4050 but it must come last in the list of options (after /sender_retain
4053 PH/05 Some modifications to the interface to the fake nameserver for the new
4061 TK/01 Added the "success_on_redirect" address verification option. See
4062 NewStuff for rationale and an example.
4064 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4066 PH/02 Patch to exigrep to allow it to work on syslog lines.
4068 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4069 fread() to skip over the body file's header line, because in Cygwin the
4070 header line is locked and is inaccessible.
4072 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4073 co-exist for some time) to make it clear that it is the Exim ID that is
4074 referenced, not the Message-ID: header line.
4076 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4077 string_format() function, because snprintf() does not exist on all
4080 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4081 ${stat: expansion item.
4083 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4084 protocol synchronization error", to keep the pedants happy.
4086 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4087 well as for IRIX systems, when gcc is being used. See the host.c source
4090 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4092 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4095 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4096 transport and to the smtp transport in LMTP mode.
4098 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4100 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4101 run a filter in a subprocess. This could lead to confusion in subsequent
4102 lookups in the parent process. There should also be a search_tidyup() at
4103 the end of the subprocess.
4105 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4106 only if the host matched helo_try_verify_hosts, which caused the
4107 verification to occur when the EHLO/HELO command was issued. The ACL just
4108 tested the remembered result. Now, if a previous verification attempt has
4109 not happened, "verify = helo" does it there and then.
4111 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4113 TK/03 Fix log output including CR from clamd.
4115 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4116 exist provoked a memory error which could cause a segfault.
4118 PH/15 Installed PCRE 6.2
4120 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4122 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4123 of the problem. Specifically, suggested +O2 rather than +O1 for the
4126 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4128 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4129 record, the port was not correctly logged when the outgoing_port log
4130 selector was set (it logged the transort's default port).
4132 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4133 fallback_hosts, and "hosts" in the smtp transport.
4135 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4136 host errors such as "Connection refused".
4138 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4139 authentication with radiusclient 0.4.9:
4141 - Error returned from rc_read_config was caught wrongly
4142 - Username/password not passed on to radius server due to wrong length.
4144 The presumption is that some radiusclient API changes for 4.51/PH/17
4145 were not taken care of correctly. The code is still untested by me (my
4146 Linux distribution still has 0.3.2 of radiusclient), but it was
4147 contributed by a Radius user.
4149 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4150 expanding the "port" option of the smtp transport.
4152 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4153 while reading a MIME header. Thanks to Tom Hughes for a patch.
4155 PH/24 Include config.h inside local_scan.h so that configuration settings are
4158 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4159 that in an ACL for RCPT (for example), you can examine exactly what was
4162 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4163 commands, but it was not correctly comparing the address with the actual
4164 client host address. Thus, it would show the EHLO address in Received:
4165 header lines when this was not necessary.
4167 PH/27 Added the % operator to ${eval:}.
4169 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4170 it should be ignoring failures (because with -C, for example, it has lost
4171 privilege). It wasn't ignoring creation failures other than "already
4174 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4177 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4178 because an input line was too long, either on its own, or by virtue of
4179 too many continuations, the temporary file was not being removed, and the
4180 return code was incorrect.
4182 PH/31 Missing "BOOL" in function definition in filtertest.c.
4184 PH/32 Applied Sieve patches from the maintainer.
4186 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4188 PH/33 Added "verify = not_blind".
4190 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4191 Local/Makefile (with some defaults set). These are used in built scripts
4192 such as exicyclog, but they have never been used in the exim_install
4193 script (though there are many overriding facilities there). I have
4194 arranged that the exim_install script now takes note of these two
4197 PH/35 Installed configuration files for Dragonfly.
4199 PH/36 When a locally submitted message by a trusted user did not contain a
4200 From: header, and the sender address was obtained from -f or from an SMTP
4201 MAIL command, and the trusted user did not use -F to supply a sender
4202 name, $originator_name was incorrectly used when constructing a From:
4203 header. Furthermore, $originator_name was used for submission mode
4204 messages from external hosts without From: headers in a similar way,
4205 which is clearly wrong.
4207 PH/37 Added control=suppress_local_fixups.
4209 PH/38 When log_selector = +received_sender was set, and the addition of the
4210 sender made the log line's construction buffer exactly full, or one byte
4211 less than full, an overflow happened when the terminating "\n" was
4214 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4215 when the result of a list match is failure because a DNS lookup failed.
4217 PH/40 RM_COMMAND is now used in the building process.
4219 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4220 the "build-* directories that it finds.
4222 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4223 address in a domain literal was a prefix of an interface address.
4225 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4226 when verifying a sender address, unless rewrite_headers is false.
4228 PH/44 Wrote a long comment about why errors_to addresses are verified as
4229 recipients, not senders.
4231 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4232 the ratelimit ACL was added.
4234 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4236 PH/46 Added extra information about PostgreSQL errors to the error string.
4238 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4239 avoiding the need to install special zones in a real server. This is
4240 backwards compatible; if it can't find the fake resolver, it drops back.
4241 Thus, both old and new test suites can be run.
4243 TF/02 Added util/ratelimit.pl
4245 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4246 clock is set back in time.
4248 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4249 Candler <B.Candler@pobox.com>.
4251 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4252 OK for addresses that are the result of redirections.
4254 PH/48 A number of further additions for the benefit of the new test suite,
4255 including a fake gethostbyname() that interfaces to the fake DNS resolver
4258 TF/06 The fix for widen_domains has also been applied to qualify_single and
4259 search_parents which are the other dnslookup options that can cause
4262 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4265 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4267 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4268 service" if it encountered an MX record with an empty target hostname.
4269 The message is now "an MX or SRV record indicated no SMTP service".
4271 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4272 if the DNS of the sending site is misconfigured. This is quite a
4273 common situation. This change restores the behaviour of treating a
4274 helo verification defer as a failure.
4276 PH/51 If self=fail was set on a router, the bounce message did not include the
4277 actual error message.
4283 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4285 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4286 command itself ended in error, the underlying message about the transport
4287 filter timeout was being overwritten with the pipe command error. Now the
4288 underlying error message should be appended to the second error message.
4290 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4292 PH/02 Reduce the amount of output that "make" produces by default. Full output
4293 can still be requested.
4295 PH/03 The warning log line about a condition test deferring for a "warn" verb
4296 was being output only once per connection, rather than after each
4297 occurrence (because it was using the same function as for successful
4298 "warn" verbs). This seems wrong, so I have changed it.
4300 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4301 it should not have, which might have caused a crash in the right
4302 circumstances, but probably never did.
4304 PH/04 Installed a modified version of Tony Finch's patch to make submission
4305 mode fix the return path as well as the Sender: header line, and to
4306 add a /name= option so that you can make the user's friendly name appear
4309 TF/03 Added the control = fakedefer ACL modifier.
4311 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4312 Mark Lowes for thorough testing.
4314 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4316 TK/03 Merged latest SRS patch from Miles Wilton.
4318 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4319 with the definition in sysexits.h (which is #included earlier).
4320 Fortunately, Exim does not actually use EX_OK. The code used to try to
4321 preserve the sysexits.h value, by assuming that macro definitions were
4322 scanned for macro replacements. I have been disabused of this notion,
4323 so now the code just undefines EX_OK before #including unistd.h.
4325 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4326 in the smtp transport. When a block could not be written in a single
4327 write() function, the timeout was being re-applied to each part-write.
4328 This seems wrong - if the receiver was accepting one byte at a time it
4329 would take for ever. The timeout is now adjusted when this happens. It
4330 doesn't have to be particularly precise.
4332 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4333 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4335 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4336 without a domain if the check to <postmaster@domain> fails.
4338 SC/01 Eximstats: added -xls and the ability to specify output files
4339 (patch written by Frank Heydlauf).
4341 SC/02 Eximstats: use FileHandles for outputting results.
4343 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4345 SC/04 Eximstats: fixed display of large numbers with -nvr option
4347 SC/05 Eximstats: fixed merging of reports with empty tables.
4349 SC/06 Eximstats: added the -include_original_destination flag
4351 SC/07 Eximstats: removed tabs and trailing whitespace.
4353 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4355 TK/06 MBOX spool code: Add real "From " MBOX separator line
4356 so the .eml file is really in mbox format (even though
4357 most programs do not really care). Patch from Alex Miller.
4359 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4360 The latter is generated from $received_to and is only set if the
4361 message has one envelope recipient. SA can use these headers,
4362 obviously out-of-the-box. Patch from Alex Miller.
4364 PH/08 The ${def test on a variable was returning false if the variable's
4365 value was "0", contrary to what the specification has always said!
4366 The result should be true unless the variable is empty.
4368 PH/09 The syntax error of a character other than { following "${if
4369 def:variable_name" (after optional whitespace) was not being diagnosed.
4370 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4371 accidental colon was present, for example, could give incorrect results.
4373 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4374 result is used (not including appendfile, where other changes are about
4377 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4378 This involved changing a lot of size variables from int to off_t. It
4379 should work with maildirs and everything.
4381 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4382 spamd dying while we are connected to it.
4384 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4387 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4388 function for BDB 4.3.
4390 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4392 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4393 the sources for PCRE have been re-arranged and more files are now
4396 PH/15 The code I had for printing potentially long long variables in PH/11
4397 above was not the best (it lost precision). The length of off_t variables
4398 is now inspected at build time, and an appropriate printing format (%ld
4399 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4400 to be "long long int" or "long int". This is needed for the internal
4401 formatting function string_vformat().
4403 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4404 the configuration file to be ":syslog", then the script "guesses" where
4405 the logs files are, rather than using the compiled in default. In our
4406 case the guess is not the same as the compiled default, so the script
4407 suddenly stopped working when I started to use syslog. The patch checks
4408 to see if log_file_path is "". If so, it attempts to read it from exim
4409 with no configuration file to get the compiled in version, before it
4410 falls back to the previous guessing code."
4412 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4413 implementing BATV in an Exim configuration. See NewStuff for the gory
4416 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4417 Makefile that are specific to HP-UX.
4419 PH/18 If the "use_postmaster" option was set for a recipient callout together
4420 with the "random" option, the postmaster address was used as the MAIL
4421 FROM address for the random test, but not for the subsequent recipient
4422 test. It is now used for both.
4424 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4425 patch removes a few documentation additions to RFC 3028, because the
4426 latest draft now contains them. It adds the new en;ascii-case comparator
4427 and a new error check for 8bit text in MIME parts. Comparator and
4428 require names are now matched exactly. I enabled the subaddress
4429 extension, but it is not well tested yet (read: it works for me)."
4431 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4432 rework some of the code of TK/09 above to avoid the hardwired use of
4433 "%lld" and "long long". Replaced the call to snprintf() with a call to
4436 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4437 records point to non-existent hosts", "retry timeout exceeded", and
4438 "retry time not reached for any host after a long failure period".
4440 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4441 experimental DomainKeys support:
4443 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4444 (2) On an error such as an illegally used "control", the wrong name for
4445 the control was given.
4447 These problems did NOT occur unless DomainKeys support was compiled.
4449 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4451 PH/24 Added ${if match_ip condition.
4453 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4454 databases so that it will be absolutely obvious if a crash occurs in the
4455 DB library. This is a regular occurrence (often caused by mis-matched
4458 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4459 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4460 picked up on a user's system that detects such things. There doesn't seem
4461 to be a gcc warning option for this - only an attribute that has to be
4462 put on the function's prototype. It seems that in Fedora Core 4 they have
4463 set this on a number of new functions. No doubt there will be more in due
4466 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4467 specify a transport. However, if an address that was verified by such a
4468 router was the subject of a callout, Exim crashed because it tried to
4469 read the rcpt_include_affixes from the non-existent transport. Now it
4470 just assumes that the setting of that option is false. This bug was
4471 introduced by 4.51/PH/31.
4473 PH/28 Changed -d+all to exclude +memory, because that information is very
4474 rarely of interest, but it makes the output a lot bigger. People tend to
4475 do -d+all out of habit.
4477 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4478 code in os-type was giving problems when libc.so lives in lib64, like on
4481 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4482 aren't the modern standard, and it seems that some systems' include files
4483 don't always have them. Exim was already checking for some of the newer
4484 ones like T_AAAA, and defining it itself. I've added checks for all the
4485 record types that Exim uses.
4487 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4488 not automatically generating a new one, as it is supposed to. This
4489 prevented TLS from working. If the file did exist, but contained invalid
4490 data, a new version was generated, as expected. It was only the case of a
4491 non-existent file that was broken.
4493 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4494 with a change in libdomainkeys > 0.64.
4496 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4497 from DNS. If the selector record carries the flag, it now has
4498 precedence over the domain-wide flag.
4500 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4502 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4503 the use of a port name that isn't defined in /etc/services) occurred, the
4504 message was deferred as in a normal delivery, and thus remained on the
4505 spool, instead of being failed because of the mua_wrapper setting. This
4506 is now fixed, and I tidied up some of the mua_wrapper messages at the
4509 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4510 the messages in a hash of arrays rather than using individual hashes.
4511 This is a bit cleaner and results in dramatic memory savings, albeit
4512 at a slight CPU cost.
4514 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4515 as requested by Marc Sherman.
4517 SC/10 Eximstats: added histograms for user specified patterns as requested
4520 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4522 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4523 fopen() in the content-scanning modules that did not already have it.
4529 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4530 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4532 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4534 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4536 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4537 to test Sieve filters that use "vacation".
4539 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4540 that changes the way the GnuTLS parameters are stored in the cache file.
4541 The new format can be generated externally. For backward compatibility,
4542 if the data in the cache doesn't make sense, Exim assumes it has read an
4543 old-format file, and it generates new data and writes a new file. This
4544 means that you can't go back to an older release without removing the
4547 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4548 work if there are any delivery delays because "one_time" forces the
4549 parent to be marked "delivered", so its unseen clone is never tried
4550 again. For this reason, Exim now forbids the simultaneous setting of
4553 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4554 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4555 entry if you want to know the details. The fix, however, neglected to
4556 consider the case when local delivery batching is involved. The test for
4557 "previously delivered" was not happening when checking to see if an
4558 address could be batched with a previous (undelivered) one; under
4559 certain circumstances this could lead to multiple deliveries to the same
4562 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4563 in its include files, and this causes problems building Exim.
4565 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4566 header_syntax) but Exim was just ignoring anything given after a slash.
4567 In particular, this caused confusion with an attempt to use "verify =
4568 reverse_host_lookup/defer_ok". An error is now given when options are
4569 supplied for verify items that do not have them. (Maybe reverse_host_
4570 lookup should have a defer_ok option, but that's a different point.)
4572 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4573 defined by RFC 821) to 2048, because there were problems with some AUTH
4574 commands, and RFC 1869 says the size should be increased for extended
4575 SMTP commands that take arguments.
4577 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4580 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4581 "unknown" error; now it says that the functionality isn't in the binary.
4583 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4584 an address' error message when a string expansion fails (syntax or
4585 whatever). Otherwise the password may appear in the log. Following change
4586 PH/42 below, there is no longer a chance of it appearing in a bounce
4589 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4591 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4592 of its addresses was ever tried. (Bugzilla bug #2.)
4594 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4595 the result incorrectly in the debug output. (It correctly added a newline
4596 to what was transported.)
4598 TF/01 Added $received_time.
4600 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4601 commented out examples of how to interface to a virus scanner and to
4602 SpamAssassin. Also added commented examples of av_scanner and
4603 spamd_address settings.
4605 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4606 and controls are allowed in which ACLs. There were a couple of minor
4607 errors. Some of the entries in the conditions table (which is a table of
4608 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4609 negation of where the condition IS allowed.
4611 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4613 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4614 header file does not have a version number, so I've had to invent a new
4615 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4616 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4617 radiusclient), but it was contributed by a Radius user.
4619 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4620 files or directories, for OpenSSL.
4622 PH/19 When an Exim process that is running as root has to create an Exim log
4623 file, it does so in a subprocess that runs as exim:exim so as to get the
4624 ownership right at creation (otherwise, other Exim processes might see
4625 the file with the wrong ownership). There was no test for failure of this
4626 fork() call, which would lead to the process getting stuck as it waited
4627 for a non-existent subprocess. Forks do occasionally fail when resources
4628 run out. I reviewed all the other calls to fork(); they all seem to check
4631 PH/20 When checking for unexpected SMTP input at connect time (before writing
4632 the banner), Exim was not dealing correctly with a non-positive return
4633 from the read() function. If the client had disconnected by this time,
4634 the result was a log entry for a synchronization error with an empty
4635 string after "input=" when read() returned zero. If read() returned -1
4636 (an event I could not check), uninitialized data bytes were printed.
4637 There were reports of junk text (parts of files, etc) appearing after
4640 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4642 PH/22 Added support for macro redefinition, and (re)definition in between
4643 driver and ACL definitions.
4645 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4646 forgetting to use the resulting value; it was using the unexpanded value.
4648 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4649 hadn't been configured. The fix is from Juergen Kreileder, who
4650 understands it better than I do:
4652 "Here's what I see happening with three configured cyrus_sasl
4653 authenticators configured (plain, login, cram-md5):
4655 On startup auth_cyrus_sasl_init() gets called for each of these.
4656 This means three calls to sasl_listmech() without a specified mech_list.
4657 => SASL tests which mechs of all available mechs actually work
4658 => three warnings about OTP not working
4659 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4661 With the patch, sasl_listmech() also gets called three times. But now
4662 SASL's mech_list option is set to the server_mech specified in the the
4663 authenticator. Or in other words, the answer from sasl_listmech()
4664 gets limited to just the mech you're testing for (which is different
4666 => the return list contains just 'plain' or 'login', 'cram-md5' or
4667 nothing depending on the value of ob->server_mech.
4669 I've just tested the patch: Authentication still works fine,
4670 unavailable mechs specified in the exim configuration are still
4671 caught, and the auth.log warnings about OTP are gone."
4673 PH/25 When debugging is enabled, the contents of the command line are added
4674 to the debugging output, even when log_selector=+arguments is not
4677 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4678 answer is "GNU", and only if the return is "GNU/something" is the answer
4681 PH/27 $acl_verify_message is now set immediately after the failure of a
4682 verification in an ACL, and so is available in subsequent modifiers. In
4683 particular, the message can be preserved by coding like this:
4685 warn !verify = sender
4686 set acl_m0 = $acl_verify_message
4688 Previously, $acl_verify_message was set only while expanding "message"
4689 and "log_message" when a very denied access.
4691 PH/28 Modified OS/os.c-Linux with
4693 -#ifndef OS_LOAD_AVERAGE
4694 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4696 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4697 nomenclature these days.)
4699 PH/29 Installed patch from the Sieve maintainer that adds the options
4700 sieve_useraddress and sieve_subaddress to the redirect router.
4702 PH/30 In these circumstances:
4703 . Two addresses routed to the same list of hosts;
4704 . First host does not offer TLS;
4705 . First host accepts first address;
4706 . First host gives temporary error to second address;
4707 . Second host offers TLS and a TLS session is established;
4708 . Second host accepts second address.
4709 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4710 and peerdn, if requested) that were in fact used only for the second
4713 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4714 attention to any local part prefix or suffix that was matched by the
4715 router that accepted the address. It now behaves in the same way as it
4716 does for delivery: the affixes are removed from the local part unless
4717 rcpt_include_affixes is set on the transport.
4719 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4720 timeout during the DATA phase of an incoming message.
4722 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4723 applied a patch sanctioned by the Sieve maintainer.
4725 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4726 the uid or gid is negative. A case of a negative gid caused this to be
4727 noticed. The fix allows for either to be negative.
4729 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4730 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4733 PH/36 The change PH/12 above was broken. Fixed it.
4735 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4736 the grounds that routing the same address twice would always produce the
4737 same answer. This might have been true once, but it is certainly no
4738 longer true now. Routing a child address may depend on the previous
4739 routing that produced that child. Some complicated redirection strategies
4740 went wrong when messages had multiple recipients, and made Exim's
4741 behaviour dependent on the order in which the addresses were given.
4743 I have moved the duplicate checking until after the routing is complete.
4744 Exim scans the addresses that are assigned to local and remote
4745 transports, and removes any duplicates. This means that more work will be
4746 done, as duplicates will always all be routed, but duplicates are
4747 presumably rare, so I don't expect this is of any significance.
4749 For deliveries to pipes, files, and autoreplies, the duplicate checking
4750 still happens during the routing process, since they are not going to be
4753 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4754 It corrects a timeout issue with spamd. This is Ian's comment: "The
4755 background is that sometimes spamd either never reads data from a
4756 connection it has accepted, or it never writes response data. The exiscan
4757 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4758 blindly assumes that writes won't block so it may never time out."
4760 PH/39 Allow G after quota size as well as K and M.
4762 PH/40 The value set for $authenticated_id in an authenticator may not contain
4763 binary zeroes or newlines because the value is written to log lines and
4764 to spool files. There was no check on this. Now the value is run through
4765 the string_printing() function so that such characters are converted to
4766 printable escape sequences.
4768 PH/41 $message_linecount is a new variable that contains the total number of
4769 lines in the message. Compare $body_linecount, which is the count for the
4772 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4773 bounce and delay warning messages, except in certain special cases, which
4776 (a) An SMTP error message from a remote host;
4777 (b) A message specified in a :fail: redirection;
4778 (c) A message specified in a "fail" command in a system filter;
4779 (d) A message specified in a FAIL return from the queryprogram router;
4780 (e) A message specified by the cannot_route_message router option.
4782 In these cases only, Exim does include the error details in bounce and
4783 warning messages. There are also a few cases where bland messages such
4784 as "unrouteable address" or "local delivery error" are given.
4786 PH/43 $value is now also set for the "else" part of a ${run expansion.
4788 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4789 being worked on, but at least Exim now implements the latest version to
4792 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4793 process to complete was treated as a delivery failure, a timeout while
4794 writing the message to the pipe was logged, but erroneously treated as a
4795 successful delivery. Such timeouts include transport filter timeouts. For
4796 consistency with the overall process timeout, these timeouts are now
4797 treated as errors, giving rise to delivery failures by default. However,
4798 there is now a new Boolean option for the pipe transport called
4799 timeout_defer, which, if set TRUE, converts the failures into defers for
4800 both kinds of timeout. A transport filter timeout is now identified in
4803 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4804 systems where "make" and "gmake" are different, calling "gmake" at top
4805 level broke things. I've arranged for the value of $(MAKE) to be passed
4806 from the Makefile to this script so that it can call the same version of
4810 A note about Exim versions 4.44 and 4.50
4811 ----------------------------------------
4813 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4814 changes of various kinds. As a consequence, a big documentation update was
4815 needed. This delayed the release for rather longer than seemed good, especially
4816 in the light of a couple of (minor) security issues. Therefore, the changes
4817 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4818 release. So 4.44 and 4.50 are in effect two different branches that both start
4821 I have left the 4.50 change log unchanged; it contains all the changes since
4822 4.43. The change log for 4.44 is below; many of its items are identical to
4823 those for 4.50. This seems to be the most sensible way to preserve the
4824 historical information.
4830 1. Minor wording change to the doc/README.SIEVE file.
4832 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4833 computation of the current number of files was incorrect.
4835 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4836 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4839 4. Give more explanation in the error message when the command for a transport
4840 filter fails to execute.
4842 5. There are several places where Exim runs a non-Exim command in a
4843 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4844 was being done only for the command run by the queryprogram router. It is
4845 now done for all such subprocesses. The other cases are: ${run, transport
4846 filters, and the commands run by the lmtp and pipe transports.
4848 6. Added CONFIGURE_GROUP build-time option.
4850 7. Some older OS have a limit of 256 on the maximum number of file
4851 descriptors. Exim was using setrlimit() to set 1000 as a large value
4852 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4853 systems. I've change it so that if it can't get 1000, it tries for 256.
4855 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4856 was an oversight, and furthermore, ever since the addition of extra
4857 controls (e.g. 4.43/32), the checks on when to allow different forms of
4858 "control" were broken. There should now be diagnostics for all cases when a
4859 control that does not make sense is encountered.
4861 9. Added the /retain_sender option to "control=submission".
4863 10. $recipients is now available in the predata ACL (oversight).
4865 11. Tidy the search cache before the fork to do a delivery from a message
4866 received from the command line. Otherwise the child will trigger a lookup
4867 failure and thereby defer the delivery if it tries to use (for example) a
4868 cached ldap connection that the parent has called unbind on.
4870 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4871 of $address_data from the recipient verification was clobbered by the
4872 sender verification.
4874 13. The value of address_data from a sender verification is now available in
4875 $sender_address_data in subsequent conditions in the ACL statement.
4877 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4879 15. Added a new option "connect=<time>" to callout options, to set a different
4882 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4883 was its contents. (It was OK if the option was not defined at all.)
4885 17. A "Completed" log line is now written for messages that are removed from
4886 the spool by the -Mrm option.
4888 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4889 information about exactly what failed.
4891 19. Added -dd to debug only the daemon process.
4893 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4894 handles timeouts, both on the server side and network timeouts. Renamed the
4895 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4897 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4898 It is now set to "smtps".
4900 22. $host_address is now set to the target address during the checking of
4901 ignore_target_hosts.
4903 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4904 being passed; this would have caused $sender_host_name to have been used if
4905 matching the list had actually called for a host name (not very likely,
4906 since this list is usually IP addresses). A host name is now passed as
4909 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4910 code that specifies a non-restarting handler (typically sigaction() in
4911 modern systems) in an attempt to fix a rare and obscure crash bug.
4913 25. Narrowed the window for a race in the daemon that could cause it to ignore
4914 SIGCHLD signals. This is not a major problem, because they are used only to
4915 wake it up if nothing else does.
4917 26. A malformed maildirsize file could cause Exim to calculate negative values
4918 for the mailbox size or file count. Odd effects could occur as a result.
4919 The maildirsize information is now recalculated if the size or filecount
4922 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4923 support for a long time. Removed HAVE_SYS_VFS_H.
4925 28. Installed the latest version of exipick from John Jetmore.
4927 29. In an address list, if the pattern was not a regular expression, an empty
4928 subject address (from a bounce message) matched only if the pattern was an
4929 empty string. Non-empty patterns were not even tested. This was the wrong
4930 because it is perfectly reasonable to use an empty address as part of a
4931 database query. An empty address is now tested by patterns that are
4932 lookups. However, all the other forms of pattern expect the subject to
4933 contain a local part and a domain, and therefore, for them, an empty
4934 address still always fails if the pattern is not itself empty.
4936 30. Exim went into a mad DNS loop when attempting to do a callout where the
4937 host was specified on an smtp transport, and looking it up yielded more
4938 than one IP address.
4940 31. Re-factored the code for checking spool and log partition space into a
4941 function that finds that data and another that does the check. The former
4942 is then used to implement four new variables: $spool_space, $log_space,
4943 $spool_inodes, and $log_inodes.
4945 32. The RFC2047 encoding function was originally intended for short strings
4946 such as real names; it was not keeping to the 75-character limit for
4947 encoded words that the RFC imposes. It now respects the limit, and
4948 generates multiple encoded words if necessary. To be on the safe side, I
4949 have increased the buffer size for the ${rfc2047: expansion operator from
4952 33. It is now permitted to omit both strings after an "if" condition; if the
4953 condition is true, the result is "true". As before, when the second string
4954 is omitted, a false condition yields an empty string. This makes it less
4955 cumbersome to write custom ACL and router conditions.
4957 34. Failure to deliver a bounce message always caused it to be frozen, even if
4958 there was an errors_to setting on the router. The errors_to setting is now
4961 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4962 canonical form (fully expanded) before being placed in
4963 $sender_host_address.
4965 36. The table in the code that translates DNS record types into text (T_A to
4966 "A" for instance) was missing entries for NS and CNAME. It is just possible
4967 that this could have caused confusion if both these types were looked up
4968 for the same domain, because the text type is used as part of Exim's
4969 per-process caching. But the chance of anyone hitting this buglet seems
4972 37. The dnsdb lookup has been extended in a number of ways.
4974 (1) There is a new type, "zns", which walks up the domain tree until it
4975 finds some nameserver records. It should be used with care.
4977 (2) There is a new type, "mxh", which is like "mx" except that it returns
4978 just the host names, not the priorities.
4980 (3) It is now possible to give a list of domains (or IP addresses) to be
4981 looked up. The behaviour when one of the lookups defers can be
4982 controlled by a keyword.
4984 (4) It is now possible to specify the separator character for use when
4985 multiple records are returned.
4987 38. The dnslists ACL condition has been extended: it is now possible to supply
4988 a list of IP addresses and/or domains to be looked up in a particular DNS
4991 39. Added log_selector=+queue_time_overall.
4993 40. When running the queue in the test harness, wait just a tad after forking a
4994 delivery process, to get repeatability of debugging output.
4996 41. Include certificate and key file names in error message when GnuTLS fails
4997 to set them up, because the GnuTLS error message doesn't include the name
4998 of the failing file when there is a problem reading it.
5000 42. Allow both -bf and -bF in the same test run.
5002 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5004 44. The "Exiscan patch" is now merged into the mainline Exim source.
5006 45. Sometimes the final signoff response after QUIT could fail to get
5007 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5008 < 0 before doing a fflush(). This bug looks as though it goes back to the
5009 introduction of TLS in release 3.20, but "sometimes" must have been rare
5010 because the tests only now provoked it.
5012 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5013 (this can affect the format of dates).
5015 47. exim_tidydb, when checking for the continued existence of a message for
5016 which it has found a message-specific retry record, was not finding
5017 messages that were in split spool directories. Consequently, it was
5018 deleting retry records that should have stayed in existence.
5020 48. Steve fixed some bugs in eximstats.
5022 49. The SPA authentication driver was not abandoning authentication and moving
5023 on to the next authenticator when an expansion was forced to fail,
5024 contradicting the general specification for all authenticators. Instead it
5025 was generating a temporary error. It now behaves as specified.
5027 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5028 (the order specifies the preference for clients). The order is now AES256,
5029 AES128, 3DES, ARCFOUR128.
5031 51. Small patch to Sieve code - explicitly set From: when generating an
5034 52. Exim crashed if a remote delivery caused a very long error message to be
5035 recorded - for instance if somebody sent an entire SpamAssassin report back
5036 as a large number of 550 error lines. This bug was coincidentally fixed by
5037 increasing the size of one of Exim's internal buffers (big_buffer) that
5038 happened as part of the Exiscan merge. However, to be on the safe side, I
5039 have made the code more robust (and fixed the comments that describe what
5042 53. Now that there can be additional text after "Completed" in log lines (if
5043 the queue_time_overall log selector is set), a one-byte patch to exigrep
5044 was needed to allow it to recognize "Completed" as not the last thing in
5047 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5048 patch that reportedly fixes this has been added. I am not expert enough to
5049 create a test for it. This is what the patch creator wrote:
5051 "I found a little strange behaviour of ldap code when working with
5052 Windows 2003 AD Domain, where users was placed in more than one
5053 Organization Units. When I tried to give exim partial DN, the exit code
5054 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5055 But simultaneously result of request was absolutely normal ldap result,
5056 so I produce this patch..."
5058 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5059 REFERENCE, so I have modified the code to exclude the patch when that macro
5062 55. Some experimental protocols are using DNS PTR records for new purposes. The
5063 keys for these records are domain names, not reversed IP addresses. The
5064 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5065 leaves it alone. Component reversal etc. now happens only for IP addresses.
5068 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5070 57. Double the size of the debug message buffer (to 2048) so that more of very
5071 long debug lines gets shown.
5073 58. The exicyclog utility now does better if the number of log files to keep
5074 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5076 59. Two changes related to the smtp_active_hostname option:
5078 (1) $smtp_active_hostname is now available as a variable.
5079 (2) The default for smtp_banner uses $smtp_active_hostname instead
5080 of $primary_hostname.
5082 60. The host_aton() function is supposed to be passed a string that is known
5083 to be a valid IP address. However, in the case of IPv6 addresses, it was
5084 not checking this. This is a hostage to fortune. Exim now panics and dies
5085 if the condition is not met. A case was found where this could be provoked
5086 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5087 components; fortuitously, this particular loophole had already been fixed
5088 by change 4.50/55 above.
5090 If there are any other similar loopholes, the new check in host_aton()
5091 itself should stop them being exploited. The report I received stated that
5092 data on the command line could provoke the exploit when Exim was running as
5093 exim, but did not say which command line option was involved. All I could
5094 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5095 running as the user.
5098 61. There was a buffer overflow vulnerability in the SPA authentication code
5099 (which came originally from the Samba project). I have added a test to the
5100 spa_base64_to_bits() function which I hope fixes it.
5103 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5104 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5106 63. The daemon start-up calls getloadavg() while still root for those OS that
5107 need the first call to be done as root, but it missed one case: when
5108 deliver_queue_load_max is set with deliver_drop_privilege. This is
5109 necessary for the benefit of the queue runner, because there is no re-exec
5110 when deliver_drop_privilege is set.
5112 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5113 This has been fixed.
5115 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5116 was in use, was not putting the data itself into the right store pool;
5117 consequently, it could be overwritten for a subsequent message in the same
5118 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5121 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5123 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5124 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5125 the function were treating the return as a boolean value, which happened to
5126 work because 0=false and not-0=true, but is not correct code.
5128 68. The host_aton() function was not handling scoped IPv6 addresses (those
5129 with, for example, "%eth0" on the end) correctly.
5131 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5132 negated items (that is, ~something) in unsigned ints. Some compilers
5133 apparently mutter when there is no cast.
5135 70. If an address verification called from an ACL failed, and did not produce a
5136 user-specific message (i.e. there was only a "system" message), nothing was
5137 put in $acl_verify_message. In this situation, it now puts the system
5140 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5141 session; change 4.31/43 added the unwanted input to the log line - except
5142 that it did not do this in the start of session case. It now does.
5144 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5145 This is wrong and can cause the other end to generate a synchronization
5146 error if it is another Exim or anything else that does the synchronization
5147 check. A QUIT command is no longer sent after a timeout.
5149 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5150 during host lookups.
5152 74. The defer_ok option of callout verification was not working if it was used
5153 when verifying addresses in header lines, that is, for this case:
5155 verify = header_sender/callout=defer_ok
5157 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5158 those file descriptors could be used for SMTP connections. If anything
5159 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5160 could be sent to the SMTP client, causing chaos. The daemon now opens
5161 stdin, stdout, and stderr to /dev/null when it puts itself into the
5164 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5165 log by default. The user can override this with suitable Perl magic.
5167 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5168 the log message when discard triggers, was not working for the DATA ACL or
5169 for the non-SMTP ACL.
5171 78. Error message wording change in sieve.c.
5173 79. If smtp_accept_max_per_host was set, the number of connections could be
5174 restricted to fewer than expected, because the daemon was trying to set up
5175 a new connection before checking whether the processes handling previous
5176 connections had finished. The check for completed processes is now done
5177 earlier. On busy systems, this bug wouldn't be noticed because something
5178 else would have woken the daemon, and it would have reaped the completed
5181 80. If a message was submitted locally by a user whose login name contained one
5182 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5183 It caused a spool format error. I have fixed the spool reading code. A
5184 related problem was that the "from" clause in the Received: line became
5185 illegal because of the space(s). It is now covered by ${quote_local_part.
5187 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5190 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5192 83. Patch from Sieve maintainer to fix unterminated string problem in
5193 "vacation" handling.
5195 84. Some minor changes to the Linux configuration files to help with other
5196 OS variants using glibc.
5198 85. One more patch for Sieve to update vacation handling to latest spec.
5201 ----------------------------------------------------
5202 See the note above about the 4.44 and 4.50 releases.
5203 ----------------------------------------------------
5209 1. Change 4.43/35 introduced a bug that caused file counts to be
5210 incorrectly computed when quota_filecount was set in an appendfile
5213 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5214 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5217 3. Give more explanation in the error message when the command for a transport
5218 filter fails to execute.
5220 4. There are several places where Exim runs a non-Exim command in a
5221 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5222 was being done only for the command run by the queryprogram router. It is
5223 now done for all such subprocesses. The other cases are: ${run, transport
5224 filters, and the commands run by the lmtp and pipe transports.
5226 5. Some older OS have a limit of 256 on the maximum number of file
5227 descriptors. Exim was using setrlimit() to set 1000 as a large value
5228 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5229 systems. I've change it so that if it can't get 1000, it tries for 256.
5231 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5232 was an oversight, and furthermore, ever since the addition of extra
5233 controls (e.g. 4.43/32), the checks on when to allow different forms of
5234 "control" were broken. There should now be diagnostics for all cases when a
5235 control that does not make sense is encountered.
5237 7. $recipients is now available in the predata ACL (oversight).
5239 8. Tidy the search cache before the fork to do a delivery from a message
5240 received from the command line. Otherwise the child will trigger a lookup
5241 failure and thereby defer the delivery if it tries to use (for example) a
5242 cached ldap connection that the parent has called unbind on.
5244 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5245 of $address_data from the recipient verification was clobbered by the
5246 sender verification.
5248 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5249 was its contents. (It was OK if the option was not defined at all.)
5251 11. A "Completed" log line is now written for messages that are removed from
5252 the spool by the -Mrm option.
5254 12. $host_address is now set to the target address during the checking of
5255 ignore_target_hosts.
5257 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5258 being passed; this would have caused $sender_host_name to have been used if
5259 matching the list had actually called for a host name (not very likely,
5260 since this list is usually IP addresses). A host name is now passed as
5263 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5264 code that specifies a non-restarting handler (typically sigaction() in
5265 modern systems) in an attempt to fix a rare and obscure crash bug.
5267 15. Narrowed the window for a race in the daemon that could cause it to ignore
5268 SIGCHLD signals. This is not a major problem, because they are used only to
5269 wake it up if nothing else does.
5271 16. A malformed maildirsize file could cause Exim to calculate negative values
5272 for the mailbox size or file count. Odd effects could occur as a result.
5273 The maildirsize information is now recalculated if the size or filecount
5276 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5277 support for a long time. Removed HAVE_SYS_VFS_H.
5279 18. Updated exipick to current release from John Jetmore.
5281 19. Allow an empty sender to be matched against a lookup in an address list.
5282 Previously the only cases considered were a regular expression, or an
5285 20. Exim went into a mad DNS lookup loop when doing a callout where the
5286 host was specified on the transport, if the DNS lookup yielded more than
5289 21. The RFC2047 encoding function was originally intended for short strings
5290 such as real names; it was not keeping to the 75-character limit for
5291 encoded words that the RFC imposes. It now respects the limit, and
5292 generates multiple encoded words if necessary. To be on the safe side, I
5293 have increased the buffer size for the ${rfc2047: expansion operator from
5296 22. Failure to deliver a bounce message always caused it to be frozen, even if
5297 there was an errors_to setting on the router. The errors_to setting is now
5300 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5301 canonical form (fully expanded) before being placed in
5302 $sender_host_address.
5304 24. Updated eximstats to version 1.33
5306 25. Include certificate and key file names in error message when GnuTLS fails
5307 to set them up, because the GnuTLS error message doesn't include the name
5308 of the failing file when there is a problem reading it.
5310 26. Expand error message when OpenSSL has problems setting up cert/key files.
5313 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5314 (this can affect the format of dates).
5316 28. exim_tidydb, when checking for the continued existence of a message for
5317 which it has found a message-specific retry record, was not finding
5318 messages that were in split spool directories. Consequently, it was
5319 deleting retry records that should have stayed in existence.
5321 29. eximstats updated to version 1.35
5322 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5323 1.35 - bugfix such that pie charts by volume are generated correctly
5325 30. The SPA authentication driver was not abandoning authentication and moving
5326 on to the next authenticator when an expansion was forced to fail,
5327 contradicting the general specification for all authenticators. Instead it
5328 was generating a temporary error. It now behaves as specified.
5330 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5331 (the order specifies the preference for clients). The order is now AES256,
5332 AES128, 3DES, ARCFOUR128.
5334 31. Small patch to Sieve code - explicitly set From: when generating an
5337 32. Exim crashed if a remote delivery caused a very long error message to be
5338 recorded - for instance if somebody sent an entire SpamAssassin report back
5339 as a large number of 550 error lines. This bug was coincidentally fixed by
5340 increasing the size of one of Exim's internal buffers (big_buffer) that
5341 happened as part of the Exiscan merge. However, to be on the safe side, I
5342 have made the code more robust (and fixed the comments that describe what
5345 33. Some experimental protocols are using DNS PTR records for new purposes. The
5346 keys for these records are domain names, not reversed IP addresses. The
5347 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5348 leaves it alone. Component reversal etc. now happens only for IP addresses.
5351 34. The host_aton() function is supposed to be passed a string that is known
5352 to be a valid IP address. However, in the case of IPv6 addresses, it was
5353 not checking this. This is a hostage to fortune. Exim now panics and dies
5354 if the condition is not met. A case was found where this could be provoked
5355 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5356 components; fortuitously, this particular loophole had already been fixed
5357 by change 4.50/55 or 4.44/33 above.
5359 If there are any other similar loopholes, the new check in host_aton()
5360 itself should stop them being exploited. The report I received stated that
5361 data on the command line could provoke the exploit when Exim was running as
5362 exim, but did not say which command line option was involved. All I could
5363 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5364 running as the user.
5367 35. There was a buffer overflow vulnerability in the SPA authentication code
5368 (which came originally from the Samba project). I have added a test to the
5369 spa_base64_to_bits() function which I hope fixes it.
5372 36. The daemon start-up calls getloadavg() while still root for those OS that
5373 need the first call to be done as root, but it missed one case: when
5374 deliver_queue_load_max is set with deliver_drop_privilege. This is
5375 necessary for the benefit of the queue runner, because there is no re-exec
5376 when deliver_drop_privilege is set.
5378 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5379 was in use, was not putting the data itself into the right store pool;
5380 consequently, it could be overwritten for a subsequent message in the same
5381 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5384 38. Sometimes the final signoff response after QUIT could fail to get
5385 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5386 < 0 before doing a fflush(). This bug looks as though it goes back to the
5387 introduction of TLS in release 3.20, but "sometimes" must have been rare
5388 because the tests only now provoked it.
5394 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5395 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5396 Now it returns an integer. A number of calls were still expecting a T/F
5397 return. Fortuitously, in all cases, the tests worked in OK situations,
5398 which is the norm. However, things would have gone wrong on any write
5399 failures on the smtp file descriptor. This function is used when sending
5400 messages over SMTP and also when doing verify callouts.
5402 2. When Exim is called to do synchronous delivery of a locally submitted
5403 message (the -odf or -odi options), it no longer closes stderr before doing
5406 3. Implemented the mua_wrapper option.
5408 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5410 5. Implemented the functions header_remove(), header_testname(),
5411 header_add_at_position(), and receive_remove_recipient(), and exported them
5414 6. If an ACL "warn" statement specified the addition of headers, Exim already
5415 inserted X-ACL-Warn: at the start if there was no header name. However, it
5416 was not making this test for the second and subsequent header lines if
5417 there were newlines in the string. This meant that an invalid header could
5418 be inserted if Exim was badly configured.
5420 7. Allow an ACL "warn" statement to add header lines at the start or after all
5421 the Received: headers, as well as at the end.
5423 8. Added the rcpt_4xx retry error code.
5425 9. Added postmaster_mailfrom=xxx to callout verification option.
5427 10. Added mailfrom=xxxx to the callout verification option, for verify=
5430 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5431 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5433 12. Inserted some casts to stop certain compilers warning when using pointer
5434 differences as field lengths or precisions in printf-type calls (mostly
5435 affecting debugging statements).
5437 13. Added optional readline() support for -be (dynamically loaded).
5439 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5440 same clock tick as a message's arrival, so that its received time was the
5441 same as the "first fail" time on the retry record, and that message
5442 remained on the queue past the ultimate address timeout, every queue runner
5443 would try a delivery (because it was past the ultimate address timeout) but
5444 after another failure, the ultimate address timeout, which should have then
5445 bounced the address, did not kick in. This was a "< instead of <=" error;
5446 in most cases the first failure would have been in the next clock tick
5447 after the received time, and all would be well.
5449 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5450 being recognized when the domain list was tested by the match_domain
5451 condition in an expansion string.
5453 16. Added the ${str2b64: operator.
5455 17. Exim was always calling setrlimit() to set a large limit for the number of
5456 processes, without checking whether the existing limit was already
5457 adequate. (It did check for the limit on file descriptors.) Furthermore,
5458 errors from getrlimit() and setrlimit() were being ignored. Now they are
5459 logged to the main and panic logs, but Exim does carry on, to try to do its
5460 job under whatever limits there are.
5462 18. Imported PCRE 5.0.
5464 19. Trivial typo in log message " temporarily refused connection" (the leading
5467 20. If the log selector return_path_on_delivery was set and an address was
5468 redirected to /dev/null, the delivery process crashed because it assumed
5469 that a return path would always be set for a "successful" delivery. In this
5470 case, the whole delivery is bypassed as an optimization, and therefore no
5473 21. Internal re-arrangement: the function for sending a challenge and reading
5474 a response while authentication was assuming a zero-terminated challenge
5475 string. It's now changed to take a pointer and a length, to allow for
5476 binary data in such strings.
5478 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5480 23. Exim was not respecting finduser_retries when seeking the login of the
5481 uid under which it was called; it was always trying 10 times. (The default
5482 setting of finduser_retries is zero.) Also, it was sleeping after the final
5483 failure, which is pointless.
5485 24. Implemented tls_on_connect_ports.
5487 25. Implemented acl_smtp_predata.
5489 26. If the domain in control=submission is set empty, Exim assumes that the
5490 authenticated id is a complete email address when it generates From: or
5491 Sender: header lines.
5493 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5494 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5495 chown and chgrp in /bin and hostname in /usr/bin.
5497 28. Exim was keeping the "process log" file open after each use, just as it
5498 does for the main log. This opens the possibility of it remaining open for
5499 long periods when the USR1 signal hits a daemon. Occasional processlog
5500 errors were reported, that could have been caused by this. Anyway, it seems
5501 much more sensible not to leave this file open at all, so that is what now
5504 29. The long-running daemon process does not normally write to the log once it
5505 has entered its main loop, and it closes the log before doing so. This is
5506 so that log files can straightforwardly be renamed and moved. However,
5507 there are a couple of unusual error situations where the daemon does write
5508 log entries, and I had neglected to close the log afterwards.
5510 30. The text of an SMTP error response that was received during a remote
5511 delivery was being truncated at 512 bytes. This is too short for some of
5512 the long messages that one sometimes sees. I've increased the limit to
5515 31. It is now possible to make retry rules that apply only when a message has a
5516 specific sender, in particular, an empty sender.
5518 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5519 it possible to be selective about when SMTP synchronization is enforced.
5521 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5523 32. Implemented hosts_connection_nolog.
5525 33. Added an ACL for QUIT.
5527 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5530 35. Added mailbox_size and mailbox_filecount to appendfile.
5532 36. Added control = no_multiline_responses to ACLs.
5534 37. There was a bug in the logic of the code that waits for the clock to tick
5535 in the case where the clock went backwards by a substantial amount such
5536 that the microsecond fraction of "now" was more than the microsecond
5537 fraction of "then" (but the whole seconds number was less).
5539 38. Added support for the libradius Radius client library this is found on
5540 FreeBSD (previously only the radiusclient library was supported).
5546 1. When certain lookups returned multiple values in the form name=value, the
5547 quoting of the values was not always being done properly. Specifically:
5548 (a) If the value started with a double quote, but contained no whitespace,
5550 (b) If the value contained whitespace other than a space character (i.e.
5551 tabs or newlines or carriage returns) it was not quoted.
5552 This fix has been applied to the mysql and pgsql lookups by writing a
5553 separate quoting function and calling it from the lookup code. The fix
5554 should probably also be applied to nisplus, ibase and oracle lookups, but
5555 since I cannot test any of those, I have not disturbed their existing code.
5557 2. A hit in the callout cache for a specific address caused a log line with no
5558 reason for rejecting RCPT. Now it says "Previous (cached) callout
5559 verification failure".
5561 3. There was an off-by-one bug in the queryprogram router. An over-long
5562 return line was truncated at 256 instead of 255 characters, thereby
5563 overflowing its buffer with the terminating zero. As well as fixing this, I
5564 have increased the buffer size to 1024 (and made a note to document this).
5566 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5567 when Exim is waiting for an SMTP response from a remote server, Exim
5568 restarts its select() call on the socket, thereby resetting its timeout.
5569 This is not a problem when such interrupts are rare. Somebody set up a cron
5570 job to run exiwhat every 2 minutes, which is less than the normal select()
5571 timeout (5 or 10 minutes). This meant that the select() timeout never
5572 kicked in because it was always reset. I have fixed this by comparing the
5573 time when an interrupt arrives with the time at the start of the first call
5574 to select(). If more time than the timeout has elapsed, the interrupt is
5575 treated as a timeout.
5577 5. Some internal re-factoring in preparation for the addition of Sieve
5578 extensions (by MH). In particular, the "personal" test is moved to a
5579 separate function, and given an option for scanning Cc: and Bcc: (which is
5580 not set for Exim filters).
5582 6. When Exim created an email address using the login of the caller as the
5583 local part (e.g. when creating a From: or Sender: header line), it was not
5584 quoting the local part when it contained special characters such as @.
5586 7. Installed new OpenBSD configuration files.
5588 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5589 try to make them clearer.
5591 9. Callout options, other than the timeout value, were being ignored when
5592 verifying sender addresses in header lines. For example, when using
5594 verify = header_sender/callout=no_cache
5596 the cache was (incorrectly) being used.
5598 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5599 only the Cygwin environment.
5601 11. When return_path_on_delivery was set as a log selector, if different remote
5602 addresses in the same message used different return paths and parallel
5603 remote delivery occurred, the wrong values would sometimes be logged.
5604 (Whenever a remote delivery process finished, the return path value from
5605 the most recently started remote delivery process was logged.)
5607 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5608 lines when AUTH and/or TLS are in use. This is the "received protocol"
5609 field. Exim used to use "asmtp" for authenticated SMTP, without any
5610 indication (in the protocol name) for TLS use. Now it follows the RFC and
5611 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5612 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5613 names appear in log lines as well as in Received: header lines.
5615 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5616 extensions, and comparison tests, and to fix some bugs.
5618 14. Changes to the "personal" filter test:
5620 (1) The test was buggy in that it was just doing the equivalent of
5621 "contains" tests on header lines. For example, if a user's address was
5622 anne@some.where, the "personal" test would incorrectly be true for
5624 To: susanne@some.where
5626 This test is now done by extracting each address from the header in turn,
5627 and checking the entire address. Other tests that are part of "personal"
5628 are now done using regular expressions (for example, to check local parts
5629 of addresses in From: header lines).
5631 (2) The list of non-personal local parts in From: addresses has been
5632 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5633 taken from the Sieve specification recommendations.
5635 (3) If the message contains any header line starting with "List-" it is
5636 treated as non-personal.
5638 (4) The test for "circular" in the Subject: header line has been removed
5639 because it now seems ill-conceived.
5641 15. Minor typos in src/EDITME comments corrected.
5643 16. Installed latest exipick from John Jetmore.
5645 17. If headers_add on a router specified a text string that was too long for
5646 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5647 of string_sprintf() is now avoided.
5649 18. $message_body_size was not set (it was always zero) when running the DATA
5650 ACL and the local_scan() function.
5652 19. For the "mail" command in an Exim filter, no default was being set for
5653 the once_repeat time, causing a random time value to be used if "once" was
5654 specified. (If the value happened to be <= 0, no repeat happened.) The
5655 default is now 0s, meaning "never repeat". The "vacation" command was OK
5656 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5657 (I found it when inspecting the code).
5659 20. There is now an overall timeout for performing a callout verification. It
5660 defaults to 4 times the callout timeout, which applies to individual SMTP
5661 commands during the callout. The overall timeout applies when there is more
5662 than one host that can be tried. The timeout is checked before trying the
5663 next host. This prevents very long delays if there are a large number of
5664 hosts and all are timing out (e.g. when the network connections are timing
5665 out). The value of the overall timeout can be changed by specifying an
5666 additional sub-option for "callout", called "maxwait". For example:
5668 verify = sender/callout=5s,maxwait=20s
5670 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5671 to the end before writing, but this should make it even safer).
5673 22. Exim was forgetting that it had advertised PIPELINING for the second and
5674 subsequent messages on an SMTP connection. It was also not resetting its
5675 memory on STARTTLS and an internal HELO.
5677 23. When Exim logs an SMTP synchronization error within a session, it now
5678 records whether PIPELINING has been advertised or not.
5680 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5681 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5682 rather than long int.
5684 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5686 26. Added the never_mail option to autoreply.
5692 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5693 crash if the getsockname() call failed; this can happen if a connection is
5694 closed very soon after it is established. The problem was simply in the
5695 order in which certain operations were done, causing Exim to try to write
5696 to the SMTP stream before it had set up the file descriptor. The bug has
5697 been fixed by making things happen in the correct order.
5703 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5704 before the connection was closed, thus losing the rejection response.
5706 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5707 some early Solaris releases, but causes trouble in current releases where
5708 socklen_t is defined.
5710 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5713 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5716 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5717 that releasing the top of it at the end releases what was used for sub-
5718 expansions (unless the block got too big). However, discard this block if
5719 the first thing is a variable or header, so that we can use its block when
5720 it is dynamic (useful for very large $message_headers, for example).
5722 6. Lookups now cache *every* query, not just the most recent. A new, separate
5723 store pool is used for this. It can be recovered when all lookup caches are
5724 flushed. Lookups now release memory at the end of their result strings.
5725 This has involved some general refactoring of the lookup sources.
5727 7. Some code has been added to the store_xxx() functions to reduce the amount
5728 of flapping under certain conditions.
5730 8. log_incoming_interface used to affect only the <= reception log lines. Now
5731 it causes the local interface and port to be added to several more SMTP log
5732 lines, for example "SMTP connection from", and rejection lines.
5734 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5736 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5738 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5739 could be overwritten at the end of the current message (or the start of a
5740 new message if it was set in a HELO ACL). The value is now preserved for
5741 the duration of the SMTP connection.
5743 12. If a transport had a headers_rewrite setting, and a matching header line
5744 contained an unqualified address, that address was qualified, even if it
5745 did not match any rewriting rules. The underlying bug was that the values
5746 of the flags that permit the existence of unqualified sender and recipient
5747 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5748 non-local messages, and by -bnq for local messages) were not being
5749 preserved with the message after it was received.
5751 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5752 "next input=" as part of the text comprising the host identity instead of
5753 the correct text. The code was using the same buffer for two different
5754 strings. However, depending on which order the printing function evaluated
5755 its arguments, the bug did not always show up. Under Linux, for example, my
5756 test suite worked just fine.
5758 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5759 doesn't work with some older versions of Perl. It has been changed to "my",
5760 which in any case is probably the better facility to use.
5762 15. A really picky compiler found some instances of statements for creating
5763 error messages that either had too many or two few arguments for the format
5766 16. The size of the buffer for calls to the DNS resolver has been increased
5767 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5768 for addresses that have a lot of PTR records. This alleviates a problem; it
5769 does not fully solve it.
5771 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5772 buffer now truncates the list and logs the incident, which is the same
5773 action as happens when Exim is looking up a host name and its aliases.
5774 Previously in this situation something unpredictable would happen;
5775 sometimes it was "internal error: store_reset failed".
5777 18. If a server dropped the connection unexpectedly when an Exim client was
5778 using GnuTLS and trying to read a response, the client delivery process
5779 crashed while trying to generate an error log message.
5781 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5782 string, for example:
5784 warn message = H1: something\nH2: something
5786 the text was added as a single header line from Exim's point of view
5787 though it ended up OK in the delivered message. However, searching for the
5788 second and subsequent header lines using $h_h2: did not work. This has been
5789 fixed. Similarly, if a system filter added multiple headers in this way,
5790 the routers could not see them.
5792 20. Expanded the error message when iplsearch is called with an invalid key to
5793 suggest using net-iplsearch in a host list.
5795 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5796 ACLs are no longer actually imposed (and a message to that effect is
5799 22. If a "gecos" field in a passwd entry contained escaped characters, in
5800 particular, if it contained a \" sequence, Exim got it wrong when building
5801 a From: or a Sender: header from that name. A second bug also caused
5802 incorrect handling when an unquoted " was present following a character
5803 that needed quoting.
5805 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5806 was not being matched caselessly.
5808 24. Arranged for all hyphens in the exim.8 source to be escaped with
5811 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5812 properly. Recipient callout cache records were still being keyed to include
5813 the sender, even when use_sender was set false. This led to far more
5814 callouts that were necessary. The sender is no longer included in the key
5815 when use_sender is false.
5817 26. Added "control = submission" modifier to ACLs.
5819 27. Added the ${base62d: operator to decode base 62 numbers.
5821 28. dnsdb lookups can now access SRV records.
5823 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5824 the configuration file.
5826 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5827 (-v) mode. This makes the output for a verbose queue run more intelligible.
5829 31. Added a use_postmaster feature to recipient callouts.
5831 32. Added the $body_zerocount variable, containing the number of binary zero
5832 bytes in the message body.
5834 33. The time of last modification of the "new" subdirectory is now used as the
5835 "mailbox time last read" when there is a quota error for a maildir
5838 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5840 35. Added +ignore_unknown as a special item in host lists.
5842 36. Code for decoding IPv6 addresses in host lists is now included, even if
5843 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5844 address was recognized as an IP address, but was then not correctly decoded
5845 into binary, causing unexpected and incorrect effects when compared with
5852 1. Very minor rewording of debugging text in manualroute to say "list of
5853 hosts" instead of "hostlist".
5855 2. If verify=header_syntax was set, and a header line with an unqualified
5856 address (no domain) and a large number of spaces between the end of the
5857 name and the colon was received, the reception process suffered a buffer
5858 overflow, and (when I tested it) crashed. This was caused by some obsolete
5859 code that should have been removed. The fix is to remove it!
5861 3. When running in the test harness, delay a bit after writing a bounce
5862 message to get a bit more predictability in the log output.
5864 4. Added a call to search_tidyup() just before forking a reception process. In
5865 theory, someone could use a lookup in the expansion of smtp_accept_max_
5866 per_host which, without the tidyup, could leave open a database connection.
5868 5. Added the variables $recipient_data and $sender_data which get set from a
5869 lookup success in an ACL "recipients" or "senders" condition, or a router
5870 "senders" option, similar to $domain_data and $local_part_data.
5872 6. Moved the writing of debug_print from before to after the "senders" test
5875 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5876 problems for message scanning, either using a data ACL, or using
5877 local_scan() because the Received: header was not generated till after they
5878 were called (in order to set the time as the time of reception completion).
5879 I have revised the way this works. The header is now generated after the
5880 body is received, but before the ACL or local_scan() are called. After they
5881 are run, the timestamp in the header is updated.
5887 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5888 before starting a queue runner without re-exec. This happened only when
5889 deliver_drop_privilege was set or when the Exim user was set to root. The
5890 effect of the bug was that timeouts during subsequent deliveries caused
5891 crashes instead of being properly handled. The handler is now left at its
5892 default (and expected) setting.
5894 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5895 message, again when deliver_drop_privilege is set or Exim is run as root.
5896 The bug described in (1) was not present in this case, but the tidying up
5897 of the other signals was missing. I have made the two cases consistent.
5899 3. The ignore_target_hosts setting on a manualroute router was being ignored
5900 for hosts that were looked up using the /MX notation.
5902 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5905 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5906 operated on the sender address. After changing the $sender_address to <>
5907 for the sender address verify, Exim was re-instated it as the original
5908 (before rewriting) address, but remembering that it had rewritten it, so it
5909 wasn't rewriting it again. This bug also had the effect of breaking the
5910 sender address verification caching when the sender address was rewritten.
5912 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5913 This has been changed so that if the ip literal address matches
5914 ignore_target_hosts, the router declines.
5916 7. Added expansion conditions match_domain, match_address, and match_local_
5917 part (NOT match_host).
5919 8. The placeholder for the Received: header didn't have a length field set.
5921 9. Added code to Exim itself and to exim_lock to test for a specific race
5922 condition that could lead to file corruption when using MBX delivery. The
5923 issue is with the lockfile that is created in /tmp. If this file is removed
5924 after a process has opened it but before that process has acquired a lock,
5925 there is the potential for a second process to recreate the file and also
5926 acquire a lock. This could lead to two Exim processes writing to the file
5927 at the same time. The added code performs the same test as UW imapd; it
5928 checks after acquiring the lock that its file descriptor still refers to
5929 the same named file.
5931 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5932 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5935 11. Added the smtp_active_hostname option. If used, this will typically be made
5936 to depend on the incoming interface address. Because $interface_address is
5937 not set up until the daemon has forked a reception process, error responses
5938 that can happen earlier (such as "too many connections") no longer contain
5941 12. If an expansion in a condition on a "warn" statement fails because a lookup
5942 defers, the "warn" statement is abandoned, and the next ACL statement is
5943 processed. Previously this caused the whole ACL to be aborted.
5945 13. Added the iplsearch lookup type.
5947 14. Added ident_timeout as a log selector.
5949 15. Added tls_certificate_verified as a log selector.
5951 16. Added a global option tls_require_ciphers (compare the smtp transport
5952 option of the same name). This controls incoming TLS connections.
5954 17. I finally figured out how to make tls_require_ciphers do a similar thing
5955 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5956 before starting the TLS session.
5958 18. Tabs are now shown as \t in -bP output.
5960 19. If the log selector return_path_on_delivery was set, Exim crashed when
5961 bouncing a message because it had too many Received: header lines.
5963 20. If two routers both had headers_remove settings, and the first one included
5964 a superfluous trailing colon, the final name in the first list and the
5965 first name in the second list were incorrectly joined into one item (with a
5966 colon in the middle).
5972 1. Added -C and -D options to the exinext utility, mainly to make it easier
5973 to include in the automated testing, but these could be helpful when
5974 multiple configurations are in use.
5976 2. The exinext utility was not formatting the output nicely when there was
5977 an alternate port involved in the retry record key, nor when there was a
5978 message id as well (for retries that were specific to a specific message
5979 and a specific host). It was also confused by IPv6 addresses, because of
5980 the additional colons they contain. I have fixed the IPv4 problem, and
5981 patched it up to do a reasonable job for IPv6.
5983 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5984 delivery, the log line now contains "pipelined" if PIPELINING was used.
5986 4. An SMTP transport process used to panic and die if the bind() call to set
5987 an explicit outgoing interface failed. This has been changed; it is now
5988 treated in the same way as a connect() failure.
5990 5. A reference to $sender_host_name in the part of a conditional expansion
5991 that was being skipped was still causing a DNS lookup. This no longer
5994 6. The def: expansion condition was not recognizing references to header lines
5995 that used bh_ and bheader_.
5997 7. Added the _cache feature to named lists.
5999 8. The code for checking quota_filecount in the appendfile transport was
6000 allowing one more file than it should have been.
6002 9. For compatibility with Sendmail, the command line option
6010 and sets the incoming protocol and host name (for trusted callers). The
6011 host name and its colon can be omitted when only the protocol is to be set.
6012 Note the Exim already has two private options, -pd and -ps, that refer to
6013 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6014 "s", but I don't think that's a major issue.
6016 10. A number of refactoring changes to the code, none of which should affect
6019 (a) The number of logging options was getting close to filling up the
6020 32-bit word that was used as a bit map. I have split them into two classes:
6021 those that are passed in the argument to log_write(), and those that are
6022 only ever tested independently outside of that function. These are now in
6023 separate 32-bit words, so there is plenty of room for expansion again.
6024 There is no change in the user interface or the logging behaviour.
6026 (b) When building, for example, log lines, the code previously used a
6027 macro that called string_cat() twice, in order to add two strings. This is
6028 not really sufficiently general. Furthermore, there was one instance where
6029 it was actually wrong because one of the argument was used twice, and in
6030 one call a function was used. (As it happened, calling the function twice
6031 did not affect the overall behaviour.) The macro has been replaced by a
6032 function that can join an arbitrary number of extra strings onto a growing
6035 (c) The code for expansion conditions now uses a table and a binary chop
6036 instead of a serial search (which was left over from when there were very
6037 few conditions). Also, it now recognizes conditions like "pam" even when
6038 the relevant support is not compiled in: a suitably worded error message is
6039 given if an attempt is made to use such a condition.
6041 11. Added ${time_interval:xxxxx}.
6043 12. A bug was causing one of the ddress fields not to be passed back correctly
6044 from remote delivery subprocesses. The field in question was not being
6045 subsequently used, so this caused to problems in practice.
6047 13. Added new log selectors queue_time and deliver_time.
6049 14. Might have fixed a bug in maildirsizefile handling that threw up
6050 "unexpected character" debug warnings, and recalculated the data
6051 unnecessarily. In any case, I expanded the warning message to give more
6054 15. Added the message "Restricted characters in address" to the statements in
6055 the default ACL that block characters like @ and % in local parts.
6057 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6058 Three changes have been made:
6060 (a) There was a serious bug; a negative response to MAIL caused the whole
6061 recipient domain to be cached as invalid, thereby blocking all messages
6062 to all local parts at the same domain, from all senders. This bug has
6063 been fixed. The domain is no longer cached after a negative response to
6064 MAIL if the sender used is not empty.
6066 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6069 (c) A new callout option, "use_sender" has been added for people who want
6070 the modified behaviour.
6076 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6079 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6080 indeed breaks things for older releases.
6082 3. Added additional logging to the case where there is a problem reading data
6083 from a filter that is running in a subprocess using a pipe, in order to
6084 try to track down a specific problem.
6086 4. Testing facility fudge: when running in the test harness and attempting
6087 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6088 getting "No route to host". Convert this to a timeout.
6090 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6093 6. Some OS don't have socklen_t but use size_t instead. This affects the
6094 fifth argument of getsockopt() amongst other things. This is now
6095 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6096 can be set for individual OS. I have set it for SunOS5, OSF1, and
6097 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6098 some earlier ones do not.
6100 7. Change 4.30/15 was not doing the test caselessly.
6102 8. The standard form for an IPv6 address literal was being rejected by address
6103 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6104 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6105 this, as well as the form without the "IPv6" on the front (but only when
6106 address literals are enabled, of course).
6108 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6110 10. Exim crashed if a message with an empty sender address specified by -f
6111 encountered a router with an errors_to setting. This could be provoked only
6112 by a command such as
6116 where an empty string was supplied; "<>" did not hit this bug.
6118 11. Installed PCRE release 4.5.
6120 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6121 remained set. It is now erased.
6123 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6124 times from message ids (which are base 36 rather than the normal 62).
6126 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6127 were being counted as actual protocol errors, and logged if the log
6128 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6129 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6130 and DATA following a set of rejected RCPTs do not count as protocol errors.
6131 In other words, Exim assumes they were pipelined, though this may not
6132 actually be the case. Of course, in all cases the client gets an
6133 appropriate error code.
6135 15. If a lookup fails in an ACL condition, a message about the failure may
6136 be available; it is used if testing the ACL cannot continue, because most
6137 such messages specify what the cause of the deferral is. However, some
6138 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6139 that caused an old message to be retained and used if a later statement
6140 caused a defer, replacing the real cause of the deferral.
6142 16. If an IP address had so many PTR records that the DNS lookup buffer
6143 was not large enough to hold them, Exim could crash while trying to process
6144 the truncated data. It now detects and logs this case.
6146 17. Further to 4.21/58, another change has been made: if (and only if) the
6147 first line of a message (the first header line) ends with CRLF, a bare LF
6148 in a subsequent header line has a space inserted after it, so as not to
6149 terminate the header.
6151 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6152 unnecessarily, used atoi() instead of strtol(), and didn't check the
6153 termination when getting file sizes from file names by regex.
6155 19. Completely re-implemented the support for maildirsize files, in the light
6156 of a number of problems with the previous contributed implementation
6157 (4.30/29). In particular:
6159 . If the quota is zero, the maildirsize file is maintained, but no quota is
6162 . If the maildir directory does not exist, it is created before any attempt
6163 to write a maildirsize file.
6165 . The quota value in the file is just a cache; if the quota is changed in
6166 the transport, the new value overrides.
6168 . A regular expression is available for excluding directories from the
6171 20. The autoreply transport checks the characters in options that define the
6172 message's headers; it allows continued headers, but it was checking with
6173 isspace() after an embedded newline instead of explicitly looking for a
6176 21. If all the "regular" hosts to which an address was routed had passed their
6177 expiry times, and had not reached their retry times, the address was
6178 bounced, even if fallback hosts were defined. Now Exim should go on to try
6181 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6182 equivalent code in the SMTP transport. Some hosts send humungous responses
6183 to HELO/EHLO, more than 1024 it seems.
6185 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6186 gives compiler warnings in some environments. I've now done it "properly",
6189 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6190 (because of problems with the built-in one) was declared to return uschar *
6191 instead of char *, causing compiler failure.
6193 25. Fixed a file descriptor leak when processing alias/forward files.
6195 26. Fixed a minor format string issue in dbfn.c.
6197 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6199 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6200 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6203 29. When a sender address is verified, it is cached, to save repeating the test
6204 when there is more than one recipient in a message. However, when the
6205 verification involves a callout, it is possible for different callout
6206 options to be set for different recipients. It is too complicated to keep
6207 track of this in the cache, so now Exim always runs a verification when a
6208 callout is required, relying on the callout cache for the optimization.
6209 The overhead is duplication of the address routing, but this should not be
6212 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6213 to be verified with callout=postmaster, and the main callout worked but the
6214 postmaster check failed, the verification correctly failed. However, if a
6215 subsequent RCPT command asked for sender verification *without* the
6216 postmaster check, incorrect caching caused this verification also to fail,
6219 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6220 it was not caching the DNS options (qualify_single, search_parents) that
6221 were used when the lookup failed. A subsequent lookup with different
6222 options therefore always gave the same answer, though there were cases
6223 where it should not have. (Example: a "domains = !$mx_any" option on a
6224 dnslookup router: the "domains" option is always processed without any
6225 widening, but the router might have qualify_single set.) Now Exim uses the
6226 cached value only when the same options are set.
6228 32. Added John Jetmore's "exipick" utility to the distribution.
6230 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6231 than a timeout (e.g. a certificate is required, and is not provided), an
6232 Exim server now closes the connection immediately. Previously it waited for
6233 the client to close - but if the client is SSL, it seems that they each
6234 wait for each other, leading to a delay before one of them times out.
6236 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6237 maintained 0.8.x compatibility because I don't think many are using it, and
6238 it is clearly obsolete.
6240 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6243 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6244 client certificate was expired. A simple patch fixes this, though I don't
6245 understand the full logic of why the verify callback is called multiple
6248 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6249 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6250 which causes problems with some clients (such as the Certicom SSL Plus
6251 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6252 disables the coutermeasure allowing Eudora to connect."
6254 38. Exim was not checking that a write() to a log file succeeded. This could
6255 lead to Bad Things if a log got too big, in particular if it hit a file
6256 size limit. Exim now panics and dies if it cannot write to a log file, just
6257 as it does if it cannot open a log file.
6259 39. Modified OS/Makefile-Linux so that it now contains
6261 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6263 The two -D definitions ensure that Exim is compiled with large file
6264 support, which makes it possible to handle log files that are bigger than
6267 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6268 instance) a domain was checked against a named list that involved a lookup,
6269 causing $domain_data to be set, then another domain was checked against the
6270 same list, then the first domain was re-checked, the value of $domain_data
6271 after the final check could be wrong. In particular, if the second check
6272 failed, it could be set empty. This bug probably also applied to
6275 41. The strip_trailing_dot option was not being applied to the address given
6276 with the -f command-line option.
6278 42. The code for reading a message's header from the spool was incrementing
6279 $received_count, but never initializing it. This meant that the value was
6280 incorrect (doubled) while delivering a message in the same process in which
6281 it was received. In the most common configuration of Exim, this never
6282 happens - a fresh exec is done - but it can happen when
6283 deliver_drop_privilege is set.
6285 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6286 it now includes up to 150 characters of the unexpected data in the log
6289 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6290 and building data strings. The size of both of these buffers was 10 000
6291 bytes - far larger than anybody would *ever* want, thought I. Needless to
6292 say, somebody hit the limit. I have increased the maximum line length to
6293 20 000 and the maximum data length of concatenated lines to 100 000. I have
6294 also fixed two bugs, because there was no checking on these buffers. Tsk,
6295 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6296 buffer is too small.
6298 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6299 lsearch lookups. Now it does.
6301 46. When parsing a route_list item in a manualroute router, a fixed-length
6302 buffer was used for the list of hosts. I made this 1024 bytes long,
6303 thinking that nobody would ever have a list of hosts that long. Wrong.
6304 Somebody had a whole pile of complicated expansion conditions, and the
6305 string was silently truncated, leading to an expansion error. It turns out
6306 that it is easier to change to an unlimited length (owing to other changes
6307 that have happened since this code was originally written) than to build
6308 structure for giving a limitation error. The length of the item that
6309 expands into the list of hosts is now unlimited.
6311 47. The lsearch lookup could not handle data where the length of text line was
6312 more than 4095 characters. Such lines were truncated, leading to shortened
6313 data being returned. It should now handle lines of any length.
6315 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6316 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6319 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6320 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6321 no longer happens in comment lines. A list of replacements is now placed
6322 at the head of all of the source files, except those whose only change is
6323 to replace PERL_COMMAND in the very first #! line.
6325 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6326 messages on the queue, with a bottom-up merge sort, using code contributed
6327 by Michael Haardt. This should make operations like -bp somewhat faster on
6328 large queues. It won't affect queue runners, except when queue_run_in_order
6331 51. Installed eximstats 1.31 in the distribution.
6333 52. Added support for SRV lookups to the dnslookup router.
6335 53. If an ACL referred to $message_body or $message_body_end, the value was not
6336 reset for any messages that followed in the same SMTP session.
6338 54. The store-handling optimization for building very long strings was not
6339 differentiating between the different store pools. I don't think this
6340 actually made any difference in practice, but I've tidied it.
6342 55. While running the routers to verify a sender address, $sender_address
6343 was still set to the sender address. This is wrong, because when routing to
6344 send a bounce to the sender, it would be empty. Therefore, I have changed
6345 it so that, while verifying a sender address, $sender_address is set to <>.
6346 (There is no change to what happens when verifying a recipient address.)
6348 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6349 target A or AAAA records (if not already returned) without resetting the
6350 qualify_single or search_parents options of the DNS resolver. These are
6351 inappropriate in this case because the targets of MX and SRV records must
6352 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6353 target that, when qualified, matched something in the local domain. These
6354 two options are now turned off when doing these lookups.
6356 57. It seems that at least some releases of Reiserfs (which does not have the
6357 concept of a fixed number of inodes) returns zero and not -1 for the
6358 number of available inodes. This interacted badly with check_spool_inodes,
6359 which assumed that -1 was the "no such thing" setting. What I have done is
6360 to check that the total number of inodes is greater than zero before doing
6361 the test of how many are available.
6363 58. When a "warn" ACL statement has a log_message modifier, the message is
6364 remembered, and not repeated. This is to avoid a lot of repetition when a
6365 message has many recipients that cause the same warning to be written.
6366 However, Exim was preserving the list of already written lines for an
6367 entire SMTP session, which doesn't seem right. The memory is now reset if a
6368 new message is started.
6370 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6371 debugging output unless log_rewrite was also set.
6373 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6374 of (int)(handle) when we know that handle contains (void *)(-1).
6376 61. The Exim daemon panic-logs an error return when it closes the incoming
6377 connection. However "connection reset by peer" seems to be common, and
6378 isn't really an error worthy of noting specially, so that particular error
6381 62. When Exim is trying to find all the local interfaces, it used to panic and
6382 die if the ioctl to get the interface flags failed. However, it seems that
6383 on at least one OS (Solaris 9) it is possible to have an interface that is
6384 included in the list of interfaces, but for which you get a failure error
6385 for this call. This happens when the interface is not "plumbed" into a
6386 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6387 failure of the "get flags" call assumes that the interface is down.
6389 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6390 makes life easier for people who are doing arithmetic on fields extracted
6391 from dates, where you often get leading zeros that should not be
6392 interpreted as octal.
6394 64. Added qualify_domain to the redirect router, to override the global
6397 65. If a pathologically long header line contained very many addresses (the
6398 report of this problem mentioned 10 000) and each of them was rewritten,
6399 Exim could use up a very large amount of memory. (It kept on making new
6400 copies of the header line as it rewrote, and never released the old ones.)
6401 At the expense of a bit more processing, the header rewriting function has
6402 been changed so that it no longer eats memory in this way.
6404 66. The generation of the Received: header has been moved from the time that a
6405 message starts to be received, to the time that it finishes. The timestamp
6406 in the Received: header should now be very close to that of the <= log
6407 line. There are two side-effects of this change:
6409 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6410 logged header lines no longer include the local Received: line, because
6411 it has not yet been created. The same applies to a copy of the message
6412 that is returned to a non-SMTP sender when a message is rejected.
6414 (b) When a filter file is tested using -bf, no additional Received: header
6415 is added to the test message. After some thought, I decided that this
6418 This change does not affect the value of $received_for. It is still set
6419 after address rewriting, but before local_scan() is called.
6421 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6423 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6424 gave an unhelpful panic error message, and a defer error. I have managed to
6425 change this behaviour so that it now rejects any supplied certificate,
6426 which seems right, as the list of acceptable certificates is empty.
6428 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6429 gave an unhelpful defer error. I have not managed to make this reject any
6430 supplied certificates, but the error message it gives is "no certificate
6431 supplied", which is not helpful.
6433 70. exigrep's output now also includes lines that are not associated with any
6434 message, but which match the given pattern. Implemented by a patch from
6435 Martin Sluka, which also tidied up the Perl a bit.
6437 71. Recipient callout verification, like sender verification, was using <> in
6438 the MAIL FROM command. This isn't really the right thing, since the actual
6439 sender may affect whether the remote host accepts the recipient or not. I
6440 have changed it to use the actual sender in the callout; this means that
6441 the cache record is now keyed on a recipient/sender pair, not just the
6442 recipient address. There doesn't seem to be a real danger of callout loops,
6443 since a callout by the remote host to check the sender would use <>.
6444 [SEE ABOVE: changed after hitting problems.]
6446 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6447 temporary errors. However, in the case of such a code being given after
6448 the end of a data transmission (i.e. after ".") Exim was failing to write
6449 a retry record for the message. (Yes, there was some broken host that was
6450 actually sending 8xx at this point.)
6452 73. An unknown lookup type in a host list could cause Exim to panic-die when
6453 the list was checked. (An example that provoked this was putting <; in the
6454 middle of a list instead of at the start.) If this happened during a DATA
6455 ACL check, a -D file could be left lying around. This kind of configuration
6456 error no longer causes Exim to die; instead it causes a defer error. The
6457 incident is still logged to the main and panic logs.
6459 74. Buglet left over from Exim 3 conversion. The message "too many messages
6460 in one connection" was written to the rejectlog but not the mainlog, except
6461 when address rewriting (yes!) was being logged.
6463 75. Added write_rejectlog option.
6465 76. When a system filter was run not as root (that is, when system_filter_user
6466 was set), the values of the $n variables were not being returned to the
6467 main process; thus, they were not subsequently available in the $sn
6470 77. Added +return_path_on_delivery log selector.
6472 78. A connection timeout was being treated differently from recipients deferred
6473 when testing hosts_max_try with a message that was older than the host's
6474 retry timeout. (The host should not be counted, thus allowing all hosts to
6475 be tried at least once before bouncing.) This may have been the cause of an
6476 occasionally reported bug whereby a message would remain on the queue
6477 longer than the retry timeout, but would be bounced if a delivery was
6478 forced. I say "may" because I never totally pinned down the problem;
6479 setting up timeout/retry tests is difficult. See also the next item.
6481 79. The ultimate address timeout was not being applied to errors that involved
6482 a combination of host plus message (for example, a timeout on a MAIL
6483 command). When an address resolved to a number of possible hosts, and they
6484 were not all tried for each delivery (e.g. because of hosts_max_try), a
6485 message could remain on the queue longer than the retry timeout.
6487 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6490 81. Fixed an obscure SMTP outgoing bug which required at least the following
6491 conditions: (a) there was another message waiting for the same server;
6492 (b) the server returned 5xx to all RCPT commands in the first message so
6493 that the message was not completed; (c) the server dropped the connection
6494 or gave a negative response to the RSET that Exim sends to abort the
6495 transaction. The observed case was a dropped connection after DATA that had
6496 been sent in pipelining mode. That is, the server had advertised PIPELINING
6497 but was not implementing it correctly. The effect of the bug was incorrect
6498 behaviour, such as trying another host, and this could lead to a crash.
6504 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6505 and daemon.c were passed as pointers to ints; they should have been
6506 pointers to socklen_t variables (which are typically unsigned ints).
6508 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6511 3. Fixed a really odd bug that affected only the testing scheme; patching a
6512 certain fixed string in the binary changed the value of another string that
6513 happened to be identical to the end of the original first string.
6515 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6516 name", it returns that address as the IP address. On some operating
6517 systems (e.g. Solaris), it also passes back the IP address string as the
6518 "host name". However, on others (e.g. Linux), it passes back an empty
6519 string. Exim wasn't checking for this, and was changing the host name to an
6520 empty string, assuming it had been canonicalized.
6522 5. Although rare, it is permitted to have more than one PTR record for a given
6523 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6524 all the names associated with an address, because they do in Solaris.
6525 However, it seems that they do not in Linux for data that comes from the
6526 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6527 I found this out when I moved to a new Linux workstation and tried to run
6528 the Exim test suite.
6530 To get round this problem I have changed the code so that it now does its
6531 own call to the DNS to look up PTR records when searching for a host name.
6532 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6533 addresses that are only in /etc/hosts are still found.
6535 This behaviour is, however, controlled by an option called host_lookup_
6536 order, which defaults to "bydns:byaddr". If people want to use the other
6537 order, or indeed, just use one or the other means of lookup, they can
6538 specify it in this variable.
6540 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6541 some operating systems, this comes back from gethostbyaddr() as an empty
6542 string, and this is what Exim used to test for. However, it seems that in
6543 other systems, "." is yielded. Exim now tests for this case too.
6545 7. The values of check_spool_space and check_log_space are now held internally
6546 as a number of kilobytes instead of an absolute number of bytes. If a
6547 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6548 kilobyte. This means that much larger values can be stored.
6550 8. Exim monitor: an attempt to get the action menu when not actually pointing
6551 at a message produces an empty menu entitled "No message selected". This
6552 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6553 no entries in it ("Shell widget menu has zero width and/or height"). So I
6554 have added a single, blank menu entry in this case.
6556 9. Added ${quote_local_part.
6558 10. MIME decoding is now applied to the contents of Subject: header lines when
6561 11. Now that a reference to $sender_host_address automatically causes a reverse
6562 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6563 host lookup before query-style lookups in lists that might use this
6564 variable. This has therefore been abolished, and the "net-" prefix is no
6565 longer necessary for query-style lookups.
6567 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6568 have been a typo for LFLAGS, so it has been changed.
6570 13. The install script calls Exim with "-C /dev/null" in order to find the
6571 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6572 to be output. However, since Exim outputs its version number before the
6573 error, it didn't break the script. It just looked ugly. I fixed this by
6574 always allowing "-C /dev/null" if the caller is root.
6576 14. Ignore overlarge ACL variable number when reading spool file - insurance
6577 against a later release with more variables having written the file.
6579 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6580 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6581 this, as well as the form without the "IPv6" on the front.
6583 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6584 OS/Makefile-Darwin file.
6586 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6587 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6589 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6590 variable. However, some OS use ints for inodes. I've added cast to long int
6591 to get rid of the compiler warning.
6593 19. I had forgotten to lock out "/../" in configuration file names when
6594 ALT_CONFIG_PREFIX was set.
6596 20. Routers used for verification do not need to specify transports. However,
6597 if such a router generated a host list, and callout was configured, Exim
6598 crashed, because it could not find a port number from the (non-existent)
6599 transport. It now assumes port 25 in this circumstance.
6601 21. Added the -t option to exigrep.
6603 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6604 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6605 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6606 them is set, LOOKUP_LSEARCH is forced.
6608 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6610 24. Added sender and host information to the "rejected by local_scan()" log
6611 line; previously there was no indication of these.
6613 25. Added .include_if_exists.
6615 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6616 Linux. It turns out that not all file systems support this. Apparently some
6617 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6618 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6619 not supported on the file descriptor, is now ignored when Exim is trying to
6620 sync a directory. This applies only to Linux.
6622 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6624 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6625 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6628 29. Added support for maildirsize files from supplied patch (modified a bit).
6630 30. The use of :fail: followed by an empty string could lead Exim to respond to
6631 sender verification failures with (e.g.):
6633 550 Verification failed for <xxx>
6634 550 Sender verify failed
6636 where the first response line was missing the '-' that indicates it is not
6637 the final line of the response.
6639 31. The loop for finding the name of the user that called Exim had a hardwired
6640 limit of 10; it now uses the value of finduser_retries, which is used for
6641 all other user lookups.
6643 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6646 33. Exim was neglecting to zero errno before one call of strtol() when
6647 expanding a string and expecting an integer value. On some systems this
6648 resulted in spurious "integer overflow" errors. Also, it was casting the
6649 result into an int without checking.
6651 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6652 did not work. The code looks as if it has *never* worked, though it appears
6653 to have been documented since at least release 1.62. I have made it work.
6655 35. The "timeout_DNS" error in retry rules, also documented since at least
6656 1.62, also never worked. As it isn't clear exactly what this means, and
6657 clearly it isn't a major issue, I have abolished the feature by treating it
6658 as "timeout", and writing a warning to the main and panic logs.
6660 36. The display of retry rules for -brt wasn't always showing the error code
6663 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6664 timeout_connect_A, timeout_connect_MX.
6666 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6667 to the empty sender.
6669 39. The daemon was not analysing the content of -oX till after it had closed
6670 stderr and disconnected from the controlling terminal. This meant that any
6671 syntax errors were only noted on the panic log, and the return code from
6672 the command was 0. By re-arranging the code a little, I've made the
6673 decoding happen first, so such errors now appear on stderr, and the return
6674 code is 1. However, the actual setting up of the sockets still happens in
6675 the disconnected process, so errors there are still only recorded on the
6678 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6679 connections (as happens on some IP stacks) was logged at start up time as
6680 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6681 it from "IPv6 and IPv4", which means that two separate sockets are being
6684 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6685 says whether AF_INET or AF_INET6 was passed as an argument.
6687 42. Exiwhat output was messed up when time zones were included in log
6690 43. Exiwhat now gives more information about the daemon's listening ports,
6691 and whether -tls-on-connect was used.
6693 44. The "port" option of the smtp transport is now expanded.
6695 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6696 silently ignored. Now an error message is written to the main and panic
6699 46. There's a new ACL modifier called "logwrite" which writes to a log file
6700 as soon as it is encountered.
6702 47. Added $local_user_uid and $local_user_gid at routing time.
6704 48. Exim crashed when trying to verify a sender address that was being
6707 49. Exim was recognizing only a space character after ".include". It now also
6708 recognizes a tab character.
6710 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6711 extracting the relevant information from the specification. The man page no
6712 longer contains scrambled data for the -d option, and I've added a section
6713 at the front about calling Exim under different names.
6715 51. Added "extra_headers" argument to the "mail" command in filter files.
6717 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6720 53. Installed eximstats 1.29.
6722 54. Added transport_filter_timeout as a generic transport option.
6724 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6725 Cc: header lines. This was required by RFC 822, but it not required by RFC
6728 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6729 incoming messages that did not have them. Now it does so only if the
6730 message originates locally, that is, if there is no associated remote host
6731 address. When Resent- header lines are present, this applies to the Resent-
6732 lines rather than the non-Resent- lines.
6734 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6735 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6737 58. Messages for configuration errors now include the name of the main
6738 configuration file - useful now that there may be more than one file in a
6739 list (.included file names were always shown).
6741 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6742 for those rare installations that do not start the daemon as root or run it
6743 setuid root. I've cut out the call to initgroups() if the daemon is not
6746 60. The Exim user and group can now be bound into the binary as text strings
6747 that are looked up at the start of Exim's processing.
6749 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6751 62. Added $mailstore_basename variable.
6753 63. Installed patch to sieve.c from Michael Haardt.
6755 64. When Exim failed to open the panic log after failing to open the main log,
6756 the original message it was trying to log was written to stderr and debug
6757 output, but if they were not available (the usual case in production), it
6758 was lost. Now it is written to syslog before the two lines that record the
6759 failures to open the logs.
6761 65. Users' Exim filters run in subprocesses under the user's uid. It is
6762 possible for a "deliver" command or an alias in a "personal" command to
6763 provoke an address rewrite. If logging of address rewriting is configured,
6764 this fails because the process is not running as root or exim. There may be
6765 a better way of dealing with this, but for the moment (because 4.30 needs
6766 to be released), I have disabled address rewrite logging when running a
6767 filter in a non-root, non-exim process.
6773 1. The buildconfig auxiliary program wasn't quoting the value set for
6774 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6775 not defined. This bug was masked in 4.22 by the effect that was fixed in
6778 2. Some messages that were rejected after a message id was allocated were
6779 shown as "incomplete" by exigrep. It no longer does this for messages that
6780 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6782 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6783 have allow_domain_literals set, the ID did not get logged in the <= line.
6784 Domain literals are now always recognized in Message-ID: header lines.
6786 4. The first argument for a ${extract expansion item is the key name or field
6787 number. Leading and trailing spaces in this item were not being ignored,
6788 causing some misleading effects.
6790 5. When deliver_drop_privilege was set, single queue runner processes started
6791 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6792 same command in the process it spins off) were not dropping privilege.
6794 6. When the daemon running as "exim" started a queue runner, it always
6795 re-executed Exim in the spun-off process. This is a waste of effort when
6796 deliver_drop_privilege is set. The new process now just calls the
6797 queue-runner function directly.
6803 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6806 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6807 ignored. Though the use of -oP was forcing the writing of a pid file, it
6808 was always written to the default place.
6810 3. If the message "no IP address found for host xxxx" is generated during
6811 incoming verification, it is now followed by identification of the incoming
6812 connection (so you can more easily find what provoked it).
6814 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6816 5. Added some features to "harden" Exim a bit more against certain attacks:
6818 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6819 be put in Local/Makefile. This is like the never_users runtime option,
6820 but it cannot be overridden. The default setting is "root".
6822 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6823 prefix string with which any file named in a -C command line option
6826 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6827 is retained for -C and -D only if the caller of Exim is root. Without
6828 it, the exim user may also use -C and -D and retain privilege.
6830 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6831 command line option is disabled.
6833 6. Macro names set by the -D option must start with an upper case letter, just
6834 like macro names defined in the configuration file.
6836 7. Added "dereference=" facility to LDAP.
6838 8. Two instances of the typo "uknown" in the source files are fixed.
6840 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6841 the Configure-Makefile script screwed up while processing it.
6843 10. Incorporated PCRE 4.4.
6845 11. The SMTP synchronization check was not operating right at the start of an
6846 SMTP session. For example, it could not catch a HELO sent before the client
6847 waited for the greeting. There is now a check for outstanding input at the
6848 point when the greeting is written. Because of the duplex, asynchronous
6849 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6850 way, but not yet received, when the check is performed.
6852 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6853 on TCP/IP sockets, because this apparently causes some broken clients to
6856 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6857 unchanged) from the Cygwin maintainer.
6859 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6860 was supported instead of testing for mbx. Effectively a typo.
6862 15. The spa authenticator server code was not checking that the input it
6863 received was valid base64.
6865 16. The debug output line for the "set" modifier in ACLs was not showing the
6866 name of the variable that was being set.
6868 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6870 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6871 lookup on its own. Something else was needed to trigger the lookup. For
6872 example, a match in host_lookup or the need for a host name in a host list.
6873 Now, if $sender_host_name is referenced and the host name has not yet been
6874 looked up, a lookup is performed. If the lookup fails, the variable remains
6875 empty, and $host_lookup_failed is set to "1".
6877 19. Added "eqi" as a case-independent comparison operator.
6879 20. The saslauthd authentication condition could segfault if neither service
6880 nor realm was specified.
6882 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6883 error message that was logged was misleading, and incoming SMTP
6884 connections were dropped. The message is now more accurate, and temporary
6885 errors are given to SMTP connections.
6887 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6888 (except RSET) with a 421 error, until QUIT is received. However, it was
6889 failing to send a response to QUIT.
6891 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6892 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6893 ACL is now run after the helo_try_verify_hosts code.
6895 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6896 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6897 case-independent, so other case variants are also recognized). Apparently
6898 some systems use these upper case variants.
6900 25. If more than two messages were waiting for the same host, and a transport
6901 filter was specified for the transport, Exim sent two messages over the
6902 same TCP/IP connection, and then failed with "socket operation on non-
6903 socket" when it tried to send the third.
6905 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6907 27. The extern definition of crypt16() in expand.c was not being excluded when
6908 the OS had its own crypt16() function.
6910 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6911 as a preferred synonym for return_size_limit, both as an option and as an
6914 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6916 30. Changed the default configuration ACL to relax the local part checking rule
6917 for addresses that are not in any local domains. For these addresses,
6918 slashes and pipe symbols are allowed within local parts, but the sequence
6919 /../ is explicitly forbidden.
6921 31. SPA server authentication was not clearing the challenge buffer before
6924 32. log_message in a "warn" ACL statement was writing to the reject log as
6925 well as to the main log, which contradicts the documentation and doesn't
6926 seem right (because no rejection is happening). So I have stopped it.
6928 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6929 However, I am unable to do any testing of this.
6931 34. Fixed an infelicity in the appendfile transport. When checking directories
6932 for a mailbox, to see if any needed to be created, it was accidentally
6933 using path names with one or more superfluous leading slashes; tracing
6934 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6936 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6937 contents are added to the log line that is written for every discarded
6938 recipient. (Previously a log_message setting was ignored.)
6940 36. The ${quote: operator now quotes the string if it is empty.
6942 37. The install script runs exim in order to find its version number. If for
6943 some reason other than non-existence or emptiness, which it checks, it
6944 could not run './exim', it was installing it with an empty version number,
6945 i.e. as "exim-". This error state is now caught, and the installation is
6948 38. An argument was missing from the function that creates an error message
6949 when Exim fails to connect to the socket for saslauthd authentication.
6950 This could cause Exim to crash, or give a corrupted message.
6952 39. Added isip, isip4, and isip6 to ${if conditions.
6954 40. The ACL variables $acl_xx are now saved with the message, and can be
6955 accessed later in routers, transports, and filters.
6957 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6958 strings in the file are not string-expanded.
6960 42. If a MAIL command specified a SIZE value that was too large to fit into an
6961 int variable, the check against message_size_limit failed. Such values are
6962 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6963 day this will have to be increased, but I don't think I want to be around
6964 when emails are that large.
6971 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6972 iconv() is not standard in FreeBSD.
6974 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6975 IPv6 enabled. The observed symptom was a segmentation fault on return from
6976 the function os_common_find_running_interfaces() in src/os.c.
6978 3. In the check_special_case() function in daemon.c I had used "errno" as an
6979 argument name, which causes warnings on some systems. This was basically a
6980 typo, since it was named "eno" in the comments!
6982 4. The code that waits for the clock to tick (at a resolution of some fraction
6983 of a second) so as to ensure message-id uniqueness was always waiting for
6984 at least one whole tick, when it could have waited for less. [This is
6985 almost certainly not relevant at current processor speeds, where it is
6986 unlikely to ever wait at all. But we try to future-proof.]
6988 5. The function that sleeps for a time interval that includes fractions of a
6989 second contained a race. It did not block SIGALRM between setting the
6990 timer, and suspending (a couple of lines later). If the interval was short
6991 and the sigsuspend() was delayed until after it had expired, the suspension
6992 never ended. On busy systems this could lead to processes getting stuck for
6995 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6996 process, before it forks any delivery processes. The open lookup caching
6997 mechanism meant that the open file or database connection was passed into
6998 the delivery process. The problem was that delivery processes always tidy
6999 up cached lookup data. This could cause a problem for the next delivery
7000 process started by the queue runner, because the external queue runner
7001 process does not know about the closure. So the next delivery process
7002 still has data in the lookup cache. In the case of a file lookup, there was
7003 no problem because closing a file descriptor in a subprocess doesn't affect
7004 the parent. However, if the lookup was caching a connection to a database,
7005 the connection was closed, and the second delivery process was likely to
7006 see errors such as "PGSQL: query failed: server closed the connection
7007 unexpectedly". The problem has been fixed by closing all cached lookups
7008 in a queue runner before running a delivery process.
7010 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7011 seem to have the "const" qualifier which it has on other OS. I've
7014 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7015 *of the same type* (client or server) with the same public name that an
7016 error should be diagnosed.
7018 9. When Exim looked up a host name for an IP address, but failed to find the
7019 original IP address when looking up the host name (a safety check), it
7020 output the message "<ip address> does not match any IP for NULL", which was
7021 confusing, to say the least. The bug was that the host name should have
7022 appeared instead of "NULL".
7024 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7025 user that is built into the binary, and the -C or -D options is used, root
7026 privilege is dropped before the configuration file is read. In addition,
7027 logging is switched to stderr instead of the normal log files. If the
7028 configuration then re-defines the Exim user, the unprivileged environment
7029 is probably not what is expected, so Exim logs a panic warning message (but
7032 However, if deliver_drop_privilege is set, the unprivileged state may well
7033 be exactly what is intended, so the warning has been cut out in that case,
7034 and Exim is allowed to try to write to its normal log files.
7040 1. smtp_return_error_details was not giving details for temporary sender
7041 or receiver verification errors.
7043 2. Diagnose a configuration error if two authenticators have the same public
7046 3. Exim used not to create the message log file for a message until the first
7047 delivery attempt. This could be confusing when incoming messages were held
7048 for policy or load reasons. The message log file is now created at the time
7049 the message is received, and an initial "Received" line is written to it.
7051 4. The automatically generated man page for command line options had a minor
7052 bug that caused no ill effects; however, a more serious problem was that
7053 the procedure for building the man page automatically didn't always
7054 operate. Consequently, release 4.20 contains an out-of-date version. This
7055 shouldn't happen again.
7057 5. When building Exim with embedded Perl support, the script that builds the
7058 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7059 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7061 6. The freeze_tell option was not being used for messages that were frozen on
7062 arrival, either by an ACL or by local_scan().
7064 7. Added the smtp_incomplete_transaction log selector.
7066 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7067 was accepting AUTH without a new EHLO.
7069 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7070 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7073 10. Logging of TCP/IP connections (when configured) now happens in the main
7074 daemon process instead of the child process, so that the TCP/IP connection
7075 count is more accurate (but it can never be perfect).
7077 11. The use of "drop" in a nested ACL was not being handled correctly in the
7078 outer ACL. Now, if condition failure induced by the nested "drop" causes
7079 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7080 or "require"), the connection is dropped.
7082 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7083 that yield "discard" can now be used with an "accept" or a "discard" verb,
7084 but an error is generated for any others (because I can't see a useful way
7085 to define what should happen).
7087 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7088 are now processed in the same way as lines in the Exim configuration file.
7089 In particular, continuation lines are supported.
7091 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7093 15. Added -ti meaning -t -i.
7095 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7096 domains, and warn by logging if others are found.
7098 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7099 returned from GIFCONF when Exim is trying to find the list of interfaces on
7100 a host. The code in os.c has been modified to copy each ifreq to an aligned
7101 structure in all cases.
7103 Also, in some cases, the returned ifreq's were being copied to a 'struct
7104 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7105 means the last couple of bytes of an IPv6 address could be chopped if the
7106 ifreq contained only a normal sockaddr (14 bytes storage).
7108 18. Named domain lists were not supported in the hosts_treat_as_local option.
7109 An entry such as +xxxx was not recognized, and was treated as a literal
7112 19. Ensure that header lines added by a DATA ACL are included in the reject log
7113 if the ACL subsequently rejects the message.
7115 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7116 MD5 (which is deprecated).
7118 21. When testing a filter file using -bf, Exim was writing a message when it
7119 took the sender from a "From " line in the message, but it was not doing so
7120 when it took $return_path from a Return-Path: header line. It now does.
7122 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7123 with a valid header line field name (a series of printing characters
7124 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7126 23. Changed "disc" in the source to "disk" to conform to the documentation and
7127 the book and for uniformity.
7129 24. Ignore Sendmail's -Ooption=value command line item.
7131 25. When execve() failed while trying to run a command in a pipe transport,
7132 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7133 could be confused with a return value of 69 from the command itself. This
7134 has been changed to 127, the value the shell returns if it is asked to run
7135 a non-existent command. The wording for the related log line suggests a
7136 non-existent command as the problem.
7138 26. If received_header_text expands to an empty string, do not add a Received:
7139 header line to the message. (Well, it adds a token one on the spool, but
7140 marks it "old" so that it doesn't get used or transmitted.)
7142 27. Installed eximstats 1.28 (addition of -nt option).
7144 28. There was no check for failure on the call to getsockname() in the daemon
7145 code. This can fail if there is a shortage of resources on the system, with
7146 ENOMEM, for example. A temporary error is now given on failure.
7148 29. Contrary to the C standard, it seems that in some environments, the
7149 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7150 program. Exim now does this explicitly; it affects the formatting of
7151 timestamps using strftime().
7153 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7154 complaints. I've now initialized all the variables, to avoid this.
7156 32. Header lines added by a system filter were not being "seen" during
7157 transport-time rewrites.
7159 33. The info_callback() function passed to OpenSSL is set up with type void
7160 (*)(SSL *, int, int), as described somewhere. However, when calling the
7161 function (actually a macro) that sets it up, the type void(*)() is
7162 expected. I've put in a cast to prevent warnings from picky compilers.
7164 34. If a DNS black list lookup found a CNAME record, but there were no A
7165 records associated with the domain it pointed at, Exim crashed.
7167 35. If a DNS black list lookup returned more than one A record, Exim ignored
7168 all but the first. It now scans all returned addresses if a particular IP
7169 value is being sought. In this situation, the contents of the
7170 $dnslist_value variable are a list of all the addresses, separated by a
7173 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7174 to accept a host name and all its aliases if the forward lookup for any of
7175 them yielded the IP address of the incoming connection. Now it accepts only
7176 those names whose forward lookup yields the correct IP address. Any other
7177 names are discarded. This closes a loophole whereby a rogue DNS
7178 administrator could create reverse DNS records to break through a
7179 wildcarded host restriction in an ACL.
7181 37. If a user filter or a system filter that ran in a subprocess used any of
7182 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7183 the wrong values were passed to the pipe command ($thisaddress had the
7184 value of $0, $0 had the value of $1, etc). This bug was introduced by
7185 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7187 38. Improved the line breaking for long SMTP error messages from ACLs.
7188 Previously, if there was no break point between 40 and 75 characters, Exim
7189 left the rest of the message alone. Two changes have been made: (a) I've
7190 reduced the minimum length to 35 characters; (b) if it can't find a break
7191 point between 35 and 75 characters, it looks ahead and uses the first one
7192 that it finds. This may give the occasional overlong line, but at least the
7193 remaining text gets split now.
7195 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7196 file descriptors might be low, and that setting 1000 would always raise it.
7197 It turns out that in some environments, the limit is already over 1000 and
7198 that lowering it causes trouble. So now Exim takes care not to decrease it.
7200 40. When delivering a message, the value of $return_path is set to $sender_
7201 address at the start of routing (routers may change the value). By an
7202 oversight, this default was not being set up when an address was tested by
7203 -bt or -bv, which affected the outcome if any router or filter referred to
7206 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7207 log only when "message" or "log_message" are set. However, if one of the
7208 conditions was an address verification, or a call to a nested ACL, the
7209 messages generated by the underlying test were being passed through. This
7210 no longer happens. The underlying message is available in $acl_verify_
7211 message for both "message" and "log_message" expansions, so it can be
7212 passed through if needed.
7214 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7215 new expansion $bh_ to give the encoded byte string without charset
7216 translation. Translation happens only if iconv() is available; HAVE_ICONV
7217 indicates this at build time. HEADERS_CHARSET gives the charset to
7218 translate to; headers_charset can change it in the configuration, and
7219 "headers charset" can change it in an individual filter file.
7221 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7222 that creates RFC 2047 encoded "words" labels them as that charset instead
7223 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7224 expansion operator; (ii) when Exim creates a From: line for a local
7225 message; (iii) when a header line is rewritten to include a "phrase" part.
7227 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7228 buggy, causing it to skip the first lines of messages whose message ID
7229 ended in 'D'. This would not have bitten before Exim release 4.14, because
7230 message IDs were unlikely to end in 'D' before then. The effect was to have
7231 incorrect size information for certain domains.
7233 45. #include "config.h" was missing at the start of the crypt16.c module. This
7234 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7237 46. If there was a timeout during a "random" callout check, Exim treated it as
7238 a failure of the random address, and carried on sending RSET and the real
7239 address. If the delay was just some slowness somewhere, the response to the
7240 original RCPT would be taken as a response to RSET and so on, causing
7241 mayhem of various kinds.
7243 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7244 when I implemented it. It didn't allow for the fact that some option values
7245 may legitimately be negative (e.g. size_addition), and it didn't even do
7246 the right test for positive values.
7248 48. Domain names in DNS records are case-independent. Exim always looks them up
7249 in lower case. Some resolvers return domain names in exactly the case they
7250 appear in the zone file, that is, they may contain uppercase letters. Not
7251 all resolvers do this - some return always lower case. Exim was treating a
7252 change of case by a resolver as a change of domain, similar to a widening
7253 of a domain abbreviation. This triggered its re-routing code and so it was
7254 trying to route what was effectively the same domain again. This normally
7255 caused routing to fail (because the router wouldn't handle the domain
7256 twice). Now Exim checks for this case specially, and just changes the
7257 casing of the domain that it ultimately uses when it transmits the message
7260 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7263 50. If a filter generated a file delivery with a non-absolute name (possible if
7264 no home directory exists for the router), the forbid_file option was not
7267 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7268 the existing equality matching.
7270 52. Exim was using ints instead of ino_t variables in some places where it was
7271 dealing with inode numbers.
7273 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7274 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7275 called TMPDIR, and if it finds it is different, it changes its value.
7277 54. The smtp_printf() function is now made available to local_scan() so
7278 additional output lines can be written before returning. There is also an
7279 smtp_fflush() function to enable the detection of a dropped connection.
7280 The variables smtp_input and smtp_batched_input are exported to
7283 55. Changed the default runtime configuration: the message "Unknown user"
7284 has been removed from the ACL, and instead placed on the localuser router,
7285 using the cannot_route_message feature. This means that any verification
7286 failures that generate their own messages won't get overridden. Similarly,
7287 the "Unrouteable address" message that was in the ACL for unverifiable
7288 relay addresses has also been removed.
7290 56. Added hosts_avoid_esmtp to the smtp transport.
7292 57. The exicyclog script was not checking for the esoteric option
7293 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7294 will work only if exicyclog is run under the appropriate euid.
7296 58. Following a discussion on the list, the rules by which Exim recognises line
7297 endings on incoming messages have been changed. The -dropcr and drop_cr
7298 options are now no-ops, retained only for backwards compatibility. The
7299 following line terminators are recognized: LF CRLF CR. However, special
7300 processing applies to CR:
7302 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7303 nor a local message in the state where . is a terminator.
7305 (ii) If a bare CR is encountered in a header line, an extra space is added
7306 after the line terminator so as not to end the header. The reasoning
7307 behind this is that bare CRs in header lines are most likely either
7308 to be mistakes, or people trying to play silly games.
7310 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7311 was being incorrectly given as 18 bytes larger than it should have been.
7312 This is a VOB (very old bug).
7314 60. This may never have affected anything current, but just in case it has:
7315 When the local host is found other than at the start of a list of hosts,
7316 the local host, those with the same MX, and any that follow, are discarded.
7317 When the list in question was part of a longer list of hosts, the following
7318 hosts (not currently being processed) were also being discarded. This no
7319 longer happens. I'm not sure if this situation could ever has previously
7322 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7325 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7326 header. This is something that is recommended in a new Internet Draft, and
7327 is something that is documented as being done by Sendmail. There are two
7328 possible values. For messages generated by the autoreply transport, Exim
7331 Auto-Submitted: auto-replied
7333 whereas for all other generated messages (e.g. bounces) it adds
7335 Auto-Submitted: auto-generated
7337 63. The "personal" condition in filters now includes a test for the
7338 Auto-Submitted: header. If it contains the string "auto-" the message it
7339 not considered personal.
7341 64. Added rcpt_include_affixes as a generic transport option.
7343 65. Added queue_only_override (default true).
7345 66. Added the syslog_duplication option.
7347 67. If what should have been the first header line of a message consisted of
7348 a space followed by a colon, Exim was mis-interpreting it as a header line.
7349 It isn't of course - it is syntactically invalid and should therefore be
7350 treated as the start of the message body. The misbehaviour could have
7351 caused a number of strange effects, including loss of data in subsequent
7352 header lines, and spool format errors.
7354 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7355 client host had authenticated. This control can now be exercised by an ACL
7356 for more flexibility.
7358 69. By default, callouts do not happen when testing with -bh. There is now a
7359 variant, -bhc, which does actually run the callout code, including
7360 consulting and updating the callout cache.
7362 70. Added support for saslauthd authentication, courtesy of Alexander
7365 71. If statvfs() failed on the spool or log directories while checking their
7366 size for availability, Exim confusingly gave the error "space shortage".
7367 Furthermore, in debugging mode it crashed with a floating point exception.
7368 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7369 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7370 serious problem, Exim now writes to the main and panic logs when this
7371 happens, with details of the failure. It then refuses to accept the
7372 incoming message, giving the message "spool directory problem" or "log
7373 directory problem" with a 421 code for SMTP messages.
7375 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7376 0, 1, and 2 exist, because some OS complain for execs without them (see
7377 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7378 descriptors. However, the code omitted to check that the open succeeded,
7379 causing mysterious errors if for some reason the permissions on /dev/null
7380 got screwed. Now Exim writes a message to the main and panic logs, and
7381 bombs out if it can't open /dev/null.
7383 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7384 interact so that it is all more flexible. It is supposed to remain
7385 backwards compatible. Also added extra_local_interfaces.
7387 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7388 to bomb out with an assertion failure - to the client this appears as a
7389 connection drop. This problem occurs in the part of the code that was taken
7390 from the Samba project. Fortunately, the assertion is in a very simple
7391 function, so I have fixed this by reproducing the function inline in the
7392 one place where it is called, and arranging for authentication to fail
7393 instead of killing the process with assert().
7395 75. The SPA client code was not working when the server requested OEM rather
7396 than Unicode encoding.
7398 76. Added code to make require_files with a specific uid setting more usable in
7399 the case where statting the file as root fails - usually a non-root-mounted
7400 NFS file system. When this happens and the failure is EACCES, Exim now
7401 forks a subprocess and does the per-uid checking as the relevant uid.
7403 77. Added process_log_path.
7405 78. If log_file_path was not explicitly set, a setting of check_log_space or
7406 check_log_inodes was ignored.
7408 79. If a space check for the spool or log partitions fails, the incident is now
7409 logged. Of course, in the latter case the data may get lost...
7411 80. Added the %p formatting code to string_format() so that it can be used to
7412 print addresses in debug_print(). Adjusted all the address printing in the
7413 debugging in store.c to use %p rather than %d.
7415 81. There was a concern that a line of code in smtp_in.c could overflow a
7416 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7417 initially expressed, the concern was not well-founded, because trailing
7418 spaces are removed early. However, if the trailing spaces were followed by
7419 a NULL, they did not get removed, so the overflow was possible. Two fixes
7422 (a) I re-wrote the offending code in a cleaner fashion.
7423 (b) If an incoming SMTP command contains a NULL character, it is rejected
7426 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7427 runs initgroups(), so that if the Exim user is in any additional groups,
7428 they will be used during message reception.
7434 The change log for 4.20 and earlier releases has been archived.