1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917, OVE-20190718-0006)
17 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
18 buffer overrun for (non-chunking) other transports.
20 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
21 TLS1.3, means that a server rejecting a client certificate is not visible
22 to the client until the first read of encrypted data (typically the
23 response to EHLO). Add detection for that case and treat it as a failed
24 TLS connection attempt, so that the normal retry-in-clear can work (if
27 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
28 and/or domain. Found and fixed by Jason Betts.
30 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
31 configuration). If a CNAME target was not a wellformed name pattern, a
34 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
35 the OS reports them interleaved with other addresses.
37 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
38 used both for input and for a verify callout, both encrypted, SMTP
39 responses being sent by the server could be lost. This resulted in
40 dropped connections and sometimes bounces generated by a peer sending
43 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
44 string. Previously it was possible to cause undefined behaviour in a
45 library routine (usually a crash). Found by "zerons".
47 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
50 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
51 API was removed, so update to use the newer ones.
58 JH/01 Remove code calling the customisable local_scan function, unless a new
59 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
61 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
62 non-signal-safe functions being used.
64 JH/03 Bug 2269: When presented with a received message having a stupidly large
65 number of DKIM-Signature headers, disable DKIM verification to avoid
66 a resource-consumption attack. The limit is set at twenty.
68 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
69 report of oldest_pass in ${authres } in consequence, and separate out
70 some descriptions of reasons for verification fail.
72 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
73 files in the spool were present and unlocked. A queue-runner could spot
74 them, resulting in a duplicate delivery. Fix that by doing the unlock
75 after the unlink. Investigation by Tim Stewart. Take the opportunity to
76 add more error-checking on spoolfile handling while that code is being
79 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
80 No known attacks, no CVE, this is defensive hardening.
82 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
83 a queue-runner could start a delivery while other operations were ongoing.
84 Cutthrough delivery was a common victim, resulting in duplicate delivery.
85 Found and investigated by Tim Stewart. Fix by using the open message data
86 file handle rather than opening another, and not locally closing it (which
87 releases a lock) for that case, while creating the temporary .eml format
88 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
90 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
91 $sender_verify_failure/$recipient_verify_failure to "random".
93 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
96 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
97 Previously this would segfault.
99 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
102 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
103 like zero, since the resolver should be doing this for us, But we need one
104 as a CNAME but no MX presence gets the CNAME returned; we need to check
105 that doesn't point to an MX to declare it "no MX returned" rather than
106 "error, loop". A new main option is added so the older capability of
107 following some limited number of chain links is maintained.
109 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
111 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
112 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
113 now-deprecated earlier definitions used only specified the range up to TLS
114 1.2 (in the older-version library docs).
116 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
118 JH/15 Rework TLS client-side context management. Stop using a global, and
119 explicitly pass a context around. This enables future use of TLS for
120 connections to service-daemons (eg. malware scanning) while a client smtp
121 connection is using TLS; with cutthrough connections this is quite likely.
123 JH/16 Fix ARC verification to do AS checks in reverse order.
125 JH/17 Support a "tls" option on the ${readsocket } expansion item.
127 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
128 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
129 Previously the "utf8" would be re-prepended for every additional message.
131 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
132 Previously thery were accepted, resulting in issues when attempting to
133 forward messages to a non-supporting MTA.
135 PP/02 Let -n work with printing macros too, not just options.
137 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
138 one parent address was copied, and bogus data was used at delivery-logging
139 time. Either a crash (after delivery) or bogus log data could result.
140 Discovery and analysis by Tim Stewart.
142 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
143 Previously if the string ended mid-character, we did not insert the
144 promised '?' replacement.
146 PP/04 Documentation: current string operators work on bytes, not codepoints.
148 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
149 should pack well giving a smaller memory footprint so better caching and
150 therefore performance. Group the declarations where this can't be done so
151 that the byte-sized flag variables are not interspersed among pointer
152 variables, giving a better chance of good packing by the compiler.
154 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
155 non-null, to avoid issues with sites running BATV. Previously reports were
156 sent with an empty envelope sender so looked like bounces.
158 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
159 The ignore_error flag wasn't being returned from the filter subprocess so
160 was not set for later routers. Investigation and fix by Matthias Kurz.
162 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
163 and a msg:complete for the whole, when a message is manually removed using
164 -Mrm. Developement by Matthias Kurz, hacked on by JH.
166 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
167 a "Gnu special" function, asprintf() in the DB utility binary builds; I
168 hope that is portable enough.
170 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
171 requiring a known-CA anchor certificate; make it now rely entirely on the
172 TLSA as an anchor. Checking the name on the leaf cert against the name
173 on the A-record for the host is still done for TA (but not for EE mode).
175 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
176 included in delivery lines for non-proxied connections, when compiled with
177 SUPPORT_SOCKS and running with proxy logging enabled.
179 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
180 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
181 move the existing event to fire before the normal logging of message
182 failure so that custom logging is bracketed by normal logging.
184 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
185 msg:fail:internal event. Developement by Matthias Kurz.
187 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
188 far too small for todays use of crypto signatures stored there. Go all
189 the way to the max DNS message size of 64kB, even though this might be
190 overmuch for IOT constrained device use.
192 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
193 copy a string over itself. The library routine is documented as not
194 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
196 JH/32 For main options check_spool_space and check_inode_space, where the
197 platform supports 64b integers, support more than the previous 2^31 kB
198 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
199 the previous G, M, k.
201 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
202 $authenticated_fail_id variable on authentication failure. Previously
205 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
206 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
207 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
208 GNUTLS_SEC_PARAM_MEDIUM.
210 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
211 side. Previously we would continue as if no SNI had been received.
213 JH/36 Harden the handling of string-lists. When a list consisted of a sole
214 "<" character, which should be a list-separator specification, we walked
215 off past the nul-terimation.
217 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
218 causes) even when the retry time is not yet met. Previously they were
219 not, meaning that when (say) an account was over-quota and temp-rejecting,
220 and multiple senders' messages were queued, only one sender would get
221 notified on each configured delay_warning cycle.
223 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
225 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
226 cipher-suites, an error can be left on the stack even for a succeeding
227 accept; this results in impossible error messages when a later operation
230 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
231 return error codes indicating retry. Under TLS1.3 this becomes required.
233 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
234 it only wrote the new authenticators, resulting in a lack of tracking of
235 peer changes of ESMTP extensions until the next cache flush.
237 JH/41 Fix the loop reading a message header line to check for integer overflow,
238 and more-often against header_maxsize. Previously a crafted message could
239 induce a crash of the recive process; now the message is cleanly rejected.
241 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
242 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
248 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
249 When redis_servers is set to a list of > 1 element, and the Redis servers
250 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
251 case of MOVED into a DEFER case instead, thus moving the query onto the
252 next server in the list. For a cluster of N elements, all N servers must
253 be defined in redis_servers.
255 GF/02 Catch and remove uninitialized value warning in exiqsumm
256 Check for existence of @ARGV before looking at $ARGV[0]
258 JH/01 Replace the store_release() internal interface with store_newblock(),
259 which internalises the check required to safely use the old one, plus
260 the allocate and data copy operations duplicated in both (!) of the
261 extant use locations.
263 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
264 modifier. This matches the restriction on the commandline.
266 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
267 Previously only the last row was returned.
269 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
270 we assumed that tags in the header were well-formed, and parsed the
271 element content after inspecting only the first char of the tag.
272 Assumptions at that stage could crash the receive process on malformed
275 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
276 While running the DKIM ACL we operate on the Permanent memory pool so that
277 variables created with "set" persist to the DATA ACL. Also (at any time)
278 DNS lookups that fail create cache records using the Permanent pool. But
279 expansions release any allocations made on the current pool - so a dnsdb
280 lookup expansion done in the DKIM ACL releases the memory used for the
281 DNS negative-cache, and bad things result. Solution is to switch to the
282 Main pool for expansions.
283 While we're in that code, add checks on the DNS cache during store_reset,
284 active in the testsuite.
285 Problem spotted, and debugging aided, by Wolfgang Breyha.
287 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
288 When none of the hosts presented to a transport match an already-open
289 connection, close it and proceed with the list. Previously we would
290 queue the message. Spotted by Lena with Yahoo, probably involving
293 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
294 Previously a spurious "250 OK id=" response was appended to the proper
297 JH/08 The "support for" informational output now, which built with Content
298 Scanning support, has a line for the malware scanner interfaces compiled
299 in. Interface can be individually included or not at build time.
301 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
302 by the template makefile "src/EDITME". The "STREAM" support for an older
303 ClamAV interface method is removed.
305 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
306 rows affected is given instead).
308 JH/11 The runtime Berkeley DB library version is now additionally output by
309 "exim -d -bV". Previously only the compile-time version was shown.
311 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
312 SMTP connection. Previously, when one had more receipients than the
313 first, an abortive onward connection was made. Move to full support for
314 multiple onward connections in sequence, handling cutthrough connection
315 for all multi-message initiating connections.
317 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
318 routers. Previously, a multi-recipient message would fail to match the
319 onward-connection opened for the first recipient, and cause its closure.
321 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
322 a timeout on read on a GnuTLS initiating connection, resulting in the
323 initiating connection being dropped. This mattered most when the callout
324 was marked defer_ok. Fix to keep the two timeout-detection methods
327 JH/15 Relax results from ACL control request to enable cutthrough, in
328 unsupported situations, from error to silently (except under debug)
329 ignoring. This covers use with PRDR, frozen messages, queue-only and
332 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
334 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
335 metadata, resulting in a crash in free().
337 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
338 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
339 Broken also in d185889f4, with init system revamp.
341 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
342 Previously we abruptly closed the connection after reading a malware-
343 found indication; now we go on to read the "scan ok" response line,
346 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
347 ACL. Previously, a crash would result.
349 JH/19 Speed up macro lookups during configuration file read, by skipping non-
350 macro text after a replacement (previously it was only once per line) and
351 by skipping builtin macros when searching for an uppercase lead character.
353 JH/20 DANE support moved from Experimental to mainline. The Makefile control
354 for the build is renamed.
356 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
357 was allocated for every new TLS startup, meaning one per message. Fix
358 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
360 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
361 reported the original. Fix to report (as far as possible) the ACL
362 result replacing the original.
364 JH/23 Fix memory leak during multi-message connections using STARTTLS under
365 OpenSSL. Certificate information is loaded for every new TLS startup,
366 and the resources needed to be freed.
368 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
370 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
373 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
374 DATA response info to the (existing) per-recipient response info for
375 the "C=" log element. It can have useful tracking info from the
376 destination system. Patch from Simon Arlott.
378 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
379 length value. Previously this would segfault.
381 HS/02 Support Avast multiline protoocol, this allows passing flags to
382 newer versions of the scanner.
384 JH/28 Ensure that variables possibly set during message acceptance are marked
385 dead before release of memory in the daemon loop. This stops complaints
386 about them when the debug_store option is enabled. Discovered specifically
387 for sender_rate_period, but applies to a whole set of variables.
388 Do the same for the queue-runner and queue-list loops, for variables set
389 from spool message files. Do the same for the SMTP per-message loop, for
390 certain variables indirectly set in ACL operations.
392 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
393 as a multi-recipient message from a mailinglist manager). The coding had
394 an arbitrary cutoff number of characters while checking for more input;
395 enforced by writing a NUL into the buffer. This corrupted long / fast
396 input. The problem was exposed more widely when more pipelineing of SMTP
397 responses was introduced, and one Exim system was feeding another.
398 The symptom is log complaints of SMTP syntax error (NUL chars) on the
399 receiving system, and refused recipients seen by the sending system
400 (propating to people being dropped from mailing lists).
401 Discovered and pinpointed by David Carter.
403 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
404 replaced by the ${authresults } expansion.
406 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
408 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
409 allows proper process termination in container environments.
411 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
412 Previously the "final dot" had a newline after it; ensure it is CR,LF.
414 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
415 and "err_perm", deprecated since 4.83 when the RFC-defined words
416 "temperror" and "permerror" were introduced.
418 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
419 transport-filters or DKIM-signing. The restriction was lost in the
420 consolidation of verify-callout and delivery SMTP handling.
421 Extend the restriction to also cover ARC-signing.
423 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
424 in defer=pass mode supply a 450 to the initiator. Previously the message
427 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
428 tls_require_ciphers is used as before.
430 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
431 "pass_unscanned". Only tmpfails from the scanner are written to
432 the paniclog, as they may require admin intervention (permission
433 denied, license issues). Other scanner errors (like decompression
434 bombs) do not cause a paniclog entry.
436 JH/36 Fix reinitialisation of DKIM logging variable between messages.
437 Previously it was possible to log spurious information in receive log
440 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
441 triggered odd behaviour from Outlook Express clients.
443 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
446 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
447 since the IETF WG has not yet settled on that versus the original
448 "bare" representation.
450 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
451 Previously the millisecond value corrupted the output.
452 Fix also for syslog_pid=no and log_selector +pid, for which the pid
453 corrupted the output.
459 JH/01 Rework error string handling in TLS interface so that the caller in
460 more cases is responsible for logging. This permits library-sourced
461 string to be attached to addresses during delivery, and collapses
462 pairs of long lines into single ones.
464 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
465 during configuration. Wildcards are allowed and expanded.
467 JH/02 Rework error string handling in DKIM to pass more info back to callers.
468 This permits better logging.
470 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
471 do not close it down and have the child transport start it up again on
472 the passed-on TCP connection. Instead, proxy the child (and any
473 subsequent ones) for TLS via a unix-domain socket channel. Logging is
474 affected: the continued delivery log lines do not have any DNSSEC, TLS
475 Certificate or OCSP information. TLS cipher information is still logged.
477 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
478 identical IP addresses on different listening ports. Will also affect
481 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
482 add noisy ifdef guards to special-case this sillyness.
483 Patch from Bernd Kuhls.
485 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
486 than 255 are no longer allowed.
488 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
489 Disable the session-cache too, which might reduce our load. Since we
490 currrectly use a new context for every connection, both as server and
491 client, there is no benefit for these.
492 GnuTLS appears to not support tickets server-side by default (we don't
493 call gnutls_session_ticket_enable_server()) but client side is enabled
494 by default on recent versions (3.1.3 +) unless the PFS priority string
497 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
498 <https://reproducible-builds.org/specs/source-date-epoch/>.
500 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
501 the check for any unsuccessful recipients did not notice the limit, and
502 erroneously found still-pending ones.
504 JH/08 Pipeline CHUNKING command and data together, on kernels that support
505 MSG_MORE. Only in-clear (not on TLS connections).
507 JH/09 Avoid using a temporary file during transport using dkim. Unless a
508 transport-filter is involved we can buffer the headers in memory for
509 creating the signature, and read the spool data file once for the
510 signature and again for transmission.
512 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
513 4.77 as the kernel support then wasn't solid, having issues in 64bit
514 mode. Now, it's been long enough. Add support for FreeBSD also.
516 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
517 case where the routing stage had gathered several addresses to send to
518 a host before calling the transport for the first, we previously failed
519 to close down TLS in the old transport process before passing the TCP
520 connection to the new process. The new one sent a STARTTLS command
521 which naturally failed, giving a failed delivery and bloating the retry
522 database. Investigation and fix prototype from Wolfgang Breyha.
524 JH/12 Fix check on SMTP command input synchronisation. Previously there were
525 false-negatives in the check that the sender had not preempted a response
526 or prompt from Exim (running as a server), due to that code's lack of
527 awareness of the SMTP input buffering.
529 PP/04 Add commandline_checks_require_admin option.
530 Exim drops privileges sanely, various checks such as -be aren't a
531 security problem, as long as you trust local users with access to their
532 own account. When invoked by services which pass untrusted data to
533 Exim, this might be an issue. Set this option in main configuration
534 AND make fixes to the calling application, such as using `--` to stop
537 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
538 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
539 responses to those, into a single TLS record each way (this usually means
540 a single packet). As a side issue, smtp_enforce_sync now works on TLS
543 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
544 affects you only if you're dancing at the edge of the param size limits.
545 If you are, and this message makes sense to you, then: raise the
546 configured limit or use OpenSSL 1.1. Nothing we can do for older
549 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
550 cmdline element to get the documented default one. Previously it was
553 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
554 are unsuable for non-message-specific reasons - eg. connection timeout,
555 banner-time rejection.
557 JH/16 Fix logging of delivery remote port, when specified by router, under
560 PP/06 Repair manualroute's ability to take options in any order, even if one
561 is the name of a transport.
564 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
566 JH/17 Change the list-building routines interface to use the expanding-string
567 triplet model, for better allocation and copying behaviour.
569 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
570 Previously it was constructed the first time a possibly-matching string
571 was met in the configuration file input during startup; now it is done
574 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
575 compatible one, to avoid the (poorly documented) possibility of a config
576 file in the working directory redirecting the DB files, possibly correpting
577 some existing file. CVE-2017-10140 assigned for BDB.
579 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
580 cache-hot. Previously, although the result was properly cached, the
581 initial verify call returned a defer.
583 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
584 the main verify for receipient in uncached-mode.
586 JH/22 Retire historical build files to an "unsupported" subdir. These are
587 defined as "ones for which we have no current evidence of testing".
589 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
590 if present. Previously it was ignored.
592 JH/24 Start using specified-initialisers in C structure init coding. This is
593 a C99 feature (it's 2017, so now considered safe).
595 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
596 if was a fixed-sized field and bitmask ops via macros; it is now more
599 PP/07 GitHub PR 56: Apply MariaDB build fix.
600 Patch provided by Jaroslav Škarvada.
602 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
603 during Coverity cleanups [4.87 JH/47]
604 Diagnosis and fix provided by Michael Fischer v. Mollard.
606 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
607 the right size to place the terminating semicolon on its own folded
608 line, the header hash was calculated to an incorrect value thanks to
609 the (relaxed) space the fold became.
611 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
612 and confused the parent.
614 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
615 which could crash as a result. This could lead to undeliverable messages.
617 JH/28 Logging: "next input sent too soon" now shows where input was truncated
620 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
621 matters on fast-turnover and PID-randomising systems, which were getting
622 out-of-order delivery.
624 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
625 a possibly-overlapping copy. The symptom was that "Remote host closed
626 connection in response to HELO" was logged instead of the actual 4xx
629 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
630 Previously only that bufferd was discarded, resulting in SYMTP command
633 JH/32 DKIM: when a message has multiple signatures matching an identity given
634 in dkim_verify_signers, run the dkim acl once for each. Previously only
635 one run was done. Bug 2189.
637 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
638 "panic the current process" to "deliberately defer". The panic log is
639 still written with the problem list name; the mail and reject logs now
640 get a temp-reject line for the message that was being handled, saying
641 something like "domains check lookup or other defer". The SMTP 451
642 message is still "Temporary local problem".
644 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
645 A crafted sequence of BDAT commands could result in in-use memory beeing
646 freed. CVE-2017-16943.
648 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
649 from SMTP input. Previously it was always done; now only done for DATA
650 and not BDAT commands. CVE-2017-16944.
652 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
653 to the message (such as an overlong header line). Previously this was
654 not done and we did not exit BDAT mode. Followon from the previous item
655 though a different problem.
661 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
662 than -2003 did; needs libidn2 in addition to libidn.
664 JH/02 The path option on a pipe transport is now expanded before use.
666 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
667 Patch provided by "Björn", documentation fix added too.
669 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
670 missing a wire-to-host endian conversion.
672 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
673 close after a BDAT command line could be taken as a following command,
674 giving a synch failure. Fix by only checking for synch immediately
675 before acknowledging the chunk.
677 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
678 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
679 macro. Patches provided by Josh Soref.
681 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
682 Previously we did not; the RFC seems ambiguous and VRFY is not listed
683 by IANA as a service extension. However, John Klensin suggests that we
686 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
687 the dkim code may be unix-mode line endings rather than smtp wire-format
688 CRLF, so prepend a CR to any bare LF.
690 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
691 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
693 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
694 the first were themselves being wrongly included in the feed into dkim
695 processing; with most chunk sizes in use this resulted in an incorrect
696 body hash calculated value.
698 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
699 DKIM signature block, for verification. Although advised against by
700 standards it is specifically not ruled illegal.
702 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
704 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
705 missing a body hash (the bh= tag).
707 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
708 It seems that HAProxy sends the Proxy Protocol information in clear and
709 only then does a TLS startup, so do the same.
711 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
712 TCP connections (such as for Spamd) unless the daemon successfully set
713 Fast Open mode on its listening sockets. This fixes breakage seen on
714 too-old kernels or those not configured for Fast Open, at the cost of
715 requiring both directions being enabled for TFO, and TFO never being used
716 by non-daemon-related Exim processes.
718 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
719 endings, at least on the first header line. Try to canonify any that get
720 past that check, despite the cost.
722 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
723 now limited to an arbitrary five deep, while parsing addresses with the
724 strip_excess_angle_brackets option enabled.
726 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
727 instead leave the unprompted TLS handshake in socket buffer for the
728 TLS library to consume.
730 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
732 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
734 JH/16 Drop variables when they go out of scope. Memory management drops a whole
735 region in one operation, for speed, and this leaves assigned pointers
736 dangling. Add checks run only under the testsuite which checks all
737 variables at a store-reset and panics on a dangling pointer; add code
738 explicitly nulling out all the variables discovered. Fixes one known
739 bug: a transport crash, where a dangling pointer for $sending_ip_address
740 originally assigned in a verify callout, is re-used.
742 PP/06 Drop '.' from @INC in various Perl scripts.
744 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
746 PP/08 Reduce a number of compilation warnings under clang; building with
747 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
748 should be warning-free.
750 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
752 HS/01 Fix portability problems introduced by PP/08 for platforms where
753 realloc(NULL) is not equivalent to malloc() [SunOS et al].
755 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
756 chunk. This allows us to accept broken chunked messages. We need a more
757 general solution here.
759 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
760 already-broken messages in the queue.
762 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
764 JH/19 Fix reference counting bug in routing-generated-address tracking.
770 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
771 supports it and a size is available (ie. the sending peer gave us one).
773 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
774 years of being deprecated). The replacements are the ACLs
775 acl_smtp_mime and acl_not_smtp_mime.
777 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
778 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
779 or even an in-clear connection were permitted. Now, if the host lookup
780 was dnssec and dane was requested then the host is only used if the
781 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
782 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
783 if one fails this test.
784 This means that a poorly-configured remote DNS will make it incommunicado;
785 but it protects against a DNS-interception attack on it.
787 JH/04 Bug 1810: make continued-use of an open smtp transport connection
788 non-noisy when a race steals the message being considered.
790 JH/05 If main configuration option tls_certificate is unset, generate a
791 self-signed certificate for inbound TLS connections.
793 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
794 in rewrites and routers.
796 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
797 and logged a warning sing 4.83; now they are a configuration file error.
799 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
800 (lacking @domain). Apply the same qualification processing as RCPT.
802 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
804 JH/10 Support ${sha256:} applied to a string (as well as the previous
807 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
808 a cutthrough deliver is pending, as we always want to make a connection.
809 This also avoids re-routing the message when later placing the cutthrough
810 connection after a verify cache hit.
811 Do not update it with the verify result either.
813 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
814 when routing results in more than one destination address.
816 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
817 signing (which inhibits the cutthrough capability). Previously only
818 the presence of an option was tested; now an expansion evaluating as
819 empty is permissible (obviously it should depend only on data available
820 when the cutthrough connection is made).
822 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
823 the relevant preceding SMTP command did not note the pipelining mode.
825 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
826 Previously they were not counted.
828 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
829 as one having no matching records. Previously we deferred the message
830 that needed the lookup.
832 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
833 distinguished as "(=".
835 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
836 for missing MX records. Previously it only worked for missing A records.
838 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
840 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
841 after the data-go-ahead and data-ack. Patch from Jason Betts.
843 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
844 even for a "none" policy. Patch from Tony Meyer.
846 JH/22 Fix continued use of a connection for further deliveries. If a port was
847 specified by a router, it must also match for the delivery to be
850 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
851 When one of the recipients of a message was unsuitable for the connection
852 (has no matching addresses), we lost track of needing to mark it
853 deferred. As a result mail would be lost.
855 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
857 JH/25 Decoding ACL controls is now done using a binary search; the source code
858 takes up less space and should be simpler to maintain. Merge the ACL
859 condition decode tables also, with similar effect.
861 JH/26 Fix problem with one_time used on a redirect router which returned the
862 parent address unchanged. A retry would see the parent address marked as
863 delivered, so not attempt the (identical) child. As a result mail would
866 JH/27 Fix a possible security hole, wherein a process operating with the Exim
867 UID can gain a root shell. Credit to http://www.halfdog.net/ for
868 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
871 JH/28 Enable {spool,log} filesystem space and inode checks as default.
872 Main config options check_{log,spool}_{inodes,space} are now
873 100 inodes, 10MB unless set otherwise in the configuration.
875 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
876 Previously it only applied to the main-section connection policy
879 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
881 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
882 by me. Added RFC7919 DH primes as an alternative.
884 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
885 are not in the system include path.
887 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
888 GnuTLS, when a session startup failed (eg because the client disconnected)
889 Exim did stdio operations after fclose. This was exposed by a recent
890 change which nulled out the file handle after the fclose.
892 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
893 signed directly by the cert-signing cert, rather than an intermediate
894 OCSP-signing cert. This is the model used by LetsEncrypt.
896 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
898 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
899 an incoming connection.
901 HS/02 Bug 1802: Do not half-close the connection after sending a request
904 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
905 fallback to "prime256v1".
907 JH/34 SECURITY: Use proper copy of DATA command in error message.
908 Could leak key material. Remotely exploitable. CVE-2016-9963.
914 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
915 and 3.4.4 - once the server is enabled to respond to an OCSP request
916 it does even when not requested, resulting in a stapling non-aware
917 client dropping the TLS connection.
919 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
920 support variable-length bit vectors. No functional change.
922 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
923 The I= interface field on outgoing lines is now after the H= remote
924 host field, same as incoming lines. There is a separate
925 outgoing_interface log selector which allows you to disable the
928 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
929 If not running log_selector +smtp_connection the mainlog would be held
930 open indefinitely after a "too many connections" event, including to a
931 deleted file after a log rotate. Leave the per net connection logging
932 leaving it open for efficiency as that will be quickly detected by the
933 check on the next write.
935 HS/01 Bug 1671: Fix post transport crash.
936 Processing the wait-<transport> messages could crash the delivery
937 process if the message IDs didn't exist for some reason. When
938 using 'split_spool_directory=yes' the construction of the spool
939 file name failed already, exposing the same netto behaviour.
941 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
942 mime_regex ACL conditions.
944 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
945 to DSN fail messages (bounces): remote IP, remote greeting, remote response
946 to HELO, local diagnostic string.
948 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
949 log line to debug. Even when configured with a tls authenticator many
950 client connections are expected to not authenticate in this way, so
951 an authenticate fail is not an error.
953 HS/02 Add the Exim version string to the process info. This way exiwhat
954 gives some more detail about the running daemon.
956 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
957 matter for fast-change records such as DNSBLs.
959 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
960 retry record, even if constant. There may be multiple transports with
961 different interface settings and the retry behaviour needs to be kept
964 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
966 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
968 JH/10 Bug 840: fix log_defer_output option of pipe transport
970 JH/11 Bug 830: use same host for all RCPTS of a message, even under
971 hosts_randomize. This matters a lot when combined with mua_wrapper.
973 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
974 ${quote_pgsql:<string>} operator.
976 JH/13 Bug 1708: avoid misaligned access in cached lookup.
978 JH/14 Change header file name for freeradius-client. Relevant if compiling
979 with Radius support; from the Gentoo tree and checked under Fedora.
981 JH/15 Bug 1712: Introduce $prdr_requested flag variable
983 JH/16 Bug 1714: Permit an empty string as expansion result for transport
984 option transport_filter, meaning no filtering.
986 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
988 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
989 defaults to "*" (all hosts). The variable is now available when not built
990 with TLS, default unset, mainly to enable keeping the testsuite sane.
991 If a server certificate is not supplied (via tls_certificate) an error is
992 logged, and clients will find TLS connections fail on startup. Presumably
993 they will retry in-clear.
994 Packagers of Exim are strongly encouraged to create a server certificate
995 at installation time.
997 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
998 with the $config_file variable.
1000 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1001 in transport context, after the attempt, and per-recipient. The latter type
1002 is per host attempted. The event data is the error message, and the errno
1003 information encodes the lookup type (A vs. MX) used for the (first) host,
1004 and the trailing two digits of the smtp 4xx response.
1006 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1007 to write to mainlog (or rejectlog, paniclog) in the window between file
1008 creation and permissions/ownership being changed. Particularly affects
1009 installations where exicyclog is run as root, rather than exim user;
1010 result is that the running daemon panics and dies.
1012 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1014 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1015 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1016 "pri" and "weight". Note that the previous implicit priority given by the
1017 list order is no longer honoured.
1019 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1020 for DKIM processing.
1022 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1023 by defining SUPPORT_SOCKS.
1025 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1026 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1027 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1028 variables are renamed to proxy_{local,external}_{address,port}.
1030 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1031 for a build by defining SUPPORT_I18N
1033 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1034 of the query string, and make ${quote_redis:} do that quoting.
1036 JH/29 Move Events support from Experimental to mainline, enabled by default
1037 and removable for a build by defining DISABLE_EVENT.
1039 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1041 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1042 cached by the daemon.
1044 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1045 by defining LOOKUP_REDIS. The libhiredis library is required.
1047 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1048 keys are given for lookup.
1050 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1051 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1052 only supported when built with TLS support. The PolarSSL SHA routines
1053 are still used when the TLS library is too old for convenient support.
1055 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1056 openssl_options), for security. OpenSSL forces this from version 1.1.0
1057 server-side so match that on older versions.
1059 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1060 allocation for $value could be released as the expansion processing
1061 concluded, but leaving the global pointer active for it.
1063 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1064 and to use the domains and local_parts ACL conditions.
1066 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1067 incorrectly not doubled on cutthrough transmission, hence seen as a
1068 body-termination at the receiving system - resulting in truncated mails.
1069 Commonly the sender saw a TCP-level error, and retransmitted the message
1070 via the normal store-and-forward channel. This could result in duplicates
1071 received - but deduplicating mailstores were liable to retain only the
1072 initial truncated version.
1074 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1076 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1078 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1079 we're in there, support oversigning also; bug 1309.
1081 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1083 HS/04 Add support for keep_environment and add_environment options.
1085 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1086 either intentional arithmetic overflow during PRNG, or testing config-
1089 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1090 delivery resulted in actual delivery. Cancel cutthrough before DATA
1093 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1094 rejects a recipient: pass the reject to the originator.
1096 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1097 Many were false-positives and ignorable, but it's worth fixing the
1100 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1101 for the new environment-manipulation done at startup. Move the routines
1102 from being local to tls.c to being global via the os.c file.
1104 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1105 an extract embedded as result-arg for a map, the first arg for extract
1106 is unavailable so we cannot tell if this is a numbered or keyed
1107 extraction. Accept either.
1113 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1116 JH/02 The smtp transport option "multi_domain" is now expanded.
1118 JH/03 The smtp transport now requests PRDR by default, if the server offers
1121 JH/04 Certificate name checking on server certificates, when exim is a client,
1122 is now done by default. The transport option tls_verify_cert_hostnames
1123 can be used to disable this per-host. The build option
1124 EXPERIMENTAL_CERTNAMES is withdrawn.
1126 JH/05 The value of the tls_verify_certificates smtp transport and main options
1127 default to the word "system" to access the system default CA bundle.
1128 For GnuTLS, only version 3.0.20 or later.
1130 JH/06 Verification of the server certificate for a TLS connection is now tried
1131 (but not required) by default. The verification status is now logged by
1132 default, for both outbound TLS and client-certificate supplying inbound
1135 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1138 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1139 Status Notification (bounce) messages are now MIME format per RFC 3464.
1140 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1141 under the control of the dsn_advertise_hosts option, and routers may
1142 have a dsn_lasthop option.
1144 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1145 default, modifiable by a malware= option. The list separator for
1146 the options can now be changed in the usual way. Bug 68.
1148 JH/10 The smtp_receive_timeout main option is now expanded before use.
1150 JH/11 The incoming_interface log option now also enables logging of the
1151 local interface on delivery outgoing connections.
1153 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1154 if the interface and destination host and port all match.
1156 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1159 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1160 Patch from Andrew Lewis.
1162 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1163 now supports optional time-restrictions, weighting, and priority
1164 modifiers per server. Patch originally by <rommer@active.by>.
1166 JH/16 The spamd_address main option now supports a mixed list of local
1167 and remote servers. Remote servers can be IPv6 addresses, and
1168 specify a port-range.
1170 JH/17 Bug 68: The spamd_address main option now supports an optional
1171 timeout value per server.
1173 JH/18 Bug 1581: Router and transport options headers_add/remove can
1174 now have the list separator specified.
1176 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1179 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1182 JH/21 Support for the A6 type of dns record is withdrawn.
1184 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1185 rather than the verbs used.
1187 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1188 from 255 to 1024 chars.
1190 JH/24 Verification callouts now attempt to use TLS by default.
1192 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1193 are generic router options now. The defaults didn't change.
1195 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1196 Original patch from Alexander Shikoff, worked over by JH.
1198 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1199 files if ZCAT_COMMAND is not executable.
1201 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1203 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1205 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1206 Normally benign, it bites when the pair was led to by a CNAME;
1207 modern usage is to not canonicalize the domain to a CNAME target
1208 (and we were inconsistent anyway for A-only vs AAAA+A).
1210 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1212 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1213 when evaluating $sender_host_dnssec.
1215 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1216 $sender_helo_dnssec variable.
1218 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1220 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1222 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1224 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1225 documented as working, but never had. Support all but $spam_report.
1227 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1228 added for tls authenticator.
1230 HS/03 Add perl_taintmode main config option
1236 TL/01 When running the test suite, the README says that variables such as
1237 no_msglog_check are global and can be placed anywhere in a specific
1238 test's script, however it was observed that placement needed to be near
1239 the beginning for it to behave that way. Changed the runtest perl
1240 script to read through the entire script once to detect and set these
1241 variables, reset to the beginning of the script, and then run through
1242 the script parsing/test process like normal.
1244 TL/02 The BSD's have an arc4random API. One of the functions to induce
1245 adding randomness was arc4random_stir(), but it has been removed in
1246 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1247 function when detected.
1249 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1250 cause callback expansion.
1252 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1253 syntax errors in an expansion can be treated as a string instead of
1254 logging or causing an error, due to the internal use of bool_lax
1255 instead of bool when processing it.
1257 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1258 server certificates when making smtp deliveries.
1260 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1262 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1264 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1266 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1267 Merged patch from Sebastian Wiedenroth.
1269 JH/05 Fix results-pipe from transport process. Several recipients, combined
1270 with certificate use, exposed issues where response data items split
1271 over buffer boundaries were not parsed properly. This eventually
1272 resulted in duplicates being sent. This issue only became common enough
1273 to notice due to the introduction of connection certificate information,
1274 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1276 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1277 size buffer was used, resulting in syntax errors when an expansion
1280 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1281 version 3.3.6 or later.
1283 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1284 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1285 both become "event_action", the variables become $event_name, $event_data
1286 and $event_defer_errno. There is a new variable $verify_mode, usable in
1287 routers, transports and related events. The tls:cert event is now also
1288 raised for inbound connections, if the main configuration event_action
1291 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1292 early OCSP support, but no stapling (appears to be less than 1.0.0).
1294 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1295 server certificate names available under the smtp transport option
1296 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1299 JH/10 Time-related extraction expansions from certificates now use the main
1300 option "timezone" setting for output formatting, and are consistent
1301 between OpenSSL and GnuTLS compilations. Bug 1541.
1303 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1304 encoded parameter in the incoming message. Bug 1558.
1306 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1307 include certificate info, eximon was claiming there were spoolfile
1310 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1312 JH/14 Log delivery-related information more consistently, using the sequence
1313 "H=<name> [<ip>]" wherever possible.
1315 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1316 are problematic for Debian distribution, omit them from the release
1319 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1321 JH/16 Fix string representation of time values on 64bit time_t architectures.
1324 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1325 output list separator was used.
1330 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1331 checkers that were complaining about end of non-void function with no
1334 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1335 This was a regression introduced in 4.83 by another bugfix.
1337 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1339 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1340 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1346 TF/01 Correctly close the server side of TLS when forking for delivery.
1348 When a message was received over SMTP with TLS, Exim failed to clear up
1349 the incoming connection properly after forking off the child process to
1350 deliver the message. In some situations the subsequent outgoing
1351 delivery connection happened to have the same fd number as the incoming
1352 connection previously had. Exim would try to use TLS and fail, logging
1353 a "Bad file descriptor" error.
1355 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1356 utilities have not been installed.
1358 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1359 temporary space as the ACL may create new global variables.
1361 TL/01 LDAP support uses per connection or global context settings, depending
1362 upon the detected version of the libraries at build time.
1364 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1365 to extract and use the src ip:port in logging and expansions as if it
1366 were a direct connection from the outside internet. PPv2 support was
1367 updated based on HAProxy spec change in May 2014.
1369 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1371 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1372 Properly escape header and check for NULL return.
1374 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1377 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1379 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1380 characters in header names, implemented as a verify condition.
1381 Contributed by Michael Fischer v. Mollard.
1383 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1384 results permerror and temperror. Previous values are deprecated but
1385 still accepted. In a future release, err_perm and err_temp will be
1386 completely removed, which will be a backward incompatibility if the
1387 ACL tests for either of these two old results. Patch contributed by
1388 user bes-internal on the mailing list.
1390 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1392 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1393 selectors, in both main and reject logs.
1395 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1398 JH/07 Add malware type "sock" for talking to simple daemon.
1400 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1402 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1403 routers/transports under cutthrough routing.
1405 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1406 numbers. Touch up "bool" conditional to keep the same definition.
1408 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1410 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1412 JH/12 Expand items in router/transport headers_add or headers_remove lists
1413 individually rather than the list as a whole. Bug 1452.
1415 Required for reasonable handling of multiple headers_ options when
1416 they may be empty; requires that headers_remove items with embedded
1417 colons must have them doubled (or the list-separator changed).
1419 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1420 view the policy declared in the DMARC record. Currently, $dmarc_status
1421 is a combined value of both the record presence and the result of the
1424 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1426 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1427 dnslookup router and the smtp transport (applying to the forward
1430 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1431 of ldap servers used for a specific lookup. Patch provided by Heiko
1434 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1435 New variable $lookup_dnssec_authenticated for observability.
1437 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1438 Patch submitted by Lars Timman.
1440 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1442 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1443 Requires trusted mode and valid format message id, aborts otherwise.
1444 Patch contributed by Heiko Schlichting.
1446 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1447 certextract with support for various fields. Bug 1358.
1449 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1450 is requested by default, modifiable by smtp transport option
1453 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1454 operate on certificate variables to give certificate fingerprints
1455 Also new ${sha256:cert_variable}.
1457 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1459 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1460 Christian Aistleitner.
1462 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1464 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1465 file. Patch from Wolfgang Breyha.
1467 JH/25 Expand the coverage of the delivery $host and $host_address to
1468 client authenticators run in verify callout. Bug 1476.
1470 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1471 align with daemon_smtp_ports. Bug 72.
1473 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1474 support and error reporting did not work properly.
1476 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1477 and is readable. Patch from Andrew Colin Kissa.
1479 TL/14 Enhance documentation of ${run expansion and how it parses the
1480 commandline after expansion, particularly in the case when an
1481 unquoted variable expansion results in an empty value.
1483 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1485 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1486 by renaming away from it. Addresses GH issue 15, reported by
1489 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1491 TL/15 SECURITY: prevent double expansion in math comparison functions
1492 (can expand unsanitized data). Not remotely exploitable.
1499 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1501 PP/02 Make -n do something, by making it not do something.
1502 When combined with -bP, the name of an option is not output.
1504 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1507 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1508 $sender_host_name and config options to manage this, and basic check
1511 PP/05 DSCP support for outbound connections and control modifier for inbound.
1513 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1514 (Only plugin which currently uses this is kerberos4, which nobody should
1515 be using, but we should make it available and other future plugins might
1516 conceivably use it, even though it would break NAT; stuff *should* be
1517 using channel bindings instead).
1519 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1520 name; added for Sendmail compatibility; requires admin caller.
1521 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1522 just ignore it); requires trusted caller.
1523 Also parse but ignore: -Ac -Am -X<logfile>
1526 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1528 TL/02 Add +smtp_confirmation as a default logging option.
1530 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1531 Patch by Magnus Holmgren from 2007-02-20.
1533 TL/04 Bugzilla 1281 - Spec typo.
1534 Bugzilla 1283 - Spec typo.
1535 Bugzilla 1290 - Spec grammar fixes.
1537 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1539 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1541 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1542 mailing list by by Dmitry Isaikin.
1544 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1546 JH/02 Support "G" suffix to numbers in ${if comparisons.
1548 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1550 NM/01 Bugzilla 1197 - Spec typo
1551 Bugzilla 1196 - Spec examples corrections
1553 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1555 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1556 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1559 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1560 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1562 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1563 "acl {{name}{arg}...}", and optional args on acl condition
1566 JH/05 Permit multiple router/transport headers_add/remove lines.
1568 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1570 JH/07 Avoid using a waiting database for a single-message-only transport.
1571 Performance patch from Paul Fisher. Bugzilla 1262.
1573 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1576 JH/09 Add $headers_added variable, with content from use of ACL modifier
1577 add_header (but not yet added to the message). Bugzilla 199.
1579 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1580 Pulled from Bugzilla 817 by Wolfgang Breyha.
1582 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1584 (nb: this is the same fix as in Exim 4.80.1)
1586 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1589 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1592 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1594 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1595 advertise SMTP AUTH mechanism to us, instead of a generic
1596 protocol violation error. Also, make Exim more robust to bad
1597 data from the Dovecot auth socket.
1599 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1601 When a queue runner is handling a message, Exim first routes the
1602 recipient addresses, during which it prunes them based on the retry
1603 hints database. After that it attempts to deliver the message to
1604 any remaining recipients. It then updates the hints database using
1607 So if a recipient address works intermittently, it can get repeatedly
1608 deferred at routing time. The retry hints record remains fresh so the
1609 address never reaches the final cutoff time.
1611 This is a fairly common occurrence when a user is bumping up against
1612 their storage quota. Exim had some logic in its local delivery code
1613 to deal with this. However it did not apply to per-recipient defers
1614 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1616 This change adds a proper retry rule check during routing so that the
1617 final cutoff time is checked against the message's age. We only do
1618 this check if there is an address retry record and there is not a
1619 domain retry record; this implies that previous attempts to handle
1620 the address had the retry_use_local_parts option turned on. We use
1621 this as an approximation for the destination being like a local
1622 delivery, as in LMTP.
1624 I suspect this new check makes the old local delivery cutoff check
1625 redundant, but I have not verified this so I left the code in place.
1627 TF/02 Correct gecos expansion when From: is a prefix of the username.
1629 Test 0254 submits a message to Exim with the header
1633 When I ran the test suite under the user fanf2, Exim expanded
1634 the header to contain my full name, whereas it should have added
1635 a Resent-Sender: header. It erroneously treats any prefix of the
1636 username as equal to the username.
1638 This change corrects that bug.
1640 GF/01 DCC debug and logging tidyup
1641 Error conditions log to paniclog rather than rejectlog.
1642 Debug lines prefixed by "DCC: " to remove any ambiguity.
1644 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1646 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1647 Bug spotted by Jeremy Harris; was flawed since initial commit.
1648 Would have resulted in OCSP responses post-SNI triggering an Exim
1649 NULL dereference and crash.
1651 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1653 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1654 Bug detection, analysis and fix by Samuel Thibault.
1655 Bugzilla 1331, Debian bug #698092.
1657 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1659 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1660 Server implementation by Todd Lyons, client by JH.
1661 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1662 config variable "prdr_enable" controls whether the server
1663 advertises the facility. If the client requests PRDR a new
1664 acl_data_smtp_prdr ACL is called once for each recipient, after
1665 the body content is received and before the acl_smtp_data ACL.
1666 The client is controlled by both of: a hosts_try_prdr option
1667 on the smtp transport, and the server advertisement.
1668 Default client logging of deliveries and rejections involving
1669 PRDR are flagged with the string "PRDR".
1671 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1672 fclose(). Diagnosis by Todd Lyons.
1674 PP/17 Update configure.default to handle IPv6 localhost better.
1675 Patch by Alain Williams (plus minor tweaks).
1678 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1679 This is now consistent with GnuTLS, and is now documented: the
1680 previous undocumented portable approach to treating the option as
1681 unset was to force an expansion failure. That still works, and
1682 an empty string is now equivalent.
1684 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1685 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1686 not performing validation itself.
1688 PP/20 Added force_command boolean option to pipe transport.
1689 Patch from Nick Koston, of cPanel Inc.
1691 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1694 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1696 PP/21 Fix eximon continuous updating with timestamped log-files.
1697 Broken in a format-string cleanup in 4.80, missed when I repaired the
1698 other false fix of the same issue.
1699 Report and fix from Heiko Schlichting.
1702 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1703 Report from Prashanth Katuri.
1705 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1706 It's SecureTransport, so affects any MacOS clients which use the
1707 system-integrated TLS libraries, including email clients.
1709 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1710 using a MIME ACL for non-SMTP local injection.
1711 Report and assistance in diagnosis by Warren Baker.
1713 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1715 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1717 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1718 last id that failed so it may be referenced in subsequent ACL's.
1720 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1723 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1724 ldap library initialization, allowing self-signed CA's to be
1725 used. Also properly sets require_cert option later in code by
1726 using NULL (global ldap config) instead of ldap handle (per
1727 session). Bug diagnosis and testing by alxgomz.
1729 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1730 the src/util/ subdirectory.
1732 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1733 renamed to Transport Post Delivery Action by Jeremy Harris, as
1736 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1737 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1738 redis_servers = needs to be configured which will be used by the redis
1739 lookup. Patch from Warren Baker, of The Packet Hub.
1741 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1743 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1744 hostname or reverse DNS when processing a host list. Used suggestions
1745 from multiple comments on this bug.
1747 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1749 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1750 Missed a few lines, added it to make the runtest require no keyboard
1753 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1754 contains upper case chars. Make router use caseful_local_part.
1756 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1757 support when GnuTLS has been built with p11-kit.
1763 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1765 This, or similar/improved, will also be change PP/11 of 4.82.
1771 PP/01 Handle short writes when writing local log-files.
1772 In practice, only affects FreeBSD (8 onwards).
1773 Bugzilla 1053, with thanks to Dmitry Isaikin.
1775 NM/01 Bugzilla 949 - Documentation tweak
1777 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1780 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1782 PP/02 Implemented gsasl authenticator.
1784 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1786 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1787 `pkg-config foo` for cflags/libs.
1789 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1790 with rest of GSASL and with heimdal_gssapi.
1792 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1793 `pkg-config foo` for cflags/libs for the TLS implementation.
1795 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1796 properties get this fed in as external SSF. A number of robustness
1797 and debugging improvements to the cyrus_sasl authenticator.
1799 PP/08 cyrus_sasl server now expands the server_realm option.
1801 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1802 Patch by Jeremy Harris.
1804 PP/10 Added dbmjz lookup type.
1806 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1808 PP/12 MAIL args handles TAB as well as SP, for better interop with
1809 non-compliant senders.
1810 Analysis and variant patch by Todd Lyons.
1812 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1813 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1814 Patch from Dirk Mueller <dmueller@suse.com>
1816 PP/13 tls_peerdn now print-escaped for spool files.
1817 Observed some $tls_peerdn in wild which contained \n, which resulted
1818 in spool file corruption.
1820 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1821 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1822 or write after TLS renegotiation, which otherwise led to messages
1825 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1826 as a tracking header (ie: a signed header comes before the signature).
1827 Patch from Wolfgang Breyha.
1829 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1830 comma-sep list; embedded commas doubled.
1832 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1834 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1836 Report and patch from Dmitry Banschikov.
1838 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1839 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1840 failures. We appear to now support TLS1.1+ with Exim.
1842 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1843 lets Exim select keys and certificates based upon TLS SNI from client.
1844 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1845 before an outbound SMTP session. New log_selector, +tls_sni.
1847 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1848 NULL dereference. Report and patch from Alun Jones.
1850 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1851 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1852 Not seeing resolver debug output on NetBSD, but suspect this is a
1853 resolver implementation change.
1855 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1856 Left warnings. Added "eximon gdb" invocation mode.
1858 PP/21 Defaulting "accept_8bitmime" to true, not false.
1860 PP/22 Added -bw for inetd wait mode support.
1862 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1863 locate the relevant includes and libraries. Made this the default.
1865 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1866 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1868 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1869 This may cause build issues on older platforms.
1871 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1872 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1873 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1874 Added SNI support via GnuTLS too.
1875 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1877 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1879 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1880 Applied second patch from Janne, implementing suggestion to default
1881 multiple-strings-in-record handling to match SPF spec.
1883 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1885 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1886 read-only, out of scope).
1887 Patch from Wolfgang Breyha, report from Stuart Northfield.
1889 PP/29 Fix three issues highlighted by clang analyser static analysis.
1890 Only crash-plausible issue would require the Cambridge-specific
1891 iplookup router and a misconfiguration.
1892 Report from Marcin Mirosław.
1894 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1896 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1897 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1898 As part of this, removing so much warning spew let me fix some minor
1899 real issues in debug logging.
1901 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1902 assignment on my part. Fixed.
1904 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1905 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1906 Janne Snabb (who went above and beyond: thank you).
1908 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1909 string otherwise requires a connection and a bunch more work and it's
1910 relatively easy to get wrong. Should also expose TLS library linkage
1913 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1914 64-bit ${eval} (JH/03).
1916 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1917 GNU libc to support some of the 64-bit stuff, should not lead to
1918 conflicts. Defined before os.h is pulled in, so if a given platform
1919 needs to override this, it can.
1921 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1922 protection layer was required, which is not implemented.
1923 Bugzilla 1254, patch from Wolfgang Breyha.
1925 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1926 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1927 tls_dhparam take prime identifiers. Also unbreak combination of
1928 OpenSSL+DH_params+TLSSNI.
1930 PP/39 Disable SSLv2 by default in OpenSSL support.
1936 PP/01 Solaris build fix for Oracle's LDAP libraries.
1937 Bugzilla 1109, patch from Stephen Usher.
1939 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1941 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1944 TF/02 Fix a couple more cases where we did not log the error message
1945 when unlink() failed. See also change 4.74-TF/03.
1947 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1948 lock up or crash if it happened to be inside a call to libc when it
1949 got a SIGUSR1 from exiwhat.
1951 The SIGUSR1 handler appends the current process status to the process
1952 log which is later printed by exiwhat. It used to use the general
1953 purpose logging code to do this, but several functions it calls are
1954 not safe for signals.
1956 The new output code in the SIGUSR1 handler is specific to the process
1957 log, and simple enough that it's easy to inspect for signal safety.
1958 Removing some special cases also simplifies the general logging code.
1959 Removing the spurious timestamps from the process log simplifies
1962 TF/04 Improved ratelimit ACL condition.
1964 The /noupdate option has been deprecated in favour of /readonly which
1965 has clearer semantics. The /leaky, /strict, and /readonly update modes
1966 are mutually exclusive. The update mode is no longer included in the
1967 database key; it just determines when the database is updated. (This
1968 means that when you upgrade Exim will forget old rate measurements.)
1970 Exim now checks that the per_* options are used with an update mode that
1971 makes sense for the current ACL. For example, when Exim is processing a
1972 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1973 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1974 must specify per_mail/readonly. If you omit the update mode it defaults to
1975 /leaky where that makes sense (as before) or /readonly where required.
1977 The /noupdate option is now undocumented but still supported for
1978 backwards compatibility. It is equivalent to /readonly except that in
1979 ACLs where /readonly is required you may specify /leaky/noupdate or
1980 /strict/noupdate which are treated the same as /readonly.
1982 A useful new feature is the /count= option. This is a generalization
1983 of the per_byte option, so that you can measure the throughput of other
1984 aggregate values. For example, the per_byte option is now equivalent
1985 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1987 The per_rcpt option has been generalized using the /count= mechanism
1988 (though it's more complicated than the per_byte equivalence). When it is
1989 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1990 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1991 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1992 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1993 Note that using per_rcpt with a non-readonly update mode in more than
1994 one ACL will cause the recipients to be double-counted. (The per_mail
1995 and per_byte options don't have this problem.)
1997 The handling of very low rates has changed slightly. If the computed rate
1998 is less than the event's count (usually one) then this event is the first
1999 after a long gap. In this case the rate is set to the same as this event's
2000 count, so that the first message of a spam run is counted properly.
2002 The major new feature is a mechanism for counting the rate of unique
2003 events. The new per_addr option counts the number of different
2004 recipients that someone has sent messages to in the last time period. It
2005 behaves like per_rcpt if all the recipient addresses are different, but
2006 duplicate recipient addresses do not increase the measured rate. Like
2007 the /count= option this is a general mechanism, so the per_addr option
2008 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2009 example, measure the rate that a client uses different sender addresses
2010 with the options per_mail/unique=$sender_address. There are further
2011 details in the main documentation.
2013 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2015 TF/06 Removed a few PCRE remnants.
2017 TF/07 Automatically extract Exim's version number from tags in the git
2018 repository when doing development or release builds.
2020 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2021 Bugzilla 879. Patch from Paul Fisher.
2023 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2024 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2027 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2029 PP/05 Variable $av_failed, true if the AV scanner deferred.
2030 Bugzilla 1078. Patch from John Horne.
2032 PP/06 Stop make process more reliably on build failure.
2033 Bugzilla 1087. Patch from Heiko Schlittermann.
2035 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2036 Bugzilla 1089. Patch from Heiko Schlittermann.
2038 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2039 Bugzilla 1131. Patch from Holger Weiß.
2041 PP/09 Handle IPv6 addresses with SPF.
2042 Bugzilla 860. Patch from Wolfgang Breyha.
2044 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2046 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2049 PP/11 match_* no longer expand right-hand-side by default.
2050 New compile-time build option, EXPAND_LISTMATCH_RHS.
2051 New expansion conditions, "inlist", "inlisti".
2053 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2055 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2057 PP/14 fix log_write() format string regression from TF/03.
2058 Bugzilla 1152. Patch from Dmitry Isaikin.
2064 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2066 PP/02 Harmonised TLS library version reporting; only show if debugging.
2067 Layout now matches that introduced for other libraries in 4.74 PP/03.
2069 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2071 PP/04 New "dns_use_edns0" global option.
2073 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2076 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2077 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2079 TK/01 Updated PolarSSL code to 0.14.2.
2080 Bugzilla 1097. Patch from Andreas Metzler.
2082 PP/07 Catch divide-by-zero in ${eval:...}.
2083 Fixes bugzilla 1102.
2085 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2088 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2089 format-string attack -- SECURITY: remote arbitrary code execution.
2091 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2092 time unintentionally subject to list matching rules, letting the header
2093 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2094 arbitrary string expansion). This allowed for information disclosure.
2096 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2097 INT_MIN/-1 -- value coerced to INT_MAX.
2103 NM/01 Workaround for PCRE version dependency in version reporting
2106 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2107 This fixes portability to compilers other than gcc, notably
2108 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2110 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2111 makefiles for portability to HP-UX and POSIX correctness.
2113 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2114 Also via indented variable definition in the Makefile.
2115 (Debugging by Oliver Heesakkers).
2117 PP/02 Restore caching of spamd results with expanded spamd_address.
2118 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2120 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2121 Improves build reliability. Fix from: Frank Elsner
2123 NM/02 Fix wide character breakage in the rfc2047 coding
2124 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2126 NM/03 Allow underscore in dnslist lookups
2127 Fixes bug 1026. Patch from Graeme Fowler
2129 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2130 Code patches from Adam Ciarcinski of NetBSD.
2132 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2135 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2136 is logged, to avoid truncation. Patch from John Horne.
2138 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2139 Patch from Jakob Hirsch.
2141 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2142 SQL string expansion failure details.
2143 Patch from Andrey Oktyabrski.
2145 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2146 Patch from Simon Arlott.
2148 PP/09 New lookups functionality failed to compile on old gcc which rejects
2149 extern declarations in function scope.
2150 Patch from Oliver Fleischmann
2152 PP/10 Use sig_atomic_t for flags set from signal handlers.
2153 Check getgroups() return and improve debugging.
2154 Fixed developed for diagnosis in bug 927 (which turned out to be
2157 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2158 Patch from Mark Zealey.
2160 PP/12 Bugzilla 1056: Improved spamd server selection.
2161 Patch from Mark Zealey.
2163 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2164 Based on patch from Heiko Schlittermann.
2166 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2167 Patch from Uwe Doering, sign-off by Michael Haardt.
2169 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2170 variable declaration deep within a block. Bug and patch from
2173 PP/15 lookups-Makefile IRIX compatibility coercion.
2175 PP/16 Make DISABLE_DKIM build knob functional.
2177 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2178 Patch by Simon Arlott
2180 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2181 variable argument macros. Our copy now differs from upstream.
2187 TF/01 Failure to get a lock on a hints database can have serious
2188 consequences so log it to the panic log.
2190 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2191 controlled using the smtp_confirmation log selector.
2193 TF/03 Include the error message when we fail to unlink a spool file.
2195 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2196 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2197 for maintaining out-of-tree patches for some time.
2199 PP/01 Bugzilla 139: Documentation and portability issues.
2200 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2201 Handle per-OS dynamic-module compilation flags.
2203 PP/02 Let /dev/null have normal permissions.
2204 The 4.73 fixes were a little too stringent and complained about the
2205 permissions on /dev/null. Exempt it from some checks.
2206 Reported by Andreas M. Kirchwitz.
2208 PP/03 Report version information for many libraries, including
2209 Exim version information for dynamically loaded libraries. Created
2210 version.h, now support a version extension string for distributors
2211 who patch heavily. Dynamic module ABI change.
2213 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2214 privilege escalation vulnerability whereby the Exim run-time user
2215 can cause root to append content of the attacker's choosing to
2218 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2221 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2222 If dropping privileges for untrusted macros, we disabled normal logging
2223 on the basis that it would fail; for the Exim run-time user, this is not
2224 the case, and it resulted in successful deliveries going unlogged.
2225 Fixed. Reported by Andreas Metzler.
2231 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2232 only prepend for the Resent-* case. Fixes regression introduced in
2233 Exim 4.70 by NM/22 for Bugzilla 607.
2235 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2236 increasing numbers of administrators be bitten by this.
2238 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2240 PP/03 Bugzilla 994: added openssl_options main configuration option.
2242 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2244 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2246 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2248 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2249 without a peer certificate, leading to a segfault because of an
2250 assumption that peers always have certificates. Be a little more
2251 paranoid. Problem reported by Martin Tscholak.
2253 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2254 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2255 NB: ClamAV planning to remove STREAM in "middle of 2010".
2256 CL also introduces -bmalware, various -d+acl logging additions and
2257 more caution in buffer sizes.
2259 PP/09 Implemented reverse_ip expansion operator.
2261 PP/10 Bugzilla 937: provide a "debug" ACL control.
2263 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2265 PP/12 Bugzilla 973: Implement --version.
2267 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2269 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2271 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2273 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2274 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2275 ignore trailing whitespace.
2277 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2279 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2282 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2283 Notification from Dr Andrew Aitchison.
2285 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2286 ExtendedDetectionInfo response format.
2287 Notification from John Horne.
2289 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2292 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2293 XSL and documented dependency on system catalogs, with examples of how
2296 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2299 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2300 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2301 configuration file which is writeable by the Exim user or group.
2303 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2304 of configuration files to cover files specified with the -C option if
2305 they are going to be used with root privileges, not just the default
2308 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2309 option (effectively making it always true).
2311 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2312 files to be used while preserving root privileges.
2314 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2315 that rogue child processes cannot use them.
2317 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2318 run-time user, instead of root.
2320 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2321 Exim run-time user without dropping privileges.
2323 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2324 result string, instead of calling string_vformat() twice with the same
2327 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2328 for other users. Others should always drop root privileges if they use
2329 -C on the command line, even for a whitelisted configure file.
2331 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2333 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2339 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2340 $data_path, and $header_path variables; fixed documentation bugs and
2343 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2344 exipick to access non-standard spools, including the "frozen" queue
2347 NM/01 Bugzilla 965: Support mysql stored procedures.
2348 Patch from Alain Williams
2350 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2352 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2353 Patch from Andreas Metzler
2355 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2356 Patch from Kirill Miazine
2358 NM/05 Bugzilla 671: Added umask to procmail example.
2360 JJ/03 installed exipick 20100323.0, fixing doc bug
2362 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2363 directory. Notification and patch from Dan Rosenberg.
2365 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2367 TK/02 Improve log output when DKIM signing operation fails.
2369 MH/01 Treat the transport option dkim_domain as a colon separated
2370 list, not as a single string, and sign the message with each element,
2371 omitting multiple occurences of the same signer.
2373 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2374 Bugzilla 985, 986. Patch by Simon Arlott
2376 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2377 Patch by Simon Arlott
2379 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2380 MBX locking. Notification from Dan Rosenberg.
2386 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2388 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2390 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2392 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2394 NM/04 Bugzilla 915: Flag broken perl installation during build.
2400 TK/01 Added patch by Johannes Berg that expands the main option
2401 "spamd_address" if it starts with a dollar sign.
2403 TK/02 Write list of recipients to X-Envelope-Sender header when building
2404 the mbox-format spool file for content scanning (suggested by Jakob
2407 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2408 (http://www.dcc-servers.net/) support via dccifd. Activated by
2409 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2411 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2412 by Mark Daniel Reidel <mr@df.eu>.
2414 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2415 When building exim an external PCRE library is now needed -
2416 PCRE is a system library on the majority of modern systems.
2417 See entry on PCRE_LIBS in EDITME file.
2419 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2420 conversation. Added nologin parameter to request.
2421 Patch contributed by Kirill Miazine.
2423 TF/01 Do not log submission mode rewrites if they do not change the address.
2425 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2427 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2428 log files in place. Contributed by Roberto Lima.
2430 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2432 TF/03 Bugzilla 615: When checking the local_parts router precondition
2433 after a local_part_suffix or local_part_prefix option, Exim now
2434 does not use the address's named list lookup cache, since this
2435 contains cached lookups for the whole local part.
2437 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2438 Robert Millan. Documentation is in experimental-spec.txt.
2440 TF/04 Bugzilla 668: Fix parallel build (make -j).
2442 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2444 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2445 Patch provided by Jan Srzednicki.
2447 TF/05 Leading white space used to be stripped from $spam_report which
2448 wrecked the formatting. Now it is preserved.
2450 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2451 that they are available at delivery time.
2453 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2455 TF/08 TLS error reporting now respects the incoming_interface and
2456 incoming_port log selectors.
2458 TF/09 Produce a more useful error message if an SMTP transport's hosts
2459 setting expands to an empty string.
2461 NM/06 Bugzilla 744: EXPN did not work under TLS.
2462 Patch provided by Phil Pennock.
2464 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2465 Patch provided by Richard Godbee.
2467 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2468 acl_smtp_notquit, added index entry.
2470 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2471 Patch provided by Eugene Bujak.
2473 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2474 accept(). Patch provided by Maxim Dounin.
2476 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2477 Patch provided by Phil Pennock.
2479 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2481 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2482 Patch provided by Brad "anomie" Jorsch.
2484 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2485 Patch provided by Dean Brooks.
2487 TK/05 Add native DKIM support (does not depend on external libraries).
2489 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2490 Patch provided by Graeme Fowler.
2492 NM/16 Bugzilla 851: Documentation example syntax fix.
2494 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2496 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2499 NM/19 Bugzilla 745: TLS version reporting.
2500 Patch provided by Phil Pennock.
2502 NM/20 Bugzilla 167: bool: condition support.
2503 Patch provided by Phil Pennock.
2505 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2506 clients. Patch provided by Phil Pennock.
2508 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2509 Patch provided by Brad "anomie" Jorsch.
2511 NM/23 Bugzilla 687: Fix misparses in eximstats.
2512 Patch provided by Heiko Schlittermann.
2514 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2515 Patch provided by Heiko Schlittermann.
2517 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2518 plus update to original patch.
2520 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2522 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2523 Patch provided by David Brownlee.
2525 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2527 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2529 NM/30 Bugzilla 888: TLS documentation bugfixes.
2531 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2533 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2534 Unlike the original bugzilla I have changed all shell scripts in src tree.
2536 NM/33 Bugzilla 898: Transport filter timeout fix.
2537 Patch by Todd Rinaldo.
2539 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2540 Patch by Serge Demonchaux.
2542 NM/35 Bugzilla 39: Base64 decode bug fixes.
2543 Patch by Jakob Hirsch.
2545 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2547 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2549 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2551 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2557 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2558 ALT-N's libdkim that I have put here:
2559 http://duncanthrax.net/exim-experimental/
2561 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2562 were called 'true' and it seems that C99 defines that as a reserved
2563 keyword to be used with 'bool' variable types. That means you could
2564 not include C99-style headers which use bools without triggering
2565 build errors in sieve.c.
2567 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2568 as mailq or other aliases. Changed the --help handling significantly
2569 to do whats expected. exim_usage() emits usage/help information.
2571 SC/01 Added the -bylocaldomain option to eximstats.
2573 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2575 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2577 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2583 PH/01 Another patch from the Sieve maintainer.
2585 PH/02 When an IPv6 address is converted to a string for single-key lookup
2586 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2587 dots are used instead of colons so that keys in lsearch files need not
2588 contain colons. This was done some time before quoting was made available
2589 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2590 (notated using the quote facility) so as to distinguish them from IPv4
2591 keys. This meant that lookups for IP addresses in host lists did not work
2592 for iplsearch lookups.
2594 This has been fixed by arranging for IPv6 addresses to be expressed with
2595 colons if the lookup type is iplsearch. This is not incompatible, because
2596 previously such lookups could never work.
2598 The situation is now rather anomalous, since one *can* have colons in
2599 ordinary lsearch keys. However, making the change in all cases is
2600 incompatible and would probably break a number of configurations.
2602 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2605 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2606 conversion specification without a maximum field width, thereby enabling
2607 a rogue spamd server to cause a buffer overflow. While nobody in their
2608 right mind would setup Exim to query an untrusted spamd server, an
2609 attacker that gains access to a server running spamd could potentially
2610 exploit this vulnerability to run arbitrary code as the Exim user.
2612 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2613 $primary_hostname instead of what libspf2 thinks the hosts name is.
2615 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2616 a directory entry by the name of the lookup key. Previously, if a
2617 symlink pointed to a non-existing file or a file in a directory that
2618 Exim lacked permissions to read, a lookup for a key matching that
2619 symlink would fail. Now it is enough that a matching directory entry
2620 exists, symlink or not. (Bugzilla 503.)
2622 PH/03 The body_linecount and body_zerocount variables are now exported in the
2625 PH/04 Added the $dnslist_matched variable.
2627 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2628 This means they are set thereafter only if the connection becomes
2631 PH/06 Added the client_condition to authenticators so that some can be skipped
2632 by clients under certain conditions.
2634 PH/07 The error message for a badly-placed control=no_multiline_responses left
2635 "_responses" off the end of the name.
2637 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2639 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2640 (without spaces) instead of just copying the configuration text.
2642 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2644 PH/11 Added $max_received_linelength.
2646 PH/12 Added +ignore_defer and +include_defer to host lists.
2648 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2649 way in which PCRE > 7.0 is built.
2651 PH/14 Implemented queue_only_load_latch.
2653 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2654 MAIL command. The effect was to mangle the value on 64-bit systems.
2656 PH/16 Another patch from the Sieve maintainer.
2658 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2660 PH/18 If a system quota error occurred while trying to create the file for
2661 a maildir delivery, the message "Mailbox is full" was not appended to the
2662 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2663 only to a quota excession during the actual writing of the file.
2665 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2666 characters?) which causes problems in log lines. The DN values are now
2667 passed through string_printing() before being added to log lines.
2669 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2670 and InterBase are left for another time.)
2672 PH/21 Added message_body_newlines option.
2674 PH/22 Guard against possible overflow in moan_check_errorcopy().
2676 PH/23 POSIX allows open() to be a macro; guard against that.
2678 PH/24 If the recipient of an error message contained an @ in the local part
2679 (suitably quoted, of course), incorrect values were put in $domain and
2680 $local_part during the evaluation of errors_copy.
2686 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2687 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2690 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2691 issue a MAIL command.
2693 PH/02 In an ACL statement such as
2695 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2697 if a client was not listed at all, or was listed with a value other than
2698 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2699 the condition was not true (as it should be), so access was not denied.
2700 The bug was that the ! inversion was incorrectly passed on to the second
2701 item. This has been fixed.
2703 PH/03 Added additional dnslists conditions == and =& which are different from
2704 = and & when the dns lookup returns more than one IP address.
2706 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2707 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2709 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2710 FSYNC, which compiles an option called disable_fsync that allows for
2711 bypassing fsync(). The documentation is heavily laced with warnings.
2713 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2715 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2716 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2717 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2718 including adding "make clean"; (3) Added -fPIC when compiling the test
2719 dynamically loaded module, to get rid of a warning.
2721 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2722 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2723 = 0s. The bug is otherwise harmless.
2725 PH/07 There was a bug in the dovecot authenticator such that the value of
2726 $auth1 could be overwritten, and so not correctly preserved, after a
2727 successful authentication. This usually meant that the value preserved by
2728 the server_setid option was incorrect.
2730 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2732 PH/09 Installed PCRE release 7.0.
2734 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2735 run for batched SMTP input. It is now run at the start of every message
2736 in the batch. While fixing this I discovered that the process information
2737 (output by running exiwhat) was not always getting set for -bs and -bS
2738 input. This is fixed, and it now also says "batched" for BSMTP.
2740 PH/11 Added control=no_pipelining.
2742 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2743 patch, slightly modified), and move the expansion of helo_data till after
2744 the connection is made in the smtp transport (so it can use these
2747 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2749 PH/14 Added log_selector = +pid.
2751 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2753 PH/16 Add ${if forany and ${if forall.
2755 PH/17 Added dsn_from option to vary the From: line in DSNs.
2757 PH/18 Flush SMTP output before performing a callout, unless control =
2758 no_callout_flush is set.
2760 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2761 was true (the default) a successful delivery failed to delete the retry
2762 item, thus causing premature timeout of the address. The bug is now
2765 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2767 PH/21 Long custom messages for fakedefer and fakereject are now split up
2768 into multiline responses in the same way that messages for "deny" and
2769 other ACL rejections are.
2771 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2772 with slight modification.
2774 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2775 draft, changing the syntax and factoring some duplicate code".
2777 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2778 for deliveries of the second and subsequent messages over the same SMTP
2781 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2782 ${reduce, with only minor "tidies".
2784 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2786 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2787 expansion side effects.
2789 PH/27 When a message times out after an over-quota error from an Exim-imposed
2790 quota, the bounce message says "mailbox is full". This message was not
2791 being given when it was a system quota that was exceeded. It now should
2794 MH/03 Made $recipients available in local_scan(). local_scan() already has
2795 better access to the recipient list through recipients_list[], but
2796 $recipients can be useful in postmaster-provided expansion strings.
2798 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2799 in the case of a MAIL command with additional options following the
2800 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2801 were accidentally chopped off.
2803 PH/29 SMTP synchronization checks are implemented when a command is read -
2804 there is a check that no more input is waiting when there shouldn't be
2805 any. However, for some commands, a delay in an ACL can mean that it is
2806 some time before the response is written. In this time, more input might
2807 arrive, invalidly. So now there are extra checks after an ACL has run for
2808 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2809 pipelining has not been advertised.
2811 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2813 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2814 returned in the SMTP response when smtp_return_error_details was false.
2815 This has been fixed.
2817 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2818 instead of the C I/O that was originally supplied, because problems were
2819 reported on Solaris.
2821 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2822 Exim which did not show up earlier: it was assuming that a call to
2823 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2824 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2825 was a macro that became an assignment, so it seemed to work. This has
2826 changed to a proper function call with a void return, hence the compile
2827 error. Exim's code has been fixed.
2829 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2832 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2834 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2836 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2837 because it turns out to be too big or there is a timeout) there is no
2838 "Completed" line in the log. When some messages of this type were
2839 selected by exigrep, they were listed as "not completed". Others were
2840 picked up by some special patterns. I have improved the selection
2841 criteria to be more general.
2843 PH/38 The host_find_failed option in the manualroute router can now be set
2844 to "ignore", to completely ignore a host whose IP address cannot be
2845 found. If all hosts are ignored, the behaviour is controlled by the new
2846 host_all_ignored option.
2848 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2849 homing or because of multiple MX records with /mx) generated more than
2850 one IP address, and the following item turned out to be the local host,
2851 all the secondary addresses of the first item were incorrectly removed
2852 from the list, along with the local host and any following hosts (which
2853 is what is supposed to happen).
2855 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2856 whoever called Exim into the -H file. In the case of the daemon it was
2857 behaving confusingly. When first started, it used values for whoever
2858 started the daemon, but after a SIGHUP it used the Exim user (because it
2859 calls itself on a restart). I have changed the code so that it now always
2862 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2863 message are rejected with the same error (e.g. no authentication or bad
2864 sender address), and a DATA command is nevertheless sent (as can happen
2865 with PIPELINING or a stupid MUA), the error message that was given to the
2866 RCPT commands is included in the rejection of the DATA command. This is
2867 intended to be helpful for MUAs that show only the final error to their
2870 PH/42 Another patch from the Sieve maintainer.
2872 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2873 Eximstats - Fixed some broken HTML links and added missing column headers
2875 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2876 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2878 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2880 PH/43 Yet another patch from the Sieve maintainer.
2882 PH/44 I found a way to check for a TCP/IP connection going away before sending
2883 the response to the final '.' that terminates a message, but only in the
2884 case where the client has not sent further data following the '.'
2885 (unfortunately, this is allowed). However, in many cases there won't be
2886 any further data because there won't be any more messages to send. A call
2887 to select() can be used: if it shows that the input is "ready", there is
2888 either input waiting, or the socket has been closed. An attempt to read
2889 the next input character can distinguish the two cases. Previously, Exim
2890 would have sent an OK response which the client would never have see.
2891 This could lead to message repetition. This fix should cure that, at
2892 least in a lot of common cases.
2894 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2895 advertised in response to EHLO.
2901 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2902 fixed by 4.65/MH/01 (is this a record?) are fixed:
2904 (i) An empty string was always treated as zero by the numeric comparison
2905 operators. This behaviour has been restored.
2907 (ii) It is documented that the numeric comparison operators always treat
2908 their arguments as decimal numbers. This was broken in that numbers
2909 starting with 0 were being interpreted as octal.
2911 While fixing these problems I realized that there was another issue that
2912 hadn't been noticed. Values of message_size_limit (both the global option
2913 and the transport option) were treated as octal if they started with 0.
2914 The documentation was vague. These values are now always treated as
2915 decimal, and I will make that clear in the documentation.
2921 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2922 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2925 MH/01 Don't check that the operands of numeric comparison operators are
2926 integers when their expansion is in "skipping" mode (fixes bug
2927 introduced by 4.64-PH/07).
2929 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2930 child addresses, Exim now panics and dies. Previously, because the count
2931 is held in a short int, deliveries were likely to be lost. As such a
2932 large number of recipients for a single message is ridiculous
2933 (performance will be very, very poor), I have chosen to impose a limit
2934 rather than extend the field.
2940 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2941 leftover -K file (the existence of which was triggered by #402).
2942 While we were at it, introduced process PID as part of the -K
2943 filename. This should rule out race conditions when creating
2946 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2947 processing considerably. Previous code took too long for large mails,
2948 triggering a timeout which in turn triggers #401.
2950 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2951 in the DK code in transports.c. sendfile() is not really portable,
2952 hence the _LINUX specificness.
2954 TF/01 In the add_headers option to the mail command in an Exim filter,
2955 there was a bug that Exim would claim a syntax error in any
2956 header after the first one which had an odd number of characters
2959 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2960 callout verification, Exim cached a "reject" for the entire domain. This
2961 is correct for most verifications, but it is not correct for a recipient
2962 verification with use_sender or use_postmaster set, because in that case
2963 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2964 case of MAIL FROM:<> rejection from other early rejections (e.g.
2965 rejection of HELO). When verifying a recipient using a non-null MAIL
2966 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2967 Whatever the result of the callout, the value of the domain cache is
2968 left unchanged (for any other kind of callout, getting as far as trying
2969 RCPT means that the domain itself is ok).
2971 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2974 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2975 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2976 session, the daemon ignores ECONNECT errors and logs others; it now
2977 ignores EPIPE as well.
2979 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2980 (quoted-printable decoding).
2982 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2983 later the small subsequent patch to fix an introduced bug.
2985 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2987 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2989 PH/08 An error is now given if message_size_limit is specified negative.
2991 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2992 to be given (somewhat) arbitrary names.
2994 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2997 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2998 miscellaneous code fixes
3000 PH/10 Added the log_reject_target ACL modifier to specify where to log
3003 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3004 hostname. This is wrong, because it relates to the incoming message (and
3005 probably the interface on which it is arriving) and not to the outgoing
3006 callout (which could be using a different interface). This has been
3007 changed to use the value of the helo_data option from the smtp transport
3008 instead - this is what is used when a message is actually being sent. If
3009 there is no remote transport (possible with a router that sets up host
3010 addresses), $smtp_active_hostname is used.
3012 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3013 tweaks were necessary in order to get it to work (see also 21 below):
3014 (a) The code assumed that strncpy() returns a negative number on buffer
3015 overflow, which isn't the case. Replaced with Exim's string_format()
3017 (b) There were several signed/unsigned issues. I just did the minimum
3018 hacking in of casts. There is scope for a larger refactoring.
3019 (c) The code used strcasecmp() which is not a standard C function.
3020 Replaced with Exim's strcmpic() function.
3021 (d) The code set only $1; it now sets $auth1 as well.
3022 (e) A simple test gave the error "authentication client didn't specify
3023 service in request". It would seem that Dovecot has changed its
3024 interface. Fortunately there's a specification; I followed it and
3025 changed what the client sends and it appears to be working now.
3027 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3030 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3031 address A is aliased to B and C, where B exists and C does not. Without
3032 -v the output is "A verified" because verification stops after a
3033 successful redirection if more than one address is generated. However,
3034 with -v the child addresses are also verified. Exim was outputting "A
3035 failed to verify" and then showing the successful verification for C,
3036 with its parentage. It now outputs "B failed to verify", showing B's
3037 parentage before showing the successful verification of C.
3039 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3040 look up a TXT record in a specific list after matching in a combined
3043 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3044 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3045 they consult the DNS. I had assumed they would set it the way they
3046 wanted; and indeed my experiments on Linux seem to show that in some
3047 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3048 To be on the safe side, however, I have now made the interface to
3049 host_find_byname() similar to host_find_bydns(), with an argument
3050 containing the DNS resolver options. The host_find_byname() function now
3051 sets these options at its start, just as host_find_bydns() does. The smtp
3052 transport options dns_qualify_single and dns_search_parents are passed to
3053 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3054 of host_find_byname() use the default settings of RES_DEFNAMES
3055 (qualify_single) but not RES_DNSRCH (search_parents).
3057 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3058 spool_read_header() do less string testing, by means of a preliminary
3059 switch on the second character of optional "-foo" lines. (This is
3060 overdue, caused by the large number of possibilities that now exist.
3061 Originally there were few.) While I was there, I also converted the
3062 str(n)cmp tests so they don't re-test the leading "-" and the first
3063 character, in the hope this might squeeze out yet more improvement.
3065 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3066 flag allowing group syntax was set by the header_syntax check but not
3067 turned off, possible causing trouble later; (2) The flag was not being
3068 set at all for the header_verify test, causing "group"-style headers to
3069 be rejected. I have now set it in this case, and also caused header_
3070 verify to ignore an empty address taken from a group. While doing this, I
3071 came across some other cases where the code for allowing group syntax
3072 while scanning a header line wasn't quite right (mostly, not resetting
3073 the flag correctly in the right place). These bugs could have caused
3074 trouble for malformed header lines. I hope it is now all correct.
3076 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3077 with the "reply" argument non-NULL. The code, however (which originally
3078 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3079 but it didn't always do it. This confused somebody who was copying the
3080 code for some other use. I have removed all the tests.
3082 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3083 feature that was used to support insecure browsers during the U.S. crypto
3084 embargo. It requires special client support, and Exim is probably the
3085 only MTA that supported it -- and would never use it because real RSA is
3086 always available. This code has been removed, because it had the bad
3087 effect of slowing Exim down by computing (never used) parameters for the
3088 RSA_EXPORT functionality.
3090 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3091 authenticator to fail if there's a tab character in the incoming data
3092 (there should never be unless someone is messing about, as it's supposed
3093 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3094 if the connection is using TLS or if the remote IP is the same as the
3095 local IP, and the "valid-client-cert option" if a client certificate has
3098 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3099 authenticators. This can be used for authorization after authentication
3100 succeeds. (In the case of plaintext, it servers for both authentication
3103 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3104 if any retry times were supplied.
3106 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3107 connection, where there is no client IP address to check. In this
3108 situation, the verify now always succeeds.
3110 PH/25 Applied John Jetmore's -Mset patch.
3112 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3114 PH/27 In a string expansion for a processed (not raw) header when multiple
3115 headers of the same name were present, leading whitespace was being
3116 removed from all of them, but trailing whitespace was being removed only
3117 from the last one. Now trailing whitespace is removed from each header
3118 before concatenation. Completely empty headers in a concatenation (as
3119 before) are ignored.
3121 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3122 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3124 PH/29 [Removed. This was a change that I later backed out, and forgot to
3125 correct the ChangeLog entry (that I had efficiently created) before
3126 committing the later change.]
3128 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3129 address errors (4xx response to RCPT) over the same connection as other
3130 messages routed to the same hosts. Such deliveries are always "forced",
3131 so retry times are not inspected. This resulted in far too many retries
3132 for the affected addresses. The effect occurred only when there were more
3133 hosts than the hosts_max_try setting in the smtp transport when it had
3134 the 4xx errors. Those hosts that it had tried were not added to the list
3135 of hosts for which the message was waiting, so if all were tried, there
3136 was no problem. Two fixes have been applied:
3138 (i) If there are any address or message errors in an SMTP delivery, none
3139 of the hosts (tried or untried) are now added to the list of hosts
3140 for which the message is waiting, so the message should not be a
3141 candidate for sending over the same connection that was used for a
3142 successful delivery of some other message. This seems entirely
3143 reasonable: after all the message is NOT "waiting for some host".
3144 This is so "obvious" that I'm not sure why it wasn't done
3145 previously. Hope I haven't missed anything, but it can't do any
3146 harm, as the worst effect is to miss an optimization.
3148 (ii) If, despite (i), such a delivery is accidentally attempted, the
3149 routing retry time is respected, so at least it doesn't keep
3150 hammering the server.
3152 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3153 in ${readsocket because some servers need this prod.
3155 PH/32 Added some extra debug output when updating a wait-xxx database.
3157 PH/33 The hint "could be header name not terminated by colon", which has been
3158 given for certain expansion errors for a long time, was not being given
3159 for the ${if def:h_colon_omitted{... case.
3161 PH/34 The spec says: "With one important exception, whenever a domain list is
3162 being scanned, $domain contains the subject domain." There was at least
3163 one case where this was not true.
3165 PH/35 The error "getsockname() failed: connection reset by peer" was being
3166 written to the panic log as well as the main log, but it isn't really
3167 panic-worthy as it just means the connection died rather early on. I have
3168 removed the panic log writing for the ECONNRESET error when getsockname()
3171 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3172 runs only) independently of the message's sender address. This meant
3173 that, if the 4xx error was in fact related to the sender, a different
3174 message to the same recipient with a different sender could confuse
3175 things. In particular, this can happen when sending to a greylisting
3176 server, but other circumstances could also provoke similar problems.
3177 I have changed the default so that the retry time for these errors is now
3178 based a combination of the sender and recipient addresses. This change
3179 can be overridden by setting address_retry_include_sender=false in the
3182 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3183 remote server are returned as part of bounce messages. This was not
3184 happening for LMTP over a pipe (the lmtp transport), but now it is the
3185 same for both kinds of LMTP.
3187 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3188 in header lines that were in fact CNAMEs. This is no longer the case.
3190 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3191 and queue runs started by the daemon processed all messages. This has
3192 been fixed so that -R and -S can now usefully be given with -q<time>.
3194 PH/40 Import PCRE release 6.7 (fixes some bugs).
3196 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3198 PH/42 Give an error if -q is specified more than once.
3200 PH/43 Renamed the variables $interface_address and $interface_port as
3201 $received_ip_address and $received_port, to make it clear that these
3202 values apply to message reception, and not to the outgoing interface when
3203 a message is delivered. (The old names remain recognized, of course.)
3205 PH/44 There was no timeout on the connect() call when using a Unix domain
3206 socket in the ${readsocket expansion. There now is.
3208 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3209 be meaningful with "accept".
3211 SC/01 Eximstats V1.43
3212 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3214 SC/02 Eximstats V1.44
3215 Use a glob alias rather than an array ref in the generated
3216 parser. This improves both readability and performance.
3218 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3219 Collect SpamAssassin and rejection statistics.
3220 Don't display local sender or destination tables unless
3221 there is data to show.
3222 Added average volumes into the top table text output.
3224 SC/04 Eximstats V1.46
3225 Collect data on the number of addresses (recipients)
3226 as well as the number of messages.
3228 SC/05 Eximstats V1.47
3229 Added 'Message too big' to the list of mail rejection
3230 reasons (thanks to Marco Gaiarin).
3232 SC/06 Eximstats V1.48
3233 Mainlog lines which have GMT offsets and are too short to
3234 have a flag are now skipped.
3236 SC/07 Eximstats V1.49 (Alain Williams)
3237 Added the -emptyok flag.
3239 SC/08 Eximstats V1.50
3240 Fixes for obtaining the IP address from reject messages.
3242 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3243 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3244 whitespace changes from 4.64-PH/27
3246 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3249 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3250 are found, allow negative numbers in numeric criteria)
3252 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3254 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3255 to match changes made in 4.64-PH/43
3257 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3259 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3260 "domain" is actually an IP address, and force a failure. This locks out
3261 those revolvers/nameservers that support "A-for-A" lookups, in
3262 contravention of the specifications.
3264 PH/48 When a host name was looked up from an IP address, and the subsequent
3265 forward lookup of the name timed out, the host name was left in
3266 $sender_host_name, contrary to the specification.
3268 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3269 restricted to single-key lookups, Exim was not diagnosing an error if
3270 * or *@ was used with a query-style lookup.
3272 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3274 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3275 long ago, but noone interested enough thought of it. Let's just say that
3276 the "1.1" means that there are some new functions that weren't there at
3277 some point in the past.
3279 PH/51 Error processing for expansion failure of helo_data from an smtp
3280 transport during callout processing was broken.
3282 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3283 tested/used via the -bh/-bhc/-bs options.
3285 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3286 bug, fixed in subsequent PCRE releases).
3288 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3289 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3291 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3297 SC/01 Use a glob alias rather than an array ref in eximstats generated
3298 parser. This improves both readability and performance.
3300 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3301 Don't display local sender or destination tables in eximstats unless
3302 there is data to show.
3303 Added average volumes into the eximstats top table text output.
3305 SC/03 Collect data on the number of addresses (recipients) as well
3306 as the number of messages in eximstats.
3308 TF/01 Correct an error in the documentation for the redirect router. Exim
3309 does (usually) call initgroups() when daemonizing.
3311 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3312 with consistent privilege compared to when running as a daemon.
3314 TF/03 Note in the spec that $authenticated_id is not set for local
3315 submissions from trusted users.
3317 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3318 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3320 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3321 by adding some example configuration directives to the default
3322 configuration file. A little bit of work is required to uncomment the
3323 directives and define how usernames and passwords are checked, but
3324 there is now a framework to start from.
3326 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3327 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3328 without this. I don't know how relevant this is to other LDAP libraries.
3330 PH/02 Add the verb name to the "unknown ACL verb" error.
3332 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3334 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3336 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3337 directory not expanded when it should be if an expanded home directory
3338 was set for the address (which is overridden by the transport).
3340 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3343 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3344 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3345 because it is too late at that time, and has no effect.
3347 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3348 security issue with \' (bugzilla #107). I could not use the
3349 PQescapeStringConn() function, because it needs a PGconn value as one of
3352 PH/08 When testing addresses using -bt, indicate those final addresses that
3353 are duplicates that would not cause an additional delivery. At least one
3354 person was confused, thinking that -bt output corresponded to deliveries.
3355 (Suppressing duplicates isn't a good idea as you lose the information
3356 about possibly different redirections that led to the duplicates.)
3358 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3359 systems where poll() doesn't work, in particular OS X.
3361 PH/10 Added more information to debugging output for retry time not reached.
3363 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3364 operations in malware.c.
3366 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3369 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3370 syslog_duplication set false, log lines that would normally be written
3371 both the the main log and to the reject log were not written to syslog at
3374 PH/14 In the default configuration, change the use of "message" in ACL warn
3375 statements to "add_header".
3377 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3378 not followed by a command (e.g. "seen endif").
3380 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3381 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3384 PH/17 Added extra conditions to the default value of delay_warning_condition
3388 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3389 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3390 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3393 The Auto-Submitted: and various List- headers are standardised, whereas I
3394 don't think Precedence: ever was.
3396 PH/18 Refactored debugging code in route_finduser() to show more information,
3397 in particular, the error code if getpwnam() issues one.
3399 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3400 This is apparently needed in addition to the PH/07 change above to avoid
3401 any possible encoding problems.
3403 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3404 but not after initializing Perl.
3406 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3407 output them only if debugging. By default they are written stderr,
3408 apparently, which is not desirable.
3410 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3413 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3416 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3418 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3419 authenticated or an ident call has been made. Suppress the default
3420 values for $authenticated_id and $authenticated_sender (but permit -oMai
3421 and -oMas) when testing with -bh.
3423 PH/24 Re-jigged the order of the tests in the default configuration so that the
3424 tests for valid domains and recipients precede the DNS black list and CSA
3425 tests, on the grounds that those ones are more expensive.
3427 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3428 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3429 This bug exists in every version of Exim that I still have, right back to
3432 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3433 However, an attempt to turn on case-sensitivity in a regex key by
3434 including (?-i) didn't work because the subject string was already
3435 lowercased, and the effects were non-intuitive. It turns out that a
3436 one-line patch can be used to allow (?-i) to work as expected.
3442 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3443 other effects) broke the use of negated acl sub-conditions.
3445 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3448 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3449 "Deny" causes Exim to reject the incoming connection with a 554 error.
3450 Unfortunately, if there is a major crisis, such as a disk failure,
3451 tcp-wrappers gives "deny", whereas what one would like would be some
3452 kind of temporary error. A kludge has been added to help with this.
3453 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3454 554 error is used if errno is still zero or contains ENOENT (which occurs
3455 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3458 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3460 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3461 errors. Otherwise a message that provokes a temporary error (when other
3462 messages do not) can cause a whole host to time out.
3464 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3465 addresses were routed directly to files or pipes from a redirect router.
3466 File deliveries just didn't batch; pipe deliveries might have suffered
3469 PH/06 A failure to get a lock for a hints database would erroneously always say
3470 "Failed to get write lock", even when it was really a read lock.
3472 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3473 of 0600. This has been changed to use the value of the lockfile_mode
3474 option (which defaults to 0600).
3476 PH/08 Applied small patch from the Sieve maintainer.
3478 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3479 folder from quota calculations, a direct delivery into this folder messed
3480 up the contents of the maildirsize file. This was because the regex was
3481 used only to exclude .Trash (or whatever) when the size of the mailbox
3482 was calculated. There was no check that a delivery was happening into an
3483 excluded directory. This bug has been fixed by ignoring all quota
3484 processing for deliveries into excluded directories.
3486 PH/10 Added the maildirfolder_create_regex option to appendfile.
3492 PH/01 The code for finding all the local interface addresses on a FreeBSD
3493 system running IPv6 was broken. This may well have applied to all BSD
3494 systems, as well as to others that have similar system calls. The broken
3495 code found IPv4 interfaces correctly, but gave incorrect values for the
3496 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3497 that it would not match correctly against @[] and not recognize the IPv6
3500 PH/02 The ipliteral router was not recognizing addresses of the form user@
3501 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3503 PH/03 Added disable_ipv6.
3505 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3506 decoded form, because it is most often used to construct To: headers
3507 lines in autoreplies, and the decoded form may well be syntactically
3508 invalid. However, $reply_address has leading white space removed, and all
3509 newlines turned into spaces so that the autoreply transport does not
3512 PH/05 If group was specified without a user on a router, and no group or user
3513 was specified on a transport, the group from the router was ignored.
3515 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3516 for visible compile-time settings that can be used to change these
3517 numbers, for those that want even more. Backwards compatibility with old
3518 spool files has been maintained. However, going back to a previous Exim
3519 release will lost any variables that are in spool files.
3521 PH/07 Two small changes when running in the test harness: increase delay when
3522 passing a TCP/IP connection to a new process, in case the original
3523 process has to generate a bounce, and remove special handling of
3524 127.0.0.2 (sic), which is no longer necessary.
3526 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3527 be the same on different OS.
3529 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3532 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3533 whether --show-vars was specified or not
3535 JJ/02 exipick: Added support for new ACL variable spool format introduced
3538 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3539 syntactically invalid From: or Reply-to: line, and a filter used this to
3540 generate an autoreply, and therefore failed to obtain an address for the
3541 autoreply, Exim could try to deliver to a non-existent relative file
3542 name, causing unrelated and misleading errors. What now happens is that
3543 it logs this as a hard delivery error, but does not attempt to create a
3546 PH/11 The exinext utility has a -C option for testing purposes, but although
3547 the given file was scanned by exinext itself; it wasn't being passed on
3548 when Exim was called.
3550 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3551 an end-of-file indication when reading a command response.
3553 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3554 compiled. In many other places in Exim, IPv6 addresses are always
3555 recognized, so I have changed this. It also means that IPv4 domain
3556 literals of the form [IPV4:n.n.n.n] are now always recognized.
3558 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3559 used if the router is not running as root, for example, when verifying at
3560 ACL time, or when using -bh. The debugging output from this situation was
3561 non-existent - all you got was a failure to exec. I have made two
3564 (a) Failures to set uid/gid, the current directory, or a process leader
3565 in a subprocess such as that created by queryprogram now generate
3566 suitable debugging output when -d is set.
3568 (b) The queryprogram router detects when it is not running as root,
3569 outputs suitable debugging information if -d is set, and then runs
3570 the subprocess without attempting to change uid/gid.
3572 PH/15 Minor change to Makefile for building test_host (undocumented testing
3575 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3576 additional section of a DNS packet that returns MX or SRV records.
3577 Instead, it always explicitly searches for A/AAAA records. This avoids
3578 major problems that occur when a DNS server includes only records of one
3579 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3580 fixed another bug: if SRV records were looked up and the corresponding
3581 address records were *not* found in the additional section, the port
3582 values from the SRV records were lost.
3584 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3585 using the correct key (the original address) when searching the retry
3586 rules in order to find which one to use for generating the retry hint.
3588 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3589 adding the default one. Similarly, if it contains a Reply-To: header, the
3590 errors_reply_to option, if set, is not used.
3592 PH/19 When calculating a retry time, Exim used to measure the "time since
3593 failure" by looking at the "first failed" field in the retry record. Now
3594 it does not use this if it is later than than the arrival time of the
3595 message. Instead it uses the arrival time. This makes for better
3596 behaviour in cases where some deliveries succeed, thus re-setting the
3597 "first failed" field. An example is a quota failure for a huge message
3598 when small messages continue to be delivered. Without this change, the
3599 "time since failure" will always be short, possible causing more frequent
3600 delivery attempts for the huge message than are intended.
3601 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3603 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3604 $1, $2, $3) because the numerical variables can be reset during some
3605 expansion items (e.g. "match"), thereby losing the authentication data.
3607 PH/21 Make -bV show the size of off_t variables so that the test suite can
3608 decide whether to run tests for quotas > 2G.
3610 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3611 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3612 filecount value is greater than 2G or if a quota value is greater than 2G
3613 on a system where the size of off_t is not greater than 4, a panic error
3616 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3617 never match. The debug and -bh output now contains an explicit error
3618 message indicating a malformed IPv4 address or mask.
3620 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3621 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3622 PH/23 above applies.
3624 PH/25 Do not write to syslog when running in the test harness. The only
3625 occasion when this arises is a failure to open the main or panic logs
3626 (for which there is an explicit test).
3628 PH/26 Added the /no_tell option to "control=freeze".
3630 PH/27 If a host name lookup failed very early in a connection, for example, if
3631 the IP address matched host_lookup and the reverse lookup yielded a name
3632 that did not have a forward lookup, an error message of the form "no IP
3633 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3634 could be logged. Now it outputs the IP address instead of "NULL".
3636 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3637 allows the sender and the authenticated sender to be set when
3638 submitting a message from within Exim. Since child_open_exim() is
3639 documented for local_scan(), the new function should be too.
3641 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3642 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3643 results in an empty string is now treated as unset.
3645 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3647 PH/31 Added sender_verify_fail logging option.
3649 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3650 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3651 tidied the source and removed it altogether.
3653 PH/33 When a queue run was abandoned because the load average was too high, a
3654 log line was always written; now it is written only if the queue_run log
3655 selector is set. In addition, the log line for abandonment now contains
3656 information about the queue run such as the pid. This is always present
3657 in "start" and "stop" lines but was omitted from the "abandon" line.
3659 PH/34 Omit spaces between a header name and the colon in the error message that
3660 is given when verify = headers_syntax fails (if there are lots of them,
3661 the message gets confusing).
3663 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3664 names, as there are now some PTR records that contain slashes. This check
3665 is only to protect against broken name servers that fall over on strange
3666 characters, so the fact that it applies to all lookups doesn't matter.
3668 PH/36 Now that the new test suite is complete, we can remove some of the
3669 special code in Exim that was needed for the old test suite. For example,
3670 sorting DNS records because real resolvers return them in an arbitrary
3671 order. The new test suite's fake resolver always returns records in the
3674 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3675 bounces) except when queue_only is set, to avoid logging races between
3676 the different processes.
3678 PH/38 Panic-die if .include specifies a non-absolute path.
3680 PH/39 A tweak to the "H" retry rule from its user.
3682 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3683 a label. They prevented compilation on older perls.
3685 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3686 a warning to be raised on newish perls.
3688 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3689 on queue. Changes to match documented behaviour of showing count of
3690 messages matching specified criteria.
3692 PH/40 Changed the default ident timeout from 30s to 5s.
3694 PH/41 Added support for the use of login_cap features, on those BSD systems
3695 that have them, for controlling the resources used by pipe deliveries.
3697 PH/42 The content-scanning code uses fopen() to create files in which to put
3698 message data. Previously it was not paying any attention to the mode of
3699 the files. Exim runs with umask(0) because the rest of the code creates
3700 files with open(), and sets the required mode explicitly. Thus, these
3701 files were ending up world-writeable. This was not a big issue, because,
3702 being within the spool directory, they were not world-accessible. I have
3703 created a function called modefopen, which takes an additional mode
3704 argument. It sets umask(777), creates the file, chmods it to the required
3705 mode, then resets the umask. All the relevant calls to fopen() in the
3706 content scanning code have been changed to use this function.
3708 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3709 to 24 hours. This avoids potential overflow problems when processing G
3710 and H retry rules. I suspect nobody ever tinkers with this value.
3712 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3714 PH/45 When the plaintext authenticator is running as a client, the server's
3715 challenges are checked to ensure they are valid base64 strings. By
3716 default, the authentication attempt is cancelled if an invalid string is
3717 received. Setting client_ignore_invalid_base64 true ignores these errors.
3718 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3719 they are received. Thus, the responses can be made to depend on the
3720 challenges. If an invalid string is ignored, an empty string is placed in
3723 PH/46 Messages that are created by the autoreply transport now contains a
3724 References: header, in accordance with RFCs 2822 and 3834.
3726 PH/47 Added authenticated_sender_force to the smtp transport.
3728 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3730 PH/49 Installed latest patch from the Sieve maintainer.
3732 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3733 was also set, the appendfile transport was unnecessarily scanning a
3734 directory of message files (e.g. for maildir delivery) to find the count
3735 of files (along with the size), even though it did not need this
3736 information. It now does the scan only if it needs to find either the
3737 size of the count of files.
3739 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3741 PH/52 Two bugs concerned with error handling when the smtp transport is
3744 (i) Exim was not creating retry information for temporary errors given
3745 for individual recipients after the DATA command when the smtp transport
3746 was used in LMTP mode. This meant that they could be retried too
3747 frequently, and not timed out correctly.
3749 (ii) Exim was setting the flag that allows error details to be returned
3750 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3751 recipients that were returned after the DATA command.
3753 PH/53 This is related to PH/52, but is more general: for any failing address,
3754 when detailed error information was permitted to be returned to the
3755 sender, but the error was temporary, then after the final timeout, only
3756 "retry timeout exceeded" was returned. Now it returns the full error as
3757 well as "retry timeout exceeded".
3759 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3760 do this, and (what is worse) MTAs that accept it.
3762 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3763 will now be deprecated.
3765 PH/56 New os.c-cygwin from the Cygwin maintainer.
3767 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3768 formats (previously only available in exim formats via -bpr, -bpru,
3769 and -bpra. Now also available in native and exiqgrep formats)
3771 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3772 with very large, slow to parse queues
3774 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3776 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3778 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3779 responses to authentication challenges, though it was showing the
3780 challenges; (ii) I've removed the CR characters from the debug output for
3783 PH/58 Allow for the insertion of a newline as well as a space when a string
3784 is turned into more than one encoded-word during RFC 2047 encoding. The
3785 Sieve code now uses this.
3787 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3788 data_4xx, lost_connection, tls_required.
3790 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3791 message was being sent as an SMTP response.
3793 PH/61 Add -l and -k options to exicyclog.
3795 PH/62 When verifying, if an address was redirected to one new address, so that
3796 verification continued, and the new address failed or deferred after
3797 having set something in $address_data, the value of $address_data was not
3798 passed back to the ACL. This was different to the case when no
3799 redirection occurred. The value is now passed back in both cases.
3801 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3802 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3803 use login_cap.h, so on its own it isn't the distinguishing feature. The
3804 new name refers directly to the setclassresources() function.
3806 PH/65 Added configuration files for NetBSD3.
3808 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3810 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3811 is preferred over IPv4.
3813 PH/68 The bounce_return_message and bounce_return_body options were not being
3814 honoured for bounces generated during the reception of non-SMTP messages.
3815 In particular, this applied to messages rejected by the ACL. This bug has
3816 been fixed. However, if bounce_return_message is true and bounce_return_
3817 body is false, the headers that are returned for a non-SMTP message
3818 include only those that have been read before the error was detected.
3819 (In the case of an ACL rejection, they have all been read.)
3821 PH/69 The HTML version of the specification is now built in a directory called
3822 spec_html instead of spec.html, because the latter looks like a path with
3823 a MIME-type, and this confuses some software.
3825 PH/70 Catch two compiler warnings in sieve.c.
3827 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3828 function verify_get_ident() calls ip_connect() to connect a socket, but
3829 if the "connect()" function timed out, ip_connect() used to close the
3830 socket. However, verify_get_ident() also closes the socket later, and in
3831 between Exim writes to the log, which may get opened at this point. When
3832 the socket was closed in ip_connect(), the log could get the same file
3833 descriptor number as the socket. This naturally causes chaos. The fix is
3834 not to close the socket in ip_connect(); the socket should be closed by
3835 the function that creates it. There was only one place in the code where
3836 this was missing, in the iplookup router, which I don't think anybody now
3837 uses, but I've fixed it anyway.
3839 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3840 well as to direct DNS lookups. Otherwise the handling of names in host
3841 lists is inconsistent and therefore confusing.
3847 PH/01 Two changes to the default runtime configuration:
3849 (1) Move the checks for relay_from_hosts and authenticated clients from
3850 after to before the (commented out) DNS black list checks.
3852 (2) Add control=submission to the relay_from_hosts and authenticated
3853 clients checks, on the grounds that messages accepted by these
3854 statements are most likely to be submissions.
3856 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3858 (1) Generate an error if the third argument for the ${prvs expansion is
3861 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3864 (3) Reset the variables that are obtained from the first argument of
3865 ${prvscheck and used in the second argument before leaving the code,
3866 because their memory is reclaimed, so using them afterwards may do
3869 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3870 one (it's much easier than Tom thought :-).
3872 (5) Because of (4), we can now allow for the use of $prvscheck_result
3873 inside the third argument.
3875 PH/03 For some reason, the default setting of PATH when running a command from
3876 a pipe transport was just "/usr/bin". I have changed it to
3879 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3880 anything to be listed in the output from -bV.
3882 PH/05 When a filter generated an autoreply, the entire To: header line was
3883 quoted in the delivery log line, like this:
3885 => >A.N.Other <ano@some.domain> <original@ddress> ...
3887 This has been changed so that it extracts the operative address. There
3888 may be more than one such address. If so, they are comma-separated, like
3891 => >ano@some.domain,ona@other.domain <original@ddress> ...
3893 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3894 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3895 not being looked up in the rDNS to get a host name, Exim was showing the
3896 IP address twice in Received: lines, even though the IP addresses were
3897 identical. For example:
3899 Received: from [1.2.3.4] (helo=[1.2.3.4])
3901 However, if the real host name was known, it was omitting the HELO data
3902 if it matched the actual IP address. This has been tidied up so that it
3903 doesn't show the same IP address twice.
3905 PH/07 When both +timestamp and +memory debugging was on, the value given by
3906 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3907 called by the debug printing, thereby overwriting the timestamp buffer.
3908 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3910 PH/08 When the original message was included in an autoreply transport, it
3911 always said "this is a copy of the message, including all the headers",
3912 even if body_only or headers_only was set. It now gives an appropriate
3915 PH/09 Applied a patch from the Sieve maintainer which:
3917 o fixes some comments
3918 o adds the (disabled) notify extension core
3919 o adds some debug output for the result of if/elsif tests
3920 o points to the current vacation draft in the documentation
3921 and documents the missing references header update
3925 o fixes a bug in processing the envelope test (when testing
3926 multiple envelope elements, the last element determined the
3929 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3930 Electronic Mail") by including:
3932 Auto-submitted: auto-generated
3934 in the messages that it generates (bounce messages and others, such as
3935 warnings). In the case of bounce messages for non-SMTP messages, there was
3936 also a typo: it was using "Auto_submitted" (underscore instead of
3937 hyphen). Since every message generated by Exim is necessarily in response
3938 to another message, thes have all been changed to:
3940 Auto-Submitted: auto-replied
3942 in accordance with these statements in the RFC:
3944 The auto-replied keyword:
3946 - SHOULD be used on messages sent in direct response to another
3947 message by an automatic process,
3949 - MUST NOT be used on manually-generated messages,
3951 - MAY be used on Delivery Status Notifications (DSNs) and Message
3952 Disposition Notifications (MDNs),
3954 - MUST NOT be used on messages generated by automatic or periodic
3955 processes, except for messages which are automatic responses to
3958 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3959 to the default Received: header definition.
3961 PH/12 Added log selector acl_warn_skipped (default on).
3963 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3964 variables because (a) they are in the wrong storage pool and (b) even if
3965 they were copied, it wouldn't work properly because of the caching.
3967 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3968 checking when decoding. Apparently there are clients that generate
3969 overlong encoded strings. Why am I not surprised?
3971 PH/15 If the first argument of "${if match_address" was not empty, but did not
3972 contain an "@" character, Exim crashed. Now it writes a panic log message
3973 and treats the condition as false.
3975 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3977 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3978 extension "envelope-auth". The code is finished and in agreement with
3979 other implementations, but there is no documentation so far and in fact,
3980 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3981 not changing the active code.
3983 Print executed "if" and "elsif" statements when debugging is used. This
3984 helps a great deal to understand what a filter does.
3986 Document more things not specified clearly in RFC3028. I had all this
3987 sorted out, when out of a sudden new issues came to my mind. Oops."
3989 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3992 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3993 canonical form (as documented). However, after a host name lookup from
3994 the IP address, check_host() was doing a simple string comparison with
3995 addresses acquired from the DNS when checking that the found name did
3996 have the original IP as one of its addresses. Since any found IPv6
3997 addresses are likely to be in abbreviated form, the comparison could
3998 fail. Luckily, there already exists a function for doing the comparison
3999 by converting both addresses to binary, so now that is used instead of
4000 the text comparison.
4002 PH/20 There was another similar case to PH/19, when a complete host name was
4003 given in a host list; looking up its IP address could give an abbreviated
4004 form, whereas the current host's name might or might not be abbreviated.
4005 The same fix has been applied.
4011 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4012 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4015 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4016 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4018 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4020 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4021 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4022 Also, for "Auto-Submitted", treat anything other than "no" as
4023 non-personal, in accordance with RFC 3834. (Previously it treated
4024 anything starting "auto-" as non-personal.)
4026 TF/01 The control=submission/name=... option had a problem with syntax
4027 errors if the name included a slash character. The /name= option
4028 now slurps the rest of the string, so it can include any characters
4029 but it must come last in the list of options (after /sender_retain
4032 PH/05 Some modifications to the interface to the fake nameserver for the new
4040 TK/01 Added the "success_on_redirect" address verification option. See
4041 NewStuff for rationale and an example.
4043 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4045 PH/02 Patch to exigrep to allow it to work on syslog lines.
4047 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4048 fread() to skip over the body file's header line, because in Cygwin the
4049 header line is locked and is inaccessible.
4051 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4052 co-exist for some time) to make it clear that it is the Exim ID that is
4053 referenced, not the Message-ID: header line.
4055 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4056 string_format() function, because snprintf() does not exist on all
4059 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4060 ${stat: expansion item.
4062 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4063 protocol synchronization error", to keep the pedants happy.
4065 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4066 well as for IRIX systems, when gcc is being used. See the host.c source
4069 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4071 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4074 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4075 transport and to the smtp transport in LMTP mode.
4077 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4079 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4080 run a filter in a subprocess. This could lead to confusion in subsequent
4081 lookups in the parent process. There should also be a search_tidyup() at
4082 the end of the subprocess.
4084 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4085 only if the host matched helo_try_verify_hosts, which caused the
4086 verification to occur when the EHLO/HELO command was issued. The ACL just
4087 tested the remembered result. Now, if a previous verification attempt has
4088 not happened, "verify = helo" does it there and then.
4090 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4092 TK/03 Fix log output including CR from clamd.
4094 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4095 exist provoked a memory error which could cause a segfault.
4097 PH/15 Installed PCRE 6.2
4099 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4101 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4102 of the problem. Specifically, suggested +O2 rather than +O1 for the
4105 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4107 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4108 record, the port was not correctly logged when the outgoing_port log
4109 selector was set (it logged the transort's default port).
4111 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4112 fallback_hosts, and "hosts" in the smtp transport.
4114 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4115 host errors such as "Connection refused".
4117 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4118 authentication with radiusclient 0.4.9:
4120 - Error returned from rc_read_config was caught wrongly
4121 - Username/password not passed on to radius server due to wrong length.
4123 The presumption is that some radiusclient API changes for 4.51/PH/17
4124 were not taken care of correctly. The code is still untested by me (my
4125 Linux distribution still has 0.3.2 of radiusclient), but it was
4126 contributed by a Radius user.
4128 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4129 expanding the "port" option of the smtp transport.
4131 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4132 while reading a MIME header. Thanks to Tom Hughes for a patch.
4134 PH/24 Include config.h inside local_scan.h so that configuration settings are
4137 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4138 that in an ACL for RCPT (for example), you can examine exactly what was
4141 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4142 commands, but it was not correctly comparing the address with the actual
4143 client host address. Thus, it would show the EHLO address in Received:
4144 header lines when this was not necessary.
4146 PH/27 Added the % operator to ${eval:}.
4148 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4149 it should be ignoring failures (because with -C, for example, it has lost
4150 privilege). It wasn't ignoring creation failures other than "already
4153 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4156 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4157 because an input line was too long, either on its own, or by virtue of
4158 too many continuations, the temporary file was not being removed, and the
4159 return code was incorrect.
4161 PH/31 Missing "BOOL" in function definition in filtertest.c.
4163 PH/32 Applied Sieve patches from the maintainer.
4165 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4167 PH/33 Added "verify = not_blind".
4169 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4170 Local/Makefile (with some defaults set). These are used in built scripts
4171 such as exicyclog, but they have never been used in the exim_install
4172 script (though there are many overriding facilities there). I have
4173 arranged that the exim_install script now takes note of these two
4176 PH/35 Installed configuration files for Dragonfly.
4178 PH/36 When a locally submitted message by a trusted user did not contain a
4179 From: header, and the sender address was obtained from -f or from an SMTP
4180 MAIL command, and the trusted user did not use -F to supply a sender
4181 name, $originator_name was incorrectly used when constructing a From:
4182 header. Furthermore, $originator_name was used for submission mode
4183 messages from external hosts without From: headers in a similar way,
4184 which is clearly wrong.
4186 PH/37 Added control=suppress_local_fixups.
4188 PH/38 When log_selector = +received_sender was set, and the addition of the
4189 sender made the log line's construction buffer exactly full, or one byte
4190 less than full, an overflow happened when the terminating "\n" was
4193 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4194 when the result of a list match is failure because a DNS lookup failed.
4196 PH/40 RM_COMMAND is now used in the building process.
4198 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4199 the "build-* directories that it finds.
4201 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4202 address in a domain literal was a prefix of an interface address.
4204 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4205 when verifying a sender address, unless rewrite_headers is false.
4207 PH/44 Wrote a long comment about why errors_to addresses are verified as
4208 recipients, not senders.
4210 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4211 the ratelimit ACL was added.
4213 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4215 PH/46 Added extra information about PostgreSQL errors to the error string.
4217 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4218 avoiding the need to install special zones in a real server. This is
4219 backwards compatible; if it can't find the fake resolver, it drops back.
4220 Thus, both old and new test suites can be run.
4222 TF/02 Added util/ratelimit.pl
4224 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4225 clock is set back in time.
4227 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4228 Candler <B.Candler@pobox.com>.
4230 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4231 OK for addresses that are the result of redirections.
4233 PH/48 A number of further additions for the benefit of the new test suite,
4234 including a fake gethostbyname() that interfaces to the fake DNS resolver
4237 TF/06 The fix for widen_domains has also been applied to qualify_single and
4238 search_parents which are the other dnslookup options that can cause
4241 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4244 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4246 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4247 service" if it encountered an MX record with an empty target hostname.
4248 The message is now "an MX or SRV record indicated no SMTP service".
4250 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4251 if the DNS of the sending site is misconfigured. This is quite a
4252 common situation. This change restores the behaviour of treating a
4253 helo verification defer as a failure.
4255 PH/51 If self=fail was set on a router, the bounce message did not include the
4256 actual error message.
4262 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4264 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4265 command itself ended in error, the underlying message about the transport
4266 filter timeout was being overwritten with the pipe command error. Now the
4267 underlying error message should be appended to the second error message.
4269 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4271 PH/02 Reduce the amount of output that "make" produces by default. Full output
4272 can still be requested.
4274 PH/03 The warning log line about a condition test deferring for a "warn" verb
4275 was being output only once per connection, rather than after each
4276 occurrence (because it was using the same function as for successful
4277 "warn" verbs). This seems wrong, so I have changed it.
4279 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4280 it should not have, which might have caused a crash in the right
4281 circumstances, but probably never did.
4283 PH/04 Installed a modified version of Tony Finch's patch to make submission
4284 mode fix the return path as well as the Sender: header line, and to
4285 add a /name= option so that you can make the user's friendly name appear
4288 TF/03 Added the control = fakedefer ACL modifier.
4290 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4291 Mark Lowes for thorough testing.
4293 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4295 TK/03 Merged latest SRS patch from Miles Wilton.
4297 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4298 with the definition in sysexits.h (which is #included earlier).
4299 Fortunately, Exim does not actually use EX_OK. The code used to try to
4300 preserve the sysexits.h value, by assuming that macro definitions were
4301 scanned for macro replacements. I have been disabused of this notion,
4302 so now the code just undefines EX_OK before #including unistd.h.
4304 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4305 in the smtp transport. When a block could not be written in a single
4306 write() function, the timeout was being re-applied to each part-write.
4307 This seems wrong - if the receiver was accepting one byte at a time it
4308 would take for ever. The timeout is now adjusted when this happens. It
4309 doesn't have to be particularly precise.
4311 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4312 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4314 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4315 without a domain if the check to <postmaster@domain> fails.
4317 SC/01 Eximstats: added -xls and the ability to specify output files
4318 (patch written by Frank Heydlauf).
4320 SC/02 Eximstats: use FileHandles for outputting results.
4322 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4324 SC/04 Eximstats: fixed display of large numbers with -nvr option
4326 SC/05 Eximstats: fixed merging of reports with empty tables.
4328 SC/06 Eximstats: added the -include_original_destination flag
4330 SC/07 Eximstats: removed tabs and trailing whitespace.
4332 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4334 TK/06 MBOX spool code: Add real "From " MBOX separator line
4335 so the .eml file is really in mbox format (even though
4336 most programs do not really care). Patch from Alex Miller.
4338 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4339 The latter is generated from $received_to and is only set if the
4340 message has one envelope recipient. SA can use these headers,
4341 obviously out-of-the-box. Patch from Alex Miller.
4343 PH/08 The ${def test on a variable was returning false if the variable's
4344 value was "0", contrary to what the specification has always said!
4345 The result should be true unless the variable is empty.
4347 PH/09 The syntax error of a character other than { following "${if
4348 def:variable_name" (after optional whitespace) was not being diagnosed.
4349 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4350 accidental colon was present, for example, could give incorrect results.
4352 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4353 result is used (not including appendfile, where other changes are about
4356 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4357 This involved changing a lot of size variables from int to off_t. It
4358 should work with maildirs and everything.
4360 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4361 spamd dying while we are connected to it.
4363 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4366 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4367 function for BDB 4.3.
4369 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4371 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4372 the sources for PCRE have been re-arranged and more files are now
4375 PH/15 The code I had for printing potentially long long variables in PH/11
4376 above was not the best (it lost precision). The length of off_t variables
4377 is now inspected at build time, and an appropriate printing format (%ld
4378 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4379 to be "long long int" or "long int". This is needed for the internal
4380 formatting function string_vformat().
4382 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4383 the configuration file to be ":syslog", then the script "guesses" where
4384 the logs files are, rather than using the compiled in default. In our
4385 case the guess is not the same as the compiled default, so the script
4386 suddenly stopped working when I started to use syslog. The patch checks
4387 to see if log_file_path is "". If so, it attempts to read it from exim
4388 with no configuration file to get the compiled in version, before it
4389 falls back to the previous guessing code."
4391 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4392 implementing BATV in an Exim configuration. See NewStuff for the gory
4395 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4396 Makefile that are specific to HP-UX.
4398 PH/18 If the "use_postmaster" option was set for a recipient callout together
4399 with the "random" option, the postmaster address was used as the MAIL
4400 FROM address for the random test, but not for the subsequent recipient
4401 test. It is now used for both.
4403 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4404 patch removes a few documentation additions to RFC 3028, because the
4405 latest draft now contains them. It adds the new en;ascii-case comparator
4406 and a new error check for 8bit text in MIME parts. Comparator and
4407 require names are now matched exactly. I enabled the subaddress
4408 extension, but it is not well tested yet (read: it works for me)."
4410 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4411 rework some of the code of TK/09 above to avoid the hardwired use of
4412 "%lld" and "long long". Replaced the call to snprintf() with a call to
4415 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4416 records point to non-existent hosts", "retry timeout exceeded", and
4417 "retry time not reached for any host after a long failure period".
4419 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4420 experimental DomainKeys support:
4422 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4423 (2) On an error such as an illegally used "control", the wrong name for
4424 the control was given.
4426 These problems did NOT occur unless DomainKeys support was compiled.
4428 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4430 PH/24 Added ${if match_ip condition.
4432 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4433 databases so that it will be absolutely obvious if a crash occurs in the
4434 DB library. This is a regular occurrence (often caused by mis-matched
4437 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4438 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4439 picked up on a user's system that detects such things. There doesn't seem
4440 to be a gcc warning option for this - only an attribute that has to be
4441 put on the function's prototype. It seems that in Fedora Core 4 they have
4442 set this on a number of new functions. No doubt there will be more in due
4445 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4446 specify a transport. However, if an address that was verified by such a
4447 router was the subject of a callout, Exim crashed because it tried to
4448 read the rcpt_include_affixes from the non-existent transport. Now it
4449 just assumes that the setting of that option is false. This bug was
4450 introduced by 4.51/PH/31.
4452 PH/28 Changed -d+all to exclude +memory, because that information is very
4453 rarely of interest, but it makes the output a lot bigger. People tend to
4454 do -d+all out of habit.
4456 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4457 code in os-type was giving problems when libc.so lives in lib64, like on
4460 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4461 aren't the modern standard, and it seems that some systems' include files
4462 don't always have them. Exim was already checking for some of the newer
4463 ones like T_AAAA, and defining it itself. I've added checks for all the
4464 record types that Exim uses.
4466 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4467 not automatically generating a new one, as it is supposed to. This
4468 prevented TLS from working. If the file did exist, but contained invalid
4469 data, a new version was generated, as expected. It was only the case of a
4470 non-existent file that was broken.
4472 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4473 with a change in libdomainkeys > 0.64.
4475 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4476 from DNS. If the selector record carries the flag, it now has
4477 precedence over the domain-wide flag.
4479 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4481 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4482 the use of a port name that isn't defined in /etc/services) occurred, the
4483 message was deferred as in a normal delivery, and thus remained on the
4484 spool, instead of being failed because of the mua_wrapper setting. This
4485 is now fixed, and I tidied up some of the mua_wrapper messages at the
4488 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4489 the messages in a hash of arrays rather than using individual hashes.
4490 This is a bit cleaner and results in dramatic memory savings, albeit
4491 at a slight CPU cost.
4493 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4494 as requested by Marc Sherman.
4496 SC/10 Eximstats: added histograms for user specified patterns as requested
4499 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4501 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4502 fopen() in the content-scanning modules that did not already have it.
4508 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4509 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4511 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4513 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4515 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4516 to test Sieve filters that use "vacation".
4518 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4519 that changes the way the GnuTLS parameters are stored in the cache file.
4520 The new format can be generated externally. For backward compatibility,
4521 if the data in the cache doesn't make sense, Exim assumes it has read an
4522 old-format file, and it generates new data and writes a new file. This
4523 means that you can't go back to an older release without removing the
4526 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4527 work if there are any delivery delays because "one_time" forces the
4528 parent to be marked "delivered", so its unseen clone is never tried
4529 again. For this reason, Exim now forbids the simultaneous setting of
4532 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4533 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4534 entry if you want to know the details. The fix, however, neglected to
4535 consider the case when local delivery batching is involved. The test for
4536 "previously delivered" was not happening when checking to see if an
4537 address could be batched with a previous (undelivered) one; under
4538 certain circumstances this could lead to multiple deliveries to the same
4541 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4542 in its include files, and this causes problems building Exim.
4544 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4545 header_syntax) but Exim was just ignoring anything given after a slash.
4546 In particular, this caused confusion with an attempt to use "verify =
4547 reverse_host_lookup/defer_ok". An error is now given when options are
4548 supplied for verify items that do not have them. (Maybe reverse_host_
4549 lookup should have a defer_ok option, but that's a different point.)
4551 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4552 defined by RFC 821) to 2048, because there were problems with some AUTH
4553 commands, and RFC 1869 says the size should be increased for extended
4554 SMTP commands that take arguments.
4556 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4559 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4560 "unknown" error; now it says that the functionality isn't in the binary.
4562 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4563 an address' error message when a string expansion fails (syntax or
4564 whatever). Otherwise the password may appear in the log. Following change
4565 PH/42 below, there is no longer a chance of it appearing in a bounce
4568 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4570 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4571 of its addresses was ever tried. (Bugzilla bug #2.)
4573 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4574 the result incorrectly in the debug output. (It correctly added a newline
4575 to what was transported.)
4577 TF/01 Added $received_time.
4579 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4580 commented out examples of how to interface to a virus scanner and to
4581 SpamAssassin. Also added commented examples of av_scanner and
4582 spamd_address settings.
4584 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4585 and controls are allowed in which ACLs. There were a couple of minor
4586 errors. Some of the entries in the conditions table (which is a table of
4587 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4588 negation of where the condition IS allowed.
4590 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4592 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4593 header file does not have a version number, so I've had to invent a new
4594 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4595 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4596 radiusclient), but it was contributed by a Radius user.
4598 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4599 files or directories, for OpenSSL.
4601 PH/19 When an Exim process that is running as root has to create an Exim log
4602 file, it does so in a subprocess that runs as exim:exim so as to get the
4603 ownership right at creation (otherwise, other Exim processes might see
4604 the file with the wrong ownership). There was no test for failure of this
4605 fork() call, which would lead to the process getting stuck as it waited
4606 for a non-existent subprocess. Forks do occasionally fail when resources
4607 run out. I reviewed all the other calls to fork(); they all seem to check
4610 PH/20 When checking for unexpected SMTP input at connect time (before writing
4611 the banner), Exim was not dealing correctly with a non-positive return
4612 from the read() function. If the client had disconnected by this time,
4613 the result was a log entry for a synchronization error with an empty
4614 string after "input=" when read() returned zero. If read() returned -1
4615 (an event I could not check), uninitialized data bytes were printed.
4616 There were reports of junk text (parts of files, etc) appearing after
4619 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4621 PH/22 Added support for macro redefinition, and (re)definition in between
4622 driver and ACL definitions.
4624 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4625 forgetting to use the resulting value; it was using the unexpanded value.
4627 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4628 hadn't been configured. The fix is from Juergen Kreileder, who
4629 understands it better than I do:
4631 "Here's what I see happening with three configured cyrus_sasl
4632 authenticators configured (plain, login, cram-md5):
4634 On startup auth_cyrus_sasl_init() gets called for each of these.
4635 This means three calls to sasl_listmech() without a specified mech_list.
4636 => SASL tests which mechs of all available mechs actually work
4637 => three warnings about OTP not working
4638 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4640 With the patch, sasl_listmech() also gets called three times. But now
4641 SASL's mech_list option is set to the server_mech specified in the the
4642 authenticator. Or in other words, the answer from sasl_listmech()
4643 gets limited to just the mech you're testing for (which is different
4645 => the return list contains just 'plain' or 'login', 'cram-md5' or
4646 nothing depending on the value of ob->server_mech.
4648 I've just tested the patch: Authentication still works fine,
4649 unavailable mechs specified in the exim configuration are still
4650 caught, and the auth.log warnings about OTP are gone."
4652 PH/25 When debugging is enabled, the contents of the command line are added
4653 to the debugging output, even when log_selector=+arguments is not
4656 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4657 answer is "GNU", and only if the return is "GNU/something" is the answer
4660 PH/27 $acl_verify_message is now set immediately after the failure of a
4661 verification in an ACL, and so is available in subsequent modifiers. In
4662 particular, the message can be preserved by coding like this:
4664 warn !verify = sender
4665 set acl_m0 = $acl_verify_message
4667 Previously, $acl_verify_message was set only while expanding "message"
4668 and "log_message" when a very denied access.
4670 PH/28 Modified OS/os.c-Linux with
4672 -#ifndef OS_LOAD_AVERAGE
4673 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4675 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4676 nomenclature these days.)
4678 PH/29 Installed patch from the Sieve maintainer that adds the options
4679 sieve_useraddress and sieve_subaddress to the redirect router.
4681 PH/30 In these circumstances:
4682 . Two addresses routed to the same list of hosts;
4683 . First host does not offer TLS;
4684 . First host accepts first address;
4685 . First host gives temporary error to second address;
4686 . Second host offers TLS and a TLS session is established;
4687 . Second host accepts second address.
4688 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4689 and peerdn, if requested) that were in fact used only for the second
4692 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4693 attention to any local part prefix or suffix that was matched by the
4694 router that accepted the address. It now behaves in the same way as it
4695 does for delivery: the affixes are removed from the local part unless
4696 rcpt_include_affixes is set on the transport.
4698 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4699 timeout during the DATA phase of an incoming message.
4701 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4702 applied a patch sanctioned by the Sieve maintainer.
4704 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4705 the uid or gid is negative. A case of a negative gid caused this to be
4706 noticed. The fix allows for either to be negative.
4708 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4709 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4712 PH/36 The change PH/12 above was broken. Fixed it.
4714 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4715 the grounds that routing the same address twice would always produce the
4716 same answer. This might have been true once, but it is certainly no
4717 longer true now. Routing a child address may depend on the previous
4718 routing that produced that child. Some complicated redirection strategies
4719 went wrong when messages had multiple recipients, and made Exim's
4720 behaviour dependent on the order in which the addresses were given.
4722 I have moved the duplicate checking until after the routing is complete.
4723 Exim scans the addresses that are assigned to local and remote
4724 transports, and removes any duplicates. This means that more work will be
4725 done, as duplicates will always all be routed, but duplicates are
4726 presumably rare, so I don't expect this is of any significance.
4728 For deliveries to pipes, files, and autoreplies, the duplicate checking
4729 still happens during the routing process, since they are not going to be
4732 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4733 It corrects a timeout issue with spamd. This is Ian's comment: "The
4734 background is that sometimes spamd either never reads data from a
4735 connection it has accepted, or it never writes response data. The exiscan
4736 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4737 blindly assumes that writes won't block so it may never time out."
4739 PH/39 Allow G after quota size as well as K and M.
4741 PH/40 The value set for $authenticated_id in an authenticator may not contain
4742 binary zeroes or newlines because the value is written to log lines and
4743 to spool files. There was no check on this. Now the value is run through
4744 the string_printing() function so that such characters are converted to
4745 printable escape sequences.
4747 PH/41 $message_linecount is a new variable that contains the total number of
4748 lines in the message. Compare $body_linecount, which is the count for the
4751 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4752 bounce and delay warning messages, except in certain special cases, which
4755 (a) An SMTP error message from a remote host;
4756 (b) A message specified in a :fail: redirection;
4757 (c) A message specified in a "fail" command in a system filter;
4758 (d) A message specified in a FAIL return from the queryprogram router;
4759 (e) A message specified by the cannot_route_message router option.
4761 In these cases only, Exim does include the error details in bounce and
4762 warning messages. There are also a few cases where bland messages such
4763 as "unrouteable address" or "local delivery error" are given.
4765 PH/43 $value is now also set for the "else" part of a ${run expansion.
4767 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4768 being worked on, but at least Exim now implements the latest version to
4771 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4772 process to complete was treated as a delivery failure, a timeout while
4773 writing the message to the pipe was logged, but erroneously treated as a
4774 successful delivery. Such timeouts include transport filter timeouts. For
4775 consistency with the overall process timeout, these timeouts are now
4776 treated as errors, giving rise to delivery failures by default. However,
4777 there is now a new Boolean option for the pipe transport called
4778 timeout_defer, which, if set TRUE, converts the failures into defers for
4779 both kinds of timeout. A transport filter timeout is now identified in
4782 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4783 systems where "make" and "gmake" are different, calling "gmake" at top
4784 level broke things. I've arranged for the value of $(MAKE) to be passed
4785 from the Makefile to this script so that it can call the same version of
4789 A note about Exim versions 4.44 and 4.50
4790 ----------------------------------------
4792 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4793 changes of various kinds. As a consequence, a big documentation update was
4794 needed. This delayed the release for rather longer than seemed good, especially
4795 in the light of a couple of (minor) security issues. Therefore, the changes
4796 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4797 release. So 4.44 and 4.50 are in effect two different branches that both start
4800 I have left the 4.50 change log unchanged; it contains all the changes since
4801 4.43. The change log for 4.44 is below; many of its items are identical to
4802 those for 4.50. This seems to be the most sensible way to preserve the
4803 historical information.
4809 1. Minor wording change to the doc/README.SIEVE file.
4811 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4812 computation of the current number of files was incorrect.
4814 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4815 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4818 4. Give more explanation in the error message when the command for a transport
4819 filter fails to execute.
4821 5. There are several places where Exim runs a non-Exim command in a
4822 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4823 was being done only for the command run by the queryprogram router. It is
4824 now done for all such subprocesses. The other cases are: ${run, transport
4825 filters, and the commands run by the lmtp and pipe transports.
4827 6. Added CONFIGURE_GROUP build-time option.
4829 7. Some older OS have a limit of 256 on the maximum number of file
4830 descriptors. Exim was using setrlimit() to set 1000 as a large value
4831 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4832 systems. I've change it so that if it can't get 1000, it tries for 256.
4834 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4835 was an oversight, and furthermore, ever since the addition of extra
4836 controls (e.g. 4.43/32), the checks on when to allow different forms of
4837 "control" were broken. There should now be diagnostics for all cases when a
4838 control that does not make sense is encountered.
4840 9. Added the /retain_sender option to "control=submission".
4842 10. $recipients is now available in the predata ACL (oversight).
4844 11. Tidy the search cache before the fork to do a delivery from a message
4845 received from the command line. Otherwise the child will trigger a lookup
4846 failure and thereby defer the delivery if it tries to use (for example) a
4847 cached ldap connection that the parent has called unbind on.
4849 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4850 of $address_data from the recipient verification was clobbered by the
4851 sender verification.
4853 13. The value of address_data from a sender verification is now available in
4854 $sender_address_data in subsequent conditions in the ACL statement.
4856 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4858 15. Added a new option "connect=<time>" to callout options, to set a different
4861 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4862 was its contents. (It was OK if the option was not defined at all.)
4864 17. A "Completed" log line is now written for messages that are removed from
4865 the spool by the -Mrm option.
4867 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4868 information about exactly what failed.
4870 19. Added -dd to debug only the daemon process.
4872 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4873 handles timeouts, both on the server side and network timeouts. Renamed the
4874 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4876 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4877 It is now set to "smtps".
4879 22. $host_address is now set to the target address during the checking of
4880 ignore_target_hosts.
4882 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4883 being passed; this would have caused $sender_host_name to have been used if
4884 matching the list had actually called for a host name (not very likely,
4885 since this list is usually IP addresses). A host name is now passed as
4888 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4889 code that specifies a non-restarting handler (typically sigaction() in
4890 modern systems) in an attempt to fix a rare and obscure crash bug.
4892 25. Narrowed the window for a race in the daemon that could cause it to ignore
4893 SIGCHLD signals. This is not a major problem, because they are used only to
4894 wake it up if nothing else does.
4896 26. A malformed maildirsize file could cause Exim to calculate negative values
4897 for the mailbox size or file count. Odd effects could occur as a result.
4898 The maildirsize information is now recalculated if the size or filecount
4901 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4902 support for a long time. Removed HAVE_SYS_VFS_H.
4904 28. Installed the latest version of exipick from John Jetmore.
4906 29. In an address list, if the pattern was not a regular expression, an empty
4907 subject address (from a bounce message) matched only if the pattern was an
4908 empty string. Non-empty patterns were not even tested. This was the wrong
4909 because it is perfectly reasonable to use an empty address as part of a
4910 database query. An empty address is now tested by patterns that are
4911 lookups. However, all the other forms of pattern expect the subject to
4912 contain a local part and a domain, and therefore, for them, an empty
4913 address still always fails if the pattern is not itself empty.
4915 30. Exim went into a mad DNS loop when attempting to do a callout where the
4916 host was specified on an smtp transport, and looking it up yielded more
4917 than one IP address.
4919 31. Re-factored the code for checking spool and log partition space into a
4920 function that finds that data and another that does the check. The former
4921 is then used to implement four new variables: $spool_space, $log_space,
4922 $spool_inodes, and $log_inodes.
4924 32. The RFC2047 encoding function was originally intended for short strings
4925 such as real names; it was not keeping to the 75-character limit for
4926 encoded words that the RFC imposes. It now respects the limit, and
4927 generates multiple encoded words if necessary. To be on the safe side, I
4928 have increased the buffer size for the ${rfc2047: expansion operator from
4931 33. It is now permitted to omit both strings after an "if" condition; if the
4932 condition is true, the result is "true". As before, when the second string
4933 is omitted, a false condition yields an empty string. This makes it less
4934 cumbersome to write custom ACL and router conditions.
4936 34. Failure to deliver a bounce message always caused it to be frozen, even if
4937 there was an errors_to setting on the router. The errors_to setting is now
4940 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4941 canonical form (fully expanded) before being placed in
4942 $sender_host_address.
4944 36. The table in the code that translates DNS record types into text (T_A to
4945 "A" for instance) was missing entries for NS and CNAME. It is just possible
4946 that this could have caused confusion if both these types were looked up
4947 for the same domain, because the text type is used as part of Exim's
4948 per-process caching. But the chance of anyone hitting this buglet seems
4951 37. The dnsdb lookup has been extended in a number of ways.
4953 (1) There is a new type, "zns", which walks up the domain tree until it
4954 finds some nameserver records. It should be used with care.
4956 (2) There is a new type, "mxh", which is like "mx" except that it returns
4957 just the host names, not the priorities.
4959 (3) It is now possible to give a list of domains (or IP addresses) to be
4960 looked up. The behaviour when one of the lookups defers can be
4961 controlled by a keyword.
4963 (4) It is now possible to specify the separator character for use when
4964 multiple records are returned.
4966 38. The dnslists ACL condition has been extended: it is now possible to supply
4967 a list of IP addresses and/or domains to be looked up in a particular DNS
4970 39. Added log_selector=+queue_time_overall.
4972 40. When running the queue in the test harness, wait just a tad after forking a
4973 delivery process, to get repeatability of debugging output.
4975 41. Include certificate and key file names in error message when GnuTLS fails
4976 to set them up, because the GnuTLS error message doesn't include the name
4977 of the failing file when there is a problem reading it.
4979 42. Allow both -bf and -bF in the same test run.
4981 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4983 44. The "Exiscan patch" is now merged into the mainline Exim source.
4985 45. Sometimes the final signoff response after QUIT could fail to get
4986 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4987 < 0 before doing a fflush(). This bug looks as though it goes back to the
4988 introduction of TLS in release 3.20, but "sometimes" must have been rare
4989 because the tests only now provoked it.
4991 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4992 (this can affect the format of dates).
4994 47. exim_tidydb, when checking for the continued existence of a message for
4995 which it has found a message-specific retry record, was not finding
4996 messages that were in split spool directories. Consequently, it was
4997 deleting retry records that should have stayed in existence.
4999 48. Steve fixed some bugs in eximstats.
5001 49. The SPA authentication driver was not abandoning authentication and moving
5002 on to the next authenticator when an expansion was forced to fail,
5003 contradicting the general specification for all authenticators. Instead it
5004 was generating a temporary error. It now behaves as specified.
5006 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5007 (the order specifies the preference for clients). The order is now AES256,
5008 AES128, 3DES, ARCFOUR128.
5010 51. Small patch to Sieve code - explicitly set From: when generating an
5013 52. Exim crashed if a remote delivery caused a very long error message to be
5014 recorded - for instance if somebody sent an entire SpamAssassin report back
5015 as a large number of 550 error lines. This bug was coincidentally fixed by
5016 increasing the size of one of Exim's internal buffers (big_buffer) that
5017 happened as part of the Exiscan merge. However, to be on the safe side, I
5018 have made the code more robust (and fixed the comments that describe what
5021 53. Now that there can be additional text after "Completed" in log lines (if
5022 the queue_time_overall log selector is set), a one-byte patch to exigrep
5023 was needed to allow it to recognize "Completed" as not the last thing in
5026 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5027 patch that reportedly fixes this has been added. I am not expert enough to
5028 create a test for it. This is what the patch creator wrote:
5030 "I found a little strange behaviour of ldap code when working with
5031 Windows 2003 AD Domain, where users was placed in more than one
5032 Organization Units. When I tried to give exim partial DN, the exit code
5033 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5034 But simultaneously result of request was absolutely normal ldap result,
5035 so I produce this patch..."
5037 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5038 REFERENCE, so I have modified the code to exclude the patch when that macro
5041 55. Some experimental protocols are using DNS PTR records for new purposes. The
5042 keys for these records are domain names, not reversed IP addresses. The
5043 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5044 leaves it alone. Component reversal etc. now happens only for IP addresses.
5047 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5049 57. Double the size of the debug message buffer (to 2048) so that more of very
5050 long debug lines gets shown.
5052 58. The exicyclog utility now does better if the number of log files to keep
5053 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5055 59. Two changes related to the smtp_active_hostname option:
5057 (1) $smtp_active_hostname is now available as a variable.
5058 (2) The default for smtp_banner uses $smtp_active_hostname instead
5059 of $primary_hostname.
5061 60. The host_aton() function is supposed to be passed a string that is known
5062 to be a valid IP address. However, in the case of IPv6 addresses, it was
5063 not checking this. This is a hostage to fortune. Exim now panics and dies
5064 if the condition is not met. A case was found where this could be provoked
5065 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5066 components; fortuitously, this particular loophole had already been fixed
5067 by change 4.50/55 above.
5069 If there are any other similar loopholes, the new check in host_aton()
5070 itself should stop them being exploited. The report I received stated that
5071 data on the command line could provoke the exploit when Exim was running as
5072 exim, but did not say which command line option was involved. All I could
5073 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5074 running as the user.
5077 61. There was a buffer overflow vulnerability in the SPA authentication code
5078 (which came originally from the Samba project). I have added a test to the
5079 spa_base64_to_bits() function which I hope fixes it.
5082 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5083 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5085 63. The daemon start-up calls getloadavg() while still root for those OS that
5086 need the first call to be done as root, but it missed one case: when
5087 deliver_queue_load_max is set with deliver_drop_privilege. This is
5088 necessary for the benefit of the queue runner, because there is no re-exec
5089 when deliver_drop_privilege is set.
5091 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5092 This has been fixed.
5094 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5095 was in use, was not putting the data itself into the right store pool;
5096 consequently, it could be overwritten for a subsequent message in the same
5097 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5100 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5102 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5103 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5104 the function were treating the return as a boolean value, which happened to
5105 work because 0=false and not-0=true, but is not correct code.
5107 68. The host_aton() function was not handling scoped IPv6 addresses (those
5108 with, for example, "%eth0" on the end) correctly.
5110 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5111 negated items (that is, ~something) in unsigned ints. Some compilers
5112 apparently mutter when there is no cast.
5114 70. If an address verification called from an ACL failed, and did not produce a
5115 user-specific message (i.e. there was only a "system" message), nothing was
5116 put in $acl_verify_message. In this situation, it now puts the system
5119 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5120 session; change 4.31/43 added the unwanted input to the log line - except
5121 that it did not do this in the start of session case. It now does.
5123 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5124 This is wrong and can cause the other end to generate a synchronization
5125 error if it is another Exim or anything else that does the synchronization
5126 check. A QUIT command is no longer sent after a timeout.
5128 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5129 during host lookups.
5131 74. The defer_ok option of callout verification was not working if it was used
5132 when verifying addresses in header lines, that is, for this case:
5134 verify = header_sender/callout=defer_ok
5136 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5137 those file descriptors could be used for SMTP connections. If anything
5138 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5139 could be sent to the SMTP client, causing chaos. The daemon now opens
5140 stdin, stdout, and stderr to /dev/null when it puts itself into the
5143 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5144 log by default. The user can override this with suitable Perl magic.
5146 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5147 the log message when discard triggers, was not working for the DATA ACL or
5148 for the non-SMTP ACL.
5150 78. Error message wording change in sieve.c.
5152 79. If smtp_accept_max_per_host was set, the number of connections could be
5153 restricted to fewer than expected, because the daemon was trying to set up
5154 a new connection before checking whether the processes handling previous
5155 connections had finished. The check for completed processes is now done
5156 earlier. On busy systems, this bug wouldn't be noticed because something
5157 else would have woken the daemon, and it would have reaped the completed
5160 80. If a message was submitted locally by a user whose login name contained one
5161 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5162 It caused a spool format error. I have fixed the spool reading code. A
5163 related problem was that the "from" clause in the Received: line became
5164 illegal because of the space(s). It is now covered by ${quote_local_part.
5166 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5169 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5171 83. Patch from Sieve maintainer to fix unterminated string problem in
5172 "vacation" handling.
5174 84. Some minor changes to the Linux configuration files to help with other
5175 OS variants using glibc.
5177 85. One more patch for Sieve to update vacation handling to latest spec.
5180 ----------------------------------------------------
5181 See the note above about the 4.44 and 4.50 releases.
5182 ----------------------------------------------------
5188 1. Change 4.43/35 introduced a bug that caused file counts to be
5189 incorrectly computed when quota_filecount was set in an appendfile
5192 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5193 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5196 3. Give more explanation in the error message when the command for a transport
5197 filter fails to execute.
5199 4. There are several places where Exim runs a non-Exim command in a
5200 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5201 was being done only for the command run by the queryprogram router. It is
5202 now done for all such subprocesses. The other cases are: ${run, transport
5203 filters, and the commands run by the lmtp and pipe transports.
5205 5. Some older OS have a limit of 256 on the maximum number of file
5206 descriptors. Exim was using setrlimit() to set 1000 as a large value
5207 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5208 systems. I've change it so that if it can't get 1000, it tries for 256.
5210 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5211 was an oversight, and furthermore, ever since the addition of extra
5212 controls (e.g. 4.43/32), the checks on when to allow different forms of
5213 "control" were broken. There should now be diagnostics for all cases when a
5214 control that does not make sense is encountered.
5216 7. $recipients is now available in the predata ACL (oversight).
5218 8. Tidy the search cache before the fork to do a delivery from a message
5219 received from the command line. Otherwise the child will trigger a lookup
5220 failure and thereby defer the delivery if it tries to use (for example) a
5221 cached ldap connection that the parent has called unbind on.
5223 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5224 of $address_data from the recipient verification was clobbered by the
5225 sender verification.
5227 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5228 was its contents. (It was OK if the option was not defined at all.)
5230 11. A "Completed" log line is now written for messages that are removed from
5231 the spool by the -Mrm option.
5233 12. $host_address is now set to the target address during the checking of
5234 ignore_target_hosts.
5236 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5237 being passed; this would have caused $sender_host_name to have been used if
5238 matching the list had actually called for a host name (not very likely,
5239 since this list is usually IP addresses). A host name is now passed as
5242 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5243 code that specifies a non-restarting handler (typically sigaction() in
5244 modern systems) in an attempt to fix a rare and obscure crash bug.
5246 15. Narrowed the window for a race in the daemon that could cause it to ignore
5247 SIGCHLD signals. This is not a major problem, because they are used only to
5248 wake it up if nothing else does.
5250 16. A malformed maildirsize file could cause Exim to calculate negative values
5251 for the mailbox size or file count. Odd effects could occur as a result.
5252 The maildirsize information is now recalculated if the size or filecount
5255 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5256 support for a long time. Removed HAVE_SYS_VFS_H.
5258 18. Updated exipick to current release from John Jetmore.
5260 19. Allow an empty sender to be matched against a lookup in an address list.
5261 Previously the only cases considered were a regular expression, or an
5264 20. Exim went into a mad DNS lookup loop when doing a callout where the
5265 host was specified on the transport, if the DNS lookup yielded more than
5268 21. The RFC2047 encoding function was originally intended for short strings
5269 such as real names; it was not keeping to the 75-character limit for
5270 encoded words that the RFC imposes. It now respects the limit, and
5271 generates multiple encoded words if necessary. To be on the safe side, I
5272 have increased the buffer size for the ${rfc2047: expansion operator from
5275 22. Failure to deliver a bounce message always caused it to be frozen, even if
5276 there was an errors_to setting on the router. The errors_to setting is now
5279 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5280 canonical form (fully expanded) before being placed in
5281 $sender_host_address.
5283 24. Updated eximstats to version 1.33
5285 25. Include certificate and key file names in error message when GnuTLS fails
5286 to set them up, because the GnuTLS error message doesn't include the name
5287 of the failing file when there is a problem reading it.
5289 26. Expand error message when OpenSSL has problems setting up cert/key files.
5292 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5293 (this can affect the format of dates).
5295 28. exim_tidydb, when checking for the continued existence of a message for
5296 which it has found a message-specific retry record, was not finding
5297 messages that were in split spool directories. Consequently, it was
5298 deleting retry records that should have stayed in existence.
5300 29. eximstats updated to version 1.35
5301 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5302 1.35 - bugfix such that pie charts by volume are generated correctly
5304 30. The SPA authentication driver was not abandoning authentication and moving
5305 on to the next authenticator when an expansion was forced to fail,
5306 contradicting the general specification for all authenticators. Instead it
5307 was generating a temporary error. It now behaves as specified.
5309 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5310 (the order specifies the preference for clients). The order is now AES256,
5311 AES128, 3DES, ARCFOUR128.
5313 31. Small patch to Sieve code - explicitly set From: when generating an
5316 32. Exim crashed if a remote delivery caused a very long error message to be
5317 recorded - for instance if somebody sent an entire SpamAssassin report back
5318 as a large number of 550 error lines. This bug was coincidentally fixed by
5319 increasing the size of one of Exim's internal buffers (big_buffer) that
5320 happened as part of the Exiscan merge. However, to be on the safe side, I
5321 have made the code more robust (and fixed the comments that describe what
5324 33. Some experimental protocols are using DNS PTR records for new purposes. The
5325 keys for these records are domain names, not reversed IP addresses. The
5326 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5327 leaves it alone. Component reversal etc. now happens only for IP addresses.
5330 34. The host_aton() function is supposed to be passed a string that is known
5331 to be a valid IP address. However, in the case of IPv6 addresses, it was
5332 not checking this. This is a hostage to fortune. Exim now panics and dies
5333 if the condition is not met. A case was found where this could be provoked
5334 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5335 components; fortuitously, this particular loophole had already been fixed
5336 by change 4.50/55 or 4.44/33 above.
5338 If there are any other similar loopholes, the new check in host_aton()
5339 itself should stop them being exploited. The report I received stated that
5340 data on the command line could provoke the exploit when Exim was running as
5341 exim, but did not say which command line option was involved. All I could
5342 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5343 running as the user.
5346 35. There was a buffer overflow vulnerability in the SPA authentication code
5347 (which came originally from the Samba project). I have added a test to the
5348 spa_base64_to_bits() function which I hope fixes it.
5351 36. The daemon start-up calls getloadavg() while still root for those OS that
5352 need the first call to be done as root, but it missed one case: when
5353 deliver_queue_load_max is set with deliver_drop_privilege. This is
5354 necessary for the benefit of the queue runner, because there is no re-exec
5355 when deliver_drop_privilege is set.
5357 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5358 was in use, was not putting the data itself into the right store pool;
5359 consequently, it could be overwritten for a subsequent message in the same
5360 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5363 38. Sometimes the final signoff response after QUIT could fail to get
5364 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5365 < 0 before doing a fflush(). This bug looks as though it goes back to the
5366 introduction of TLS in release 3.20, but "sometimes" must have been rare
5367 because the tests only now provoked it.
5373 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5374 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5375 Now it returns an integer. A number of calls were still expecting a T/F
5376 return. Fortuitously, in all cases, the tests worked in OK situations,
5377 which is the norm. However, things would have gone wrong on any write
5378 failures on the smtp file descriptor. This function is used when sending
5379 messages over SMTP and also when doing verify callouts.
5381 2. When Exim is called to do synchronous delivery of a locally submitted
5382 message (the -odf or -odi options), it no longer closes stderr before doing
5385 3. Implemented the mua_wrapper option.
5387 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5389 5. Implemented the functions header_remove(), header_testname(),
5390 header_add_at_position(), and receive_remove_recipient(), and exported them
5393 6. If an ACL "warn" statement specified the addition of headers, Exim already
5394 inserted X-ACL-Warn: at the start if there was no header name. However, it
5395 was not making this test for the second and subsequent header lines if
5396 there were newlines in the string. This meant that an invalid header could
5397 be inserted if Exim was badly configured.
5399 7. Allow an ACL "warn" statement to add header lines at the start or after all
5400 the Received: headers, as well as at the end.
5402 8. Added the rcpt_4xx retry error code.
5404 9. Added postmaster_mailfrom=xxx to callout verification option.
5406 10. Added mailfrom=xxxx to the callout verification option, for verify=
5409 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5410 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5412 12. Inserted some casts to stop certain compilers warning when using pointer
5413 differences as field lengths or precisions in printf-type calls (mostly
5414 affecting debugging statements).
5416 13. Added optional readline() support for -be (dynamically loaded).
5418 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5419 same clock tick as a message's arrival, so that its received time was the
5420 same as the "first fail" time on the retry record, and that message
5421 remained on the queue past the ultimate address timeout, every queue runner
5422 would try a delivery (because it was past the ultimate address timeout) but
5423 after another failure, the ultimate address timeout, which should have then
5424 bounced the address, did not kick in. This was a "< instead of <=" error;
5425 in most cases the first failure would have been in the next clock tick
5426 after the received time, and all would be well.
5428 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5429 being recognized when the domain list was tested by the match_domain
5430 condition in an expansion string.
5432 16. Added the ${str2b64: operator.
5434 17. Exim was always calling setrlimit() to set a large limit for the number of
5435 processes, without checking whether the existing limit was already
5436 adequate. (It did check for the limit on file descriptors.) Furthermore,
5437 errors from getrlimit() and setrlimit() were being ignored. Now they are
5438 logged to the main and panic logs, but Exim does carry on, to try to do its
5439 job under whatever limits there are.
5441 18. Imported PCRE 5.0.
5443 19. Trivial typo in log message " temporarily refused connection" (the leading
5446 20. If the log selector return_path_on_delivery was set and an address was
5447 redirected to /dev/null, the delivery process crashed because it assumed
5448 that a return path would always be set for a "successful" delivery. In this
5449 case, the whole delivery is bypassed as an optimization, and therefore no
5452 21. Internal re-arrangement: the function for sending a challenge and reading
5453 a response while authentication was assuming a zero-terminated challenge
5454 string. It's now changed to take a pointer and a length, to allow for
5455 binary data in such strings.
5457 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5459 23. Exim was not respecting finduser_retries when seeking the login of the
5460 uid under which it was called; it was always trying 10 times. (The default
5461 setting of finduser_retries is zero.) Also, it was sleeping after the final
5462 failure, which is pointless.
5464 24. Implemented tls_on_connect_ports.
5466 25. Implemented acl_smtp_predata.
5468 26. If the domain in control=submission is set empty, Exim assumes that the
5469 authenticated id is a complete email address when it generates From: or
5470 Sender: header lines.
5472 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5473 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5474 chown and chgrp in /bin and hostname in /usr/bin.
5476 28. Exim was keeping the "process log" file open after each use, just as it
5477 does for the main log. This opens the possibility of it remaining open for
5478 long periods when the USR1 signal hits a daemon. Occasional processlog
5479 errors were reported, that could have been caused by this. Anyway, it seems
5480 much more sensible not to leave this file open at all, so that is what now
5483 29. The long-running daemon process does not normally write to the log once it
5484 has entered its main loop, and it closes the log before doing so. This is
5485 so that log files can straightforwardly be renamed and moved. However,
5486 there are a couple of unusual error situations where the daemon does write
5487 log entries, and I had neglected to close the log afterwards.
5489 30. The text of an SMTP error response that was received during a remote
5490 delivery was being truncated at 512 bytes. This is too short for some of
5491 the long messages that one sometimes sees. I've increased the limit to
5494 31. It is now possible to make retry rules that apply only when a message has a
5495 specific sender, in particular, an empty sender.
5497 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5498 it possible to be selective about when SMTP synchronization is enforced.
5500 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5502 32. Implemented hosts_connection_nolog.
5504 33. Added an ACL for QUIT.
5506 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5509 35. Added mailbox_size and mailbox_filecount to appendfile.
5511 36. Added control = no_multiline_responses to ACLs.
5513 37. There was a bug in the logic of the code that waits for the clock to tick
5514 in the case where the clock went backwards by a substantial amount such
5515 that the microsecond fraction of "now" was more than the microsecond
5516 fraction of "then" (but the whole seconds number was less).
5518 38. Added support for the libradius Radius client library this is found on
5519 FreeBSD (previously only the radiusclient library was supported).
5525 1. When certain lookups returned multiple values in the form name=value, the
5526 quoting of the values was not always being done properly. Specifically:
5527 (a) If the value started with a double quote, but contained no whitespace,
5529 (b) If the value contained whitespace other than a space character (i.e.
5530 tabs or newlines or carriage returns) it was not quoted.
5531 This fix has been applied to the mysql and pgsql lookups by writing a
5532 separate quoting function and calling it from the lookup code. The fix
5533 should probably also be applied to nisplus, ibase and oracle lookups, but
5534 since I cannot test any of those, I have not disturbed their existing code.
5536 2. A hit in the callout cache for a specific address caused a log line with no
5537 reason for rejecting RCPT. Now it says "Previous (cached) callout
5538 verification failure".
5540 3. There was an off-by-one bug in the queryprogram router. An over-long
5541 return line was truncated at 256 instead of 255 characters, thereby
5542 overflowing its buffer with the terminating zero. As well as fixing this, I
5543 have increased the buffer size to 1024 (and made a note to document this).
5545 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5546 when Exim is waiting for an SMTP response from a remote server, Exim
5547 restarts its select() call on the socket, thereby resetting its timeout.
5548 This is not a problem when such interrupts are rare. Somebody set up a cron
5549 job to run exiwhat every 2 minutes, which is less than the normal select()
5550 timeout (5 or 10 minutes). This meant that the select() timeout never
5551 kicked in because it was always reset. I have fixed this by comparing the
5552 time when an interrupt arrives with the time at the start of the first call
5553 to select(). If more time than the timeout has elapsed, the interrupt is
5554 treated as a timeout.
5556 5. Some internal re-factoring in preparation for the addition of Sieve
5557 extensions (by MH). In particular, the "personal" test is moved to a
5558 separate function, and given an option for scanning Cc: and Bcc: (which is
5559 not set for Exim filters).
5561 6. When Exim created an email address using the login of the caller as the
5562 local part (e.g. when creating a From: or Sender: header line), it was not
5563 quoting the local part when it contained special characters such as @.
5565 7. Installed new OpenBSD configuration files.
5567 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5568 try to make them clearer.
5570 9. Callout options, other than the timeout value, were being ignored when
5571 verifying sender addresses in header lines. For example, when using
5573 verify = header_sender/callout=no_cache
5575 the cache was (incorrectly) being used.
5577 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5578 only the Cygwin environment.
5580 11. When return_path_on_delivery was set as a log selector, if different remote
5581 addresses in the same message used different return paths and parallel
5582 remote delivery occurred, the wrong values would sometimes be logged.
5583 (Whenever a remote delivery process finished, the return path value from
5584 the most recently started remote delivery process was logged.)
5586 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5587 lines when AUTH and/or TLS are in use. This is the "received protocol"
5588 field. Exim used to use "asmtp" for authenticated SMTP, without any
5589 indication (in the protocol name) for TLS use. Now it follows the RFC and
5590 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5591 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5592 names appear in log lines as well as in Received: header lines.
5594 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5595 extensions, and comparison tests, and to fix some bugs.
5597 14. Changes to the "personal" filter test:
5599 (1) The test was buggy in that it was just doing the equivalent of
5600 "contains" tests on header lines. For example, if a user's address was
5601 anne@some.where, the "personal" test would incorrectly be true for
5603 To: susanne@some.where
5605 This test is now done by extracting each address from the header in turn,
5606 and checking the entire address. Other tests that are part of "personal"
5607 are now done using regular expressions (for example, to check local parts
5608 of addresses in From: header lines).
5610 (2) The list of non-personal local parts in From: addresses has been
5611 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5612 taken from the Sieve specification recommendations.
5614 (3) If the message contains any header line starting with "List-" it is
5615 treated as non-personal.
5617 (4) The test for "circular" in the Subject: header line has been removed
5618 because it now seems ill-conceived.
5620 15. Minor typos in src/EDITME comments corrected.
5622 16. Installed latest exipick from John Jetmore.
5624 17. If headers_add on a router specified a text string that was too long for
5625 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5626 of string_sprintf() is now avoided.
5628 18. $message_body_size was not set (it was always zero) when running the DATA
5629 ACL and the local_scan() function.
5631 19. For the "mail" command in an Exim filter, no default was being set for
5632 the once_repeat time, causing a random time value to be used if "once" was
5633 specified. (If the value happened to be <= 0, no repeat happened.) The
5634 default is now 0s, meaning "never repeat". The "vacation" command was OK
5635 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5636 (I found it when inspecting the code).
5638 20. There is now an overall timeout for performing a callout verification. It
5639 defaults to 4 times the callout timeout, which applies to individual SMTP
5640 commands during the callout. The overall timeout applies when there is more
5641 than one host that can be tried. The timeout is checked before trying the
5642 next host. This prevents very long delays if there are a large number of
5643 hosts and all are timing out (e.g. when the network connections are timing
5644 out). The value of the overall timeout can be changed by specifying an
5645 additional sub-option for "callout", called "maxwait". For example:
5647 verify = sender/callout=5s,maxwait=20s
5649 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5650 to the end before writing, but this should make it even safer).
5652 22. Exim was forgetting that it had advertised PIPELINING for the second and
5653 subsequent messages on an SMTP connection. It was also not resetting its
5654 memory on STARTTLS and an internal HELO.
5656 23. When Exim logs an SMTP synchronization error within a session, it now
5657 records whether PIPELINING has been advertised or not.
5659 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5660 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5661 rather than long int.
5663 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5665 26. Added the never_mail option to autoreply.
5671 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5672 crash if the getsockname() call failed; this can happen if a connection is
5673 closed very soon after it is established. The problem was simply in the
5674 order in which certain operations were done, causing Exim to try to write
5675 to the SMTP stream before it had set up the file descriptor. The bug has
5676 been fixed by making things happen in the correct order.
5682 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5683 before the connection was closed, thus losing the rejection response.
5685 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5686 some early Solaris releases, but causes trouble in current releases where
5687 socklen_t is defined.
5689 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5692 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5695 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5696 that releasing the top of it at the end releases what was used for sub-
5697 expansions (unless the block got too big). However, discard this block if
5698 the first thing is a variable or header, so that we can use its block when
5699 it is dynamic (useful for very large $message_headers, for example).
5701 6. Lookups now cache *every* query, not just the most recent. A new, separate
5702 store pool is used for this. It can be recovered when all lookup caches are
5703 flushed. Lookups now release memory at the end of their result strings.
5704 This has involved some general refactoring of the lookup sources.
5706 7. Some code has been added to the store_xxx() functions to reduce the amount
5707 of flapping under certain conditions.
5709 8. log_incoming_interface used to affect only the <= reception log lines. Now
5710 it causes the local interface and port to be added to several more SMTP log
5711 lines, for example "SMTP connection from", and rejection lines.
5713 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5715 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5717 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5718 could be overwritten at the end of the current message (or the start of a
5719 new message if it was set in a HELO ACL). The value is now preserved for
5720 the duration of the SMTP connection.
5722 12. If a transport had a headers_rewrite setting, and a matching header line
5723 contained an unqualified address, that address was qualified, even if it
5724 did not match any rewriting rules. The underlying bug was that the values
5725 of the flags that permit the existence of unqualified sender and recipient
5726 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5727 non-local messages, and by -bnq for local messages) were not being
5728 preserved with the message after it was received.
5730 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5731 "next input=" as part of the text comprising the host identity instead of
5732 the correct text. The code was using the same buffer for two different
5733 strings. However, depending on which order the printing function evaluated
5734 its arguments, the bug did not always show up. Under Linux, for example, my
5735 test suite worked just fine.
5737 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5738 doesn't work with some older versions of Perl. It has been changed to "my",
5739 which in any case is probably the better facility to use.
5741 15. A really picky compiler found some instances of statements for creating
5742 error messages that either had too many or two few arguments for the format
5745 16. The size of the buffer for calls to the DNS resolver has been increased
5746 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5747 for addresses that have a lot of PTR records. This alleviates a problem; it
5748 does not fully solve it.
5750 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5751 buffer now truncates the list and logs the incident, which is the same
5752 action as happens when Exim is looking up a host name and its aliases.
5753 Previously in this situation something unpredictable would happen;
5754 sometimes it was "internal error: store_reset failed".
5756 18. If a server dropped the connection unexpectedly when an Exim client was
5757 using GnuTLS and trying to read a response, the client delivery process
5758 crashed while trying to generate an error log message.
5760 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5761 string, for example:
5763 warn message = H1: something\nH2: something
5765 the text was added as a single header line from Exim's point of view
5766 though it ended up OK in the delivered message. However, searching for the
5767 second and subsequent header lines using $h_h2: did not work. This has been
5768 fixed. Similarly, if a system filter added multiple headers in this way,
5769 the routers could not see them.
5771 20. Expanded the error message when iplsearch is called with an invalid key to
5772 suggest using net-iplsearch in a host list.
5774 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5775 ACLs are no longer actually imposed (and a message to that effect is
5778 22. If a "gecos" field in a passwd entry contained escaped characters, in
5779 particular, if it contained a \" sequence, Exim got it wrong when building
5780 a From: or a Sender: header from that name. A second bug also caused
5781 incorrect handling when an unquoted " was present following a character
5782 that needed quoting.
5784 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5785 was not being matched caselessly.
5787 24. Arranged for all hyphens in the exim.8 source to be escaped with
5790 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5791 properly. Recipient callout cache records were still being keyed to include
5792 the sender, even when use_sender was set false. This led to far more
5793 callouts that were necessary. The sender is no longer included in the key
5794 when use_sender is false.
5796 26. Added "control = submission" modifier to ACLs.
5798 27. Added the ${base62d: operator to decode base 62 numbers.
5800 28. dnsdb lookups can now access SRV records.
5802 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5803 the configuration file.
5805 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5806 (-v) mode. This makes the output for a verbose queue run more intelligible.
5808 31. Added a use_postmaster feature to recipient callouts.
5810 32. Added the $body_zerocount variable, containing the number of binary zero
5811 bytes in the message body.
5813 33. The time of last modification of the "new" subdirectory is now used as the
5814 "mailbox time last read" when there is a quota error for a maildir
5817 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5819 35. Added +ignore_unknown as a special item in host lists.
5821 36. Code for decoding IPv6 addresses in host lists is now included, even if
5822 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5823 address was recognized as an IP address, but was then not correctly decoded
5824 into binary, causing unexpected and incorrect effects when compared with
5831 1. Very minor rewording of debugging text in manualroute to say "list of
5832 hosts" instead of "hostlist".
5834 2. If verify=header_syntax was set, and a header line with an unqualified
5835 address (no domain) and a large number of spaces between the end of the
5836 name and the colon was received, the reception process suffered a buffer
5837 overflow, and (when I tested it) crashed. This was caused by some obsolete
5838 code that should have been removed. The fix is to remove it!
5840 3. When running in the test harness, delay a bit after writing a bounce
5841 message to get a bit more predictability in the log output.
5843 4. Added a call to search_tidyup() just before forking a reception process. In
5844 theory, someone could use a lookup in the expansion of smtp_accept_max_
5845 per_host which, without the tidyup, could leave open a database connection.
5847 5. Added the variables $recipient_data and $sender_data which get set from a
5848 lookup success in an ACL "recipients" or "senders" condition, or a router
5849 "senders" option, similar to $domain_data and $local_part_data.
5851 6. Moved the writing of debug_print from before to after the "senders" test
5854 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5855 problems for message scanning, either using a data ACL, or using
5856 local_scan() because the Received: header was not generated till after they
5857 were called (in order to set the time as the time of reception completion).
5858 I have revised the way this works. The header is now generated after the
5859 body is received, but before the ACL or local_scan() are called. After they
5860 are run, the timestamp in the header is updated.
5866 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5867 before starting a queue runner without re-exec. This happened only when
5868 deliver_drop_privilege was set or when the Exim user was set to root. The
5869 effect of the bug was that timeouts during subsequent deliveries caused
5870 crashes instead of being properly handled. The handler is now left at its
5871 default (and expected) setting.
5873 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5874 message, again when deliver_drop_privilege is set or Exim is run as root.
5875 The bug described in (1) was not present in this case, but the tidying up
5876 of the other signals was missing. I have made the two cases consistent.
5878 3. The ignore_target_hosts setting on a manualroute router was being ignored
5879 for hosts that were looked up using the /MX notation.
5881 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5884 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5885 operated on the sender address. After changing the $sender_address to <>
5886 for the sender address verify, Exim was re-instated it as the original
5887 (before rewriting) address, but remembering that it had rewritten it, so it
5888 wasn't rewriting it again. This bug also had the effect of breaking the
5889 sender address verification caching when the sender address was rewritten.
5891 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5892 This has been changed so that if the ip literal address matches
5893 ignore_target_hosts, the router declines.
5895 7. Added expansion conditions match_domain, match_address, and match_local_
5896 part (NOT match_host).
5898 8. The placeholder for the Received: header didn't have a length field set.
5900 9. Added code to Exim itself and to exim_lock to test for a specific race
5901 condition that could lead to file corruption when using MBX delivery. The
5902 issue is with the lockfile that is created in /tmp. If this file is removed
5903 after a process has opened it but before that process has acquired a lock,
5904 there is the potential for a second process to recreate the file and also
5905 acquire a lock. This could lead to two Exim processes writing to the file
5906 at the same time. The added code performs the same test as UW imapd; it
5907 checks after acquiring the lock that its file descriptor still refers to
5908 the same named file.
5910 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5911 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5914 11. Added the smtp_active_hostname option. If used, this will typically be made
5915 to depend on the incoming interface address. Because $interface_address is
5916 not set up until the daemon has forked a reception process, error responses
5917 that can happen earlier (such as "too many connections") no longer contain
5920 12. If an expansion in a condition on a "warn" statement fails because a lookup
5921 defers, the "warn" statement is abandoned, and the next ACL statement is
5922 processed. Previously this caused the whole ACL to be aborted.
5924 13. Added the iplsearch lookup type.
5926 14. Added ident_timeout as a log selector.
5928 15. Added tls_certificate_verified as a log selector.
5930 16. Added a global option tls_require_ciphers (compare the smtp transport
5931 option of the same name). This controls incoming TLS connections.
5933 17. I finally figured out how to make tls_require_ciphers do a similar thing
5934 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5935 before starting the TLS session.
5937 18. Tabs are now shown as \t in -bP output.
5939 19. If the log selector return_path_on_delivery was set, Exim crashed when
5940 bouncing a message because it had too many Received: header lines.
5942 20. If two routers both had headers_remove settings, and the first one included
5943 a superfluous trailing colon, the final name in the first list and the
5944 first name in the second list were incorrectly joined into one item (with a
5945 colon in the middle).
5951 1. Added -C and -D options to the exinext utility, mainly to make it easier
5952 to include in the automated testing, but these could be helpful when
5953 multiple configurations are in use.
5955 2. The exinext utility was not formatting the output nicely when there was
5956 an alternate port involved in the retry record key, nor when there was a
5957 message id as well (for retries that were specific to a specific message
5958 and a specific host). It was also confused by IPv6 addresses, because of
5959 the additional colons they contain. I have fixed the IPv4 problem, and
5960 patched it up to do a reasonable job for IPv6.
5962 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5963 delivery, the log line now contains "pipelined" if PIPELINING was used.
5965 4. An SMTP transport process used to panic and die if the bind() call to set
5966 an explicit outgoing interface failed. This has been changed; it is now
5967 treated in the same way as a connect() failure.
5969 5. A reference to $sender_host_name in the part of a conditional expansion
5970 that was being skipped was still causing a DNS lookup. This no longer
5973 6. The def: expansion condition was not recognizing references to header lines
5974 that used bh_ and bheader_.
5976 7. Added the _cache feature to named lists.
5978 8. The code for checking quota_filecount in the appendfile transport was
5979 allowing one more file than it should have been.
5981 9. For compatibility with Sendmail, the command line option
5989 and sets the incoming protocol and host name (for trusted callers). The
5990 host name and its colon can be omitted when only the protocol is to be set.
5991 Note the Exim already has two private options, -pd and -ps, that refer to
5992 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5993 "s", but I don't think that's a major issue.
5995 10. A number of refactoring changes to the code, none of which should affect
5998 (a) The number of logging options was getting close to filling up the
5999 32-bit word that was used as a bit map. I have split them into two classes:
6000 those that are passed in the argument to log_write(), and those that are
6001 only ever tested independently outside of that function. These are now in
6002 separate 32-bit words, so there is plenty of room for expansion again.
6003 There is no change in the user interface or the logging behaviour.
6005 (b) When building, for example, log lines, the code previously used a
6006 macro that called string_cat() twice, in order to add two strings. This is
6007 not really sufficiently general. Furthermore, there was one instance where
6008 it was actually wrong because one of the argument was used twice, and in
6009 one call a function was used. (As it happened, calling the function twice
6010 did not affect the overall behaviour.) The macro has been replaced by a
6011 function that can join an arbitrary number of extra strings onto a growing
6014 (c) The code for expansion conditions now uses a table and a binary chop
6015 instead of a serial search (which was left over from when there were very
6016 few conditions). Also, it now recognizes conditions like "pam" even when
6017 the relevant support is not compiled in: a suitably worded error message is
6018 given if an attempt is made to use such a condition.
6020 11. Added ${time_interval:xxxxx}.
6022 12. A bug was causing one of the ddress fields not to be passed back correctly
6023 from remote delivery subprocesses. The field in question was not being
6024 subsequently used, so this caused to problems in practice.
6026 13. Added new log selectors queue_time and deliver_time.
6028 14. Might have fixed a bug in maildirsizefile handling that threw up
6029 "unexpected character" debug warnings, and recalculated the data
6030 unnecessarily. In any case, I expanded the warning message to give more
6033 15. Added the message "Restricted characters in address" to the statements in
6034 the default ACL that block characters like @ and % in local parts.
6036 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6037 Three changes have been made:
6039 (a) There was a serious bug; a negative response to MAIL caused the whole
6040 recipient domain to be cached as invalid, thereby blocking all messages
6041 to all local parts at the same domain, from all senders. This bug has
6042 been fixed. The domain is no longer cached after a negative response to
6043 MAIL if the sender used is not empty.
6045 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6048 (c) A new callout option, "use_sender" has been added for people who want
6049 the modified behaviour.
6055 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6058 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6059 indeed breaks things for older releases.
6061 3. Added additional logging to the case where there is a problem reading data
6062 from a filter that is running in a subprocess using a pipe, in order to
6063 try to track down a specific problem.
6065 4. Testing facility fudge: when running in the test harness and attempting
6066 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6067 getting "No route to host". Convert this to a timeout.
6069 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6072 6. Some OS don't have socklen_t but use size_t instead. This affects the
6073 fifth argument of getsockopt() amongst other things. This is now
6074 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6075 can be set for individual OS. I have set it for SunOS5, OSF1, and
6076 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6077 some earlier ones do not.
6079 7. Change 4.30/15 was not doing the test caselessly.
6081 8. The standard form for an IPv6 address literal was being rejected by address
6082 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6083 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6084 this, as well as the form without the "IPv6" on the front (but only when
6085 address literals are enabled, of course).
6087 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6089 10. Exim crashed if a message with an empty sender address specified by -f
6090 encountered a router with an errors_to setting. This could be provoked only
6091 by a command such as
6095 where an empty string was supplied; "<>" did not hit this bug.
6097 11. Installed PCRE release 4.5.
6099 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6100 remained set. It is now erased.
6102 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6103 times from message ids (which are base 36 rather than the normal 62).
6105 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6106 were being counted as actual protocol errors, and logged if the log
6107 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6108 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6109 and DATA following a set of rejected RCPTs do not count as protocol errors.
6110 In other words, Exim assumes they were pipelined, though this may not
6111 actually be the case. Of course, in all cases the client gets an
6112 appropriate error code.
6114 15. If a lookup fails in an ACL condition, a message about the failure may
6115 be available; it is used if testing the ACL cannot continue, because most
6116 such messages specify what the cause of the deferral is. However, some
6117 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6118 that caused an old message to be retained and used if a later statement
6119 caused a defer, replacing the real cause of the deferral.
6121 16. If an IP address had so many PTR records that the DNS lookup buffer
6122 was not large enough to hold them, Exim could crash while trying to process
6123 the truncated data. It now detects and logs this case.
6125 17. Further to 4.21/58, another change has been made: if (and only if) the
6126 first line of a message (the first header line) ends with CRLF, a bare LF
6127 in a subsequent header line has a space inserted after it, so as not to
6128 terminate the header.
6130 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6131 unnecessarily, used atoi() instead of strtol(), and didn't check the
6132 termination when getting file sizes from file names by regex.
6134 19. Completely re-implemented the support for maildirsize files, in the light
6135 of a number of problems with the previous contributed implementation
6136 (4.30/29). In particular:
6138 . If the quota is zero, the maildirsize file is maintained, but no quota is
6141 . If the maildir directory does not exist, it is created before any attempt
6142 to write a maildirsize file.
6144 . The quota value in the file is just a cache; if the quota is changed in
6145 the transport, the new value overrides.
6147 . A regular expression is available for excluding directories from the
6150 20. The autoreply transport checks the characters in options that define the
6151 message's headers; it allows continued headers, but it was checking with
6152 isspace() after an embedded newline instead of explicitly looking for a
6155 21. If all the "regular" hosts to which an address was routed had passed their
6156 expiry times, and had not reached their retry times, the address was
6157 bounced, even if fallback hosts were defined. Now Exim should go on to try
6160 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6161 equivalent code in the SMTP transport. Some hosts send humungous responses
6162 to HELO/EHLO, more than 1024 it seems.
6164 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6165 gives compiler warnings in some environments. I've now done it "properly",
6168 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6169 (because of problems with the built-in one) was declared to return uschar *
6170 instead of char *, causing compiler failure.
6172 25. Fixed a file descriptor leak when processing alias/forward files.
6174 26. Fixed a minor format string issue in dbfn.c.
6176 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6178 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6179 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6182 29. When a sender address is verified, it is cached, to save repeating the test
6183 when there is more than one recipient in a message. However, when the
6184 verification involves a callout, it is possible for different callout
6185 options to be set for different recipients. It is too complicated to keep
6186 track of this in the cache, so now Exim always runs a verification when a
6187 callout is required, relying on the callout cache for the optimization.
6188 The overhead is duplication of the address routing, but this should not be
6191 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6192 to be verified with callout=postmaster, and the main callout worked but the
6193 postmaster check failed, the verification correctly failed. However, if a
6194 subsequent RCPT command asked for sender verification *without* the
6195 postmaster check, incorrect caching caused this verification also to fail,
6198 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6199 it was not caching the DNS options (qualify_single, search_parents) that
6200 were used when the lookup failed. A subsequent lookup with different
6201 options therefore always gave the same answer, though there were cases
6202 where it should not have. (Example: a "domains = !$mx_any" option on a
6203 dnslookup router: the "domains" option is always processed without any
6204 widening, but the router might have qualify_single set.) Now Exim uses the
6205 cached value only when the same options are set.
6207 32. Added John Jetmore's "exipick" utility to the distribution.
6209 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6210 than a timeout (e.g. a certificate is required, and is not provided), an
6211 Exim server now closes the connection immediately. Previously it waited for
6212 the client to close - but if the client is SSL, it seems that they each
6213 wait for each other, leading to a delay before one of them times out.
6215 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6216 maintained 0.8.x compatibility because I don't think many are using it, and
6217 it is clearly obsolete.
6219 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6222 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6223 client certificate was expired. A simple patch fixes this, though I don't
6224 understand the full logic of why the verify callback is called multiple
6227 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6228 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6229 which causes problems with some clients (such as the Certicom SSL Plus
6230 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6231 disables the coutermeasure allowing Eudora to connect."
6233 38. Exim was not checking that a write() to a log file succeeded. This could
6234 lead to Bad Things if a log got too big, in particular if it hit a file
6235 size limit. Exim now panics and dies if it cannot write to a log file, just
6236 as it does if it cannot open a log file.
6238 39. Modified OS/Makefile-Linux so that it now contains
6240 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6242 The two -D definitions ensure that Exim is compiled with large file
6243 support, which makes it possible to handle log files that are bigger than
6246 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6247 instance) a domain was checked against a named list that involved a lookup,
6248 causing $domain_data to be set, then another domain was checked against the
6249 same list, then the first domain was re-checked, the value of $domain_data
6250 after the final check could be wrong. In particular, if the second check
6251 failed, it could be set empty. This bug probably also applied to
6254 41. The strip_trailing_dot option was not being applied to the address given
6255 with the -f command-line option.
6257 42. The code for reading a message's header from the spool was incrementing
6258 $received_count, but never initializing it. This meant that the value was
6259 incorrect (doubled) while delivering a message in the same process in which
6260 it was received. In the most common configuration of Exim, this never
6261 happens - a fresh exec is done - but it can happen when
6262 deliver_drop_privilege is set.
6264 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6265 it now includes up to 150 characters of the unexpected data in the log
6268 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6269 and building data strings. The size of both of these buffers was 10 000
6270 bytes - far larger than anybody would *ever* want, thought I. Needless to
6271 say, somebody hit the limit. I have increased the maximum line length to
6272 20 000 and the maximum data length of concatenated lines to 100 000. I have
6273 also fixed two bugs, because there was no checking on these buffers. Tsk,
6274 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6275 buffer is too small.
6277 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6278 lsearch lookups. Now it does.
6280 46. When parsing a route_list item in a manualroute router, a fixed-length
6281 buffer was used for the list of hosts. I made this 1024 bytes long,
6282 thinking that nobody would ever have a list of hosts that long. Wrong.
6283 Somebody had a whole pile of complicated expansion conditions, and the
6284 string was silently truncated, leading to an expansion error. It turns out
6285 that it is easier to change to an unlimited length (owing to other changes
6286 that have happened since this code was originally written) than to build
6287 structure for giving a limitation error. The length of the item that
6288 expands into the list of hosts is now unlimited.
6290 47. The lsearch lookup could not handle data where the length of text line was
6291 more than 4095 characters. Such lines were truncated, leading to shortened
6292 data being returned. It should now handle lines of any length.
6294 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6295 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6298 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6299 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6300 no longer happens in comment lines. A list of replacements is now placed
6301 at the head of all of the source files, except those whose only change is
6302 to replace PERL_COMMAND in the very first #! line.
6304 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6305 messages on the queue, with a bottom-up merge sort, using code contributed
6306 by Michael Haardt. This should make operations like -bp somewhat faster on
6307 large queues. It won't affect queue runners, except when queue_run_in_order
6310 51. Installed eximstats 1.31 in the distribution.
6312 52. Added support for SRV lookups to the dnslookup router.
6314 53. If an ACL referred to $message_body or $message_body_end, the value was not
6315 reset for any messages that followed in the same SMTP session.
6317 54. The store-handling optimization for building very long strings was not
6318 differentiating between the different store pools. I don't think this
6319 actually made any difference in practice, but I've tidied it.
6321 55. While running the routers to verify a sender address, $sender_address
6322 was still set to the sender address. This is wrong, because when routing to
6323 send a bounce to the sender, it would be empty. Therefore, I have changed
6324 it so that, while verifying a sender address, $sender_address is set to <>.
6325 (There is no change to what happens when verifying a recipient address.)
6327 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6328 target A or AAAA records (if not already returned) without resetting the
6329 qualify_single or search_parents options of the DNS resolver. These are
6330 inappropriate in this case because the targets of MX and SRV records must
6331 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6332 target that, when qualified, matched something in the local domain. These
6333 two options are now turned off when doing these lookups.
6335 57. It seems that at least some releases of Reiserfs (which does not have the
6336 concept of a fixed number of inodes) returns zero and not -1 for the
6337 number of available inodes. This interacted badly with check_spool_inodes,
6338 which assumed that -1 was the "no such thing" setting. What I have done is
6339 to check that the total number of inodes is greater than zero before doing
6340 the test of how many are available.
6342 58. When a "warn" ACL statement has a log_message modifier, the message is
6343 remembered, and not repeated. This is to avoid a lot of repetition when a
6344 message has many recipients that cause the same warning to be written.
6345 However, Exim was preserving the list of already written lines for an
6346 entire SMTP session, which doesn't seem right. The memory is now reset if a
6347 new message is started.
6349 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6350 debugging output unless log_rewrite was also set.
6352 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6353 of (int)(handle) when we know that handle contains (void *)(-1).
6355 61. The Exim daemon panic-logs an error return when it closes the incoming
6356 connection. However "connection reset by peer" seems to be common, and
6357 isn't really an error worthy of noting specially, so that particular error
6360 62. When Exim is trying to find all the local interfaces, it used to panic and
6361 die if the ioctl to get the interface flags failed. However, it seems that
6362 on at least one OS (Solaris 9) it is possible to have an interface that is
6363 included in the list of interfaces, but for which you get a failure error
6364 for this call. This happens when the interface is not "plumbed" into a
6365 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6366 failure of the "get flags" call assumes that the interface is down.
6368 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6369 makes life easier for people who are doing arithmetic on fields extracted
6370 from dates, where you often get leading zeros that should not be
6371 interpreted as octal.
6373 64. Added qualify_domain to the redirect router, to override the global
6376 65. If a pathologically long header line contained very many addresses (the
6377 report of this problem mentioned 10 000) and each of them was rewritten,
6378 Exim could use up a very large amount of memory. (It kept on making new
6379 copies of the header line as it rewrote, and never released the old ones.)
6380 At the expense of a bit more processing, the header rewriting function has
6381 been changed so that it no longer eats memory in this way.
6383 66. The generation of the Received: header has been moved from the time that a
6384 message starts to be received, to the time that it finishes. The timestamp
6385 in the Received: header should now be very close to that of the <= log
6386 line. There are two side-effects of this change:
6388 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6389 logged header lines no longer include the local Received: line, because
6390 it has not yet been created. The same applies to a copy of the message
6391 that is returned to a non-SMTP sender when a message is rejected.
6393 (b) When a filter file is tested using -bf, no additional Received: header
6394 is added to the test message. After some thought, I decided that this
6397 This change does not affect the value of $received_for. It is still set
6398 after address rewriting, but before local_scan() is called.
6400 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6402 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6403 gave an unhelpful panic error message, and a defer error. I have managed to
6404 change this behaviour so that it now rejects any supplied certificate,
6405 which seems right, as the list of acceptable certificates is empty.
6407 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6408 gave an unhelpful defer error. I have not managed to make this reject any
6409 supplied certificates, but the error message it gives is "no certificate
6410 supplied", which is not helpful.
6412 70. exigrep's output now also includes lines that are not associated with any
6413 message, but which match the given pattern. Implemented by a patch from
6414 Martin Sluka, which also tidied up the Perl a bit.
6416 71. Recipient callout verification, like sender verification, was using <> in
6417 the MAIL FROM command. This isn't really the right thing, since the actual
6418 sender may affect whether the remote host accepts the recipient or not. I
6419 have changed it to use the actual sender in the callout; this means that
6420 the cache record is now keyed on a recipient/sender pair, not just the
6421 recipient address. There doesn't seem to be a real danger of callout loops,
6422 since a callout by the remote host to check the sender would use <>.
6423 [SEE ABOVE: changed after hitting problems.]
6425 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6426 temporary errors. However, in the case of such a code being given after
6427 the end of a data transmission (i.e. after ".") Exim was failing to write
6428 a retry record for the message. (Yes, there was some broken host that was
6429 actually sending 8xx at this point.)
6431 73. An unknown lookup type in a host list could cause Exim to panic-die when
6432 the list was checked. (An example that provoked this was putting <; in the
6433 middle of a list instead of at the start.) If this happened during a DATA
6434 ACL check, a -D file could be left lying around. This kind of configuration
6435 error no longer causes Exim to die; instead it causes a defer error. The
6436 incident is still logged to the main and panic logs.
6438 74. Buglet left over from Exim 3 conversion. The message "too many messages
6439 in one connection" was written to the rejectlog but not the mainlog, except
6440 when address rewriting (yes!) was being logged.
6442 75. Added write_rejectlog option.
6444 76. When a system filter was run not as root (that is, when system_filter_user
6445 was set), the values of the $n variables were not being returned to the
6446 main process; thus, they were not subsequently available in the $sn
6449 77. Added +return_path_on_delivery log selector.
6451 78. A connection timeout was being treated differently from recipients deferred
6452 when testing hosts_max_try with a message that was older than the host's
6453 retry timeout. (The host should not be counted, thus allowing all hosts to
6454 be tried at least once before bouncing.) This may have been the cause of an
6455 occasionally reported bug whereby a message would remain on the queue
6456 longer than the retry timeout, but would be bounced if a delivery was
6457 forced. I say "may" because I never totally pinned down the problem;
6458 setting up timeout/retry tests is difficult. See also the next item.
6460 79. The ultimate address timeout was not being applied to errors that involved
6461 a combination of host plus message (for example, a timeout on a MAIL
6462 command). When an address resolved to a number of possible hosts, and they
6463 were not all tried for each delivery (e.g. because of hosts_max_try), a
6464 message could remain on the queue longer than the retry timeout.
6466 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6469 81. Fixed an obscure SMTP outgoing bug which required at least the following
6470 conditions: (a) there was another message waiting for the same server;
6471 (b) the server returned 5xx to all RCPT commands in the first message so
6472 that the message was not completed; (c) the server dropped the connection
6473 or gave a negative response to the RSET that Exim sends to abort the
6474 transaction. The observed case was a dropped connection after DATA that had
6475 been sent in pipelining mode. That is, the server had advertised PIPELINING
6476 but was not implementing it correctly. The effect of the bug was incorrect
6477 behaviour, such as trying another host, and this could lead to a crash.
6483 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6484 and daemon.c were passed as pointers to ints; they should have been
6485 pointers to socklen_t variables (which are typically unsigned ints).
6487 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6490 3. Fixed a really odd bug that affected only the testing scheme; patching a
6491 certain fixed string in the binary changed the value of another string that
6492 happened to be identical to the end of the original first string.
6494 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6495 name", it returns that address as the IP address. On some operating
6496 systems (e.g. Solaris), it also passes back the IP address string as the
6497 "host name". However, on others (e.g. Linux), it passes back an empty
6498 string. Exim wasn't checking for this, and was changing the host name to an
6499 empty string, assuming it had been canonicalized.
6501 5. Although rare, it is permitted to have more than one PTR record for a given
6502 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6503 all the names associated with an address, because they do in Solaris.
6504 However, it seems that they do not in Linux for data that comes from the
6505 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6506 I found this out when I moved to a new Linux workstation and tried to run
6507 the Exim test suite.
6509 To get round this problem I have changed the code so that it now does its
6510 own call to the DNS to look up PTR records when searching for a host name.
6511 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6512 addresses that are only in /etc/hosts are still found.
6514 This behaviour is, however, controlled by an option called host_lookup_
6515 order, which defaults to "bydns:byaddr". If people want to use the other
6516 order, or indeed, just use one or the other means of lookup, they can
6517 specify it in this variable.
6519 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6520 some operating systems, this comes back from gethostbyaddr() as an empty
6521 string, and this is what Exim used to test for. However, it seems that in
6522 other systems, "." is yielded. Exim now tests for this case too.
6524 7. The values of check_spool_space and check_log_space are now held internally
6525 as a number of kilobytes instead of an absolute number of bytes. If a
6526 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6527 kilobyte. This means that much larger values can be stored.
6529 8. Exim monitor: an attempt to get the action menu when not actually pointing
6530 at a message produces an empty menu entitled "No message selected". This
6531 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6532 no entries in it ("Shell widget menu has zero width and/or height"). So I
6533 have added a single, blank menu entry in this case.
6535 9. Added ${quote_local_part.
6537 10. MIME decoding is now applied to the contents of Subject: header lines when
6540 11. Now that a reference to $sender_host_address automatically causes a reverse
6541 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6542 host lookup before query-style lookups in lists that might use this
6543 variable. This has therefore been abolished, and the "net-" prefix is no
6544 longer necessary for query-style lookups.
6546 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6547 have been a typo for LFLAGS, so it has been changed.
6549 13. The install script calls Exim with "-C /dev/null" in order to find the
6550 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6551 to be output. However, since Exim outputs its version number before the
6552 error, it didn't break the script. It just looked ugly. I fixed this by
6553 always allowing "-C /dev/null" if the caller is root.
6555 14. Ignore overlarge ACL variable number when reading spool file - insurance
6556 against a later release with more variables having written the file.
6558 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6559 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6560 this, as well as the form without the "IPv6" on the front.
6562 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6563 OS/Makefile-Darwin file.
6565 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6566 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6568 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6569 variable. However, some OS use ints for inodes. I've added cast to long int
6570 to get rid of the compiler warning.
6572 19. I had forgotten to lock out "/../" in configuration file names when
6573 ALT_CONFIG_PREFIX was set.
6575 20. Routers used for verification do not need to specify transports. However,
6576 if such a router generated a host list, and callout was configured, Exim
6577 crashed, because it could not find a port number from the (non-existent)
6578 transport. It now assumes port 25 in this circumstance.
6580 21. Added the -t option to exigrep.
6582 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6583 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6584 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6585 them is set, LOOKUP_LSEARCH is forced.
6587 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6589 24. Added sender and host information to the "rejected by local_scan()" log
6590 line; previously there was no indication of these.
6592 25. Added .include_if_exists.
6594 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6595 Linux. It turns out that not all file systems support this. Apparently some
6596 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6597 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6598 not supported on the file descriptor, is now ignored when Exim is trying to
6599 sync a directory. This applies only to Linux.
6601 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6603 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6604 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6607 29. Added support for maildirsize files from supplied patch (modified a bit).
6609 30. The use of :fail: followed by an empty string could lead Exim to respond to
6610 sender verification failures with (e.g.):
6612 550 Verification failed for <xxx>
6613 550 Sender verify failed
6615 where the first response line was missing the '-' that indicates it is not
6616 the final line of the response.
6618 31. The loop for finding the name of the user that called Exim had a hardwired
6619 limit of 10; it now uses the value of finduser_retries, which is used for
6620 all other user lookups.
6622 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6625 33. Exim was neglecting to zero errno before one call of strtol() when
6626 expanding a string and expecting an integer value. On some systems this
6627 resulted in spurious "integer overflow" errors. Also, it was casting the
6628 result into an int without checking.
6630 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6631 did not work. The code looks as if it has *never* worked, though it appears
6632 to have been documented since at least release 1.62. I have made it work.
6634 35. The "timeout_DNS" error in retry rules, also documented since at least
6635 1.62, also never worked. As it isn't clear exactly what this means, and
6636 clearly it isn't a major issue, I have abolished the feature by treating it
6637 as "timeout", and writing a warning to the main and panic logs.
6639 36. The display of retry rules for -brt wasn't always showing the error code
6642 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6643 timeout_connect_A, timeout_connect_MX.
6645 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6646 to the empty sender.
6648 39. The daemon was not analysing the content of -oX till after it had closed
6649 stderr and disconnected from the controlling terminal. This meant that any
6650 syntax errors were only noted on the panic log, and the return code from
6651 the command was 0. By re-arranging the code a little, I've made the
6652 decoding happen first, so such errors now appear on stderr, and the return
6653 code is 1. However, the actual setting up of the sockets still happens in
6654 the disconnected process, so errors there are still only recorded on the
6657 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6658 connections (as happens on some IP stacks) was logged at start up time as
6659 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6660 it from "IPv6 and IPv4", which means that two separate sockets are being
6663 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6664 says whether AF_INET or AF_INET6 was passed as an argument.
6666 42. Exiwhat output was messed up when time zones were included in log
6669 43. Exiwhat now gives more information about the daemon's listening ports,
6670 and whether -tls-on-connect was used.
6672 44. The "port" option of the smtp transport is now expanded.
6674 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6675 silently ignored. Now an error message is written to the main and panic
6678 46. There's a new ACL modifier called "logwrite" which writes to a log file
6679 as soon as it is encountered.
6681 47. Added $local_user_uid and $local_user_gid at routing time.
6683 48. Exim crashed when trying to verify a sender address that was being
6686 49. Exim was recognizing only a space character after ".include". It now also
6687 recognizes a tab character.
6689 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6690 extracting the relevant information from the specification. The man page no
6691 longer contains scrambled data for the -d option, and I've added a section
6692 at the front about calling Exim under different names.
6694 51. Added "extra_headers" argument to the "mail" command in filter files.
6696 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6699 53. Installed eximstats 1.29.
6701 54. Added transport_filter_timeout as a generic transport option.
6703 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6704 Cc: header lines. This was required by RFC 822, but it not required by RFC
6707 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6708 incoming messages that did not have them. Now it does so only if the
6709 message originates locally, that is, if there is no associated remote host
6710 address. When Resent- header lines are present, this applies to the Resent-
6711 lines rather than the non-Resent- lines.
6713 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6714 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6716 58. Messages for configuration errors now include the name of the main
6717 configuration file - useful now that there may be more than one file in a
6718 list (.included file names were always shown).
6720 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6721 for those rare installations that do not start the daemon as root or run it
6722 setuid root. I've cut out the call to initgroups() if the daemon is not
6725 60. The Exim user and group can now be bound into the binary as text strings
6726 that are looked up at the start of Exim's processing.
6728 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6730 62. Added $mailstore_basename variable.
6732 63. Installed patch to sieve.c from Michael Haardt.
6734 64. When Exim failed to open the panic log after failing to open the main log,
6735 the original message it was trying to log was written to stderr and debug
6736 output, but if they were not available (the usual case in production), it
6737 was lost. Now it is written to syslog before the two lines that record the
6738 failures to open the logs.
6740 65. Users' Exim filters run in subprocesses under the user's uid. It is
6741 possible for a "deliver" command or an alias in a "personal" command to
6742 provoke an address rewrite. If logging of address rewriting is configured,
6743 this fails because the process is not running as root or exim. There may be
6744 a better way of dealing with this, but for the moment (because 4.30 needs
6745 to be released), I have disabled address rewrite logging when running a
6746 filter in a non-root, non-exim process.
6752 1. The buildconfig auxiliary program wasn't quoting the value set for
6753 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6754 not defined. This bug was masked in 4.22 by the effect that was fixed in
6757 2. Some messages that were rejected after a message id was allocated were
6758 shown as "incomplete" by exigrep. It no longer does this for messages that
6759 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6761 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6762 have allow_domain_literals set, the ID did not get logged in the <= line.
6763 Domain literals are now always recognized in Message-ID: header lines.
6765 4. The first argument for a ${extract expansion item is the key name or field
6766 number. Leading and trailing spaces in this item were not being ignored,
6767 causing some misleading effects.
6769 5. When deliver_drop_privilege was set, single queue runner processes started
6770 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6771 same command in the process it spins off) were not dropping privilege.
6773 6. When the daemon running as "exim" started a queue runner, it always
6774 re-executed Exim in the spun-off process. This is a waste of effort when
6775 deliver_drop_privilege is set. The new process now just calls the
6776 queue-runner function directly.
6782 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6785 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6786 ignored. Though the use of -oP was forcing the writing of a pid file, it
6787 was always written to the default place.
6789 3. If the message "no IP address found for host xxxx" is generated during
6790 incoming verification, it is now followed by identification of the incoming
6791 connection (so you can more easily find what provoked it).
6793 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6795 5. Added some features to "harden" Exim a bit more against certain attacks:
6797 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6798 be put in Local/Makefile. This is like the never_users runtime option,
6799 but it cannot be overridden. The default setting is "root".
6801 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6802 prefix string with which any file named in a -C command line option
6805 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6806 is retained for -C and -D only if the caller of Exim is root. Without
6807 it, the exim user may also use -C and -D and retain privilege.
6809 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6810 command line option is disabled.
6812 6. Macro names set by the -D option must start with an upper case letter, just
6813 like macro names defined in the configuration file.
6815 7. Added "dereference=" facility to LDAP.
6817 8. Two instances of the typo "uknown" in the source files are fixed.
6819 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6820 the Configure-Makefile script screwed up while processing it.
6822 10. Incorporated PCRE 4.4.
6824 11. The SMTP synchronization check was not operating right at the start of an
6825 SMTP session. For example, it could not catch a HELO sent before the client
6826 waited for the greeting. There is now a check for outstanding input at the
6827 point when the greeting is written. Because of the duplex, asynchronous
6828 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6829 way, but not yet received, when the check is performed.
6831 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6832 on TCP/IP sockets, because this apparently causes some broken clients to
6835 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6836 unchanged) from the Cygwin maintainer.
6838 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6839 was supported instead of testing for mbx. Effectively a typo.
6841 15. The spa authenticator server code was not checking that the input it
6842 received was valid base64.
6844 16. The debug output line for the "set" modifier in ACLs was not showing the
6845 name of the variable that was being set.
6847 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6849 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6850 lookup on its own. Something else was needed to trigger the lookup. For
6851 example, a match in host_lookup or the need for a host name in a host list.
6852 Now, if $sender_host_name is referenced and the host name has not yet been
6853 looked up, a lookup is performed. If the lookup fails, the variable remains
6854 empty, and $host_lookup_failed is set to "1".
6856 19. Added "eqi" as a case-independent comparison operator.
6858 20. The saslauthd authentication condition could segfault if neither service
6859 nor realm was specified.
6861 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6862 error message that was logged was misleading, and incoming SMTP
6863 connections were dropped. The message is now more accurate, and temporary
6864 errors are given to SMTP connections.
6866 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6867 (except RSET) with a 421 error, until QUIT is received. However, it was
6868 failing to send a response to QUIT.
6870 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6871 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6872 ACL is now run after the helo_try_verify_hosts code.
6874 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6875 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6876 case-independent, so other case variants are also recognized). Apparently
6877 some systems use these upper case variants.
6879 25. If more than two messages were waiting for the same host, and a transport
6880 filter was specified for the transport, Exim sent two messages over the
6881 same TCP/IP connection, and then failed with "socket operation on non-
6882 socket" when it tried to send the third.
6884 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6886 27. The extern definition of crypt16() in expand.c was not being excluded when
6887 the OS had its own crypt16() function.
6889 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6890 as a preferred synonym for return_size_limit, both as an option and as an
6893 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6895 30. Changed the default configuration ACL to relax the local part checking rule
6896 for addresses that are not in any local domains. For these addresses,
6897 slashes and pipe symbols are allowed within local parts, but the sequence
6898 /../ is explicitly forbidden.
6900 31. SPA server authentication was not clearing the challenge buffer before
6903 32. log_message in a "warn" ACL statement was writing to the reject log as
6904 well as to the main log, which contradicts the documentation and doesn't
6905 seem right (because no rejection is happening). So I have stopped it.
6907 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6908 However, I am unable to do any testing of this.
6910 34. Fixed an infelicity in the appendfile transport. When checking directories
6911 for a mailbox, to see if any needed to be created, it was accidentally
6912 using path names with one or more superfluous leading slashes; tracing
6913 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6915 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6916 contents are added to the log line that is written for every discarded
6917 recipient. (Previously a log_message setting was ignored.)
6919 36. The ${quote: operator now quotes the string if it is empty.
6921 37. The install script runs exim in order to find its version number. If for
6922 some reason other than non-existence or emptiness, which it checks, it
6923 could not run './exim', it was installing it with an empty version number,
6924 i.e. as "exim-". This error state is now caught, and the installation is
6927 38. An argument was missing from the function that creates an error message
6928 when Exim fails to connect to the socket for saslauthd authentication.
6929 This could cause Exim to crash, or give a corrupted message.
6931 39. Added isip, isip4, and isip6 to ${if conditions.
6933 40. The ACL variables $acl_xx are now saved with the message, and can be
6934 accessed later in routers, transports, and filters.
6936 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6937 strings in the file are not string-expanded.
6939 42. If a MAIL command specified a SIZE value that was too large to fit into an
6940 int variable, the check against message_size_limit failed. Such values are
6941 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6942 day this will have to be increased, but I don't think I want to be around
6943 when emails are that large.
6950 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6951 iconv() is not standard in FreeBSD.
6953 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6954 IPv6 enabled. The observed symptom was a segmentation fault on return from
6955 the function os_common_find_running_interfaces() in src/os.c.
6957 3. In the check_special_case() function in daemon.c I had used "errno" as an
6958 argument name, which causes warnings on some systems. This was basically a
6959 typo, since it was named "eno" in the comments!
6961 4. The code that waits for the clock to tick (at a resolution of some fraction
6962 of a second) so as to ensure message-id uniqueness was always waiting for
6963 at least one whole tick, when it could have waited for less. [This is
6964 almost certainly not relevant at current processor speeds, where it is
6965 unlikely to ever wait at all. But we try to future-proof.]
6967 5. The function that sleeps for a time interval that includes fractions of a
6968 second contained a race. It did not block SIGALRM between setting the
6969 timer, and suspending (a couple of lines later). If the interval was short
6970 and the sigsuspend() was delayed until after it had expired, the suspension
6971 never ended. On busy systems this could lead to processes getting stuck for
6974 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6975 process, before it forks any delivery processes. The open lookup caching
6976 mechanism meant that the open file or database connection was passed into
6977 the delivery process. The problem was that delivery processes always tidy
6978 up cached lookup data. This could cause a problem for the next delivery
6979 process started by the queue runner, because the external queue runner
6980 process does not know about the closure. So the next delivery process
6981 still has data in the lookup cache. In the case of a file lookup, there was
6982 no problem because closing a file descriptor in a subprocess doesn't affect
6983 the parent. However, if the lookup was caching a connection to a database,
6984 the connection was closed, and the second delivery process was likely to
6985 see errors such as "PGSQL: query failed: server closed the connection
6986 unexpectedly". The problem has been fixed by closing all cached lookups
6987 in a queue runner before running a delivery process.
6989 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6990 seem to have the "const" qualifier which it has on other OS. I've
6993 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6994 *of the same type* (client or server) with the same public name that an
6995 error should be diagnosed.
6997 9. When Exim looked up a host name for an IP address, but failed to find the
6998 original IP address when looking up the host name (a safety check), it
6999 output the message "<ip address> does not match any IP for NULL", which was
7000 confusing, to say the least. The bug was that the host name should have
7001 appeared instead of "NULL".
7003 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7004 user that is built into the binary, and the -C or -D options is used, root
7005 privilege is dropped before the configuration file is read. In addition,
7006 logging is switched to stderr instead of the normal log files. If the
7007 configuration then re-defines the Exim user, the unprivileged environment
7008 is probably not what is expected, so Exim logs a panic warning message (but
7011 However, if deliver_drop_privilege is set, the unprivileged state may well
7012 be exactly what is intended, so the warning has been cut out in that case,
7013 and Exim is allowed to try to write to its normal log files.
7019 1. smtp_return_error_details was not giving details for temporary sender
7020 or receiver verification errors.
7022 2. Diagnose a configuration error if two authenticators have the same public
7025 3. Exim used not to create the message log file for a message until the first
7026 delivery attempt. This could be confusing when incoming messages were held
7027 for policy or load reasons. The message log file is now created at the time
7028 the message is received, and an initial "Received" line is written to it.
7030 4. The automatically generated man page for command line options had a minor
7031 bug that caused no ill effects; however, a more serious problem was that
7032 the procedure for building the man page automatically didn't always
7033 operate. Consequently, release 4.20 contains an out-of-date version. This
7034 shouldn't happen again.
7036 5. When building Exim with embedded Perl support, the script that builds the
7037 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7038 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7040 6. The freeze_tell option was not being used for messages that were frozen on
7041 arrival, either by an ACL or by local_scan().
7043 7. Added the smtp_incomplete_transaction log selector.
7045 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7046 was accepting AUTH without a new EHLO.
7048 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7049 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7052 10. Logging of TCP/IP connections (when configured) now happens in the main
7053 daemon process instead of the child process, so that the TCP/IP connection
7054 count is more accurate (but it can never be perfect).
7056 11. The use of "drop" in a nested ACL was not being handled correctly in the
7057 outer ACL. Now, if condition failure induced by the nested "drop" causes
7058 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7059 or "require"), the connection is dropped.
7061 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7062 that yield "discard" can now be used with an "accept" or a "discard" verb,
7063 but an error is generated for any others (because I can't see a useful way
7064 to define what should happen).
7066 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7067 are now processed in the same way as lines in the Exim configuration file.
7068 In particular, continuation lines are supported.
7070 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7072 15. Added -ti meaning -t -i.
7074 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7075 domains, and warn by logging if others are found.
7077 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7078 returned from GIFCONF when Exim is trying to find the list of interfaces on
7079 a host. The code in os.c has been modified to copy each ifreq to an aligned
7080 structure in all cases.
7082 Also, in some cases, the returned ifreq's were being copied to a 'struct
7083 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7084 means the last couple of bytes of an IPv6 address could be chopped if the
7085 ifreq contained only a normal sockaddr (14 bytes storage).
7087 18. Named domain lists were not supported in the hosts_treat_as_local option.
7088 An entry such as +xxxx was not recognized, and was treated as a literal
7091 19. Ensure that header lines added by a DATA ACL are included in the reject log
7092 if the ACL subsequently rejects the message.
7094 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7095 MD5 (which is deprecated).
7097 21. When testing a filter file using -bf, Exim was writing a message when it
7098 took the sender from a "From " line in the message, but it was not doing so
7099 when it took $return_path from a Return-Path: header line. It now does.
7101 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7102 with a valid header line field name (a series of printing characters
7103 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7105 23. Changed "disc" in the source to "disk" to conform to the documentation and
7106 the book and for uniformity.
7108 24. Ignore Sendmail's -Ooption=value command line item.
7110 25. When execve() failed while trying to run a command in a pipe transport,
7111 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7112 could be confused with a return value of 69 from the command itself. This
7113 has been changed to 127, the value the shell returns if it is asked to run
7114 a non-existent command. The wording for the related log line suggests a
7115 non-existent command as the problem.
7117 26. If received_header_text expands to an empty string, do not add a Received:
7118 header line to the message. (Well, it adds a token one on the spool, but
7119 marks it "old" so that it doesn't get used or transmitted.)
7121 27. Installed eximstats 1.28 (addition of -nt option).
7123 28. There was no check for failure on the call to getsockname() in the daemon
7124 code. This can fail if there is a shortage of resources on the system, with
7125 ENOMEM, for example. A temporary error is now given on failure.
7127 29. Contrary to the C standard, it seems that in some environments, the
7128 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7129 program. Exim now does this explicitly; it affects the formatting of
7130 timestamps using strftime().
7132 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7133 complaints. I've now initialized all the variables, to avoid this.
7135 32. Header lines added by a system filter were not being "seen" during
7136 transport-time rewrites.
7138 33. The info_callback() function passed to OpenSSL is set up with type void
7139 (*)(SSL *, int, int), as described somewhere. However, when calling the
7140 function (actually a macro) that sets it up, the type void(*)() is
7141 expected. I've put in a cast to prevent warnings from picky compilers.
7143 34. If a DNS black list lookup found a CNAME record, but there were no A
7144 records associated with the domain it pointed at, Exim crashed.
7146 35. If a DNS black list lookup returned more than one A record, Exim ignored
7147 all but the first. It now scans all returned addresses if a particular IP
7148 value is being sought. In this situation, the contents of the
7149 $dnslist_value variable are a list of all the addresses, separated by a
7152 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7153 to accept a host name and all its aliases if the forward lookup for any of
7154 them yielded the IP address of the incoming connection. Now it accepts only
7155 those names whose forward lookup yields the correct IP address. Any other
7156 names are discarded. This closes a loophole whereby a rogue DNS
7157 administrator could create reverse DNS records to break through a
7158 wildcarded host restriction in an ACL.
7160 37. If a user filter or a system filter that ran in a subprocess used any of
7161 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7162 the wrong values were passed to the pipe command ($thisaddress had the
7163 value of $0, $0 had the value of $1, etc). This bug was introduced by
7164 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7166 38. Improved the line breaking for long SMTP error messages from ACLs.
7167 Previously, if there was no break point between 40 and 75 characters, Exim
7168 left the rest of the message alone. Two changes have been made: (a) I've
7169 reduced the minimum length to 35 characters; (b) if it can't find a break
7170 point between 35 and 75 characters, it looks ahead and uses the first one
7171 that it finds. This may give the occasional overlong line, but at least the
7172 remaining text gets split now.
7174 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7175 file descriptors might be low, and that setting 1000 would always raise it.
7176 It turns out that in some environments, the limit is already over 1000 and
7177 that lowering it causes trouble. So now Exim takes care not to decrease it.
7179 40. When delivering a message, the value of $return_path is set to $sender_
7180 address at the start of routing (routers may change the value). By an
7181 oversight, this default was not being set up when an address was tested by
7182 -bt or -bv, which affected the outcome if any router or filter referred to
7185 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7186 log only when "message" or "log_message" are set. However, if one of the
7187 conditions was an address verification, or a call to a nested ACL, the
7188 messages generated by the underlying test were being passed through. This
7189 no longer happens. The underlying message is available in $acl_verify_
7190 message for both "message" and "log_message" expansions, so it can be
7191 passed through if needed.
7193 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7194 new expansion $bh_ to give the encoded byte string without charset
7195 translation. Translation happens only if iconv() is available; HAVE_ICONV
7196 indicates this at build time. HEADERS_CHARSET gives the charset to
7197 translate to; headers_charset can change it in the configuration, and
7198 "headers charset" can change it in an individual filter file.
7200 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7201 that creates RFC 2047 encoded "words" labels them as that charset instead
7202 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7203 expansion operator; (ii) when Exim creates a From: line for a local
7204 message; (iii) when a header line is rewritten to include a "phrase" part.
7206 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7207 buggy, causing it to skip the first lines of messages whose message ID
7208 ended in 'D'. This would not have bitten before Exim release 4.14, because
7209 message IDs were unlikely to end in 'D' before then. The effect was to have
7210 incorrect size information for certain domains.
7212 45. #include "config.h" was missing at the start of the crypt16.c module. This
7213 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7216 46. If there was a timeout during a "random" callout check, Exim treated it as
7217 a failure of the random address, and carried on sending RSET and the real
7218 address. If the delay was just some slowness somewhere, the response to the
7219 original RCPT would be taken as a response to RSET and so on, causing
7220 mayhem of various kinds.
7222 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7223 when I implemented it. It didn't allow for the fact that some option values
7224 may legitimately be negative (e.g. size_addition), and it didn't even do
7225 the right test for positive values.
7227 48. Domain names in DNS records are case-independent. Exim always looks them up
7228 in lower case. Some resolvers return domain names in exactly the case they
7229 appear in the zone file, that is, they may contain uppercase letters. Not
7230 all resolvers do this - some return always lower case. Exim was treating a
7231 change of case by a resolver as a change of domain, similar to a widening
7232 of a domain abbreviation. This triggered its re-routing code and so it was
7233 trying to route what was effectively the same domain again. This normally
7234 caused routing to fail (because the router wouldn't handle the domain
7235 twice). Now Exim checks for this case specially, and just changes the
7236 casing of the domain that it ultimately uses when it transmits the message
7239 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7242 50. If a filter generated a file delivery with a non-absolute name (possible if
7243 no home directory exists for the router), the forbid_file option was not
7246 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7247 the existing equality matching.
7249 52. Exim was using ints instead of ino_t variables in some places where it was
7250 dealing with inode numbers.
7252 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7253 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7254 called TMPDIR, and if it finds it is different, it changes its value.
7256 54. The smtp_printf() function is now made available to local_scan() so
7257 additional output lines can be written before returning. There is also an
7258 smtp_fflush() function to enable the detection of a dropped connection.
7259 The variables smtp_input and smtp_batched_input are exported to
7262 55. Changed the default runtime configuration: the message "Unknown user"
7263 has been removed from the ACL, and instead placed on the localuser router,
7264 using the cannot_route_message feature. This means that any verification
7265 failures that generate their own messages won't get overridden. Similarly,
7266 the "Unrouteable address" message that was in the ACL for unverifiable
7267 relay addresses has also been removed.
7269 56. Added hosts_avoid_esmtp to the smtp transport.
7271 57. The exicyclog script was not checking for the esoteric option
7272 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7273 will work only if exicyclog is run under the appropriate euid.
7275 58. Following a discussion on the list, the rules by which Exim recognises line
7276 endings on incoming messages have been changed. The -dropcr and drop_cr
7277 options are now no-ops, retained only for backwards compatibility. The
7278 following line terminators are recognized: LF CRLF CR. However, special
7279 processing applies to CR:
7281 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7282 nor a local message in the state where . is a terminator.
7284 (ii) If a bare CR is encountered in a header line, an extra space is added
7285 after the line terminator so as not to end the header. The reasoning
7286 behind this is that bare CRs in header lines are most likely either
7287 to be mistakes, or people trying to play silly games.
7289 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7290 was being incorrectly given as 18 bytes larger than it should have been.
7291 This is a VOB (very old bug).
7293 60. This may never have affected anything current, but just in case it has:
7294 When the local host is found other than at the start of a list of hosts,
7295 the local host, those with the same MX, and any that follow, are discarded.
7296 When the list in question was part of a longer list of hosts, the following
7297 hosts (not currently being processed) were also being discarded. This no
7298 longer happens. I'm not sure if this situation could ever has previously
7301 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7304 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7305 header. This is something that is recommended in a new Internet Draft, and
7306 is something that is documented as being done by Sendmail. There are two
7307 possible values. For messages generated by the autoreply transport, Exim
7310 Auto-Submitted: auto-replied
7312 whereas for all other generated messages (e.g. bounces) it adds
7314 Auto-Submitted: auto-generated
7316 63. The "personal" condition in filters now includes a test for the
7317 Auto-Submitted: header. If it contains the string "auto-" the message it
7318 not considered personal.
7320 64. Added rcpt_include_affixes as a generic transport option.
7322 65. Added queue_only_override (default true).
7324 66. Added the syslog_duplication option.
7326 67. If what should have been the first header line of a message consisted of
7327 a space followed by a colon, Exim was mis-interpreting it as a header line.
7328 It isn't of course - it is syntactically invalid and should therefore be
7329 treated as the start of the message body. The misbehaviour could have
7330 caused a number of strange effects, including loss of data in subsequent
7331 header lines, and spool format errors.
7333 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7334 client host had authenticated. This control can now be exercised by an ACL
7335 for more flexibility.
7337 69. By default, callouts do not happen when testing with -bh. There is now a
7338 variant, -bhc, which does actually run the callout code, including
7339 consulting and updating the callout cache.
7341 70. Added support for saslauthd authentication, courtesy of Alexander
7344 71. If statvfs() failed on the spool or log directories while checking their
7345 size for availability, Exim confusingly gave the error "space shortage".
7346 Furthermore, in debugging mode it crashed with a floating point exception.
7347 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7348 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7349 serious problem, Exim now writes to the main and panic logs when this
7350 happens, with details of the failure. It then refuses to accept the
7351 incoming message, giving the message "spool directory problem" or "log
7352 directory problem" with a 421 code for SMTP messages.
7354 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7355 0, 1, and 2 exist, because some OS complain for execs without them (see
7356 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7357 descriptors. However, the code omitted to check that the open succeeded,
7358 causing mysterious errors if for some reason the permissions on /dev/null
7359 got screwed. Now Exim writes a message to the main and panic logs, and
7360 bombs out if it can't open /dev/null.
7362 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7363 interact so that it is all more flexible. It is supposed to remain
7364 backwards compatible. Also added extra_local_interfaces.
7366 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7367 to bomb out with an assertion failure - to the client this appears as a
7368 connection drop. This problem occurs in the part of the code that was taken
7369 from the Samba project. Fortunately, the assertion is in a very simple
7370 function, so I have fixed this by reproducing the function inline in the
7371 one place where it is called, and arranging for authentication to fail
7372 instead of killing the process with assert().
7374 75. The SPA client code was not working when the server requested OEM rather
7375 than Unicode encoding.
7377 76. Added code to make require_files with a specific uid setting more usable in
7378 the case where statting the file as root fails - usually a non-root-mounted
7379 NFS file system. When this happens and the failure is EACCES, Exim now
7380 forks a subprocess and does the per-uid checking as the relevant uid.
7382 77. Added process_log_path.
7384 78. If log_file_path was not explicitly set, a setting of check_log_space or
7385 check_log_inodes was ignored.
7387 79. If a space check for the spool or log partitions fails, the incident is now
7388 logged. Of course, in the latter case the data may get lost...
7390 80. Added the %p formatting code to string_format() so that it can be used to
7391 print addresses in debug_print(). Adjusted all the address printing in the
7392 debugging in store.c to use %p rather than %d.
7394 81. There was a concern that a line of code in smtp_in.c could overflow a
7395 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7396 initially expressed, the concern was not well-founded, because trailing
7397 spaces are removed early. However, if the trailing spaces were followed by
7398 a NULL, they did not get removed, so the overflow was possible. Two fixes
7401 (a) I re-wrote the offending code in a cleaner fashion.
7402 (b) If an incoming SMTP command contains a NULL character, it is rejected
7405 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7406 runs initgroups(), so that if the Exim user is in any additional groups,
7407 they will be used during message reception.
7413 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob