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.
54 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
55 to changes apparently associated with TLS1.3 handling some of the APIs
56 previously used were either nonfunctional or inappropriate. Strings
57 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
58 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
59 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
60 This affects log line X= elements, the $tls_{in,out}_cipher variables,
61 and the use of specific cipher names in the encrypted= ACL condition.
63 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
64 verification result was not updated unless hosts_require_ocsp applied.
66 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
67 queue_list_requires_admin set to false, non-admin users were denied the
70 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
71 directory-of-certs mode. Previously they were advertised despite the
74 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
75 success-DSN messages. Previously the From: header was always the default
76 one for these; the option was ignored.
78 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
84 JH/01 Remove code calling the customisable local_scan function, unless a new
85 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
87 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
88 non-signal-safe functions being used.
90 JH/03 Bug 2269: When presented with a received message having a stupidly large
91 number of DKIM-Signature headers, disable DKIM verification to avoid
92 a resource-consumption attack. The limit is set at twenty.
94 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
95 report of oldest_pass in ${authres } in consequence, and separate out
96 some descriptions of reasons for verification fail.
98 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
99 files in the spool were present and unlocked. A queue-runner could spot
100 them, resulting in a duplicate delivery. Fix that by doing the unlock
101 after the unlink. Investigation by Tim Stewart. Take the opportunity to
102 add more error-checking on spoolfile handling while that code is being
105 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
106 No known attacks, no CVE, this is defensive hardening.
108 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
109 a queue-runner could start a delivery while other operations were ongoing.
110 Cutthrough delivery was a common victim, resulting in duplicate delivery.
111 Found and investigated by Tim Stewart. Fix by using the open message data
112 file handle rather than opening another, and not locally closing it (which
113 releases a lock) for that case, while creating the temporary .eml format
114 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
116 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
117 $sender_verify_failure/$recipient_verify_failure to "random".
119 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
122 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
123 Previously this would segfault.
125 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
128 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
129 like zero, since the resolver should be doing this for us, But we need one
130 as a CNAME but no MX presence gets the CNAME returned; we need to check
131 that doesn't point to an MX to declare it "no MX returned" rather than
132 "error, loop". A new main option is added so the older capability of
133 following some limited number of chain links is maintained.
135 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
137 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
138 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
139 now-deprecated earlier definitions used only specified the range up to TLS
140 1.2 (in the older-version library docs).
142 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
144 JH/15 Rework TLS client-side context management. Stop using a global, and
145 explicitly pass a context around. This enables future use of TLS for
146 connections to service-daemons (eg. malware scanning) while a client smtp
147 connection is using TLS; with cutthrough connections this is quite likely.
149 JH/16 Fix ARC verification to do AS checks in reverse order.
151 JH/17 Support a "tls" option on the ${readsocket } expansion item.
153 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
154 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
155 Previously the "utf8" would be re-prepended for every additional message.
157 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
158 Previously thery were accepted, resulting in issues when attempting to
159 forward messages to a non-supporting MTA.
161 PP/02 Let -n work with printing macros too, not just options.
163 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
164 one parent address was copied, and bogus data was used at delivery-logging
165 time. Either a crash (after delivery) or bogus log data could result.
166 Discovery and analysis by Tim Stewart.
168 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
169 Previously if the string ended mid-character, we did not insert the
170 promised '?' replacement.
172 PP/04 Documentation: current string operators work on bytes, not codepoints.
174 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
175 should pack well giving a smaller memory footprint so better caching and
176 therefore performance. Group the declarations where this can't be done so
177 that the byte-sized flag variables are not interspersed among pointer
178 variables, giving a better chance of good packing by the compiler.
180 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
181 non-null, to avoid issues with sites running BATV. Previously reports were
182 sent with an empty envelope sender so looked like bounces.
184 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
185 The ignore_error flag wasn't being returned from the filter subprocess so
186 was not set for later routers. Investigation and fix by Matthias Kurz.
188 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
189 and a msg:complete for the whole, when a message is manually removed using
190 -Mrm. Developement by Matthias Kurz, hacked on by JH.
192 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
193 a "Gnu special" function, asprintf() in the DB utility binary builds; I
194 hope that is portable enough.
196 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
197 requiring a known-CA anchor certificate; make it now rely entirely on the
198 TLSA as an anchor. Checking the name on the leaf cert against the name
199 on the A-record for the host is still done for TA (but not for EE mode).
201 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
202 included in delivery lines for non-proxied connections, when compiled with
203 SUPPORT_SOCKS and running with proxy logging enabled.
205 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
206 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
207 move the existing event to fire before the normal logging of message
208 failure so that custom logging is bracketed by normal logging.
210 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
211 msg:fail:internal event. Developement by Matthias Kurz.
213 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
214 far too small for todays use of crypto signatures stored there. Go all
215 the way to the max DNS message size of 64kB, even though this might be
216 overmuch for IOT constrained device use.
218 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
219 copy a string over itself. The library routine is documented as not
220 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
222 JH/32 For main options check_spool_space and check_inode_space, where the
223 platform supports 64b integers, support more than the previous 2^31 kB
224 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
225 the previous G, M, k.
227 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
228 $authenticated_fail_id variable on authentication failure. Previously
231 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
232 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
233 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
234 GNUTLS_SEC_PARAM_MEDIUM.
236 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
237 side. Previously we would continue as if no SNI had been received.
239 JH/36 Harden the handling of string-lists. When a list consisted of a sole
240 "<" character, which should be a list-separator specification, we walked
241 off past the nul-terimation.
243 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
244 causes) even when the retry time is not yet met. Previously they were
245 not, meaning that when (say) an account was over-quota and temp-rejecting,
246 and multiple senders' messages were queued, only one sender would get
247 notified on each configured delay_warning cycle.
249 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
251 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
252 cipher-suites, an error can be left on the stack even for a succeeding
253 accept; this results in impossible error messages when a later operation
256 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
257 return error codes indicating retry. Under TLS1.3 this becomes required.
259 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
260 it only wrote the new authenticators, resulting in a lack of tracking of
261 peer changes of ESMTP extensions until the next cache flush.
263 JH/41 Fix the loop reading a message header line to check for integer overflow,
264 and more-often against header_maxsize. Previously a crafted message could
265 induce a crash of the recive process; now the message is cleanly rejected.
267 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
268 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
274 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
275 When redis_servers is set to a list of > 1 element, and the Redis servers
276 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
277 case of MOVED into a DEFER case instead, thus moving the query onto the
278 next server in the list. For a cluster of N elements, all N servers must
279 be defined in redis_servers.
281 GF/02 Catch and remove uninitialized value warning in exiqsumm
282 Check for existence of @ARGV before looking at $ARGV[0]
284 JH/01 Replace the store_release() internal interface with store_newblock(),
285 which internalises the check required to safely use the old one, plus
286 the allocate and data copy operations duplicated in both (!) of the
287 extant use locations.
289 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
290 modifier. This matches the restriction on the commandline.
292 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
293 Previously only the last row was returned.
295 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
296 we assumed that tags in the header were well-formed, and parsed the
297 element content after inspecting only the first char of the tag.
298 Assumptions at that stage could crash the receive process on malformed
301 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
302 While running the DKIM ACL we operate on the Permanent memory pool so that
303 variables created with "set" persist to the DATA ACL. Also (at any time)
304 DNS lookups that fail create cache records using the Permanent pool. But
305 expansions release any allocations made on the current pool - so a dnsdb
306 lookup expansion done in the DKIM ACL releases the memory used for the
307 DNS negative-cache, and bad things result. Solution is to switch to the
308 Main pool for expansions.
309 While we're in that code, add checks on the DNS cache during store_reset,
310 active in the testsuite.
311 Problem spotted, and debugging aided, by Wolfgang Breyha.
313 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
314 When none of the hosts presented to a transport match an already-open
315 connection, close it and proceed with the list. Previously we would
316 queue the message. Spotted by Lena with Yahoo, probably involving
319 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
320 Previously a spurious "250 OK id=" response was appended to the proper
323 JH/08 The "support for" informational output now, which built with Content
324 Scanning support, has a line for the malware scanner interfaces compiled
325 in. Interface can be individually included or not at build time.
327 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
328 by the template makefile "src/EDITME". The "STREAM" support for an older
329 ClamAV interface method is removed.
331 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
332 rows affected is given instead).
334 JH/11 The runtime Berkeley DB library version is now additionally output by
335 "exim -d -bV". Previously only the compile-time version was shown.
337 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
338 SMTP connection. Previously, when one had more receipients than the
339 first, an abortive onward connection was made. Move to full support for
340 multiple onward connections in sequence, handling cutthrough connection
341 for all multi-message initiating connections.
343 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
344 routers. Previously, a multi-recipient message would fail to match the
345 onward-connection opened for the first recipient, and cause its closure.
347 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
348 a timeout on read on a GnuTLS initiating connection, resulting in the
349 initiating connection being dropped. This mattered most when the callout
350 was marked defer_ok. Fix to keep the two timeout-detection methods
353 JH/15 Relax results from ACL control request to enable cutthrough, in
354 unsupported situations, from error to silently (except under debug)
355 ignoring. This covers use with PRDR, frozen messages, queue-only and
358 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
360 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
361 metadata, resulting in a crash in free().
363 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
364 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
365 Broken also in d185889f4, with init system revamp.
367 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
368 Previously we abruptly closed the connection after reading a malware-
369 found indication; now we go on to read the "scan ok" response line,
372 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
373 ACL. Previously, a crash would result.
375 JH/19 Speed up macro lookups during configuration file read, by skipping non-
376 macro text after a replacement (previously it was only once per line) and
377 by skipping builtin macros when searching for an uppercase lead character.
379 JH/20 DANE support moved from Experimental to mainline. The Makefile control
380 for the build is renamed.
382 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
383 was allocated for every new TLS startup, meaning one per message. Fix
384 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
386 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
387 reported the original. Fix to report (as far as possible) the ACL
388 result replacing the original.
390 JH/23 Fix memory leak during multi-message connections using STARTTLS under
391 OpenSSL. Certificate information is loaded for every new TLS startup,
392 and the resources needed to be freed.
394 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
396 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
399 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
400 DATA response info to the (existing) per-recipient response info for
401 the "C=" log element. It can have useful tracking info from the
402 destination system. Patch from Simon Arlott.
404 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
405 length value. Previously this would segfault.
407 HS/02 Support Avast multiline protoocol, this allows passing flags to
408 newer versions of the scanner.
410 JH/28 Ensure that variables possibly set during message acceptance are marked
411 dead before release of memory in the daemon loop. This stops complaints
412 about them when the debug_store option is enabled. Discovered specifically
413 for sender_rate_period, but applies to a whole set of variables.
414 Do the same for the queue-runner and queue-list loops, for variables set
415 from spool message files. Do the same for the SMTP per-message loop, for
416 certain variables indirectly set in ACL operations.
418 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
419 as a multi-recipient message from a mailinglist manager). The coding had
420 an arbitrary cutoff number of characters while checking for more input;
421 enforced by writing a NUL into the buffer. This corrupted long / fast
422 input. The problem was exposed more widely when more pipelineing of SMTP
423 responses was introduced, and one Exim system was feeding another.
424 The symptom is log complaints of SMTP syntax error (NUL chars) on the
425 receiving system, and refused recipients seen by the sending system
426 (propating to people being dropped from mailing lists).
427 Discovered and pinpointed by David Carter.
429 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
430 replaced by the ${authresults } expansion.
432 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
434 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
435 allows proper process termination in container environments.
437 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
438 Previously the "final dot" had a newline after it; ensure it is CR,LF.
440 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
441 and "err_perm", deprecated since 4.83 when the RFC-defined words
442 "temperror" and "permerror" were introduced.
444 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
445 transport-filters or DKIM-signing. The restriction was lost in the
446 consolidation of verify-callout and delivery SMTP handling.
447 Extend the restriction to also cover ARC-signing.
449 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
450 in defer=pass mode supply a 450 to the initiator. Previously the message
453 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
454 tls_require_ciphers is used as before.
456 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
457 "pass_unscanned". Only tmpfails from the scanner are written to
458 the paniclog, as they may require admin intervention (permission
459 denied, license issues). Other scanner errors (like decompression
460 bombs) do not cause a paniclog entry.
462 JH/36 Fix reinitialisation of DKIM logging variable between messages.
463 Previously it was possible to log spurious information in receive log
466 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
467 triggered odd behaviour from Outlook Express clients.
469 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
472 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
473 since the IETF WG has not yet settled on that versus the original
474 "bare" representation.
476 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
477 Previously the millisecond value corrupted the output.
478 Fix also for syslog_pid=no and log_selector +pid, for which the pid
479 corrupted the output.
485 JH/01 Rework error string handling in TLS interface so that the caller in
486 more cases is responsible for logging. This permits library-sourced
487 string to be attached to addresses during delivery, and collapses
488 pairs of long lines into single ones.
490 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
491 during configuration. Wildcards are allowed and expanded.
493 JH/02 Rework error string handling in DKIM to pass more info back to callers.
494 This permits better logging.
496 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
497 do not close it down and have the child transport start it up again on
498 the passed-on TCP connection. Instead, proxy the child (and any
499 subsequent ones) for TLS via a unix-domain socket channel. Logging is
500 affected: the continued delivery log lines do not have any DNSSEC, TLS
501 Certificate or OCSP information. TLS cipher information is still logged.
503 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
504 identical IP addresses on different listening ports. Will also affect
507 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
508 add noisy ifdef guards to special-case this sillyness.
509 Patch from Bernd Kuhls.
511 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
512 than 255 are no longer allowed.
514 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
515 Disable the session-cache too, which might reduce our load. Since we
516 currrectly use a new context for every connection, both as server and
517 client, there is no benefit for these.
518 GnuTLS appears to not support tickets server-side by default (we don't
519 call gnutls_session_ticket_enable_server()) but client side is enabled
520 by default on recent versions (3.1.3 +) unless the PFS priority string
523 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
524 <https://reproducible-builds.org/specs/source-date-epoch/>.
526 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
527 the check for any unsuccessful recipients did not notice the limit, and
528 erroneously found still-pending ones.
530 JH/08 Pipeline CHUNKING command and data together, on kernels that support
531 MSG_MORE. Only in-clear (not on TLS connections).
533 JH/09 Avoid using a temporary file during transport using dkim. Unless a
534 transport-filter is involved we can buffer the headers in memory for
535 creating the signature, and read the spool data file once for the
536 signature and again for transmission.
538 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
539 4.77 as the kernel support then wasn't solid, having issues in 64bit
540 mode. Now, it's been long enough. Add support for FreeBSD also.
542 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
543 case where the routing stage had gathered several addresses to send to
544 a host before calling the transport for the first, we previously failed
545 to close down TLS in the old transport process before passing the TCP
546 connection to the new process. The new one sent a STARTTLS command
547 which naturally failed, giving a failed delivery and bloating the retry
548 database. Investigation and fix prototype from Wolfgang Breyha.
550 JH/12 Fix check on SMTP command input synchronisation. Previously there were
551 false-negatives in the check that the sender had not preempted a response
552 or prompt from Exim (running as a server), due to that code's lack of
553 awareness of the SMTP input buffering.
555 PP/04 Add commandline_checks_require_admin option.
556 Exim drops privileges sanely, various checks such as -be aren't a
557 security problem, as long as you trust local users with access to their
558 own account. When invoked by services which pass untrusted data to
559 Exim, this might be an issue. Set this option in main configuration
560 AND make fixes to the calling application, such as using `--` to stop
563 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
564 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
565 responses to those, into a single TLS record each way (this usually means
566 a single packet). As a side issue, smtp_enforce_sync now works on TLS
569 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
570 affects you only if you're dancing at the edge of the param size limits.
571 If you are, and this message makes sense to you, then: raise the
572 configured limit or use OpenSSL 1.1. Nothing we can do for older
575 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
576 cmdline element to get the documented default one. Previously it was
579 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
580 are unsuable for non-message-specific reasons - eg. connection timeout,
581 banner-time rejection.
583 JH/16 Fix logging of delivery remote port, when specified by router, under
586 PP/06 Repair manualroute's ability to take options in any order, even if one
587 is the name of a transport.
590 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
592 JH/17 Change the list-building routines interface to use the expanding-string
593 triplet model, for better allocation and copying behaviour.
595 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
596 Previously it was constructed the first time a possibly-matching string
597 was met in the configuration file input during startup; now it is done
600 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
601 compatible one, to avoid the (poorly documented) possibility of a config
602 file in the working directory redirecting the DB files, possibly correpting
603 some existing file. CVE-2017-10140 assigned for BDB.
605 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
606 cache-hot. Previously, although the result was properly cached, the
607 initial verify call returned a defer.
609 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
610 the main verify for receipient in uncached-mode.
612 JH/22 Retire historical build files to an "unsupported" subdir. These are
613 defined as "ones for which we have no current evidence of testing".
615 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
616 if present. Previously it was ignored.
618 JH/24 Start using specified-initialisers in C structure init coding. This is
619 a C99 feature (it's 2017, so now considered safe).
621 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
622 if was a fixed-sized field and bitmask ops via macros; it is now more
625 PP/07 GitHub PR 56: Apply MariaDB build fix.
626 Patch provided by Jaroslav Škarvada.
628 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
629 during Coverity cleanups [4.87 JH/47]
630 Diagnosis and fix provided by Michael Fischer v. Mollard.
632 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
633 the right size to place the terminating semicolon on its own folded
634 line, the header hash was calculated to an incorrect value thanks to
635 the (relaxed) space the fold became.
637 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
638 and confused the parent.
640 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
641 which could crash as a result. This could lead to undeliverable messages.
643 JH/28 Logging: "next input sent too soon" now shows where input was truncated
646 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
647 matters on fast-turnover and PID-randomising systems, which were getting
648 out-of-order delivery.
650 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
651 a possibly-overlapping copy. The symptom was that "Remote host closed
652 connection in response to HELO" was logged instead of the actual 4xx
655 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
656 Previously only that bufferd was discarded, resulting in SYMTP command
659 JH/32 DKIM: when a message has multiple signatures matching an identity given
660 in dkim_verify_signers, run the dkim acl once for each. Previously only
661 one run was done. Bug 2189.
663 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
664 "panic the current process" to "deliberately defer". The panic log is
665 still written with the problem list name; the mail and reject logs now
666 get a temp-reject line for the message that was being handled, saying
667 something like "domains check lookup or other defer". The SMTP 451
668 message is still "Temporary local problem".
670 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
671 A crafted sequence of BDAT commands could result in in-use memory beeing
672 freed. CVE-2017-16943.
674 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
675 from SMTP input. Previously it was always done; now only done for DATA
676 and not BDAT commands. CVE-2017-16944.
678 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
679 to the message (such as an overlong header line). Previously this was
680 not done and we did not exit BDAT mode. Followon from the previous item
681 though a different problem.
687 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
688 than -2003 did; needs libidn2 in addition to libidn.
690 JH/02 The path option on a pipe transport is now expanded before use.
692 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
693 Patch provided by "Björn", documentation fix added too.
695 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
696 missing a wire-to-host endian conversion.
698 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
699 close after a BDAT command line could be taken as a following command,
700 giving a synch failure. Fix by only checking for synch immediately
701 before acknowledging the chunk.
703 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
704 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
705 macro. Patches provided by Josh Soref.
707 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
708 Previously we did not; the RFC seems ambiguous and VRFY is not listed
709 by IANA as a service extension. However, John Klensin suggests that we
712 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
713 the dkim code may be unix-mode line endings rather than smtp wire-format
714 CRLF, so prepend a CR to any bare LF.
716 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
717 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
719 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
720 the first were themselves being wrongly included in the feed into dkim
721 processing; with most chunk sizes in use this resulted in an incorrect
722 body hash calculated value.
724 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
725 DKIM signature block, for verification. Although advised against by
726 standards it is specifically not ruled illegal.
728 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
730 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
731 missing a body hash (the bh= tag).
733 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
734 It seems that HAProxy sends the Proxy Protocol information in clear and
735 only then does a TLS startup, so do the same.
737 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
738 TCP connections (such as for Spamd) unless the daemon successfully set
739 Fast Open mode on its listening sockets. This fixes breakage seen on
740 too-old kernels or those not configured for Fast Open, at the cost of
741 requiring both directions being enabled for TFO, and TFO never being used
742 by non-daemon-related Exim processes.
744 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
745 endings, at least on the first header line. Try to canonify any that get
746 past that check, despite the cost.
748 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
749 now limited to an arbitrary five deep, while parsing addresses with the
750 strip_excess_angle_brackets option enabled.
752 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
753 instead leave the unprompted TLS handshake in socket buffer for the
754 TLS library to consume.
756 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
758 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
760 JH/16 Drop variables when they go out of scope. Memory management drops a whole
761 region in one operation, for speed, and this leaves assigned pointers
762 dangling. Add checks run only under the testsuite which checks all
763 variables at a store-reset and panics on a dangling pointer; add code
764 explicitly nulling out all the variables discovered. Fixes one known
765 bug: a transport crash, where a dangling pointer for $sending_ip_address
766 originally assigned in a verify callout, is re-used.
768 PP/06 Drop '.' from @INC in various Perl scripts.
770 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
772 PP/08 Reduce a number of compilation warnings under clang; building with
773 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
774 should be warning-free.
776 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
778 HS/01 Fix portability problems introduced by PP/08 for platforms where
779 realloc(NULL) is not equivalent to malloc() [SunOS et al].
781 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
782 chunk. This allows us to accept broken chunked messages. We need a more
783 general solution here.
785 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
786 already-broken messages in the queue.
788 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
790 JH/19 Fix reference counting bug in routing-generated-address tracking.
796 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
797 supports it and a size is available (ie. the sending peer gave us one).
799 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
800 years of being deprecated). The replacements are the ACLs
801 acl_smtp_mime and acl_not_smtp_mime.
803 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
804 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
805 or even an in-clear connection were permitted. Now, if the host lookup
806 was dnssec and dane was requested then the host is only used if the
807 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
808 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
809 if one fails this test.
810 This means that a poorly-configured remote DNS will make it incommunicado;
811 but it protects against a DNS-interception attack on it.
813 JH/04 Bug 1810: make continued-use of an open smtp transport connection
814 non-noisy when a race steals the message being considered.
816 JH/05 If main configuration option tls_certificate is unset, generate a
817 self-signed certificate for inbound TLS connections.
819 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
820 in rewrites and routers.
822 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
823 and logged a warning sing 4.83; now they are a configuration file error.
825 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
826 (lacking @domain). Apply the same qualification processing as RCPT.
828 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
830 JH/10 Support ${sha256:} applied to a string (as well as the previous
833 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
834 a cutthrough deliver is pending, as we always want to make a connection.
835 This also avoids re-routing the message when later placing the cutthrough
836 connection after a verify cache hit.
837 Do not update it with the verify result either.
839 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
840 when routing results in more than one destination address.
842 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
843 signing (which inhibits the cutthrough capability). Previously only
844 the presence of an option was tested; now an expansion evaluating as
845 empty is permissible (obviously it should depend only on data available
846 when the cutthrough connection is made).
848 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
849 the relevant preceding SMTP command did not note the pipelining mode.
851 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
852 Previously they were not counted.
854 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
855 as one having no matching records. Previously we deferred the message
856 that needed the lookup.
858 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
859 distinguished as "(=".
861 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
862 for missing MX records. Previously it only worked for missing A records.
864 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
866 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
867 after the data-go-ahead and data-ack. Patch from Jason Betts.
869 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
870 even for a "none" policy. Patch from Tony Meyer.
872 JH/22 Fix continued use of a connection for further deliveries. If a port was
873 specified by a router, it must also match for the delivery to be
876 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
877 When one of the recipients of a message was unsuitable for the connection
878 (has no matching addresses), we lost track of needing to mark it
879 deferred. As a result mail would be lost.
881 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
883 JH/25 Decoding ACL controls is now done using a binary search; the source code
884 takes up less space and should be simpler to maintain. Merge the ACL
885 condition decode tables also, with similar effect.
887 JH/26 Fix problem with one_time used on a redirect router which returned the
888 parent address unchanged. A retry would see the parent address marked as
889 delivered, so not attempt the (identical) child. As a result mail would
892 JH/27 Fix a possible security hole, wherein a process operating with the Exim
893 UID can gain a root shell. Credit to http://www.halfdog.net/ for
894 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
897 JH/28 Enable {spool,log} filesystem space and inode checks as default.
898 Main config options check_{log,spool}_{inodes,space} are now
899 100 inodes, 10MB unless set otherwise in the configuration.
901 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
902 Previously it only applied to the main-section connection policy
905 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
907 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
908 by me. Added RFC7919 DH primes as an alternative.
910 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
911 are not in the system include path.
913 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
914 GnuTLS, when a session startup failed (eg because the client disconnected)
915 Exim did stdio operations after fclose. This was exposed by a recent
916 change which nulled out the file handle after the fclose.
918 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
919 signed directly by the cert-signing cert, rather than an intermediate
920 OCSP-signing cert. This is the model used by LetsEncrypt.
922 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
924 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
925 an incoming connection.
927 HS/02 Bug 1802: Do not half-close the connection after sending a request
930 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
931 fallback to "prime256v1".
933 JH/34 SECURITY: Use proper copy of DATA command in error message.
934 Could leak key material. Remotely exploitable. CVE-2016-9963.
940 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
941 and 3.4.4 - once the server is enabled to respond to an OCSP request
942 it does even when not requested, resulting in a stapling non-aware
943 client dropping the TLS connection.
945 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
946 support variable-length bit vectors. No functional change.
948 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
949 The I= interface field on outgoing lines is now after the H= remote
950 host field, same as incoming lines. There is a separate
951 outgoing_interface log selector which allows you to disable the
954 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
955 If not running log_selector +smtp_connection the mainlog would be held
956 open indefinitely after a "too many connections" event, including to a
957 deleted file after a log rotate. Leave the per net connection logging
958 leaving it open for efficiency as that will be quickly detected by the
959 check on the next write.
961 HS/01 Bug 1671: Fix post transport crash.
962 Processing the wait-<transport> messages could crash the delivery
963 process if the message IDs didn't exist for some reason. When
964 using 'split_spool_directory=yes' the construction of the spool
965 file name failed already, exposing the same netto behaviour.
967 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
968 mime_regex ACL conditions.
970 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
971 to DSN fail messages (bounces): remote IP, remote greeting, remote response
972 to HELO, local diagnostic string.
974 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
975 log line to debug. Even when configured with a tls authenticator many
976 client connections are expected to not authenticate in this way, so
977 an authenticate fail is not an error.
979 HS/02 Add the Exim version string to the process info. This way exiwhat
980 gives some more detail about the running daemon.
982 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
983 matter for fast-change records such as DNSBLs.
985 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
986 retry record, even if constant. There may be multiple transports with
987 different interface settings and the retry behaviour needs to be kept
990 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
992 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
994 JH/10 Bug 840: fix log_defer_output option of pipe transport
996 JH/11 Bug 830: use same host for all RCPTS of a message, even under
997 hosts_randomize. This matters a lot when combined with mua_wrapper.
999 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1000 ${quote_pgsql:<string>} operator.
1002 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1004 JH/14 Change header file name for freeradius-client. Relevant if compiling
1005 with Radius support; from the Gentoo tree and checked under Fedora.
1007 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1009 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1010 option transport_filter, meaning no filtering.
1012 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1014 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1015 defaults to "*" (all hosts). The variable is now available when not built
1016 with TLS, default unset, mainly to enable keeping the testsuite sane.
1017 If a server certificate is not supplied (via tls_certificate) an error is
1018 logged, and clients will find TLS connections fail on startup. Presumably
1019 they will retry in-clear.
1020 Packagers of Exim are strongly encouraged to create a server certificate
1021 at installation time.
1023 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1024 with the $config_file variable.
1026 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1027 in transport context, after the attempt, and per-recipient. The latter type
1028 is per host attempted. The event data is the error message, and the errno
1029 information encodes the lookup type (A vs. MX) used for the (first) host,
1030 and the trailing two digits of the smtp 4xx response.
1032 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1033 to write to mainlog (or rejectlog, paniclog) in the window between file
1034 creation and permissions/ownership being changed. Particularly affects
1035 installations where exicyclog is run as root, rather than exim user;
1036 result is that the running daemon panics and dies.
1038 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1040 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1041 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1042 "pri" and "weight". Note that the previous implicit priority given by the
1043 list order is no longer honoured.
1045 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1046 for DKIM processing.
1048 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1049 by defining SUPPORT_SOCKS.
1051 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1052 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1053 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1054 variables are renamed to proxy_{local,external}_{address,port}.
1056 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1057 for a build by defining SUPPORT_I18N
1059 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1060 of the query string, and make ${quote_redis:} do that quoting.
1062 JH/29 Move Events support from Experimental to mainline, enabled by default
1063 and removable for a build by defining DISABLE_EVENT.
1065 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1067 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1068 cached by the daemon.
1070 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1071 by defining LOOKUP_REDIS. The libhiredis library is required.
1073 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1074 keys are given for lookup.
1076 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1077 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1078 only supported when built with TLS support. The PolarSSL SHA routines
1079 are still used when the TLS library is too old for convenient support.
1081 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1082 openssl_options), for security. OpenSSL forces this from version 1.1.0
1083 server-side so match that on older versions.
1085 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1086 allocation for $value could be released as the expansion processing
1087 concluded, but leaving the global pointer active for it.
1089 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1090 and to use the domains and local_parts ACL conditions.
1092 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1093 incorrectly not doubled on cutthrough transmission, hence seen as a
1094 body-termination at the receiving system - resulting in truncated mails.
1095 Commonly the sender saw a TCP-level error, and retransmitted the message
1096 via the normal store-and-forward channel. This could result in duplicates
1097 received - but deduplicating mailstores were liable to retain only the
1098 initial truncated version.
1100 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1102 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1104 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1105 we're in there, support oversigning also; bug 1309.
1107 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1109 HS/04 Add support for keep_environment and add_environment options.
1111 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1112 either intentional arithmetic overflow during PRNG, or testing config-
1115 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1116 delivery resulted in actual delivery. Cancel cutthrough before DATA
1119 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1120 rejects a recipient: pass the reject to the originator.
1122 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1123 Many were false-positives and ignorable, but it's worth fixing the
1126 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1127 for the new environment-manipulation done at startup. Move the routines
1128 from being local to tls.c to being global via the os.c file.
1130 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1131 an extract embedded as result-arg for a map, the first arg for extract
1132 is unavailable so we cannot tell if this is a numbered or keyed
1133 extraction. Accept either.
1139 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1142 JH/02 The smtp transport option "multi_domain" is now expanded.
1144 JH/03 The smtp transport now requests PRDR by default, if the server offers
1147 JH/04 Certificate name checking on server certificates, when exim is a client,
1148 is now done by default. The transport option tls_verify_cert_hostnames
1149 can be used to disable this per-host. The build option
1150 EXPERIMENTAL_CERTNAMES is withdrawn.
1152 JH/05 The value of the tls_verify_certificates smtp transport and main options
1153 default to the word "system" to access the system default CA bundle.
1154 For GnuTLS, only version 3.0.20 or later.
1156 JH/06 Verification of the server certificate for a TLS connection is now tried
1157 (but not required) by default. The verification status is now logged by
1158 default, for both outbound TLS and client-certificate supplying inbound
1161 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1164 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1165 Status Notification (bounce) messages are now MIME format per RFC 3464.
1166 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1167 under the control of the dsn_advertise_hosts option, and routers may
1168 have a dsn_lasthop option.
1170 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1171 default, modifiable by a malware= option. The list separator for
1172 the options can now be changed in the usual way. Bug 68.
1174 JH/10 The smtp_receive_timeout main option is now expanded before use.
1176 JH/11 The incoming_interface log option now also enables logging of the
1177 local interface on delivery outgoing connections.
1179 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1180 if the interface and destination host and port all match.
1182 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1185 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1186 Patch from Andrew Lewis.
1188 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1189 now supports optional time-restrictions, weighting, and priority
1190 modifiers per server. Patch originally by <rommer@active.by>.
1192 JH/16 The spamd_address main option now supports a mixed list of local
1193 and remote servers. Remote servers can be IPv6 addresses, and
1194 specify a port-range.
1196 JH/17 Bug 68: The spamd_address main option now supports an optional
1197 timeout value per server.
1199 JH/18 Bug 1581: Router and transport options headers_add/remove can
1200 now have the list separator specified.
1202 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1205 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1208 JH/21 Support for the A6 type of dns record is withdrawn.
1210 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1211 rather than the verbs used.
1213 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1214 from 255 to 1024 chars.
1216 JH/24 Verification callouts now attempt to use TLS by default.
1218 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1219 are generic router options now. The defaults didn't change.
1221 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1222 Original patch from Alexander Shikoff, worked over by JH.
1224 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1225 files if ZCAT_COMMAND is not executable.
1227 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1229 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1231 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1232 Normally benign, it bites when the pair was led to by a CNAME;
1233 modern usage is to not canonicalize the domain to a CNAME target
1234 (and we were inconsistent anyway for A-only vs AAAA+A).
1236 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1238 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1239 when evaluating $sender_host_dnssec.
1241 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1242 $sender_helo_dnssec variable.
1244 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1246 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1248 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1250 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1251 documented as working, but never had. Support all but $spam_report.
1253 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1254 added for tls authenticator.
1256 HS/03 Add perl_taintmode main config option
1262 TL/01 When running the test suite, the README says that variables such as
1263 no_msglog_check are global and can be placed anywhere in a specific
1264 test's script, however it was observed that placement needed to be near
1265 the beginning for it to behave that way. Changed the runtest perl
1266 script to read through the entire script once to detect and set these
1267 variables, reset to the beginning of the script, and then run through
1268 the script parsing/test process like normal.
1270 TL/02 The BSD's have an arc4random API. One of the functions to induce
1271 adding randomness was arc4random_stir(), but it has been removed in
1272 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1273 function when detected.
1275 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1276 cause callback expansion.
1278 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1279 syntax errors in an expansion can be treated as a string instead of
1280 logging or causing an error, due to the internal use of bool_lax
1281 instead of bool when processing it.
1283 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1284 server certificates when making smtp deliveries.
1286 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1288 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1290 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1292 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1293 Merged patch from Sebastian Wiedenroth.
1295 JH/05 Fix results-pipe from transport process. Several recipients, combined
1296 with certificate use, exposed issues where response data items split
1297 over buffer boundaries were not parsed properly. This eventually
1298 resulted in duplicates being sent. This issue only became common enough
1299 to notice due to the introduction of connection certificate information,
1300 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1302 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1303 size buffer was used, resulting in syntax errors when an expansion
1306 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1307 version 3.3.6 or later.
1309 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1310 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1311 both become "event_action", the variables become $event_name, $event_data
1312 and $event_defer_errno. There is a new variable $verify_mode, usable in
1313 routers, transports and related events. The tls:cert event is now also
1314 raised for inbound connections, if the main configuration event_action
1317 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1318 early OCSP support, but no stapling (appears to be less than 1.0.0).
1320 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1321 server certificate names available under the smtp transport option
1322 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1325 JH/10 Time-related extraction expansions from certificates now use the main
1326 option "timezone" setting for output formatting, and are consistent
1327 between OpenSSL and GnuTLS compilations. Bug 1541.
1329 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1330 encoded parameter in the incoming message. Bug 1558.
1332 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1333 include certificate info, eximon was claiming there were spoolfile
1336 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1338 JH/14 Log delivery-related information more consistently, using the sequence
1339 "H=<name> [<ip>]" wherever possible.
1341 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1342 are problematic for Debian distribution, omit them from the release
1345 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1347 JH/16 Fix string representation of time values on 64bit time_t architectures.
1350 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1351 output list separator was used.
1356 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1357 checkers that were complaining about end of non-void function with no
1360 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1361 This was a regression introduced in 4.83 by another bugfix.
1363 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1365 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1366 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1372 TF/01 Correctly close the server side of TLS when forking for delivery.
1374 When a message was received over SMTP with TLS, Exim failed to clear up
1375 the incoming connection properly after forking off the child process to
1376 deliver the message. In some situations the subsequent outgoing
1377 delivery connection happened to have the same fd number as the incoming
1378 connection previously had. Exim would try to use TLS and fail, logging
1379 a "Bad file descriptor" error.
1381 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1382 utilities have not been installed.
1384 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1385 temporary space as the ACL may create new global variables.
1387 TL/01 LDAP support uses per connection or global context settings, depending
1388 upon the detected version of the libraries at build time.
1390 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1391 to extract and use the src ip:port in logging and expansions as if it
1392 were a direct connection from the outside internet. PPv2 support was
1393 updated based on HAProxy spec change in May 2014.
1395 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1397 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1398 Properly escape header and check for NULL return.
1400 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1403 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1405 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1406 characters in header names, implemented as a verify condition.
1407 Contributed by Michael Fischer v. Mollard.
1409 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1410 results permerror and temperror. Previous values are deprecated but
1411 still accepted. In a future release, err_perm and err_temp will be
1412 completely removed, which will be a backward incompatibility if the
1413 ACL tests for either of these two old results. Patch contributed by
1414 user bes-internal on the mailing list.
1416 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1418 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1419 selectors, in both main and reject logs.
1421 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1424 JH/07 Add malware type "sock" for talking to simple daemon.
1426 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1428 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1429 routers/transports under cutthrough routing.
1431 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1432 numbers. Touch up "bool" conditional to keep the same definition.
1434 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1436 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1438 JH/12 Expand items in router/transport headers_add or headers_remove lists
1439 individually rather than the list as a whole. Bug 1452.
1441 Required for reasonable handling of multiple headers_ options when
1442 they may be empty; requires that headers_remove items with embedded
1443 colons must have them doubled (or the list-separator changed).
1445 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1446 view the policy declared in the DMARC record. Currently, $dmarc_status
1447 is a combined value of both the record presence and the result of the
1450 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1452 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1453 dnslookup router and the smtp transport (applying to the forward
1456 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1457 of ldap servers used for a specific lookup. Patch provided by Heiko
1460 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1461 New variable $lookup_dnssec_authenticated for observability.
1463 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1464 Patch submitted by Lars Timman.
1466 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1468 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1469 Requires trusted mode and valid format message id, aborts otherwise.
1470 Patch contributed by Heiko Schlichting.
1472 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1473 certextract with support for various fields. Bug 1358.
1475 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1476 is requested by default, modifiable by smtp transport option
1479 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1480 operate on certificate variables to give certificate fingerprints
1481 Also new ${sha256:cert_variable}.
1483 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1485 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1486 Christian Aistleitner.
1488 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1490 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1491 file. Patch from Wolfgang Breyha.
1493 JH/25 Expand the coverage of the delivery $host and $host_address to
1494 client authenticators run in verify callout. Bug 1476.
1496 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1497 align with daemon_smtp_ports. Bug 72.
1499 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1500 support and error reporting did not work properly.
1502 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1503 and is readable. Patch from Andrew Colin Kissa.
1505 TL/14 Enhance documentation of ${run expansion and how it parses the
1506 commandline after expansion, particularly in the case when an
1507 unquoted variable expansion results in an empty value.
1509 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1511 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1512 by renaming away from it. Addresses GH issue 15, reported by
1515 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1517 TL/15 SECURITY: prevent double expansion in math comparison functions
1518 (can expand unsanitized data). Not remotely exploitable.
1525 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1527 PP/02 Make -n do something, by making it not do something.
1528 When combined with -bP, the name of an option is not output.
1530 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1533 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1534 $sender_host_name and config options to manage this, and basic check
1537 PP/05 DSCP support for outbound connections and control modifier for inbound.
1539 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1540 (Only plugin which currently uses this is kerberos4, which nobody should
1541 be using, but we should make it available and other future plugins might
1542 conceivably use it, even though it would break NAT; stuff *should* be
1543 using channel bindings instead).
1545 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1546 name; added for Sendmail compatibility; requires admin caller.
1547 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1548 just ignore it); requires trusted caller.
1549 Also parse but ignore: -Ac -Am -X<logfile>
1552 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1554 TL/02 Add +smtp_confirmation as a default logging option.
1556 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1557 Patch by Magnus Holmgren from 2007-02-20.
1559 TL/04 Bugzilla 1281 - Spec typo.
1560 Bugzilla 1283 - Spec typo.
1561 Bugzilla 1290 - Spec grammar fixes.
1563 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1565 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1567 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1568 mailing list by by Dmitry Isaikin.
1570 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1572 JH/02 Support "G" suffix to numbers in ${if comparisons.
1574 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1576 NM/01 Bugzilla 1197 - Spec typo
1577 Bugzilla 1196 - Spec examples corrections
1579 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1581 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1582 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1585 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1586 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1588 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1589 "acl {{name}{arg}...}", and optional args on acl condition
1592 JH/05 Permit multiple router/transport headers_add/remove lines.
1594 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1596 JH/07 Avoid using a waiting database for a single-message-only transport.
1597 Performance patch from Paul Fisher. Bugzilla 1262.
1599 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1602 JH/09 Add $headers_added variable, with content from use of ACL modifier
1603 add_header (but not yet added to the message). Bugzilla 199.
1605 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1606 Pulled from Bugzilla 817 by Wolfgang Breyha.
1608 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1610 (nb: this is the same fix as in Exim 4.80.1)
1612 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1615 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1618 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1620 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1621 advertise SMTP AUTH mechanism to us, instead of a generic
1622 protocol violation error. Also, make Exim more robust to bad
1623 data from the Dovecot auth socket.
1625 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1627 When a queue runner is handling a message, Exim first routes the
1628 recipient addresses, during which it prunes them based on the retry
1629 hints database. After that it attempts to deliver the message to
1630 any remaining recipients. It then updates the hints database using
1633 So if a recipient address works intermittently, it can get repeatedly
1634 deferred at routing time. The retry hints record remains fresh so the
1635 address never reaches the final cutoff time.
1637 This is a fairly common occurrence when a user is bumping up against
1638 their storage quota. Exim had some logic in its local delivery code
1639 to deal with this. However it did not apply to per-recipient defers
1640 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1642 This change adds a proper retry rule check during routing so that the
1643 final cutoff time is checked against the message's age. We only do
1644 this check if there is an address retry record and there is not a
1645 domain retry record; this implies that previous attempts to handle
1646 the address had the retry_use_local_parts option turned on. We use
1647 this as an approximation for the destination being like a local
1648 delivery, as in LMTP.
1650 I suspect this new check makes the old local delivery cutoff check
1651 redundant, but I have not verified this so I left the code in place.
1653 TF/02 Correct gecos expansion when From: is a prefix of the username.
1655 Test 0254 submits a message to Exim with the header
1659 When I ran the test suite under the user fanf2, Exim expanded
1660 the header to contain my full name, whereas it should have added
1661 a Resent-Sender: header. It erroneously treats any prefix of the
1662 username as equal to the username.
1664 This change corrects that bug.
1666 GF/01 DCC debug and logging tidyup
1667 Error conditions log to paniclog rather than rejectlog.
1668 Debug lines prefixed by "DCC: " to remove any ambiguity.
1670 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1672 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1673 Bug spotted by Jeremy Harris; was flawed since initial commit.
1674 Would have resulted in OCSP responses post-SNI triggering an Exim
1675 NULL dereference and crash.
1677 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1679 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1680 Bug detection, analysis and fix by Samuel Thibault.
1681 Bugzilla 1331, Debian bug #698092.
1683 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1685 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1686 Server implementation by Todd Lyons, client by JH.
1687 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1688 config variable "prdr_enable" controls whether the server
1689 advertises the facility. If the client requests PRDR a new
1690 acl_data_smtp_prdr ACL is called once for each recipient, after
1691 the body content is received and before the acl_smtp_data ACL.
1692 The client is controlled by both of: a hosts_try_prdr option
1693 on the smtp transport, and the server advertisement.
1694 Default client logging of deliveries and rejections involving
1695 PRDR are flagged with the string "PRDR".
1697 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1698 fclose(). Diagnosis by Todd Lyons.
1700 PP/17 Update configure.default to handle IPv6 localhost better.
1701 Patch by Alain Williams (plus minor tweaks).
1704 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1705 This is now consistent with GnuTLS, and is now documented: the
1706 previous undocumented portable approach to treating the option as
1707 unset was to force an expansion failure. That still works, and
1708 an empty string is now equivalent.
1710 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1711 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1712 not performing validation itself.
1714 PP/20 Added force_command boolean option to pipe transport.
1715 Patch from Nick Koston, of cPanel Inc.
1717 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1720 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1722 PP/21 Fix eximon continuous updating with timestamped log-files.
1723 Broken in a format-string cleanup in 4.80, missed when I repaired the
1724 other false fix of the same issue.
1725 Report and fix from Heiko Schlichting.
1728 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1729 Report from Prashanth Katuri.
1731 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1732 It's SecureTransport, so affects any MacOS clients which use the
1733 system-integrated TLS libraries, including email clients.
1735 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1736 using a MIME ACL for non-SMTP local injection.
1737 Report and assistance in diagnosis by Warren Baker.
1739 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1741 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1743 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1744 last id that failed so it may be referenced in subsequent ACL's.
1746 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1749 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1750 ldap library initialization, allowing self-signed CA's to be
1751 used. Also properly sets require_cert option later in code by
1752 using NULL (global ldap config) instead of ldap handle (per
1753 session). Bug diagnosis and testing by alxgomz.
1755 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1756 the src/util/ subdirectory.
1758 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1759 renamed to Transport Post Delivery Action by Jeremy Harris, as
1762 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1763 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1764 redis_servers = needs to be configured which will be used by the redis
1765 lookup. Patch from Warren Baker, of The Packet Hub.
1767 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1769 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1770 hostname or reverse DNS when processing a host list. Used suggestions
1771 from multiple comments on this bug.
1773 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1775 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1776 Missed a few lines, added it to make the runtest require no keyboard
1779 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1780 contains upper case chars. Make router use caseful_local_part.
1782 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1783 support when GnuTLS has been built with p11-kit.
1789 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1791 This, or similar/improved, will also be change PP/11 of 4.82.
1797 PP/01 Handle short writes when writing local log-files.
1798 In practice, only affects FreeBSD (8 onwards).
1799 Bugzilla 1053, with thanks to Dmitry Isaikin.
1801 NM/01 Bugzilla 949 - Documentation tweak
1803 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1806 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1808 PP/02 Implemented gsasl authenticator.
1810 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1812 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1813 `pkg-config foo` for cflags/libs.
1815 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1816 with rest of GSASL and with heimdal_gssapi.
1818 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1819 `pkg-config foo` for cflags/libs for the TLS implementation.
1821 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1822 properties get this fed in as external SSF. A number of robustness
1823 and debugging improvements to the cyrus_sasl authenticator.
1825 PP/08 cyrus_sasl server now expands the server_realm option.
1827 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1828 Patch by Jeremy Harris.
1830 PP/10 Added dbmjz lookup type.
1832 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1834 PP/12 MAIL args handles TAB as well as SP, for better interop with
1835 non-compliant senders.
1836 Analysis and variant patch by Todd Lyons.
1838 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1839 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1840 Patch from Dirk Mueller <dmueller@suse.com>
1842 PP/13 tls_peerdn now print-escaped for spool files.
1843 Observed some $tls_peerdn in wild which contained \n, which resulted
1844 in spool file corruption.
1846 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1847 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1848 or write after TLS renegotiation, which otherwise led to messages
1851 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1852 as a tracking header (ie: a signed header comes before the signature).
1853 Patch from Wolfgang Breyha.
1855 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1856 comma-sep list; embedded commas doubled.
1858 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1860 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1862 Report and patch from Dmitry Banschikov.
1864 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1865 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1866 failures. We appear to now support TLS1.1+ with Exim.
1868 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1869 lets Exim select keys and certificates based upon TLS SNI from client.
1870 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1871 before an outbound SMTP session. New log_selector, +tls_sni.
1873 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1874 NULL dereference. Report and patch from Alun Jones.
1876 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1877 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1878 Not seeing resolver debug output on NetBSD, but suspect this is a
1879 resolver implementation change.
1881 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1882 Left warnings. Added "eximon gdb" invocation mode.
1884 PP/21 Defaulting "accept_8bitmime" to true, not false.
1886 PP/22 Added -bw for inetd wait mode support.
1888 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1889 locate the relevant includes and libraries. Made this the default.
1891 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1892 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1894 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1895 This may cause build issues on older platforms.
1897 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1898 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1899 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1900 Added SNI support via GnuTLS too.
1901 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1903 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1905 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1906 Applied second patch from Janne, implementing suggestion to default
1907 multiple-strings-in-record handling to match SPF spec.
1909 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1911 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1912 read-only, out of scope).
1913 Patch from Wolfgang Breyha, report from Stuart Northfield.
1915 PP/29 Fix three issues highlighted by clang analyser static analysis.
1916 Only crash-plausible issue would require the Cambridge-specific
1917 iplookup router and a misconfiguration.
1918 Report from Marcin Mirosław.
1920 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1922 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1923 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1924 As part of this, removing so much warning spew let me fix some minor
1925 real issues in debug logging.
1927 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1928 assignment on my part. Fixed.
1930 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1931 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1932 Janne Snabb (who went above and beyond: thank you).
1934 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1935 string otherwise requires a connection and a bunch more work and it's
1936 relatively easy to get wrong. Should also expose TLS library linkage
1939 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1940 64-bit ${eval} (JH/03).
1942 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1943 GNU libc to support some of the 64-bit stuff, should not lead to
1944 conflicts. Defined before os.h is pulled in, so if a given platform
1945 needs to override this, it can.
1947 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1948 protection layer was required, which is not implemented.
1949 Bugzilla 1254, patch from Wolfgang Breyha.
1951 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1952 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1953 tls_dhparam take prime identifiers. Also unbreak combination of
1954 OpenSSL+DH_params+TLSSNI.
1956 PP/39 Disable SSLv2 by default in OpenSSL support.
1962 PP/01 Solaris build fix for Oracle's LDAP libraries.
1963 Bugzilla 1109, patch from Stephen Usher.
1965 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1967 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1970 TF/02 Fix a couple more cases where we did not log the error message
1971 when unlink() failed. See also change 4.74-TF/03.
1973 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1974 lock up or crash if it happened to be inside a call to libc when it
1975 got a SIGUSR1 from exiwhat.
1977 The SIGUSR1 handler appends the current process status to the process
1978 log which is later printed by exiwhat. It used to use the general
1979 purpose logging code to do this, but several functions it calls are
1980 not safe for signals.
1982 The new output code in the SIGUSR1 handler is specific to the process
1983 log, and simple enough that it's easy to inspect for signal safety.
1984 Removing some special cases also simplifies the general logging code.
1985 Removing the spurious timestamps from the process log simplifies
1988 TF/04 Improved ratelimit ACL condition.
1990 The /noupdate option has been deprecated in favour of /readonly which
1991 has clearer semantics. The /leaky, /strict, and /readonly update modes
1992 are mutually exclusive. The update mode is no longer included in the
1993 database key; it just determines when the database is updated. (This
1994 means that when you upgrade Exim will forget old rate measurements.)
1996 Exim now checks that the per_* options are used with an update mode that
1997 makes sense for the current ACL. For example, when Exim is processing a
1998 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1999 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2000 must specify per_mail/readonly. If you omit the update mode it defaults to
2001 /leaky where that makes sense (as before) or /readonly where required.
2003 The /noupdate option is now undocumented but still supported for
2004 backwards compatibility. It is equivalent to /readonly except that in
2005 ACLs where /readonly is required you may specify /leaky/noupdate or
2006 /strict/noupdate which are treated the same as /readonly.
2008 A useful new feature is the /count= option. This is a generalization
2009 of the per_byte option, so that you can measure the throughput of other
2010 aggregate values. For example, the per_byte option is now equivalent
2011 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2013 The per_rcpt option has been generalized using the /count= mechanism
2014 (though it's more complicated than the per_byte equivalence). When it is
2015 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2016 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2017 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2018 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2019 Note that using per_rcpt with a non-readonly update mode in more than
2020 one ACL will cause the recipients to be double-counted. (The per_mail
2021 and per_byte options don't have this problem.)
2023 The handling of very low rates has changed slightly. If the computed rate
2024 is less than the event's count (usually one) then this event is the first
2025 after a long gap. In this case the rate is set to the same as this event's
2026 count, so that the first message of a spam run is counted properly.
2028 The major new feature is a mechanism for counting the rate of unique
2029 events. The new per_addr option counts the number of different
2030 recipients that someone has sent messages to in the last time period. It
2031 behaves like per_rcpt if all the recipient addresses are different, but
2032 duplicate recipient addresses do not increase the measured rate. Like
2033 the /count= option this is a general mechanism, so the per_addr option
2034 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2035 example, measure the rate that a client uses different sender addresses
2036 with the options per_mail/unique=$sender_address. There are further
2037 details in the main documentation.
2039 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2041 TF/06 Removed a few PCRE remnants.
2043 TF/07 Automatically extract Exim's version number from tags in the git
2044 repository when doing development or release builds.
2046 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2047 Bugzilla 879. Patch from Paul Fisher.
2049 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2050 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2053 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2055 PP/05 Variable $av_failed, true if the AV scanner deferred.
2056 Bugzilla 1078. Patch from John Horne.
2058 PP/06 Stop make process more reliably on build failure.
2059 Bugzilla 1087. Patch from Heiko Schlittermann.
2061 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2062 Bugzilla 1089. Patch from Heiko Schlittermann.
2064 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2065 Bugzilla 1131. Patch from Holger Weiß.
2067 PP/09 Handle IPv6 addresses with SPF.
2068 Bugzilla 860. Patch from Wolfgang Breyha.
2070 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2072 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2075 PP/11 match_* no longer expand right-hand-side by default.
2076 New compile-time build option, EXPAND_LISTMATCH_RHS.
2077 New expansion conditions, "inlist", "inlisti".
2079 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2081 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2083 PP/14 fix log_write() format string regression from TF/03.
2084 Bugzilla 1152. Patch from Dmitry Isaikin.
2090 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2092 PP/02 Harmonised TLS library version reporting; only show if debugging.
2093 Layout now matches that introduced for other libraries in 4.74 PP/03.
2095 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2097 PP/04 New "dns_use_edns0" global option.
2099 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2102 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2103 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2105 TK/01 Updated PolarSSL code to 0.14.2.
2106 Bugzilla 1097. Patch from Andreas Metzler.
2108 PP/07 Catch divide-by-zero in ${eval:...}.
2109 Fixes bugzilla 1102.
2111 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2114 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2115 format-string attack -- SECURITY: remote arbitrary code execution.
2117 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2118 time unintentionally subject to list matching rules, letting the header
2119 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2120 arbitrary string expansion). This allowed for information disclosure.
2122 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2123 INT_MIN/-1 -- value coerced to INT_MAX.
2129 NM/01 Workaround for PCRE version dependency in version reporting
2132 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2133 This fixes portability to compilers other than gcc, notably
2134 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2136 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2137 makefiles for portability to HP-UX and POSIX correctness.
2139 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2140 Also via indented variable definition in the Makefile.
2141 (Debugging by Oliver Heesakkers).
2143 PP/02 Restore caching of spamd results with expanded spamd_address.
2144 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2146 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2147 Improves build reliability. Fix from: Frank Elsner
2149 NM/02 Fix wide character breakage in the rfc2047 coding
2150 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2152 NM/03 Allow underscore in dnslist lookups
2153 Fixes bug 1026. Patch from Graeme Fowler
2155 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2156 Code patches from Adam Ciarcinski of NetBSD.
2158 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2161 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2162 is logged, to avoid truncation. Patch from John Horne.
2164 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2165 Patch from Jakob Hirsch.
2167 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2168 SQL string expansion failure details.
2169 Patch from Andrey Oktyabrski.
2171 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2172 Patch from Simon Arlott.
2174 PP/09 New lookups functionality failed to compile on old gcc which rejects
2175 extern declarations in function scope.
2176 Patch from Oliver Fleischmann
2178 PP/10 Use sig_atomic_t for flags set from signal handlers.
2179 Check getgroups() return and improve debugging.
2180 Fixed developed for diagnosis in bug 927 (which turned out to be
2183 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2184 Patch from Mark Zealey.
2186 PP/12 Bugzilla 1056: Improved spamd server selection.
2187 Patch from Mark Zealey.
2189 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2190 Based on patch from Heiko Schlittermann.
2192 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2193 Patch from Uwe Doering, sign-off by Michael Haardt.
2195 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2196 variable declaration deep within a block. Bug and patch from
2199 PP/15 lookups-Makefile IRIX compatibility coercion.
2201 PP/16 Make DISABLE_DKIM build knob functional.
2203 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2204 Patch by Simon Arlott
2206 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2207 variable argument macros. Our copy now differs from upstream.
2213 TF/01 Failure to get a lock on a hints database can have serious
2214 consequences so log it to the panic log.
2216 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2217 controlled using the smtp_confirmation log selector.
2219 TF/03 Include the error message when we fail to unlink a spool file.
2221 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2222 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2223 for maintaining out-of-tree patches for some time.
2225 PP/01 Bugzilla 139: Documentation and portability issues.
2226 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2227 Handle per-OS dynamic-module compilation flags.
2229 PP/02 Let /dev/null have normal permissions.
2230 The 4.73 fixes were a little too stringent and complained about the
2231 permissions on /dev/null. Exempt it from some checks.
2232 Reported by Andreas M. Kirchwitz.
2234 PP/03 Report version information for many libraries, including
2235 Exim version information for dynamically loaded libraries. Created
2236 version.h, now support a version extension string for distributors
2237 who patch heavily. Dynamic module ABI change.
2239 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2240 privilege escalation vulnerability whereby the Exim run-time user
2241 can cause root to append content of the attacker's choosing to
2244 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2247 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2248 If dropping privileges for untrusted macros, we disabled normal logging
2249 on the basis that it would fail; for the Exim run-time user, this is not
2250 the case, and it resulted in successful deliveries going unlogged.
2251 Fixed. Reported by Andreas Metzler.
2257 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2258 only prepend for the Resent-* case. Fixes regression introduced in
2259 Exim 4.70 by NM/22 for Bugzilla 607.
2261 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2262 increasing numbers of administrators be bitten by this.
2264 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2266 PP/03 Bugzilla 994: added openssl_options main configuration option.
2268 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2270 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2272 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2274 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2275 without a peer certificate, leading to a segfault because of an
2276 assumption that peers always have certificates. Be a little more
2277 paranoid. Problem reported by Martin Tscholak.
2279 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2280 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2281 NB: ClamAV planning to remove STREAM in "middle of 2010".
2282 CL also introduces -bmalware, various -d+acl logging additions and
2283 more caution in buffer sizes.
2285 PP/09 Implemented reverse_ip expansion operator.
2287 PP/10 Bugzilla 937: provide a "debug" ACL control.
2289 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2291 PP/12 Bugzilla 973: Implement --version.
2293 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2295 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2297 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2299 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2300 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2301 ignore trailing whitespace.
2303 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2305 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2308 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2309 Notification from Dr Andrew Aitchison.
2311 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2312 ExtendedDetectionInfo response format.
2313 Notification from John Horne.
2315 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2318 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2319 XSL and documented dependency on system catalogs, with examples of how
2322 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2325 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2326 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2327 configuration file which is writeable by the Exim user or group.
2329 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2330 of configuration files to cover files specified with the -C option if
2331 they are going to be used with root privileges, not just the default
2334 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2335 option (effectively making it always true).
2337 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2338 files to be used while preserving root privileges.
2340 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2341 that rogue child processes cannot use them.
2343 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2344 run-time user, instead of root.
2346 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2347 Exim run-time user without dropping privileges.
2349 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2350 result string, instead of calling string_vformat() twice with the same
2353 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2354 for other users. Others should always drop root privileges if they use
2355 -C on the command line, even for a whitelisted configure file.
2357 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2359 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2365 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2366 $data_path, and $header_path variables; fixed documentation bugs and
2369 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2370 exipick to access non-standard spools, including the "frozen" queue
2373 NM/01 Bugzilla 965: Support mysql stored procedures.
2374 Patch from Alain Williams
2376 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2378 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2379 Patch from Andreas Metzler
2381 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2382 Patch from Kirill Miazine
2384 NM/05 Bugzilla 671: Added umask to procmail example.
2386 JJ/03 installed exipick 20100323.0, fixing doc bug
2388 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2389 directory. Notification and patch from Dan Rosenberg.
2391 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2393 TK/02 Improve log output when DKIM signing operation fails.
2395 MH/01 Treat the transport option dkim_domain as a colon separated
2396 list, not as a single string, and sign the message with each element,
2397 omitting multiple occurences of the same signer.
2399 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2400 Bugzilla 985, 986. Patch by Simon Arlott
2402 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2403 Patch by Simon Arlott
2405 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2406 MBX locking. Notification from Dan Rosenberg.
2412 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2414 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2416 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2418 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2420 NM/04 Bugzilla 915: Flag broken perl installation during build.
2426 TK/01 Added patch by Johannes Berg that expands the main option
2427 "spamd_address" if it starts with a dollar sign.
2429 TK/02 Write list of recipients to X-Envelope-Sender header when building
2430 the mbox-format spool file for content scanning (suggested by Jakob
2433 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2434 (http://www.dcc-servers.net/) support via dccifd. Activated by
2435 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2437 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2438 by Mark Daniel Reidel <mr@df.eu>.
2440 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2441 When building exim an external PCRE library is now needed -
2442 PCRE is a system library on the majority of modern systems.
2443 See entry on PCRE_LIBS in EDITME file.
2445 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2446 conversation. Added nologin parameter to request.
2447 Patch contributed by Kirill Miazine.
2449 TF/01 Do not log submission mode rewrites if they do not change the address.
2451 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2453 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2454 log files in place. Contributed by Roberto Lima.
2456 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2458 TF/03 Bugzilla 615: When checking the local_parts router precondition
2459 after a local_part_suffix or local_part_prefix option, Exim now
2460 does not use the address's named list lookup cache, since this
2461 contains cached lookups for the whole local part.
2463 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2464 Robert Millan. Documentation is in experimental-spec.txt.
2466 TF/04 Bugzilla 668: Fix parallel build (make -j).
2468 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2470 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2471 Patch provided by Jan Srzednicki.
2473 TF/05 Leading white space used to be stripped from $spam_report which
2474 wrecked the formatting. Now it is preserved.
2476 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2477 that they are available at delivery time.
2479 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2481 TF/08 TLS error reporting now respects the incoming_interface and
2482 incoming_port log selectors.
2484 TF/09 Produce a more useful error message if an SMTP transport's hosts
2485 setting expands to an empty string.
2487 NM/06 Bugzilla 744: EXPN did not work under TLS.
2488 Patch provided by Phil Pennock.
2490 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2491 Patch provided by Richard Godbee.
2493 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2494 acl_smtp_notquit, added index entry.
2496 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2497 Patch provided by Eugene Bujak.
2499 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2500 accept(). Patch provided by Maxim Dounin.
2502 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2503 Patch provided by Phil Pennock.
2505 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2507 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2508 Patch provided by Brad "anomie" Jorsch.
2510 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2511 Patch provided by Dean Brooks.
2513 TK/05 Add native DKIM support (does not depend on external libraries).
2515 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2516 Patch provided by Graeme Fowler.
2518 NM/16 Bugzilla 851: Documentation example syntax fix.
2520 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2522 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2525 NM/19 Bugzilla 745: TLS version reporting.
2526 Patch provided by Phil Pennock.
2528 NM/20 Bugzilla 167: bool: condition support.
2529 Patch provided by Phil Pennock.
2531 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2532 clients. Patch provided by Phil Pennock.
2534 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2535 Patch provided by Brad "anomie" Jorsch.
2537 NM/23 Bugzilla 687: Fix misparses in eximstats.
2538 Patch provided by Heiko Schlittermann.
2540 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2541 Patch provided by Heiko Schlittermann.
2543 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2544 plus update to original patch.
2546 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2548 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2549 Patch provided by David Brownlee.
2551 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2553 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2555 NM/30 Bugzilla 888: TLS documentation bugfixes.
2557 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2559 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2560 Unlike the original bugzilla I have changed all shell scripts in src tree.
2562 NM/33 Bugzilla 898: Transport filter timeout fix.
2563 Patch by Todd Rinaldo.
2565 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2566 Patch by Serge Demonchaux.
2568 NM/35 Bugzilla 39: Base64 decode bug fixes.
2569 Patch by Jakob Hirsch.
2571 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2573 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2575 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2577 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2583 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2584 ALT-N's libdkim that I have put here:
2585 http://duncanthrax.net/exim-experimental/
2587 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2588 were called 'true' and it seems that C99 defines that as a reserved
2589 keyword to be used with 'bool' variable types. That means you could
2590 not include C99-style headers which use bools without triggering
2591 build errors in sieve.c.
2593 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2594 as mailq or other aliases. Changed the --help handling significantly
2595 to do whats expected. exim_usage() emits usage/help information.
2597 SC/01 Added the -bylocaldomain option to eximstats.
2599 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2601 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2603 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2609 PH/01 Another patch from the Sieve maintainer.
2611 PH/02 When an IPv6 address is converted to a string for single-key lookup
2612 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2613 dots are used instead of colons so that keys in lsearch files need not
2614 contain colons. This was done some time before quoting was made available
2615 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2616 (notated using the quote facility) so as to distinguish them from IPv4
2617 keys. This meant that lookups for IP addresses in host lists did not work
2618 for iplsearch lookups.
2620 This has been fixed by arranging for IPv6 addresses to be expressed with
2621 colons if the lookup type is iplsearch. This is not incompatible, because
2622 previously such lookups could never work.
2624 The situation is now rather anomalous, since one *can* have colons in
2625 ordinary lsearch keys. However, making the change in all cases is
2626 incompatible and would probably break a number of configurations.
2628 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2631 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2632 conversion specification without a maximum field width, thereby enabling
2633 a rogue spamd server to cause a buffer overflow. While nobody in their
2634 right mind would setup Exim to query an untrusted spamd server, an
2635 attacker that gains access to a server running spamd could potentially
2636 exploit this vulnerability to run arbitrary code as the Exim user.
2638 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2639 $primary_hostname instead of what libspf2 thinks the hosts name is.
2641 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2642 a directory entry by the name of the lookup key. Previously, if a
2643 symlink pointed to a non-existing file or a file in a directory that
2644 Exim lacked permissions to read, a lookup for a key matching that
2645 symlink would fail. Now it is enough that a matching directory entry
2646 exists, symlink or not. (Bugzilla 503.)
2648 PH/03 The body_linecount and body_zerocount variables are now exported in the
2651 PH/04 Added the $dnslist_matched variable.
2653 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2654 This means they are set thereafter only if the connection becomes
2657 PH/06 Added the client_condition to authenticators so that some can be skipped
2658 by clients under certain conditions.
2660 PH/07 The error message for a badly-placed control=no_multiline_responses left
2661 "_responses" off the end of the name.
2663 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2665 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2666 (without spaces) instead of just copying the configuration text.
2668 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2670 PH/11 Added $max_received_linelength.
2672 PH/12 Added +ignore_defer and +include_defer to host lists.
2674 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2675 way in which PCRE > 7.0 is built.
2677 PH/14 Implemented queue_only_load_latch.
2679 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2680 MAIL command. The effect was to mangle the value on 64-bit systems.
2682 PH/16 Another patch from the Sieve maintainer.
2684 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2686 PH/18 If a system quota error occurred while trying to create the file for
2687 a maildir delivery, the message "Mailbox is full" was not appended to the
2688 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2689 only to a quota excession during the actual writing of the file.
2691 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2692 characters?) which causes problems in log lines. The DN values are now
2693 passed through string_printing() before being added to log lines.
2695 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2696 and InterBase are left for another time.)
2698 PH/21 Added message_body_newlines option.
2700 PH/22 Guard against possible overflow in moan_check_errorcopy().
2702 PH/23 POSIX allows open() to be a macro; guard against that.
2704 PH/24 If the recipient of an error message contained an @ in the local part
2705 (suitably quoted, of course), incorrect values were put in $domain and
2706 $local_part during the evaluation of errors_copy.
2712 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2713 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2716 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2717 issue a MAIL command.
2719 PH/02 In an ACL statement such as
2721 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2723 if a client was not listed at all, or was listed with a value other than
2724 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2725 the condition was not true (as it should be), so access was not denied.
2726 The bug was that the ! inversion was incorrectly passed on to the second
2727 item. This has been fixed.
2729 PH/03 Added additional dnslists conditions == and =& which are different from
2730 = and & when the dns lookup returns more than one IP address.
2732 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2733 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2735 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2736 FSYNC, which compiles an option called disable_fsync that allows for
2737 bypassing fsync(). The documentation is heavily laced with warnings.
2739 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2741 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2742 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2743 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2744 including adding "make clean"; (3) Added -fPIC when compiling the test
2745 dynamically loaded module, to get rid of a warning.
2747 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2748 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2749 = 0s. The bug is otherwise harmless.
2751 PH/07 There was a bug in the dovecot authenticator such that the value of
2752 $auth1 could be overwritten, and so not correctly preserved, after a
2753 successful authentication. This usually meant that the value preserved by
2754 the server_setid option was incorrect.
2756 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2758 PH/09 Installed PCRE release 7.0.
2760 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2761 run for batched SMTP input. It is now run at the start of every message
2762 in the batch. While fixing this I discovered that the process information
2763 (output by running exiwhat) was not always getting set for -bs and -bS
2764 input. This is fixed, and it now also says "batched" for BSMTP.
2766 PH/11 Added control=no_pipelining.
2768 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2769 patch, slightly modified), and move the expansion of helo_data till after
2770 the connection is made in the smtp transport (so it can use these
2773 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2775 PH/14 Added log_selector = +pid.
2777 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2779 PH/16 Add ${if forany and ${if forall.
2781 PH/17 Added dsn_from option to vary the From: line in DSNs.
2783 PH/18 Flush SMTP output before performing a callout, unless control =
2784 no_callout_flush is set.
2786 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2787 was true (the default) a successful delivery failed to delete the retry
2788 item, thus causing premature timeout of the address. The bug is now
2791 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2793 PH/21 Long custom messages for fakedefer and fakereject are now split up
2794 into multiline responses in the same way that messages for "deny" and
2795 other ACL rejections are.
2797 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2798 with slight modification.
2800 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2801 draft, changing the syntax and factoring some duplicate code".
2803 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2804 for deliveries of the second and subsequent messages over the same SMTP
2807 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2808 ${reduce, with only minor "tidies".
2810 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2812 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2813 expansion side effects.
2815 PH/27 When a message times out after an over-quota error from an Exim-imposed
2816 quota, the bounce message says "mailbox is full". This message was not
2817 being given when it was a system quota that was exceeded. It now should
2820 MH/03 Made $recipients available in local_scan(). local_scan() already has
2821 better access to the recipient list through recipients_list[], but
2822 $recipients can be useful in postmaster-provided expansion strings.
2824 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2825 in the case of a MAIL command with additional options following the
2826 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2827 were accidentally chopped off.
2829 PH/29 SMTP synchronization checks are implemented when a command is read -
2830 there is a check that no more input is waiting when there shouldn't be
2831 any. However, for some commands, a delay in an ACL can mean that it is
2832 some time before the response is written. In this time, more input might
2833 arrive, invalidly. So now there are extra checks after an ACL has run for
2834 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2835 pipelining has not been advertised.
2837 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2839 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2840 returned in the SMTP response when smtp_return_error_details was false.
2841 This has been fixed.
2843 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2844 instead of the C I/O that was originally supplied, because problems were
2845 reported on Solaris.
2847 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2848 Exim which did not show up earlier: it was assuming that a call to
2849 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2850 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2851 was a macro that became an assignment, so it seemed to work. This has
2852 changed to a proper function call with a void return, hence the compile
2853 error. Exim's code has been fixed.
2855 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2858 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2860 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2862 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2863 because it turns out to be too big or there is a timeout) there is no
2864 "Completed" line in the log. When some messages of this type were
2865 selected by exigrep, they were listed as "not completed". Others were
2866 picked up by some special patterns. I have improved the selection
2867 criteria to be more general.
2869 PH/38 The host_find_failed option in the manualroute router can now be set
2870 to "ignore", to completely ignore a host whose IP address cannot be
2871 found. If all hosts are ignored, the behaviour is controlled by the new
2872 host_all_ignored option.
2874 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2875 homing or because of multiple MX records with /mx) generated more than
2876 one IP address, and the following item turned out to be the local host,
2877 all the secondary addresses of the first item were incorrectly removed
2878 from the list, along with the local host and any following hosts (which
2879 is what is supposed to happen).
2881 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2882 whoever called Exim into the -H file. In the case of the daemon it was
2883 behaving confusingly. When first started, it used values for whoever
2884 started the daemon, but after a SIGHUP it used the Exim user (because it
2885 calls itself on a restart). I have changed the code so that it now always
2888 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2889 message are rejected with the same error (e.g. no authentication or bad
2890 sender address), and a DATA command is nevertheless sent (as can happen
2891 with PIPELINING or a stupid MUA), the error message that was given to the
2892 RCPT commands is included in the rejection of the DATA command. This is
2893 intended to be helpful for MUAs that show only the final error to their
2896 PH/42 Another patch from the Sieve maintainer.
2898 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2899 Eximstats - Fixed some broken HTML links and added missing column headers
2901 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2902 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2904 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2906 PH/43 Yet another patch from the Sieve maintainer.
2908 PH/44 I found a way to check for a TCP/IP connection going away before sending
2909 the response to the final '.' that terminates a message, but only in the
2910 case where the client has not sent further data following the '.'
2911 (unfortunately, this is allowed). However, in many cases there won't be
2912 any further data because there won't be any more messages to send. A call
2913 to select() can be used: if it shows that the input is "ready", there is
2914 either input waiting, or the socket has been closed. An attempt to read
2915 the next input character can distinguish the two cases. Previously, Exim
2916 would have sent an OK response which the client would never have see.
2917 This could lead to message repetition. This fix should cure that, at
2918 least in a lot of common cases.
2920 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2921 advertised in response to EHLO.
2927 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2928 fixed by 4.65/MH/01 (is this a record?) are fixed:
2930 (i) An empty string was always treated as zero by the numeric comparison
2931 operators. This behaviour has been restored.
2933 (ii) It is documented that the numeric comparison operators always treat
2934 their arguments as decimal numbers. This was broken in that numbers
2935 starting with 0 were being interpreted as octal.
2937 While fixing these problems I realized that there was another issue that
2938 hadn't been noticed. Values of message_size_limit (both the global option
2939 and the transport option) were treated as octal if they started with 0.
2940 The documentation was vague. These values are now always treated as
2941 decimal, and I will make that clear in the documentation.
2947 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2948 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2951 MH/01 Don't check that the operands of numeric comparison operators are
2952 integers when their expansion is in "skipping" mode (fixes bug
2953 introduced by 4.64-PH/07).
2955 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2956 child addresses, Exim now panics and dies. Previously, because the count
2957 is held in a short int, deliveries were likely to be lost. As such a
2958 large number of recipients for a single message is ridiculous
2959 (performance will be very, very poor), I have chosen to impose a limit
2960 rather than extend the field.
2966 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2967 leftover -K file (the existence of which was triggered by #402).
2968 While we were at it, introduced process PID as part of the -K
2969 filename. This should rule out race conditions when creating
2972 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2973 processing considerably. Previous code took too long for large mails,
2974 triggering a timeout which in turn triggers #401.
2976 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2977 in the DK code in transports.c. sendfile() is not really portable,
2978 hence the _LINUX specificness.
2980 TF/01 In the add_headers option to the mail command in an Exim filter,
2981 there was a bug that Exim would claim a syntax error in any
2982 header after the first one which had an odd number of characters
2985 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2986 callout verification, Exim cached a "reject" for the entire domain. This
2987 is correct for most verifications, but it is not correct for a recipient
2988 verification with use_sender or use_postmaster set, because in that case
2989 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2990 case of MAIL FROM:<> rejection from other early rejections (e.g.
2991 rejection of HELO). When verifying a recipient using a non-null MAIL
2992 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2993 Whatever the result of the callout, the value of the domain cache is
2994 left unchanged (for any other kind of callout, getting as far as trying
2995 RCPT means that the domain itself is ok).
2997 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3000 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3001 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3002 session, the daemon ignores ECONNECT errors and logs others; it now
3003 ignores EPIPE as well.
3005 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3006 (quoted-printable decoding).
3008 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3009 later the small subsequent patch to fix an introduced bug.
3011 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3013 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3015 PH/08 An error is now given if message_size_limit is specified negative.
3017 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3018 to be given (somewhat) arbitrary names.
3020 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3023 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3024 miscellaneous code fixes
3026 PH/10 Added the log_reject_target ACL modifier to specify where to log
3029 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3030 hostname. This is wrong, because it relates to the incoming message (and
3031 probably the interface on which it is arriving) and not to the outgoing
3032 callout (which could be using a different interface). This has been
3033 changed to use the value of the helo_data option from the smtp transport
3034 instead - this is what is used when a message is actually being sent. If
3035 there is no remote transport (possible with a router that sets up host
3036 addresses), $smtp_active_hostname is used.
3038 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3039 tweaks were necessary in order to get it to work (see also 21 below):
3040 (a) The code assumed that strncpy() returns a negative number on buffer
3041 overflow, which isn't the case. Replaced with Exim's string_format()
3043 (b) There were several signed/unsigned issues. I just did the minimum
3044 hacking in of casts. There is scope for a larger refactoring.
3045 (c) The code used strcasecmp() which is not a standard C function.
3046 Replaced with Exim's strcmpic() function.
3047 (d) The code set only $1; it now sets $auth1 as well.
3048 (e) A simple test gave the error "authentication client didn't specify
3049 service in request". It would seem that Dovecot has changed its
3050 interface. Fortunately there's a specification; I followed it and
3051 changed what the client sends and it appears to be working now.
3053 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3056 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3057 address A is aliased to B and C, where B exists and C does not. Without
3058 -v the output is "A verified" because verification stops after a
3059 successful redirection if more than one address is generated. However,
3060 with -v the child addresses are also verified. Exim was outputting "A
3061 failed to verify" and then showing the successful verification for C,
3062 with its parentage. It now outputs "B failed to verify", showing B's
3063 parentage before showing the successful verification of C.
3065 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3066 look up a TXT record in a specific list after matching in a combined
3069 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3070 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3071 they consult the DNS. I had assumed they would set it the way they
3072 wanted; and indeed my experiments on Linux seem to show that in some
3073 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3074 To be on the safe side, however, I have now made the interface to
3075 host_find_byname() similar to host_find_bydns(), with an argument
3076 containing the DNS resolver options. The host_find_byname() function now
3077 sets these options at its start, just as host_find_bydns() does. The smtp
3078 transport options dns_qualify_single and dns_search_parents are passed to
3079 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3080 of host_find_byname() use the default settings of RES_DEFNAMES
3081 (qualify_single) but not RES_DNSRCH (search_parents).
3083 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3084 spool_read_header() do less string testing, by means of a preliminary
3085 switch on the second character of optional "-foo" lines. (This is
3086 overdue, caused by the large number of possibilities that now exist.
3087 Originally there were few.) While I was there, I also converted the
3088 str(n)cmp tests so they don't re-test the leading "-" and the first
3089 character, in the hope this might squeeze out yet more improvement.
3091 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3092 flag allowing group syntax was set by the header_syntax check but not
3093 turned off, possible causing trouble later; (2) The flag was not being
3094 set at all for the header_verify test, causing "group"-style headers to
3095 be rejected. I have now set it in this case, and also caused header_
3096 verify to ignore an empty address taken from a group. While doing this, I
3097 came across some other cases where the code for allowing group syntax
3098 while scanning a header line wasn't quite right (mostly, not resetting
3099 the flag correctly in the right place). These bugs could have caused
3100 trouble for malformed header lines. I hope it is now all correct.
3102 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3103 with the "reply" argument non-NULL. The code, however (which originally
3104 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3105 but it didn't always do it. This confused somebody who was copying the
3106 code for some other use. I have removed all the tests.
3108 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3109 feature that was used to support insecure browsers during the U.S. crypto
3110 embargo. It requires special client support, and Exim is probably the
3111 only MTA that supported it -- and would never use it because real RSA is
3112 always available. This code has been removed, because it had the bad
3113 effect of slowing Exim down by computing (never used) parameters for the
3114 RSA_EXPORT functionality.
3116 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3117 authenticator to fail if there's a tab character in the incoming data
3118 (there should never be unless someone is messing about, as it's supposed
3119 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3120 if the connection is using TLS or if the remote IP is the same as the
3121 local IP, and the "valid-client-cert option" if a client certificate has
3124 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3125 authenticators. This can be used for authorization after authentication
3126 succeeds. (In the case of plaintext, it servers for both authentication
3129 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3130 if any retry times were supplied.
3132 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3133 connection, where there is no client IP address to check. In this
3134 situation, the verify now always succeeds.
3136 PH/25 Applied John Jetmore's -Mset patch.
3138 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3140 PH/27 In a string expansion for a processed (not raw) header when multiple
3141 headers of the same name were present, leading whitespace was being
3142 removed from all of them, but trailing whitespace was being removed only
3143 from the last one. Now trailing whitespace is removed from each header
3144 before concatenation. Completely empty headers in a concatenation (as
3145 before) are ignored.
3147 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3148 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3150 PH/29 [Removed. This was a change that I later backed out, and forgot to
3151 correct the ChangeLog entry (that I had efficiently created) before
3152 committing the later change.]
3154 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3155 address errors (4xx response to RCPT) over the same connection as other
3156 messages routed to the same hosts. Such deliveries are always "forced",
3157 so retry times are not inspected. This resulted in far too many retries
3158 for the affected addresses. The effect occurred only when there were more
3159 hosts than the hosts_max_try setting in the smtp transport when it had
3160 the 4xx errors. Those hosts that it had tried were not added to the list
3161 of hosts for which the message was waiting, so if all were tried, there
3162 was no problem. Two fixes have been applied:
3164 (i) If there are any address or message errors in an SMTP delivery, none
3165 of the hosts (tried or untried) are now added to the list of hosts
3166 for which the message is waiting, so the message should not be a
3167 candidate for sending over the same connection that was used for a
3168 successful delivery of some other message. This seems entirely
3169 reasonable: after all the message is NOT "waiting for some host".
3170 This is so "obvious" that I'm not sure why it wasn't done
3171 previously. Hope I haven't missed anything, but it can't do any
3172 harm, as the worst effect is to miss an optimization.
3174 (ii) If, despite (i), such a delivery is accidentally attempted, the
3175 routing retry time is respected, so at least it doesn't keep
3176 hammering the server.
3178 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3179 in ${readsocket because some servers need this prod.
3181 PH/32 Added some extra debug output when updating a wait-xxx database.
3183 PH/33 The hint "could be header name not terminated by colon", which has been
3184 given for certain expansion errors for a long time, was not being given
3185 for the ${if def:h_colon_omitted{... case.
3187 PH/34 The spec says: "With one important exception, whenever a domain list is
3188 being scanned, $domain contains the subject domain." There was at least
3189 one case where this was not true.
3191 PH/35 The error "getsockname() failed: connection reset by peer" was being
3192 written to the panic log as well as the main log, but it isn't really
3193 panic-worthy as it just means the connection died rather early on. I have
3194 removed the panic log writing for the ECONNRESET error when getsockname()
3197 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3198 runs only) independently of the message's sender address. This meant
3199 that, if the 4xx error was in fact related to the sender, a different
3200 message to the same recipient with a different sender could confuse
3201 things. In particular, this can happen when sending to a greylisting
3202 server, but other circumstances could also provoke similar problems.
3203 I have changed the default so that the retry time for these errors is now
3204 based a combination of the sender and recipient addresses. This change
3205 can be overridden by setting address_retry_include_sender=false in the
3208 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3209 remote server are returned as part of bounce messages. This was not
3210 happening for LMTP over a pipe (the lmtp transport), but now it is the
3211 same for both kinds of LMTP.
3213 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3214 in header lines that were in fact CNAMEs. This is no longer the case.
3216 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3217 and queue runs started by the daemon processed all messages. This has
3218 been fixed so that -R and -S can now usefully be given with -q<time>.
3220 PH/40 Import PCRE release 6.7 (fixes some bugs).
3222 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3224 PH/42 Give an error if -q is specified more than once.
3226 PH/43 Renamed the variables $interface_address and $interface_port as
3227 $received_ip_address and $received_port, to make it clear that these
3228 values apply to message reception, and not to the outgoing interface when
3229 a message is delivered. (The old names remain recognized, of course.)
3231 PH/44 There was no timeout on the connect() call when using a Unix domain
3232 socket in the ${readsocket expansion. There now is.
3234 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3235 be meaningful with "accept".
3237 SC/01 Eximstats V1.43
3238 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3240 SC/02 Eximstats V1.44
3241 Use a glob alias rather than an array ref in the generated
3242 parser. This improves both readability and performance.
3244 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3245 Collect SpamAssassin and rejection statistics.
3246 Don't display local sender or destination tables unless
3247 there is data to show.
3248 Added average volumes into the top table text output.
3250 SC/04 Eximstats V1.46
3251 Collect data on the number of addresses (recipients)
3252 as well as the number of messages.
3254 SC/05 Eximstats V1.47
3255 Added 'Message too big' to the list of mail rejection
3256 reasons (thanks to Marco Gaiarin).
3258 SC/06 Eximstats V1.48
3259 Mainlog lines which have GMT offsets and are too short to
3260 have a flag are now skipped.
3262 SC/07 Eximstats V1.49 (Alain Williams)
3263 Added the -emptyok flag.
3265 SC/08 Eximstats V1.50
3266 Fixes for obtaining the IP address from reject messages.
3268 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3269 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3270 whitespace changes from 4.64-PH/27
3272 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3275 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3276 are found, allow negative numbers in numeric criteria)
3278 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3280 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3281 to match changes made in 4.64-PH/43
3283 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3285 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3286 "domain" is actually an IP address, and force a failure. This locks out
3287 those revolvers/nameservers that support "A-for-A" lookups, in
3288 contravention of the specifications.
3290 PH/48 When a host name was looked up from an IP address, and the subsequent
3291 forward lookup of the name timed out, the host name was left in
3292 $sender_host_name, contrary to the specification.
3294 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3295 restricted to single-key lookups, Exim was not diagnosing an error if
3296 * or *@ was used with a query-style lookup.
3298 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3300 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3301 long ago, but noone interested enough thought of it. Let's just say that
3302 the "1.1" means that there are some new functions that weren't there at
3303 some point in the past.
3305 PH/51 Error processing for expansion failure of helo_data from an smtp
3306 transport during callout processing was broken.
3308 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3309 tested/used via the -bh/-bhc/-bs options.
3311 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3312 bug, fixed in subsequent PCRE releases).
3314 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3315 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3317 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3323 SC/01 Use a glob alias rather than an array ref in eximstats generated
3324 parser. This improves both readability and performance.
3326 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3327 Don't display local sender or destination tables in eximstats unless
3328 there is data to show.
3329 Added average volumes into the eximstats top table text output.
3331 SC/03 Collect data on the number of addresses (recipients) as well
3332 as the number of messages in eximstats.
3334 TF/01 Correct an error in the documentation for the redirect router. Exim
3335 does (usually) call initgroups() when daemonizing.
3337 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3338 with consistent privilege compared to when running as a daemon.
3340 TF/03 Note in the spec that $authenticated_id is not set for local
3341 submissions from trusted users.
3343 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3344 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3346 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3347 by adding some example configuration directives to the default
3348 configuration file. A little bit of work is required to uncomment the
3349 directives and define how usernames and passwords are checked, but
3350 there is now a framework to start from.
3352 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3353 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3354 without this. I don't know how relevant this is to other LDAP libraries.
3356 PH/02 Add the verb name to the "unknown ACL verb" error.
3358 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3360 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3362 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3363 directory not expanded when it should be if an expanded home directory
3364 was set for the address (which is overridden by the transport).
3366 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3369 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3370 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3371 because it is too late at that time, and has no effect.
3373 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3374 security issue with \' (bugzilla #107). I could not use the
3375 PQescapeStringConn() function, because it needs a PGconn value as one of
3378 PH/08 When testing addresses using -bt, indicate those final addresses that
3379 are duplicates that would not cause an additional delivery. At least one
3380 person was confused, thinking that -bt output corresponded to deliveries.
3381 (Suppressing duplicates isn't a good idea as you lose the information
3382 about possibly different redirections that led to the duplicates.)
3384 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3385 systems where poll() doesn't work, in particular OS X.
3387 PH/10 Added more information to debugging output for retry time not reached.
3389 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3390 operations in malware.c.
3392 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3395 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3396 syslog_duplication set false, log lines that would normally be written
3397 both the the main log and to the reject log were not written to syslog at
3400 PH/14 In the default configuration, change the use of "message" in ACL warn
3401 statements to "add_header".
3403 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3404 not followed by a command (e.g. "seen endif").
3406 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3407 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3410 PH/17 Added extra conditions to the default value of delay_warning_condition
3414 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3415 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3416 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3419 The Auto-Submitted: and various List- headers are standardised, whereas I
3420 don't think Precedence: ever was.
3422 PH/18 Refactored debugging code in route_finduser() to show more information,
3423 in particular, the error code if getpwnam() issues one.
3425 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3426 This is apparently needed in addition to the PH/07 change above to avoid
3427 any possible encoding problems.
3429 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3430 but not after initializing Perl.
3432 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3433 output them only if debugging. By default they are written stderr,
3434 apparently, which is not desirable.
3436 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3439 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3442 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3444 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3445 authenticated or an ident call has been made. Suppress the default
3446 values for $authenticated_id and $authenticated_sender (but permit -oMai
3447 and -oMas) when testing with -bh.
3449 PH/24 Re-jigged the order of the tests in the default configuration so that the
3450 tests for valid domains and recipients precede the DNS black list and CSA
3451 tests, on the grounds that those ones are more expensive.
3453 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3454 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3455 This bug exists in every version of Exim that I still have, right back to
3458 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3459 However, an attempt to turn on case-sensitivity in a regex key by
3460 including (?-i) didn't work because the subject string was already
3461 lowercased, and the effects were non-intuitive. It turns out that a
3462 one-line patch can be used to allow (?-i) to work as expected.
3468 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3469 other effects) broke the use of negated acl sub-conditions.
3471 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3474 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3475 "Deny" causes Exim to reject the incoming connection with a 554 error.
3476 Unfortunately, if there is a major crisis, such as a disk failure,
3477 tcp-wrappers gives "deny", whereas what one would like would be some
3478 kind of temporary error. A kludge has been added to help with this.
3479 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3480 554 error is used if errno is still zero or contains ENOENT (which occurs
3481 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3484 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3486 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3487 errors. Otherwise a message that provokes a temporary error (when other
3488 messages do not) can cause a whole host to time out.
3490 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3491 addresses were routed directly to files or pipes from a redirect router.
3492 File deliveries just didn't batch; pipe deliveries might have suffered
3495 PH/06 A failure to get a lock for a hints database would erroneously always say
3496 "Failed to get write lock", even when it was really a read lock.
3498 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3499 of 0600. This has been changed to use the value of the lockfile_mode
3500 option (which defaults to 0600).
3502 PH/08 Applied small patch from the Sieve maintainer.
3504 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3505 folder from quota calculations, a direct delivery into this folder messed
3506 up the contents of the maildirsize file. This was because the regex was
3507 used only to exclude .Trash (or whatever) when the size of the mailbox
3508 was calculated. There was no check that a delivery was happening into an
3509 excluded directory. This bug has been fixed by ignoring all quota
3510 processing for deliveries into excluded directories.
3512 PH/10 Added the maildirfolder_create_regex option to appendfile.
3518 PH/01 The code for finding all the local interface addresses on a FreeBSD
3519 system running IPv6 was broken. This may well have applied to all BSD
3520 systems, as well as to others that have similar system calls. The broken
3521 code found IPv4 interfaces correctly, but gave incorrect values for the
3522 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3523 that it would not match correctly against @[] and not recognize the IPv6
3526 PH/02 The ipliteral router was not recognizing addresses of the form user@
3527 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3529 PH/03 Added disable_ipv6.
3531 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3532 decoded form, because it is most often used to construct To: headers
3533 lines in autoreplies, and the decoded form may well be syntactically
3534 invalid. However, $reply_address has leading white space removed, and all
3535 newlines turned into spaces so that the autoreply transport does not
3538 PH/05 If group was specified without a user on a router, and no group or user
3539 was specified on a transport, the group from the router was ignored.
3541 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3542 for visible compile-time settings that can be used to change these
3543 numbers, for those that want even more. Backwards compatibility with old
3544 spool files has been maintained. However, going back to a previous Exim
3545 release will lost any variables that are in spool files.
3547 PH/07 Two small changes when running in the test harness: increase delay when
3548 passing a TCP/IP connection to a new process, in case the original
3549 process has to generate a bounce, and remove special handling of
3550 127.0.0.2 (sic), which is no longer necessary.
3552 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3553 be the same on different OS.
3555 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3558 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3559 whether --show-vars was specified or not
3561 JJ/02 exipick: Added support for new ACL variable spool format introduced
3564 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3565 syntactically invalid From: or Reply-to: line, and a filter used this to
3566 generate an autoreply, and therefore failed to obtain an address for the
3567 autoreply, Exim could try to deliver to a non-existent relative file
3568 name, causing unrelated and misleading errors. What now happens is that
3569 it logs this as a hard delivery error, but does not attempt to create a
3572 PH/11 The exinext utility has a -C option for testing purposes, but although
3573 the given file was scanned by exinext itself; it wasn't being passed on
3574 when Exim was called.
3576 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3577 an end-of-file indication when reading a command response.
3579 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3580 compiled. In many other places in Exim, IPv6 addresses are always
3581 recognized, so I have changed this. It also means that IPv4 domain
3582 literals of the form [IPV4:n.n.n.n] are now always recognized.
3584 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3585 used if the router is not running as root, for example, when verifying at
3586 ACL time, or when using -bh. The debugging output from this situation was
3587 non-existent - all you got was a failure to exec. I have made two
3590 (a) Failures to set uid/gid, the current directory, or a process leader
3591 in a subprocess such as that created by queryprogram now generate
3592 suitable debugging output when -d is set.
3594 (b) The queryprogram router detects when it is not running as root,
3595 outputs suitable debugging information if -d is set, and then runs
3596 the subprocess without attempting to change uid/gid.
3598 PH/15 Minor change to Makefile for building test_host (undocumented testing
3601 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3602 additional section of a DNS packet that returns MX or SRV records.
3603 Instead, it always explicitly searches for A/AAAA records. This avoids
3604 major problems that occur when a DNS server includes only records of one
3605 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3606 fixed another bug: if SRV records were looked up and the corresponding
3607 address records were *not* found in the additional section, the port
3608 values from the SRV records were lost.
3610 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3611 using the correct key (the original address) when searching the retry
3612 rules in order to find which one to use for generating the retry hint.
3614 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3615 adding the default one. Similarly, if it contains a Reply-To: header, the
3616 errors_reply_to option, if set, is not used.
3618 PH/19 When calculating a retry time, Exim used to measure the "time since
3619 failure" by looking at the "first failed" field in the retry record. Now
3620 it does not use this if it is later than than the arrival time of the
3621 message. Instead it uses the arrival time. This makes for better
3622 behaviour in cases where some deliveries succeed, thus re-setting the
3623 "first failed" field. An example is a quota failure for a huge message
3624 when small messages continue to be delivered. Without this change, the
3625 "time since failure" will always be short, possible causing more frequent
3626 delivery attempts for the huge message than are intended.
3627 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3629 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3630 $1, $2, $3) because the numerical variables can be reset during some
3631 expansion items (e.g. "match"), thereby losing the authentication data.
3633 PH/21 Make -bV show the size of off_t variables so that the test suite can
3634 decide whether to run tests for quotas > 2G.
3636 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3637 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3638 filecount value is greater than 2G or if a quota value is greater than 2G
3639 on a system where the size of off_t is not greater than 4, a panic error
3642 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3643 never match. The debug and -bh output now contains an explicit error
3644 message indicating a malformed IPv4 address or mask.
3646 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3647 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3648 PH/23 above applies.
3650 PH/25 Do not write to syslog when running in the test harness. The only
3651 occasion when this arises is a failure to open the main or panic logs
3652 (for which there is an explicit test).
3654 PH/26 Added the /no_tell option to "control=freeze".
3656 PH/27 If a host name lookup failed very early in a connection, for example, if
3657 the IP address matched host_lookup and the reverse lookup yielded a name
3658 that did not have a forward lookup, an error message of the form "no IP
3659 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3660 could be logged. Now it outputs the IP address instead of "NULL".
3662 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3663 allows the sender and the authenticated sender to be set when
3664 submitting a message from within Exim. Since child_open_exim() is
3665 documented for local_scan(), the new function should be too.
3667 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3668 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3669 results in an empty string is now treated as unset.
3671 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3673 PH/31 Added sender_verify_fail logging option.
3675 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3676 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3677 tidied the source and removed it altogether.
3679 PH/33 When a queue run was abandoned because the load average was too high, a
3680 log line was always written; now it is written only if the queue_run log
3681 selector is set. In addition, the log line for abandonment now contains
3682 information about the queue run such as the pid. This is always present
3683 in "start" and "stop" lines but was omitted from the "abandon" line.
3685 PH/34 Omit spaces between a header name and the colon in the error message that
3686 is given when verify = headers_syntax fails (if there are lots of them,
3687 the message gets confusing).
3689 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3690 names, as there are now some PTR records that contain slashes. This check
3691 is only to protect against broken name servers that fall over on strange
3692 characters, so the fact that it applies to all lookups doesn't matter.
3694 PH/36 Now that the new test suite is complete, we can remove some of the
3695 special code in Exim that was needed for the old test suite. For example,
3696 sorting DNS records because real resolvers return them in an arbitrary
3697 order. The new test suite's fake resolver always returns records in the
3700 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3701 bounces) except when queue_only is set, to avoid logging races between
3702 the different processes.
3704 PH/38 Panic-die if .include specifies a non-absolute path.
3706 PH/39 A tweak to the "H" retry rule from its user.
3708 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3709 a label. They prevented compilation on older perls.
3711 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3712 a warning to be raised on newish perls.
3714 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3715 on queue. Changes to match documented behaviour of showing count of
3716 messages matching specified criteria.
3718 PH/40 Changed the default ident timeout from 30s to 5s.
3720 PH/41 Added support for the use of login_cap features, on those BSD systems
3721 that have them, for controlling the resources used by pipe deliveries.
3723 PH/42 The content-scanning code uses fopen() to create files in which to put
3724 message data. Previously it was not paying any attention to the mode of
3725 the files. Exim runs with umask(0) because the rest of the code creates
3726 files with open(), and sets the required mode explicitly. Thus, these
3727 files were ending up world-writeable. This was not a big issue, because,
3728 being within the spool directory, they were not world-accessible. I have
3729 created a function called modefopen, which takes an additional mode
3730 argument. It sets umask(777), creates the file, chmods it to the required
3731 mode, then resets the umask. All the relevant calls to fopen() in the
3732 content scanning code have been changed to use this function.
3734 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3735 to 24 hours. This avoids potential overflow problems when processing G
3736 and H retry rules. I suspect nobody ever tinkers with this value.
3738 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3740 PH/45 When the plaintext authenticator is running as a client, the server's
3741 challenges are checked to ensure they are valid base64 strings. By
3742 default, the authentication attempt is cancelled if an invalid string is
3743 received. Setting client_ignore_invalid_base64 true ignores these errors.
3744 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3745 they are received. Thus, the responses can be made to depend on the
3746 challenges. If an invalid string is ignored, an empty string is placed in
3749 PH/46 Messages that are created by the autoreply transport now contains a
3750 References: header, in accordance with RFCs 2822 and 3834.
3752 PH/47 Added authenticated_sender_force to the smtp transport.
3754 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3756 PH/49 Installed latest patch from the Sieve maintainer.
3758 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3759 was also set, the appendfile transport was unnecessarily scanning a
3760 directory of message files (e.g. for maildir delivery) to find the count
3761 of files (along with the size), even though it did not need this
3762 information. It now does the scan only if it needs to find either the
3763 size of the count of files.
3765 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3767 PH/52 Two bugs concerned with error handling when the smtp transport is
3770 (i) Exim was not creating retry information for temporary errors given
3771 for individual recipients after the DATA command when the smtp transport
3772 was used in LMTP mode. This meant that they could be retried too
3773 frequently, and not timed out correctly.
3775 (ii) Exim was setting the flag that allows error details to be returned
3776 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3777 recipients that were returned after the DATA command.
3779 PH/53 This is related to PH/52, but is more general: for any failing address,
3780 when detailed error information was permitted to be returned to the
3781 sender, but the error was temporary, then after the final timeout, only
3782 "retry timeout exceeded" was returned. Now it returns the full error as
3783 well as "retry timeout exceeded".
3785 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3786 do this, and (what is worse) MTAs that accept it.
3788 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3789 will now be deprecated.
3791 PH/56 New os.c-cygwin from the Cygwin maintainer.
3793 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3794 formats (previously only available in exim formats via -bpr, -bpru,
3795 and -bpra. Now also available in native and exiqgrep formats)
3797 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3798 with very large, slow to parse queues
3800 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3802 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3804 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3805 responses to authentication challenges, though it was showing the
3806 challenges; (ii) I've removed the CR characters from the debug output for
3809 PH/58 Allow for the insertion of a newline as well as a space when a string
3810 is turned into more than one encoded-word during RFC 2047 encoding. The
3811 Sieve code now uses this.
3813 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3814 data_4xx, lost_connection, tls_required.
3816 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3817 message was being sent as an SMTP response.
3819 PH/61 Add -l and -k options to exicyclog.
3821 PH/62 When verifying, if an address was redirected to one new address, so that
3822 verification continued, and the new address failed or deferred after
3823 having set something in $address_data, the value of $address_data was not
3824 passed back to the ACL. This was different to the case when no
3825 redirection occurred. The value is now passed back in both cases.
3827 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3828 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3829 use login_cap.h, so on its own it isn't the distinguishing feature. The
3830 new name refers directly to the setclassresources() function.
3832 PH/65 Added configuration files for NetBSD3.
3834 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3836 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3837 is preferred over IPv4.
3839 PH/68 The bounce_return_message and bounce_return_body options were not being
3840 honoured for bounces generated during the reception of non-SMTP messages.
3841 In particular, this applied to messages rejected by the ACL. This bug has
3842 been fixed. However, if bounce_return_message is true and bounce_return_
3843 body is false, the headers that are returned for a non-SMTP message
3844 include only those that have been read before the error was detected.
3845 (In the case of an ACL rejection, they have all been read.)
3847 PH/69 The HTML version of the specification is now built in a directory called
3848 spec_html instead of spec.html, because the latter looks like a path with
3849 a MIME-type, and this confuses some software.
3851 PH/70 Catch two compiler warnings in sieve.c.
3853 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3854 function verify_get_ident() calls ip_connect() to connect a socket, but
3855 if the "connect()" function timed out, ip_connect() used to close the
3856 socket. However, verify_get_ident() also closes the socket later, and in
3857 between Exim writes to the log, which may get opened at this point. When
3858 the socket was closed in ip_connect(), the log could get the same file
3859 descriptor number as the socket. This naturally causes chaos. The fix is
3860 not to close the socket in ip_connect(); the socket should be closed by
3861 the function that creates it. There was only one place in the code where
3862 this was missing, in the iplookup router, which I don't think anybody now
3863 uses, but I've fixed it anyway.
3865 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3866 well as to direct DNS lookups. Otherwise the handling of names in host
3867 lists is inconsistent and therefore confusing.
3873 PH/01 Two changes to the default runtime configuration:
3875 (1) Move the checks for relay_from_hosts and authenticated clients from
3876 after to before the (commented out) DNS black list checks.
3878 (2) Add control=submission to the relay_from_hosts and authenticated
3879 clients checks, on the grounds that messages accepted by these
3880 statements are most likely to be submissions.
3882 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3884 (1) Generate an error if the third argument for the ${prvs expansion is
3887 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3890 (3) Reset the variables that are obtained from the first argument of
3891 ${prvscheck and used in the second argument before leaving the code,
3892 because their memory is reclaimed, so using them afterwards may do
3895 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3896 one (it's much easier than Tom thought :-).
3898 (5) Because of (4), we can now allow for the use of $prvscheck_result
3899 inside the third argument.
3901 PH/03 For some reason, the default setting of PATH when running a command from
3902 a pipe transport was just "/usr/bin". I have changed it to
3905 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3906 anything to be listed in the output from -bV.
3908 PH/05 When a filter generated an autoreply, the entire To: header line was
3909 quoted in the delivery log line, like this:
3911 => >A.N.Other <ano@some.domain> <original@ddress> ...
3913 This has been changed so that it extracts the operative address. There
3914 may be more than one such address. If so, they are comma-separated, like
3917 => >ano@some.domain,ona@other.domain <original@ddress> ...
3919 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3920 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3921 not being looked up in the rDNS to get a host name, Exim was showing the
3922 IP address twice in Received: lines, even though the IP addresses were
3923 identical. For example:
3925 Received: from [1.2.3.4] (helo=[1.2.3.4])
3927 However, if the real host name was known, it was omitting the HELO data
3928 if it matched the actual IP address. This has been tidied up so that it
3929 doesn't show the same IP address twice.
3931 PH/07 When both +timestamp and +memory debugging was on, the value given by
3932 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3933 called by the debug printing, thereby overwriting the timestamp buffer.
3934 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3936 PH/08 When the original message was included in an autoreply transport, it
3937 always said "this is a copy of the message, including all the headers",
3938 even if body_only or headers_only was set. It now gives an appropriate
3941 PH/09 Applied a patch from the Sieve maintainer which:
3943 o fixes some comments
3944 o adds the (disabled) notify extension core
3945 o adds some debug output for the result of if/elsif tests
3946 o points to the current vacation draft in the documentation
3947 and documents the missing references header update
3951 o fixes a bug in processing the envelope test (when testing
3952 multiple envelope elements, the last element determined the
3955 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3956 Electronic Mail") by including:
3958 Auto-submitted: auto-generated
3960 in the messages that it generates (bounce messages and others, such as
3961 warnings). In the case of bounce messages for non-SMTP messages, there was
3962 also a typo: it was using "Auto_submitted" (underscore instead of
3963 hyphen). Since every message generated by Exim is necessarily in response
3964 to another message, thes have all been changed to:
3966 Auto-Submitted: auto-replied
3968 in accordance with these statements in the RFC:
3970 The auto-replied keyword:
3972 - SHOULD be used on messages sent in direct response to another
3973 message by an automatic process,
3975 - MUST NOT be used on manually-generated messages,
3977 - MAY be used on Delivery Status Notifications (DSNs) and Message
3978 Disposition Notifications (MDNs),
3980 - MUST NOT be used on messages generated by automatic or periodic
3981 processes, except for messages which are automatic responses to
3984 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3985 to the default Received: header definition.
3987 PH/12 Added log selector acl_warn_skipped (default on).
3989 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3990 variables because (a) they are in the wrong storage pool and (b) even if
3991 they were copied, it wouldn't work properly because of the caching.
3993 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3994 checking when decoding. Apparently there are clients that generate
3995 overlong encoded strings. Why am I not surprised?
3997 PH/15 If the first argument of "${if match_address" was not empty, but did not
3998 contain an "@" character, Exim crashed. Now it writes a panic log message
3999 and treats the condition as false.
4001 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4003 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4004 extension "envelope-auth". The code is finished and in agreement with
4005 other implementations, but there is no documentation so far and in fact,
4006 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4007 not changing the active code.
4009 Print executed "if" and "elsif" statements when debugging is used. This
4010 helps a great deal to understand what a filter does.
4012 Document more things not specified clearly in RFC3028. I had all this
4013 sorted out, when out of a sudden new issues came to my mind. Oops."
4015 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4018 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4019 canonical form (as documented). However, after a host name lookup from
4020 the IP address, check_host() was doing a simple string comparison with
4021 addresses acquired from the DNS when checking that the found name did
4022 have the original IP as one of its addresses. Since any found IPv6
4023 addresses are likely to be in abbreviated form, the comparison could
4024 fail. Luckily, there already exists a function for doing the comparison
4025 by converting both addresses to binary, so now that is used instead of
4026 the text comparison.
4028 PH/20 There was another similar case to PH/19, when a complete host name was
4029 given in a host list; looking up its IP address could give an abbreviated
4030 form, whereas the current host's name might or might not be abbreviated.
4031 The same fix has been applied.
4037 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4038 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4041 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4042 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4044 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4046 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4047 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4048 Also, for "Auto-Submitted", treat anything other than "no" as
4049 non-personal, in accordance with RFC 3834. (Previously it treated
4050 anything starting "auto-" as non-personal.)
4052 TF/01 The control=submission/name=... option had a problem with syntax
4053 errors if the name included a slash character. The /name= option
4054 now slurps the rest of the string, so it can include any characters
4055 but it must come last in the list of options (after /sender_retain
4058 PH/05 Some modifications to the interface to the fake nameserver for the new
4066 TK/01 Added the "success_on_redirect" address verification option. See
4067 NewStuff for rationale and an example.
4069 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4071 PH/02 Patch to exigrep to allow it to work on syslog lines.
4073 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4074 fread() to skip over the body file's header line, because in Cygwin the
4075 header line is locked and is inaccessible.
4077 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4078 co-exist for some time) to make it clear that it is the Exim ID that is
4079 referenced, not the Message-ID: header line.
4081 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4082 string_format() function, because snprintf() does not exist on all
4085 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4086 ${stat: expansion item.
4088 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4089 protocol synchronization error", to keep the pedants happy.
4091 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4092 well as for IRIX systems, when gcc is being used. See the host.c source
4095 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4097 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4100 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4101 transport and to the smtp transport in LMTP mode.
4103 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4105 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4106 run a filter in a subprocess. This could lead to confusion in subsequent
4107 lookups in the parent process. There should also be a search_tidyup() at
4108 the end of the subprocess.
4110 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4111 only if the host matched helo_try_verify_hosts, which caused the
4112 verification to occur when the EHLO/HELO command was issued. The ACL just
4113 tested the remembered result. Now, if a previous verification attempt has
4114 not happened, "verify = helo" does it there and then.
4116 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4118 TK/03 Fix log output including CR from clamd.
4120 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4121 exist provoked a memory error which could cause a segfault.
4123 PH/15 Installed PCRE 6.2
4125 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4127 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4128 of the problem. Specifically, suggested +O2 rather than +O1 for the
4131 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4133 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4134 record, the port was not correctly logged when the outgoing_port log
4135 selector was set (it logged the transort's default port).
4137 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4138 fallback_hosts, and "hosts" in the smtp transport.
4140 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4141 host errors such as "Connection refused".
4143 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4144 authentication with radiusclient 0.4.9:
4146 - Error returned from rc_read_config was caught wrongly
4147 - Username/password not passed on to radius server due to wrong length.
4149 The presumption is that some radiusclient API changes for 4.51/PH/17
4150 were not taken care of correctly. The code is still untested by me (my
4151 Linux distribution still has 0.3.2 of radiusclient), but it was
4152 contributed by a Radius user.
4154 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4155 expanding the "port" option of the smtp transport.
4157 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4158 while reading a MIME header. Thanks to Tom Hughes for a patch.
4160 PH/24 Include config.h inside local_scan.h so that configuration settings are
4163 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4164 that in an ACL for RCPT (for example), you can examine exactly what was
4167 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4168 commands, but it was not correctly comparing the address with the actual
4169 client host address. Thus, it would show the EHLO address in Received:
4170 header lines when this was not necessary.
4172 PH/27 Added the % operator to ${eval:}.
4174 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4175 it should be ignoring failures (because with -C, for example, it has lost
4176 privilege). It wasn't ignoring creation failures other than "already
4179 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4182 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4183 because an input line was too long, either on its own, or by virtue of
4184 too many continuations, the temporary file was not being removed, and the
4185 return code was incorrect.
4187 PH/31 Missing "BOOL" in function definition in filtertest.c.
4189 PH/32 Applied Sieve patches from the maintainer.
4191 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4193 PH/33 Added "verify = not_blind".
4195 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4196 Local/Makefile (with some defaults set). These are used in built scripts
4197 such as exicyclog, but they have never been used in the exim_install
4198 script (though there are many overriding facilities there). I have
4199 arranged that the exim_install script now takes note of these two
4202 PH/35 Installed configuration files for Dragonfly.
4204 PH/36 When a locally submitted message by a trusted user did not contain a
4205 From: header, and the sender address was obtained from -f or from an SMTP
4206 MAIL command, and the trusted user did not use -F to supply a sender
4207 name, $originator_name was incorrectly used when constructing a From:
4208 header. Furthermore, $originator_name was used for submission mode
4209 messages from external hosts without From: headers in a similar way,
4210 which is clearly wrong.
4212 PH/37 Added control=suppress_local_fixups.
4214 PH/38 When log_selector = +received_sender was set, and the addition of the
4215 sender made the log line's construction buffer exactly full, or one byte
4216 less than full, an overflow happened when the terminating "\n" was
4219 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4220 when the result of a list match is failure because a DNS lookup failed.
4222 PH/40 RM_COMMAND is now used in the building process.
4224 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4225 the "build-* directories that it finds.
4227 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4228 address in a domain literal was a prefix of an interface address.
4230 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4231 when verifying a sender address, unless rewrite_headers is false.
4233 PH/44 Wrote a long comment about why errors_to addresses are verified as
4234 recipients, not senders.
4236 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4237 the ratelimit ACL was added.
4239 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4241 PH/46 Added extra information about PostgreSQL errors to the error string.
4243 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4244 avoiding the need to install special zones in a real server. This is
4245 backwards compatible; if it can't find the fake resolver, it drops back.
4246 Thus, both old and new test suites can be run.
4248 TF/02 Added util/ratelimit.pl
4250 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4251 clock is set back in time.
4253 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4254 Candler <B.Candler@pobox.com>.
4256 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4257 OK for addresses that are the result of redirections.
4259 PH/48 A number of further additions for the benefit of the new test suite,
4260 including a fake gethostbyname() that interfaces to the fake DNS resolver
4263 TF/06 The fix for widen_domains has also been applied to qualify_single and
4264 search_parents which are the other dnslookup options that can cause
4267 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4270 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4272 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4273 service" if it encountered an MX record with an empty target hostname.
4274 The message is now "an MX or SRV record indicated no SMTP service".
4276 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4277 if the DNS of the sending site is misconfigured. This is quite a
4278 common situation. This change restores the behaviour of treating a
4279 helo verification defer as a failure.
4281 PH/51 If self=fail was set on a router, the bounce message did not include the
4282 actual error message.
4288 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4290 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4291 command itself ended in error, the underlying message about the transport
4292 filter timeout was being overwritten with the pipe command error. Now the
4293 underlying error message should be appended to the second error message.
4295 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4297 PH/02 Reduce the amount of output that "make" produces by default. Full output
4298 can still be requested.
4300 PH/03 The warning log line about a condition test deferring for a "warn" verb
4301 was being output only once per connection, rather than after each
4302 occurrence (because it was using the same function as for successful
4303 "warn" verbs). This seems wrong, so I have changed it.
4305 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4306 it should not have, which might have caused a crash in the right
4307 circumstances, but probably never did.
4309 PH/04 Installed a modified version of Tony Finch's patch to make submission
4310 mode fix the return path as well as the Sender: header line, and to
4311 add a /name= option so that you can make the user's friendly name appear
4314 TF/03 Added the control = fakedefer ACL modifier.
4316 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4317 Mark Lowes for thorough testing.
4319 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4321 TK/03 Merged latest SRS patch from Miles Wilton.
4323 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4324 with the definition in sysexits.h (which is #included earlier).
4325 Fortunately, Exim does not actually use EX_OK. The code used to try to
4326 preserve the sysexits.h value, by assuming that macro definitions were
4327 scanned for macro replacements. I have been disabused of this notion,
4328 so now the code just undefines EX_OK before #including unistd.h.
4330 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4331 in the smtp transport. When a block could not be written in a single
4332 write() function, the timeout was being re-applied to each part-write.
4333 This seems wrong - if the receiver was accepting one byte at a time it
4334 would take for ever. The timeout is now adjusted when this happens. It
4335 doesn't have to be particularly precise.
4337 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4338 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4340 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4341 without a domain if the check to <postmaster@domain> fails.
4343 SC/01 Eximstats: added -xls and the ability to specify output files
4344 (patch written by Frank Heydlauf).
4346 SC/02 Eximstats: use FileHandles for outputting results.
4348 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4350 SC/04 Eximstats: fixed display of large numbers with -nvr option
4352 SC/05 Eximstats: fixed merging of reports with empty tables.
4354 SC/06 Eximstats: added the -include_original_destination flag
4356 SC/07 Eximstats: removed tabs and trailing whitespace.
4358 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4360 TK/06 MBOX spool code: Add real "From " MBOX separator line
4361 so the .eml file is really in mbox format (even though
4362 most programs do not really care). Patch from Alex Miller.
4364 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4365 The latter is generated from $received_to and is only set if the
4366 message has one envelope recipient. SA can use these headers,
4367 obviously out-of-the-box. Patch from Alex Miller.
4369 PH/08 The ${def test on a variable was returning false if the variable's
4370 value was "0", contrary to what the specification has always said!
4371 The result should be true unless the variable is empty.
4373 PH/09 The syntax error of a character other than { following "${if
4374 def:variable_name" (after optional whitespace) was not being diagnosed.
4375 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4376 accidental colon was present, for example, could give incorrect results.
4378 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4379 result is used (not including appendfile, where other changes are about
4382 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4383 This involved changing a lot of size variables from int to off_t. It
4384 should work with maildirs and everything.
4386 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4387 spamd dying while we are connected to it.
4389 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4392 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4393 function for BDB 4.3.
4395 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4397 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4398 the sources for PCRE have been re-arranged and more files are now
4401 PH/15 The code I had for printing potentially long long variables in PH/11
4402 above was not the best (it lost precision). The length of off_t variables
4403 is now inspected at build time, and an appropriate printing format (%ld
4404 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4405 to be "long long int" or "long int". This is needed for the internal
4406 formatting function string_vformat().
4408 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4409 the configuration file to be ":syslog", then the script "guesses" where
4410 the logs files are, rather than using the compiled in default. In our
4411 case the guess is not the same as the compiled default, so the script
4412 suddenly stopped working when I started to use syslog. The patch checks
4413 to see if log_file_path is "". If so, it attempts to read it from exim
4414 with no configuration file to get the compiled in version, before it
4415 falls back to the previous guessing code."
4417 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4418 implementing BATV in an Exim configuration. See NewStuff for the gory
4421 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4422 Makefile that are specific to HP-UX.
4424 PH/18 If the "use_postmaster" option was set for a recipient callout together
4425 with the "random" option, the postmaster address was used as the MAIL
4426 FROM address for the random test, but not for the subsequent recipient
4427 test. It is now used for both.
4429 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4430 patch removes a few documentation additions to RFC 3028, because the
4431 latest draft now contains them. It adds the new en;ascii-case comparator
4432 and a new error check for 8bit text in MIME parts. Comparator and
4433 require names are now matched exactly. I enabled the subaddress
4434 extension, but it is not well tested yet (read: it works for me)."
4436 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4437 rework some of the code of TK/09 above to avoid the hardwired use of
4438 "%lld" and "long long". Replaced the call to snprintf() with a call to
4441 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4442 records point to non-existent hosts", "retry timeout exceeded", and
4443 "retry time not reached for any host after a long failure period".
4445 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4446 experimental DomainKeys support:
4448 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4449 (2) On an error such as an illegally used "control", the wrong name for
4450 the control was given.
4452 These problems did NOT occur unless DomainKeys support was compiled.
4454 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4456 PH/24 Added ${if match_ip condition.
4458 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4459 databases so that it will be absolutely obvious if a crash occurs in the
4460 DB library. This is a regular occurrence (often caused by mis-matched
4463 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4464 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4465 picked up on a user's system that detects such things. There doesn't seem
4466 to be a gcc warning option for this - only an attribute that has to be
4467 put on the function's prototype. It seems that in Fedora Core 4 they have
4468 set this on a number of new functions. No doubt there will be more in due
4471 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4472 specify a transport. However, if an address that was verified by such a
4473 router was the subject of a callout, Exim crashed because it tried to
4474 read the rcpt_include_affixes from the non-existent transport. Now it
4475 just assumes that the setting of that option is false. This bug was
4476 introduced by 4.51/PH/31.
4478 PH/28 Changed -d+all to exclude +memory, because that information is very
4479 rarely of interest, but it makes the output a lot bigger. People tend to
4480 do -d+all out of habit.
4482 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4483 code in os-type was giving problems when libc.so lives in lib64, like on
4486 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4487 aren't the modern standard, and it seems that some systems' include files
4488 don't always have them. Exim was already checking for some of the newer
4489 ones like T_AAAA, and defining it itself. I've added checks for all the
4490 record types that Exim uses.
4492 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4493 not automatically generating a new one, as it is supposed to. This
4494 prevented TLS from working. If the file did exist, but contained invalid
4495 data, a new version was generated, as expected. It was only the case of a
4496 non-existent file that was broken.
4498 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4499 with a change in libdomainkeys > 0.64.
4501 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4502 from DNS. If the selector record carries the flag, it now has
4503 precedence over the domain-wide flag.
4505 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4507 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4508 the use of a port name that isn't defined in /etc/services) occurred, the
4509 message was deferred as in a normal delivery, and thus remained on the
4510 spool, instead of being failed because of the mua_wrapper setting. This
4511 is now fixed, and I tidied up some of the mua_wrapper messages at the
4514 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4515 the messages in a hash of arrays rather than using individual hashes.
4516 This is a bit cleaner and results in dramatic memory savings, albeit
4517 at a slight CPU cost.
4519 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4520 as requested by Marc Sherman.
4522 SC/10 Eximstats: added histograms for user specified patterns as requested
4525 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4527 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4528 fopen() in the content-scanning modules that did not already have it.
4534 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4535 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4537 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4539 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4541 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4542 to test Sieve filters that use "vacation".
4544 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4545 that changes the way the GnuTLS parameters are stored in the cache file.
4546 The new format can be generated externally. For backward compatibility,
4547 if the data in the cache doesn't make sense, Exim assumes it has read an
4548 old-format file, and it generates new data and writes a new file. This
4549 means that you can't go back to an older release without removing the
4552 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4553 work if there are any delivery delays because "one_time" forces the
4554 parent to be marked "delivered", so its unseen clone is never tried
4555 again. For this reason, Exim now forbids the simultaneous setting of
4558 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4559 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4560 entry if you want to know the details. The fix, however, neglected to
4561 consider the case when local delivery batching is involved. The test for
4562 "previously delivered" was not happening when checking to see if an
4563 address could be batched with a previous (undelivered) one; under
4564 certain circumstances this could lead to multiple deliveries to the same
4567 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4568 in its include files, and this causes problems building Exim.
4570 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4571 header_syntax) but Exim was just ignoring anything given after a slash.
4572 In particular, this caused confusion with an attempt to use "verify =
4573 reverse_host_lookup/defer_ok". An error is now given when options are
4574 supplied for verify items that do not have them. (Maybe reverse_host_
4575 lookup should have a defer_ok option, but that's a different point.)
4577 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4578 defined by RFC 821) to 2048, because there were problems with some AUTH
4579 commands, and RFC 1869 says the size should be increased for extended
4580 SMTP commands that take arguments.
4582 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4585 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4586 "unknown" error; now it says that the functionality isn't in the binary.
4588 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4589 an address' error message when a string expansion fails (syntax or
4590 whatever). Otherwise the password may appear in the log. Following change
4591 PH/42 below, there is no longer a chance of it appearing in a bounce
4594 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4596 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4597 of its addresses was ever tried. (Bugzilla bug #2.)
4599 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4600 the result incorrectly in the debug output. (It correctly added a newline
4601 to what was transported.)
4603 TF/01 Added $received_time.
4605 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4606 commented out examples of how to interface to a virus scanner and to
4607 SpamAssassin. Also added commented examples of av_scanner and
4608 spamd_address settings.
4610 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4611 and controls are allowed in which ACLs. There were a couple of minor
4612 errors. Some of the entries in the conditions table (which is a table of
4613 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4614 negation of where the condition IS allowed.
4616 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4618 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4619 header file does not have a version number, so I've had to invent a new
4620 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4621 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4622 radiusclient), but it was contributed by a Radius user.
4624 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4625 files or directories, for OpenSSL.
4627 PH/19 When an Exim process that is running as root has to create an Exim log
4628 file, it does so in a subprocess that runs as exim:exim so as to get the
4629 ownership right at creation (otherwise, other Exim processes might see
4630 the file with the wrong ownership). There was no test for failure of this
4631 fork() call, which would lead to the process getting stuck as it waited
4632 for a non-existent subprocess. Forks do occasionally fail when resources
4633 run out. I reviewed all the other calls to fork(); they all seem to check
4636 PH/20 When checking for unexpected SMTP input at connect time (before writing
4637 the banner), Exim was not dealing correctly with a non-positive return
4638 from the read() function. If the client had disconnected by this time,
4639 the result was a log entry for a synchronization error with an empty
4640 string after "input=" when read() returned zero. If read() returned -1
4641 (an event I could not check), uninitialized data bytes were printed.
4642 There were reports of junk text (parts of files, etc) appearing after
4645 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4647 PH/22 Added support for macro redefinition, and (re)definition in between
4648 driver and ACL definitions.
4650 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4651 forgetting to use the resulting value; it was using the unexpanded value.
4653 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4654 hadn't been configured. The fix is from Juergen Kreileder, who
4655 understands it better than I do:
4657 "Here's what I see happening with three configured cyrus_sasl
4658 authenticators configured (plain, login, cram-md5):
4660 On startup auth_cyrus_sasl_init() gets called for each of these.
4661 This means three calls to sasl_listmech() without a specified mech_list.
4662 => SASL tests which mechs of all available mechs actually work
4663 => three warnings about OTP not working
4664 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4666 With the patch, sasl_listmech() also gets called three times. But now
4667 SASL's mech_list option is set to the server_mech specified in the the
4668 authenticator. Or in other words, the answer from sasl_listmech()
4669 gets limited to just the mech you're testing for (which is different
4671 => the return list contains just 'plain' or 'login', 'cram-md5' or
4672 nothing depending on the value of ob->server_mech.
4674 I've just tested the patch: Authentication still works fine,
4675 unavailable mechs specified in the exim configuration are still
4676 caught, and the auth.log warnings about OTP are gone."
4678 PH/25 When debugging is enabled, the contents of the command line are added
4679 to the debugging output, even when log_selector=+arguments is not
4682 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4683 answer is "GNU", and only if the return is "GNU/something" is the answer
4686 PH/27 $acl_verify_message is now set immediately after the failure of a
4687 verification in an ACL, and so is available in subsequent modifiers. In
4688 particular, the message can be preserved by coding like this:
4690 warn !verify = sender
4691 set acl_m0 = $acl_verify_message
4693 Previously, $acl_verify_message was set only while expanding "message"
4694 and "log_message" when a very denied access.
4696 PH/28 Modified OS/os.c-Linux with
4698 -#ifndef OS_LOAD_AVERAGE
4699 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4701 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4702 nomenclature these days.)
4704 PH/29 Installed patch from the Sieve maintainer that adds the options
4705 sieve_useraddress and sieve_subaddress to the redirect router.
4707 PH/30 In these circumstances:
4708 . Two addresses routed to the same list of hosts;
4709 . First host does not offer TLS;
4710 . First host accepts first address;
4711 . First host gives temporary error to second address;
4712 . Second host offers TLS and a TLS session is established;
4713 . Second host accepts second address.
4714 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4715 and peerdn, if requested) that were in fact used only for the second
4718 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4719 attention to any local part prefix or suffix that was matched by the
4720 router that accepted the address. It now behaves in the same way as it
4721 does for delivery: the affixes are removed from the local part unless
4722 rcpt_include_affixes is set on the transport.
4724 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4725 timeout during the DATA phase of an incoming message.
4727 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4728 applied a patch sanctioned by the Sieve maintainer.
4730 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4731 the uid or gid is negative. A case of a negative gid caused this to be
4732 noticed. The fix allows for either to be negative.
4734 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4735 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4738 PH/36 The change PH/12 above was broken. Fixed it.
4740 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4741 the grounds that routing the same address twice would always produce the
4742 same answer. This might have been true once, but it is certainly no
4743 longer true now. Routing a child address may depend on the previous
4744 routing that produced that child. Some complicated redirection strategies
4745 went wrong when messages had multiple recipients, and made Exim's
4746 behaviour dependent on the order in which the addresses were given.
4748 I have moved the duplicate checking until after the routing is complete.
4749 Exim scans the addresses that are assigned to local and remote
4750 transports, and removes any duplicates. This means that more work will be
4751 done, as duplicates will always all be routed, but duplicates are
4752 presumably rare, so I don't expect this is of any significance.
4754 For deliveries to pipes, files, and autoreplies, the duplicate checking
4755 still happens during the routing process, since they are not going to be
4758 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4759 It corrects a timeout issue with spamd. This is Ian's comment: "The
4760 background is that sometimes spamd either never reads data from a
4761 connection it has accepted, or it never writes response data. The exiscan
4762 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4763 blindly assumes that writes won't block so it may never time out."
4765 PH/39 Allow G after quota size as well as K and M.
4767 PH/40 The value set for $authenticated_id in an authenticator may not contain
4768 binary zeroes or newlines because the value is written to log lines and
4769 to spool files. There was no check on this. Now the value is run through
4770 the string_printing() function so that such characters are converted to
4771 printable escape sequences.
4773 PH/41 $message_linecount is a new variable that contains the total number of
4774 lines in the message. Compare $body_linecount, which is the count for the
4777 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4778 bounce and delay warning messages, except in certain special cases, which
4781 (a) An SMTP error message from a remote host;
4782 (b) A message specified in a :fail: redirection;
4783 (c) A message specified in a "fail" command in a system filter;
4784 (d) A message specified in a FAIL return from the queryprogram router;
4785 (e) A message specified by the cannot_route_message router option.
4787 In these cases only, Exim does include the error details in bounce and
4788 warning messages. There are also a few cases where bland messages such
4789 as "unrouteable address" or "local delivery error" are given.
4791 PH/43 $value is now also set for the "else" part of a ${run expansion.
4793 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4794 being worked on, but at least Exim now implements the latest version to
4797 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4798 process to complete was treated as a delivery failure, a timeout while
4799 writing the message to the pipe was logged, but erroneously treated as a
4800 successful delivery. Such timeouts include transport filter timeouts. For
4801 consistency with the overall process timeout, these timeouts are now
4802 treated as errors, giving rise to delivery failures by default. However,
4803 there is now a new Boolean option for the pipe transport called
4804 timeout_defer, which, if set TRUE, converts the failures into defers for
4805 both kinds of timeout. A transport filter timeout is now identified in
4808 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4809 systems where "make" and "gmake" are different, calling "gmake" at top
4810 level broke things. I've arranged for the value of $(MAKE) to be passed
4811 from the Makefile to this script so that it can call the same version of
4815 A note about Exim versions 4.44 and 4.50
4816 ----------------------------------------
4818 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4819 changes of various kinds. As a consequence, a big documentation update was
4820 needed. This delayed the release for rather longer than seemed good, especially
4821 in the light of a couple of (minor) security issues. Therefore, the changes
4822 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4823 release. So 4.44 and 4.50 are in effect two different branches that both start
4826 I have left the 4.50 change log unchanged; it contains all the changes since
4827 4.43. The change log for 4.44 is below; many of its items are identical to
4828 those for 4.50. This seems to be the most sensible way to preserve the
4829 historical information.
4835 1. Minor wording change to the doc/README.SIEVE file.
4837 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4838 computation of the current number of files was incorrect.
4840 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4841 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4844 4. Give more explanation in the error message when the command for a transport
4845 filter fails to execute.
4847 5. There are several places where Exim runs a non-Exim command in a
4848 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4849 was being done only for the command run by the queryprogram router. It is
4850 now done for all such subprocesses. The other cases are: ${run, transport
4851 filters, and the commands run by the lmtp and pipe transports.
4853 6. Added CONFIGURE_GROUP build-time option.
4855 7. Some older OS have a limit of 256 on the maximum number of file
4856 descriptors. Exim was using setrlimit() to set 1000 as a large value
4857 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4858 systems. I've change it so that if it can't get 1000, it tries for 256.
4860 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4861 was an oversight, and furthermore, ever since the addition of extra
4862 controls (e.g. 4.43/32), the checks on when to allow different forms of
4863 "control" were broken. There should now be diagnostics for all cases when a
4864 control that does not make sense is encountered.
4866 9. Added the /retain_sender option to "control=submission".
4868 10. $recipients is now available in the predata ACL (oversight).
4870 11. Tidy the search cache before the fork to do a delivery from a message
4871 received from the command line. Otherwise the child will trigger a lookup
4872 failure and thereby defer the delivery if it tries to use (for example) a
4873 cached ldap connection that the parent has called unbind on.
4875 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4876 of $address_data from the recipient verification was clobbered by the
4877 sender verification.
4879 13. The value of address_data from a sender verification is now available in
4880 $sender_address_data in subsequent conditions in the ACL statement.
4882 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4884 15. Added a new option "connect=<time>" to callout options, to set a different
4887 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4888 was its contents. (It was OK if the option was not defined at all.)
4890 17. A "Completed" log line is now written for messages that are removed from
4891 the spool by the -Mrm option.
4893 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4894 information about exactly what failed.
4896 19. Added -dd to debug only the daemon process.
4898 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4899 handles timeouts, both on the server side and network timeouts. Renamed the
4900 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4902 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4903 It is now set to "smtps".
4905 22. $host_address is now set to the target address during the checking of
4906 ignore_target_hosts.
4908 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4909 being passed; this would have caused $sender_host_name to have been used if
4910 matching the list had actually called for a host name (not very likely,
4911 since this list is usually IP addresses). A host name is now passed as
4914 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4915 code that specifies a non-restarting handler (typically sigaction() in
4916 modern systems) in an attempt to fix a rare and obscure crash bug.
4918 25. Narrowed the window for a race in the daemon that could cause it to ignore
4919 SIGCHLD signals. This is not a major problem, because they are used only to
4920 wake it up if nothing else does.
4922 26. A malformed maildirsize file could cause Exim to calculate negative values
4923 for the mailbox size or file count. Odd effects could occur as a result.
4924 The maildirsize information is now recalculated if the size or filecount
4927 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4928 support for a long time. Removed HAVE_SYS_VFS_H.
4930 28. Installed the latest version of exipick from John Jetmore.
4932 29. In an address list, if the pattern was not a regular expression, an empty
4933 subject address (from a bounce message) matched only if the pattern was an
4934 empty string. Non-empty patterns were not even tested. This was the wrong
4935 because it is perfectly reasonable to use an empty address as part of a
4936 database query. An empty address is now tested by patterns that are
4937 lookups. However, all the other forms of pattern expect the subject to
4938 contain a local part and a domain, and therefore, for them, an empty
4939 address still always fails if the pattern is not itself empty.
4941 30. Exim went into a mad DNS loop when attempting to do a callout where the
4942 host was specified on an smtp transport, and looking it up yielded more
4943 than one IP address.
4945 31. Re-factored the code for checking spool and log partition space into a
4946 function that finds that data and another that does the check. The former
4947 is then used to implement four new variables: $spool_space, $log_space,
4948 $spool_inodes, and $log_inodes.
4950 32. The RFC2047 encoding function was originally intended for short strings
4951 such as real names; it was not keeping to the 75-character limit for
4952 encoded words that the RFC imposes. It now respects the limit, and
4953 generates multiple encoded words if necessary. To be on the safe side, I
4954 have increased the buffer size for the ${rfc2047: expansion operator from
4957 33. It is now permitted to omit both strings after an "if" condition; if the
4958 condition is true, the result is "true". As before, when the second string
4959 is omitted, a false condition yields an empty string. This makes it less
4960 cumbersome to write custom ACL and router conditions.
4962 34. Failure to deliver a bounce message always caused it to be frozen, even if
4963 there was an errors_to setting on the router. The errors_to setting is now
4966 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4967 canonical form (fully expanded) before being placed in
4968 $sender_host_address.
4970 36. The table in the code that translates DNS record types into text (T_A to
4971 "A" for instance) was missing entries for NS and CNAME. It is just possible
4972 that this could have caused confusion if both these types were looked up
4973 for the same domain, because the text type is used as part of Exim's
4974 per-process caching. But the chance of anyone hitting this buglet seems
4977 37. The dnsdb lookup has been extended in a number of ways.
4979 (1) There is a new type, "zns", which walks up the domain tree until it
4980 finds some nameserver records. It should be used with care.
4982 (2) There is a new type, "mxh", which is like "mx" except that it returns
4983 just the host names, not the priorities.
4985 (3) It is now possible to give a list of domains (or IP addresses) to be
4986 looked up. The behaviour when one of the lookups defers can be
4987 controlled by a keyword.
4989 (4) It is now possible to specify the separator character for use when
4990 multiple records are returned.
4992 38. The dnslists ACL condition has been extended: it is now possible to supply
4993 a list of IP addresses and/or domains to be looked up in a particular DNS
4996 39. Added log_selector=+queue_time_overall.
4998 40. When running the queue in the test harness, wait just a tad after forking a
4999 delivery process, to get repeatability of debugging output.
5001 41. Include certificate and key file names in error message when GnuTLS fails
5002 to set them up, because the GnuTLS error message doesn't include the name
5003 of the failing file when there is a problem reading it.
5005 42. Allow both -bf and -bF in the same test run.
5007 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5009 44. The "Exiscan patch" is now merged into the mainline Exim source.
5011 45. Sometimes the final signoff response after QUIT could fail to get
5012 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5013 < 0 before doing a fflush(). This bug looks as though it goes back to the
5014 introduction of TLS in release 3.20, but "sometimes" must have been rare
5015 because the tests only now provoked it.
5017 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5018 (this can affect the format of dates).
5020 47. exim_tidydb, when checking for the continued existence of a message for
5021 which it has found a message-specific retry record, was not finding
5022 messages that were in split spool directories. Consequently, it was
5023 deleting retry records that should have stayed in existence.
5025 48. Steve fixed some bugs in eximstats.
5027 49. The SPA authentication driver was not abandoning authentication and moving
5028 on to the next authenticator when an expansion was forced to fail,
5029 contradicting the general specification for all authenticators. Instead it
5030 was generating a temporary error. It now behaves as specified.
5032 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5033 (the order specifies the preference for clients). The order is now AES256,
5034 AES128, 3DES, ARCFOUR128.
5036 51. Small patch to Sieve code - explicitly set From: when generating an
5039 52. Exim crashed if a remote delivery caused a very long error message to be
5040 recorded - for instance if somebody sent an entire SpamAssassin report back
5041 as a large number of 550 error lines. This bug was coincidentally fixed by
5042 increasing the size of one of Exim's internal buffers (big_buffer) that
5043 happened as part of the Exiscan merge. However, to be on the safe side, I
5044 have made the code more robust (and fixed the comments that describe what
5047 53. Now that there can be additional text after "Completed" in log lines (if
5048 the queue_time_overall log selector is set), a one-byte patch to exigrep
5049 was needed to allow it to recognize "Completed" as not the last thing in
5052 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5053 patch that reportedly fixes this has been added. I am not expert enough to
5054 create a test for it. This is what the patch creator wrote:
5056 "I found a little strange behaviour of ldap code when working with
5057 Windows 2003 AD Domain, where users was placed in more than one
5058 Organization Units. When I tried to give exim partial DN, the exit code
5059 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5060 But simultaneously result of request was absolutely normal ldap result,
5061 so I produce this patch..."
5063 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5064 REFERENCE, so I have modified the code to exclude the patch when that macro
5067 55. Some experimental protocols are using DNS PTR records for new purposes. The
5068 keys for these records are domain names, not reversed IP addresses. The
5069 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5070 leaves it alone. Component reversal etc. now happens only for IP addresses.
5073 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5075 57. Double the size of the debug message buffer (to 2048) so that more of very
5076 long debug lines gets shown.
5078 58. The exicyclog utility now does better if the number of log files to keep
5079 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5081 59. Two changes related to the smtp_active_hostname option:
5083 (1) $smtp_active_hostname is now available as a variable.
5084 (2) The default for smtp_banner uses $smtp_active_hostname instead
5085 of $primary_hostname.
5087 60. The host_aton() function is supposed to be passed a string that is known
5088 to be a valid IP address. However, in the case of IPv6 addresses, it was
5089 not checking this. This is a hostage to fortune. Exim now panics and dies
5090 if the condition is not met. A case was found where this could be provoked
5091 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5092 components; fortuitously, this particular loophole had already been fixed
5093 by change 4.50/55 above.
5095 If there are any other similar loopholes, the new check in host_aton()
5096 itself should stop them being exploited. The report I received stated that
5097 data on the command line could provoke the exploit when Exim was running as
5098 exim, but did not say which command line option was involved. All I could
5099 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5100 running as the user.
5103 61. There was a buffer overflow vulnerability in the SPA authentication code
5104 (which came originally from the Samba project). I have added a test to the
5105 spa_base64_to_bits() function which I hope fixes it.
5108 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5109 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5111 63. The daemon start-up calls getloadavg() while still root for those OS that
5112 need the first call to be done as root, but it missed one case: when
5113 deliver_queue_load_max is set with deliver_drop_privilege. This is
5114 necessary for the benefit of the queue runner, because there is no re-exec
5115 when deliver_drop_privilege is set.
5117 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5118 This has been fixed.
5120 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5121 was in use, was not putting the data itself into the right store pool;
5122 consequently, it could be overwritten for a subsequent message in the same
5123 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5126 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5128 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5129 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5130 the function were treating the return as a boolean value, which happened to
5131 work because 0=false and not-0=true, but is not correct code.
5133 68. The host_aton() function was not handling scoped IPv6 addresses (those
5134 with, for example, "%eth0" on the end) correctly.
5136 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5137 negated items (that is, ~something) in unsigned ints. Some compilers
5138 apparently mutter when there is no cast.
5140 70. If an address verification called from an ACL failed, and did not produce a
5141 user-specific message (i.e. there was only a "system" message), nothing was
5142 put in $acl_verify_message. In this situation, it now puts the system
5145 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5146 session; change 4.31/43 added the unwanted input to the log line - except
5147 that it did not do this in the start of session case. It now does.
5149 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5150 This is wrong and can cause the other end to generate a synchronization
5151 error if it is another Exim or anything else that does the synchronization
5152 check. A QUIT command is no longer sent after a timeout.
5154 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5155 during host lookups.
5157 74. The defer_ok option of callout verification was not working if it was used
5158 when verifying addresses in header lines, that is, for this case:
5160 verify = header_sender/callout=defer_ok
5162 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5163 those file descriptors could be used for SMTP connections. If anything
5164 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5165 could be sent to the SMTP client, causing chaos. The daemon now opens
5166 stdin, stdout, and stderr to /dev/null when it puts itself into the
5169 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5170 log by default. The user can override this with suitable Perl magic.
5172 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5173 the log message when discard triggers, was not working for the DATA ACL or
5174 for the non-SMTP ACL.
5176 78. Error message wording change in sieve.c.
5178 79. If smtp_accept_max_per_host was set, the number of connections could be
5179 restricted to fewer than expected, because the daemon was trying to set up
5180 a new connection before checking whether the processes handling previous
5181 connections had finished. The check for completed processes is now done
5182 earlier. On busy systems, this bug wouldn't be noticed because something
5183 else would have woken the daemon, and it would have reaped the completed
5186 80. If a message was submitted locally by a user whose login name contained one
5187 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5188 It caused a spool format error. I have fixed the spool reading code. A
5189 related problem was that the "from" clause in the Received: line became
5190 illegal because of the space(s). It is now covered by ${quote_local_part.
5192 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5195 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5197 83. Patch from Sieve maintainer to fix unterminated string problem in
5198 "vacation" handling.
5200 84. Some minor changes to the Linux configuration files to help with other
5201 OS variants using glibc.
5203 85. One more patch for Sieve to update vacation handling to latest spec.
5206 ----------------------------------------------------
5207 See the note above about the 4.44 and 4.50 releases.
5208 ----------------------------------------------------
5214 1. Change 4.43/35 introduced a bug that caused file counts to be
5215 incorrectly computed when quota_filecount was set in an appendfile
5218 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5219 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5222 3. Give more explanation in the error message when the command for a transport
5223 filter fails to execute.
5225 4. There are several places where Exim runs a non-Exim command in a
5226 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5227 was being done only for the command run by the queryprogram router. It is
5228 now done for all such subprocesses. The other cases are: ${run, transport
5229 filters, and the commands run by the lmtp and pipe transports.
5231 5. Some older OS have a limit of 256 on the maximum number of file
5232 descriptors. Exim was using setrlimit() to set 1000 as a large value
5233 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5234 systems. I've change it so that if it can't get 1000, it tries for 256.
5236 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5237 was an oversight, and furthermore, ever since the addition of extra
5238 controls (e.g. 4.43/32), the checks on when to allow different forms of
5239 "control" were broken. There should now be diagnostics for all cases when a
5240 control that does not make sense is encountered.
5242 7. $recipients is now available in the predata ACL (oversight).
5244 8. Tidy the search cache before the fork to do a delivery from a message
5245 received from the command line. Otherwise the child will trigger a lookup
5246 failure and thereby defer the delivery if it tries to use (for example) a
5247 cached ldap connection that the parent has called unbind on.
5249 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5250 of $address_data from the recipient verification was clobbered by the
5251 sender verification.
5253 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5254 was its contents. (It was OK if the option was not defined at all.)
5256 11. A "Completed" log line is now written for messages that are removed from
5257 the spool by the -Mrm option.
5259 12. $host_address is now set to the target address during the checking of
5260 ignore_target_hosts.
5262 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5263 being passed; this would have caused $sender_host_name to have been used if
5264 matching the list had actually called for a host name (not very likely,
5265 since this list is usually IP addresses). A host name is now passed as
5268 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5269 code that specifies a non-restarting handler (typically sigaction() in
5270 modern systems) in an attempt to fix a rare and obscure crash bug.
5272 15. Narrowed the window for a race in the daemon that could cause it to ignore
5273 SIGCHLD signals. This is not a major problem, because they are used only to
5274 wake it up if nothing else does.
5276 16. A malformed maildirsize file could cause Exim to calculate negative values
5277 for the mailbox size or file count. Odd effects could occur as a result.
5278 The maildirsize information is now recalculated if the size or filecount
5281 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5282 support for a long time. Removed HAVE_SYS_VFS_H.
5284 18. Updated exipick to current release from John Jetmore.
5286 19. Allow an empty sender to be matched against a lookup in an address list.
5287 Previously the only cases considered were a regular expression, or an
5290 20. Exim went into a mad DNS lookup loop when doing a callout where the
5291 host was specified on the transport, if the DNS lookup yielded more than
5294 21. The RFC2047 encoding function was originally intended for short strings
5295 such as real names; it was not keeping to the 75-character limit for
5296 encoded words that the RFC imposes. It now respects the limit, and
5297 generates multiple encoded words if necessary. To be on the safe side, I
5298 have increased the buffer size for the ${rfc2047: expansion operator from
5301 22. Failure to deliver a bounce message always caused it to be frozen, even if
5302 there was an errors_to setting on the router. The errors_to setting is now
5305 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5306 canonical form (fully expanded) before being placed in
5307 $sender_host_address.
5309 24. Updated eximstats to version 1.33
5311 25. Include certificate and key file names in error message when GnuTLS fails
5312 to set them up, because the GnuTLS error message doesn't include the name
5313 of the failing file when there is a problem reading it.
5315 26. Expand error message when OpenSSL has problems setting up cert/key files.
5318 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5319 (this can affect the format of dates).
5321 28. exim_tidydb, when checking for the continued existence of a message for
5322 which it has found a message-specific retry record, was not finding
5323 messages that were in split spool directories. Consequently, it was
5324 deleting retry records that should have stayed in existence.
5326 29. eximstats updated to version 1.35
5327 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5328 1.35 - bugfix such that pie charts by volume are generated correctly
5330 30. The SPA authentication driver was not abandoning authentication and moving
5331 on to the next authenticator when an expansion was forced to fail,
5332 contradicting the general specification for all authenticators. Instead it
5333 was generating a temporary error. It now behaves as specified.
5335 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5336 (the order specifies the preference for clients). The order is now AES256,
5337 AES128, 3DES, ARCFOUR128.
5339 31. Small patch to Sieve code - explicitly set From: when generating an
5342 32. Exim crashed if a remote delivery caused a very long error message to be
5343 recorded - for instance if somebody sent an entire SpamAssassin report back
5344 as a large number of 550 error lines. This bug was coincidentally fixed by
5345 increasing the size of one of Exim's internal buffers (big_buffer) that
5346 happened as part of the Exiscan merge. However, to be on the safe side, I
5347 have made the code more robust (and fixed the comments that describe what
5350 33. Some experimental protocols are using DNS PTR records for new purposes. The
5351 keys for these records are domain names, not reversed IP addresses. The
5352 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5353 leaves it alone. Component reversal etc. now happens only for IP addresses.
5356 34. The host_aton() function is supposed to be passed a string that is known
5357 to be a valid IP address. However, in the case of IPv6 addresses, it was
5358 not checking this. This is a hostage to fortune. Exim now panics and dies
5359 if the condition is not met. A case was found where this could be provoked
5360 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5361 components; fortuitously, this particular loophole had already been fixed
5362 by change 4.50/55 or 4.44/33 above.
5364 If there are any other similar loopholes, the new check in host_aton()
5365 itself should stop them being exploited. The report I received stated that
5366 data on the command line could provoke the exploit when Exim was running as
5367 exim, but did not say which command line option was involved. All I could
5368 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5369 running as the user.
5372 35. There was a buffer overflow vulnerability in the SPA authentication code
5373 (which came originally from the Samba project). I have added a test to the
5374 spa_base64_to_bits() function which I hope fixes it.
5377 36. The daemon start-up calls getloadavg() while still root for those OS that
5378 need the first call to be done as root, but it missed one case: when
5379 deliver_queue_load_max is set with deliver_drop_privilege. This is
5380 necessary for the benefit of the queue runner, because there is no re-exec
5381 when deliver_drop_privilege is set.
5383 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5384 was in use, was not putting the data itself into the right store pool;
5385 consequently, it could be overwritten for a subsequent message in the same
5386 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5389 38. Sometimes the final signoff response after QUIT could fail to get
5390 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5391 < 0 before doing a fflush(). This bug looks as though it goes back to the
5392 introduction of TLS in release 3.20, but "sometimes" must have been rare
5393 because the tests only now provoked it.
5399 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5400 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5401 Now it returns an integer. A number of calls were still expecting a T/F
5402 return. Fortuitously, in all cases, the tests worked in OK situations,
5403 which is the norm. However, things would have gone wrong on any write
5404 failures on the smtp file descriptor. This function is used when sending
5405 messages over SMTP and also when doing verify callouts.
5407 2. When Exim is called to do synchronous delivery of a locally submitted
5408 message (the -odf or -odi options), it no longer closes stderr before doing
5411 3. Implemented the mua_wrapper option.
5413 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5415 5. Implemented the functions header_remove(), header_testname(),
5416 header_add_at_position(), and receive_remove_recipient(), and exported them
5419 6. If an ACL "warn" statement specified the addition of headers, Exim already
5420 inserted X-ACL-Warn: at the start if there was no header name. However, it
5421 was not making this test for the second and subsequent header lines if
5422 there were newlines in the string. This meant that an invalid header could
5423 be inserted if Exim was badly configured.
5425 7. Allow an ACL "warn" statement to add header lines at the start or after all
5426 the Received: headers, as well as at the end.
5428 8. Added the rcpt_4xx retry error code.
5430 9. Added postmaster_mailfrom=xxx to callout verification option.
5432 10. Added mailfrom=xxxx to the callout verification option, for verify=
5435 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5436 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5438 12. Inserted some casts to stop certain compilers warning when using pointer
5439 differences as field lengths or precisions in printf-type calls (mostly
5440 affecting debugging statements).
5442 13. Added optional readline() support for -be (dynamically loaded).
5444 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5445 same clock tick as a message's arrival, so that its received time was the
5446 same as the "first fail" time on the retry record, and that message
5447 remained on the queue past the ultimate address timeout, every queue runner
5448 would try a delivery (because it was past the ultimate address timeout) but
5449 after another failure, the ultimate address timeout, which should have then
5450 bounced the address, did not kick in. This was a "< instead of <=" error;
5451 in most cases the first failure would have been in the next clock tick
5452 after the received time, and all would be well.
5454 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5455 being recognized when the domain list was tested by the match_domain
5456 condition in an expansion string.
5458 16. Added the ${str2b64: operator.
5460 17. Exim was always calling setrlimit() to set a large limit for the number of
5461 processes, without checking whether the existing limit was already
5462 adequate. (It did check for the limit on file descriptors.) Furthermore,
5463 errors from getrlimit() and setrlimit() were being ignored. Now they are
5464 logged to the main and panic logs, but Exim does carry on, to try to do its
5465 job under whatever limits there are.
5467 18. Imported PCRE 5.0.
5469 19. Trivial typo in log message " temporarily refused connection" (the leading
5472 20. If the log selector return_path_on_delivery was set and an address was
5473 redirected to /dev/null, the delivery process crashed because it assumed
5474 that a return path would always be set for a "successful" delivery. In this
5475 case, the whole delivery is bypassed as an optimization, and therefore no
5478 21. Internal re-arrangement: the function for sending a challenge and reading
5479 a response while authentication was assuming a zero-terminated challenge
5480 string. It's now changed to take a pointer and a length, to allow for
5481 binary data in such strings.
5483 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5485 23. Exim was not respecting finduser_retries when seeking the login of the
5486 uid under which it was called; it was always trying 10 times. (The default
5487 setting of finduser_retries is zero.) Also, it was sleeping after the final
5488 failure, which is pointless.
5490 24. Implemented tls_on_connect_ports.
5492 25. Implemented acl_smtp_predata.
5494 26. If the domain in control=submission is set empty, Exim assumes that the
5495 authenticated id is a complete email address when it generates From: or
5496 Sender: header lines.
5498 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5499 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5500 chown and chgrp in /bin and hostname in /usr/bin.
5502 28. Exim was keeping the "process log" file open after each use, just as it
5503 does for the main log. This opens the possibility of it remaining open for
5504 long periods when the USR1 signal hits a daemon. Occasional processlog
5505 errors were reported, that could have been caused by this. Anyway, it seems
5506 much more sensible not to leave this file open at all, so that is what now
5509 29. The long-running daemon process does not normally write to the log once it
5510 has entered its main loop, and it closes the log before doing so. This is
5511 so that log files can straightforwardly be renamed and moved. However,
5512 there are a couple of unusual error situations where the daemon does write
5513 log entries, and I had neglected to close the log afterwards.
5515 30. The text of an SMTP error response that was received during a remote
5516 delivery was being truncated at 512 bytes. This is too short for some of
5517 the long messages that one sometimes sees. I've increased the limit to
5520 31. It is now possible to make retry rules that apply only when a message has a
5521 specific sender, in particular, an empty sender.
5523 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5524 it possible to be selective about when SMTP synchronization is enforced.
5526 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5528 32. Implemented hosts_connection_nolog.
5530 33. Added an ACL for QUIT.
5532 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5535 35. Added mailbox_size and mailbox_filecount to appendfile.
5537 36. Added control = no_multiline_responses to ACLs.
5539 37. There was a bug in the logic of the code that waits for the clock to tick
5540 in the case where the clock went backwards by a substantial amount such
5541 that the microsecond fraction of "now" was more than the microsecond
5542 fraction of "then" (but the whole seconds number was less).
5544 38. Added support for the libradius Radius client library this is found on
5545 FreeBSD (previously only the radiusclient library was supported).
5551 1. When certain lookups returned multiple values in the form name=value, the
5552 quoting of the values was not always being done properly. Specifically:
5553 (a) If the value started with a double quote, but contained no whitespace,
5555 (b) If the value contained whitespace other than a space character (i.e.
5556 tabs or newlines or carriage returns) it was not quoted.
5557 This fix has been applied to the mysql and pgsql lookups by writing a
5558 separate quoting function and calling it from the lookup code. The fix
5559 should probably also be applied to nisplus, ibase and oracle lookups, but
5560 since I cannot test any of those, I have not disturbed their existing code.
5562 2. A hit in the callout cache for a specific address caused a log line with no
5563 reason for rejecting RCPT. Now it says "Previous (cached) callout
5564 verification failure".
5566 3. There was an off-by-one bug in the queryprogram router. An over-long
5567 return line was truncated at 256 instead of 255 characters, thereby
5568 overflowing its buffer with the terminating zero. As well as fixing this, I
5569 have increased the buffer size to 1024 (and made a note to document this).
5571 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5572 when Exim is waiting for an SMTP response from a remote server, Exim
5573 restarts its select() call on the socket, thereby resetting its timeout.
5574 This is not a problem when such interrupts are rare. Somebody set up a cron
5575 job to run exiwhat every 2 minutes, which is less than the normal select()
5576 timeout (5 or 10 minutes). This meant that the select() timeout never
5577 kicked in because it was always reset. I have fixed this by comparing the
5578 time when an interrupt arrives with the time at the start of the first call
5579 to select(). If more time than the timeout has elapsed, the interrupt is
5580 treated as a timeout.
5582 5. Some internal re-factoring in preparation for the addition of Sieve
5583 extensions (by MH). In particular, the "personal" test is moved to a
5584 separate function, and given an option for scanning Cc: and Bcc: (which is
5585 not set for Exim filters).
5587 6. When Exim created an email address using the login of the caller as the
5588 local part (e.g. when creating a From: or Sender: header line), it was not
5589 quoting the local part when it contained special characters such as @.
5591 7. Installed new OpenBSD configuration files.
5593 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5594 try to make them clearer.
5596 9. Callout options, other than the timeout value, were being ignored when
5597 verifying sender addresses in header lines. For example, when using
5599 verify = header_sender/callout=no_cache
5601 the cache was (incorrectly) being used.
5603 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5604 only the Cygwin environment.
5606 11. When return_path_on_delivery was set as a log selector, if different remote
5607 addresses in the same message used different return paths and parallel
5608 remote delivery occurred, the wrong values would sometimes be logged.
5609 (Whenever a remote delivery process finished, the return path value from
5610 the most recently started remote delivery process was logged.)
5612 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5613 lines when AUTH and/or TLS are in use. This is the "received protocol"
5614 field. Exim used to use "asmtp" for authenticated SMTP, without any
5615 indication (in the protocol name) for TLS use. Now it follows the RFC and
5616 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5617 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5618 names appear in log lines as well as in Received: header lines.
5620 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5621 extensions, and comparison tests, and to fix some bugs.
5623 14. Changes to the "personal" filter test:
5625 (1) The test was buggy in that it was just doing the equivalent of
5626 "contains" tests on header lines. For example, if a user's address was
5627 anne@some.where, the "personal" test would incorrectly be true for
5629 To: susanne@some.where
5631 This test is now done by extracting each address from the header in turn,
5632 and checking the entire address. Other tests that are part of "personal"
5633 are now done using regular expressions (for example, to check local parts
5634 of addresses in From: header lines).
5636 (2) The list of non-personal local parts in From: addresses has been
5637 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5638 taken from the Sieve specification recommendations.
5640 (3) If the message contains any header line starting with "List-" it is
5641 treated as non-personal.
5643 (4) The test for "circular" in the Subject: header line has been removed
5644 because it now seems ill-conceived.
5646 15. Minor typos in src/EDITME comments corrected.
5648 16. Installed latest exipick from John Jetmore.
5650 17. If headers_add on a router specified a text string that was too long for
5651 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5652 of string_sprintf() is now avoided.
5654 18. $message_body_size was not set (it was always zero) when running the DATA
5655 ACL and the local_scan() function.
5657 19. For the "mail" command in an Exim filter, no default was being set for
5658 the once_repeat time, causing a random time value to be used if "once" was
5659 specified. (If the value happened to be <= 0, no repeat happened.) The
5660 default is now 0s, meaning "never repeat". The "vacation" command was OK
5661 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5662 (I found it when inspecting the code).
5664 20. There is now an overall timeout for performing a callout verification. It
5665 defaults to 4 times the callout timeout, which applies to individual SMTP
5666 commands during the callout. The overall timeout applies when there is more
5667 than one host that can be tried. The timeout is checked before trying the
5668 next host. This prevents very long delays if there are a large number of
5669 hosts and all are timing out (e.g. when the network connections are timing
5670 out). The value of the overall timeout can be changed by specifying an
5671 additional sub-option for "callout", called "maxwait". For example:
5673 verify = sender/callout=5s,maxwait=20s
5675 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5676 to the end before writing, but this should make it even safer).
5678 22. Exim was forgetting that it had advertised PIPELINING for the second and
5679 subsequent messages on an SMTP connection. It was also not resetting its
5680 memory on STARTTLS and an internal HELO.
5682 23. When Exim logs an SMTP synchronization error within a session, it now
5683 records whether PIPELINING has been advertised or not.
5685 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5686 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5687 rather than long int.
5689 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5691 26. Added the never_mail option to autoreply.
5697 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5698 crash if the getsockname() call failed; this can happen if a connection is
5699 closed very soon after it is established. The problem was simply in the
5700 order in which certain operations were done, causing Exim to try to write
5701 to the SMTP stream before it had set up the file descriptor. The bug has
5702 been fixed by making things happen in the correct order.
5708 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5709 before the connection was closed, thus losing the rejection response.
5711 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5712 some early Solaris releases, but causes trouble in current releases where
5713 socklen_t is defined.
5715 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5718 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5721 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5722 that releasing the top of it at the end releases what was used for sub-
5723 expansions (unless the block got too big). However, discard this block if
5724 the first thing is a variable or header, so that we can use its block when
5725 it is dynamic (useful for very large $message_headers, for example).
5727 6. Lookups now cache *every* query, not just the most recent. A new, separate
5728 store pool is used for this. It can be recovered when all lookup caches are
5729 flushed. Lookups now release memory at the end of their result strings.
5730 This has involved some general refactoring of the lookup sources.
5732 7. Some code has been added to the store_xxx() functions to reduce the amount
5733 of flapping under certain conditions.
5735 8. log_incoming_interface used to affect only the <= reception log lines. Now
5736 it causes the local interface and port to be added to several more SMTP log
5737 lines, for example "SMTP connection from", and rejection lines.
5739 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5741 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5743 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5744 could be overwritten at the end of the current message (or the start of a
5745 new message if it was set in a HELO ACL). The value is now preserved for
5746 the duration of the SMTP connection.
5748 12. If a transport had a headers_rewrite setting, and a matching header line
5749 contained an unqualified address, that address was qualified, even if it
5750 did not match any rewriting rules. The underlying bug was that the values
5751 of the flags that permit the existence of unqualified sender and recipient
5752 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5753 non-local messages, and by -bnq for local messages) were not being
5754 preserved with the message after it was received.
5756 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5757 "next input=" as part of the text comprising the host identity instead of
5758 the correct text. The code was using the same buffer for two different
5759 strings. However, depending on which order the printing function evaluated
5760 its arguments, the bug did not always show up. Under Linux, for example, my
5761 test suite worked just fine.
5763 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5764 doesn't work with some older versions of Perl. It has been changed to "my",
5765 which in any case is probably the better facility to use.
5767 15. A really picky compiler found some instances of statements for creating
5768 error messages that either had too many or two few arguments for the format
5771 16. The size of the buffer for calls to the DNS resolver has been increased
5772 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5773 for addresses that have a lot of PTR records. This alleviates a problem; it
5774 does not fully solve it.
5776 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5777 buffer now truncates the list and logs the incident, which is the same
5778 action as happens when Exim is looking up a host name and its aliases.
5779 Previously in this situation something unpredictable would happen;
5780 sometimes it was "internal error: store_reset failed".
5782 18. If a server dropped the connection unexpectedly when an Exim client was
5783 using GnuTLS and trying to read a response, the client delivery process
5784 crashed while trying to generate an error log message.
5786 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5787 string, for example:
5789 warn message = H1: something\nH2: something
5791 the text was added as a single header line from Exim's point of view
5792 though it ended up OK in the delivered message. However, searching for the
5793 second and subsequent header lines using $h_h2: did not work. This has been
5794 fixed. Similarly, if a system filter added multiple headers in this way,
5795 the routers could not see them.
5797 20. Expanded the error message when iplsearch is called with an invalid key to
5798 suggest using net-iplsearch in a host list.
5800 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5801 ACLs are no longer actually imposed (and a message to that effect is
5804 22. If a "gecos" field in a passwd entry contained escaped characters, in
5805 particular, if it contained a \" sequence, Exim got it wrong when building
5806 a From: or a Sender: header from that name. A second bug also caused
5807 incorrect handling when an unquoted " was present following a character
5808 that needed quoting.
5810 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5811 was not being matched caselessly.
5813 24. Arranged for all hyphens in the exim.8 source to be escaped with
5816 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5817 properly. Recipient callout cache records were still being keyed to include
5818 the sender, even when use_sender was set false. This led to far more
5819 callouts that were necessary. The sender is no longer included in the key
5820 when use_sender is false.
5822 26. Added "control = submission" modifier to ACLs.
5824 27. Added the ${base62d: operator to decode base 62 numbers.
5826 28. dnsdb lookups can now access SRV records.
5828 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5829 the configuration file.
5831 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5832 (-v) mode. This makes the output for a verbose queue run more intelligible.
5834 31. Added a use_postmaster feature to recipient callouts.
5836 32. Added the $body_zerocount variable, containing the number of binary zero
5837 bytes in the message body.
5839 33. The time of last modification of the "new" subdirectory is now used as the
5840 "mailbox time last read" when there is a quota error for a maildir
5843 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5845 35. Added +ignore_unknown as a special item in host lists.
5847 36. Code for decoding IPv6 addresses in host lists is now included, even if
5848 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5849 address was recognized as an IP address, but was then not correctly decoded
5850 into binary, causing unexpected and incorrect effects when compared with
5857 1. Very minor rewording of debugging text in manualroute to say "list of
5858 hosts" instead of "hostlist".
5860 2. If verify=header_syntax was set, and a header line with an unqualified
5861 address (no domain) and a large number of spaces between the end of the
5862 name and the colon was received, the reception process suffered a buffer
5863 overflow, and (when I tested it) crashed. This was caused by some obsolete
5864 code that should have been removed. The fix is to remove it!
5866 3. When running in the test harness, delay a bit after writing a bounce
5867 message to get a bit more predictability in the log output.
5869 4. Added a call to search_tidyup() just before forking a reception process. In
5870 theory, someone could use a lookup in the expansion of smtp_accept_max_
5871 per_host which, without the tidyup, could leave open a database connection.
5873 5. Added the variables $recipient_data and $sender_data which get set from a
5874 lookup success in an ACL "recipients" or "senders" condition, or a router
5875 "senders" option, similar to $domain_data and $local_part_data.
5877 6. Moved the writing of debug_print from before to after the "senders" test
5880 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5881 problems for message scanning, either using a data ACL, or using
5882 local_scan() because the Received: header was not generated till after they
5883 were called (in order to set the time as the time of reception completion).
5884 I have revised the way this works. The header is now generated after the
5885 body is received, but before the ACL or local_scan() are called. After they
5886 are run, the timestamp in the header is updated.
5892 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5893 before starting a queue runner without re-exec. This happened only when
5894 deliver_drop_privilege was set or when the Exim user was set to root. The
5895 effect of the bug was that timeouts during subsequent deliveries caused
5896 crashes instead of being properly handled. The handler is now left at its
5897 default (and expected) setting.
5899 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5900 message, again when deliver_drop_privilege is set or Exim is run as root.
5901 The bug described in (1) was not present in this case, but the tidying up
5902 of the other signals was missing. I have made the two cases consistent.
5904 3. The ignore_target_hosts setting on a manualroute router was being ignored
5905 for hosts that were looked up using the /MX notation.
5907 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5910 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5911 operated on the sender address. After changing the $sender_address to <>
5912 for the sender address verify, Exim was re-instated it as the original
5913 (before rewriting) address, but remembering that it had rewritten it, so it
5914 wasn't rewriting it again. This bug also had the effect of breaking the
5915 sender address verification caching when the sender address was rewritten.
5917 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5918 This has been changed so that if the ip literal address matches
5919 ignore_target_hosts, the router declines.
5921 7. Added expansion conditions match_domain, match_address, and match_local_
5922 part (NOT match_host).
5924 8. The placeholder for the Received: header didn't have a length field set.
5926 9. Added code to Exim itself and to exim_lock to test for a specific race
5927 condition that could lead to file corruption when using MBX delivery. The
5928 issue is with the lockfile that is created in /tmp. If this file is removed
5929 after a process has opened it but before that process has acquired a lock,
5930 there is the potential for a second process to recreate the file and also
5931 acquire a lock. This could lead to two Exim processes writing to the file
5932 at the same time. The added code performs the same test as UW imapd; it
5933 checks after acquiring the lock that its file descriptor still refers to
5934 the same named file.
5936 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5937 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5940 11. Added the smtp_active_hostname option. If used, this will typically be made
5941 to depend on the incoming interface address. Because $interface_address is
5942 not set up until the daemon has forked a reception process, error responses
5943 that can happen earlier (such as "too many connections") no longer contain
5946 12. If an expansion in a condition on a "warn" statement fails because a lookup
5947 defers, the "warn" statement is abandoned, and the next ACL statement is
5948 processed. Previously this caused the whole ACL to be aborted.
5950 13. Added the iplsearch lookup type.
5952 14. Added ident_timeout as a log selector.
5954 15. Added tls_certificate_verified as a log selector.
5956 16. Added a global option tls_require_ciphers (compare the smtp transport
5957 option of the same name). This controls incoming TLS connections.
5959 17. I finally figured out how to make tls_require_ciphers do a similar thing
5960 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5961 before starting the TLS session.
5963 18. Tabs are now shown as \t in -bP output.
5965 19. If the log selector return_path_on_delivery was set, Exim crashed when
5966 bouncing a message because it had too many Received: header lines.
5968 20. If two routers both had headers_remove settings, and the first one included
5969 a superfluous trailing colon, the final name in the first list and the
5970 first name in the second list were incorrectly joined into one item (with a
5971 colon in the middle).
5977 1. Added -C and -D options to the exinext utility, mainly to make it easier
5978 to include in the automated testing, but these could be helpful when
5979 multiple configurations are in use.
5981 2. The exinext utility was not formatting the output nicely when there was
5982 an alternate port involved in the retry record key, nor when there was a
5983 message id as well (for retries that were specific to a specific message
5984 and a specific host). It was also confused by IPv6 addresses, because of
5985 the additional colons they contain. I have fixed the IPv4 problem, and
5986 patched it up to do a reasonable job for IPv6.
5988 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5989 delivery, the log line now contains "pipelined" if PIPELINING was used.
5991 4. An SMTP transport process used to panic and die if the bind() call to set
5992 an explicit outgoing interface failed. This has been changed; it is now
5993 treated in the same way as a connect() failure.
5995 5. A reference to $sender_host_name in the part of a conditional expansion
5996 that was being skipped was still causing a DNS lookup. This no longer
5999 6. The def: expansion condition was not recognizing references to header lines
6000 that used bh_ and bheader_.
6002 7. Added the _cache feature to named lists.
6004 8. The code for checking quota_filecount in the appendfile transport was
6005 allowing one more file than it should have been.
6007 9. For compatibility with Sendmail, the command line option
6015 and sets the incoming protocol and host name (for trusted callers). The
6016 host name and its colon can be omitted when only the protocol is to be set.
6017 Note the Exim already has two private options, -pd and -ps, that refer to
6018 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6019 "s", but I don't think that's a major issue.
6021 10. A number of refactoring changes to the code, none of which should affect
6024 (a) The number of logging options was getting close to filling up the
6025 32-bit word that was used as a bit map. I have split them into two classes:
6026 those that are passed in the argument to log_write(), and those that are
6027 only ever tested independently outside of that function. These are now in
6028 separate 32-bit words, so there is plenty of room for expansion again.
6029 There is no change in the user interface or the logging behaviour.
6031 (b) When building, for example, log lines, the code previously used a
6032 macro that called string_cat() twice, in order to add two strings. This is
6033 not really sufficiently general. Furthermore, there was one instance where
6034 it was actually wrong because one of the argument was used twice, and in
6035 one call a function was used. (As it happened, calling the function twice
6036 did not affect the overall behaviour.) The macro has been replaced by a
6037 function that can join an arbitrary number of extra strings onto a growing
6040 (c) The code for expansion conditions now uses a table and a binary chop
6041 instead of a serial search (which was left over from when there were very
6042 few conditions). Also, it now recognizes conditions like "pam" even when
6043 the relevant support is not compiled in: a suitably worded error message is
6044 given if an attempt is made to use such a condition.
6046 11. Added ${time_interval:xxxxx}.
6048 12. A bug was causing one of the ddress fields not to be passed back correctly
6049 from remote delivery subprocesses. The field in question was not being
6050 subsequently used, so this caused to problems in practice.
6052 13. Added new log selectors queue_time and deliver_time.
6054 14. Might have fixed a bug in maildirsizefile handling that threw up
6055 "unexpected character" debug warnings, and recalculated the data
6056 unnecessarily. In any case, I expanded the warning message to give more
6059 15. Added the message "Restricted characters in address" to the statements in
6060 the default ACL that block characters like @ and % in local parts.
6062 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6063 Three changes have been made:
6065 (a) There was a serious bug; a negative response to MAIL caused the whole
6066 recipient domain to be cached as invalid, thereby blocking all messages
6067 to all local parts at the same domain, from all senders. This bug has
6068 been fixed. The domain is no longer cached after a negative response to
6069 MAIL if the sender used is not empty.
6071 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6074 (c) A new callout option, "use_sender" has been added for people who want
6075 the modified behaviour.
6081 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6084 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6085 indeed breaks things for older releases.
6087 3. Added additional logging to the case where there is a problem reading data
6088 from a filter that is running in a subprocess using a pipe, in order to
6089 try to track down a specific problem.
6091 4. Testing facility fudge: when running in the test harness and attempting
6092 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6093 getting "No route to host". Convert this to a timeout.
6095 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6098 6. Some OS don't have socklen_t but use size_t instead. This affects the
6099 fifth argument of getsockopt() amongst other things. This is now
6100 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6101 can be set for individual OS. I have set it for SunOS5, OSF1, and
6102 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6103 some earlier ones do not.
6105 7. Change 4.30/15 was not doing the test caselessly.
6107 8. The standard form for an IPv6 address literal was being rejected by address
6108 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6109 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6110 this, as well as the form without the "IPv6" on the front (but only when
6111 address literals are enabled, of course).
6113 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6115 10. Exim crashed if a message with an empty sender address specified by -f
6116 encountered a router with an errors_to setting. This could be provoked only
6117 by a command such as
6121 where an empty string was supplied; "<>" did not hit this bug.
6123 11. Installed PCRE release 4.5.
6125 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6126 remained set. It is now erased.
6128 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6129 times from message ids (which are base 36 rather than the normal 62).
6131 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6132 were being counted as actual protocol errors, and logged if the log
6133 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6134 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6135 and DATA following a set of rejected RCPTs do not count as protocol errors.
6136 In other words, Exim assumes they were pipelined, though this may not
6137 actually be the case. Of course, in all cases the client gets an
6138 appropriate error code.
6140 15. If a lookup fails in an ACL condition, a message about the failure may
6141 be available; it is used if testing the ACL cannot continue, because most
6142 such messages specify what the cause of the deferral is. However, some
6143 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6144 that caused an old message to be retained and used if a later statement
6145 caused a defer, replacing the real cause of the deferral.
6147 16. If an IP address had so many PTR records that the DNS lookup buffer
6148 was not large enough to hold them, Exim could crash while trying to process
6149 the truncated data. It now detects and logs this case.
6151 17. Further to 4.21/58, another change has been made: if (and only if) the
6152 first line of a message (the first header line) ends with CRLF, a bare LF
6153 in a subsequent header line has a space inserted after it, so as not to
6154 terminate the header.
6156 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6157 unnecessarily, used atoi() instead of strtol(), and didn't check the
6158 termination when getting file sizes from file names by regex.
6160 19. Completely re-implemented the support for maildirsize files, in the light
6161 of a number of problems with the previous contributed implementation
6162 (4.30/29). In particular:
6164 . If the quota is zero, the maildirsize file is maintained, but no quota is
6167 . If the maildir directory does not exist, it is created before any attempt
6168 to write a maildirsize file.
6170 . The quota value in the file is just a cache; if the quota is changed in
6171 the transport, the new value overrides.
6173 . A regular expression is available for excluding directories from the
6176 20. The autoreply transport checks the characters in options that define the
6177 message's headers; it allows continued headers, but it was checking with
6178 isspace() after an embedded newline instead of explicitly looking for a
6181 21. If all the "regular" hosts to which an address was routed had passed their
6182 expiry times, and had not reached their retry times, the address was
6183 bounced, even if fallback hosts were defined. Now Exim should go on to try
6186 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6187 equivalent code in the SMTP transport. Some hosts send humungous responses
6188 to HELO/EHLO, more than 1024 it seems.
6190 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6191 gives compiler warnings in some environments. I've now done it "properly",
6194 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6195 (because of problems with the built-in one) was declared to return uschar *
6196 instead of char *, causing compiler failure.
6198 25. Fixed a file descriptor leak when processing alias/forward files.
6200 26. Fixed a minor format string issue in dbfn.c.
6202 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6204 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6205 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6208 29. When a sender address is verified, it is cached, to save repeating the test
6209 when there is more than one recipient in a message. However, when the
6210 verification involves a callout, it is possible for different callout
6211 options to be set for different recipients. It is too complicated to keep
6212 track of this in the cache, so now Exim always runs a verification when a
6213 callout is required, relying on the callout cache for the optimization.
6214 The overhead is duplication of the address routing, but this should not be
6217 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6218 to be verified with callout=postmaster, and the main callout worked but the
6219 postmaster check failed, the verification correctly failed. However, if a
6220 subsequent RCPT command asked for sender verification *without* the
6221 postmaster check, incorrect caching caused this verification also to fail,
6224 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6225 it was not caching the DNS options (qualify_single, search_parents) that
6226 were used when the lookup failed. A subsequent lookup with different
6227 options therefore always gave the same answer, though there were cases
6228 where it should not have. (Example: a "domains = !$mx_any" option on a
6229 dnslookup router: the "domains" option is always processed without any
6230 widening, but the router might have qualify_single set.) Now Exim uses the
6231 cached value only when the same options are set.
6233 32. Added John Jetmore's "exipick" utility to the distribution.
6235 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6236 than a timeout (e.g. a certificate is required, and is not provided), an
6237 Exim server now closes the connection immediately. Previously it waited for
6238 the client to close - but if the client is SSL, it seems that they each
6239 wait for each other, leading to a delay before one of them times out.
6241 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6242 maintained 0.8.x compatibility because I don't think many are using it, and
6243 it is clearly obsolete.
6245 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6248 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6249 client certificate was expired. A simple patch fixes this, though I don't
6250 understand the full logic of why the verify callback is called multiple
6253 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6254 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6255 which causes problems with some clients (such as the Certicom SSL Plus
6256 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6257 disables the coutermeasure allowing Eudora to connect."
6259 38. Exim was not checking that a write() to a log file succeeded. This could
6260 lead to Bad Things if a log got too big, in particular if it hit a file
6261 size limit. Exim now panics and dies if it cannot write to a log file, just
6262 as it does if it cannot open a log file.
6264 39. Modified OS/Makefile-Linux so that it now contains
6266 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6268 The two -D definitions ensure that Exim is compiled with large file
6269 support, which makes it possible to handle log files that are bigger than
6272 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6273 instance) a domain was checked against a named list that involved a lookup,
6274 causing $domain_data to be set, then another domain was checked against the
6275 same list, then the first domain was re-checked, the value of $domain_data
6276 after the final check could be wrong. In particular, if the second check
6277 failed, it could be set empty. This bug probably also applied to
6280 41. The strip_trailing_dot option was not being applied to the address given
6281 with the -f command-line option.
6283 42. The code for reading a message's header from the spool was incrementing
6284 $received_count, but never initializing it. This meant that the value was
6285 incorrect (doubled) while delivering a message in the same process in which
6286 it was received. In the most common configuration of Exim, this never
6287 happens - a fresh exec is done - but it can happen when
6288 deliver_drop_privilege is set.
6290 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6291 it now includes up to 150 characters of the unexpected data in the log
6294 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6295 and building data strings. The size of both of these buffers was 10 000
6296 bytes - far larger than anybody would *ever* want, thought I. Needless to
6297 say, somebody hit the limit. I have increased the maximum line length to
6298 20 000 and the maximum data length of concatenated lines to 100 000. I have
6299 also fixed two bugs, because there was no checking on these buffers. Tsk,
6300 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6301 buffer is too small.
6303 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6304 lsearch lookups. Now it does.
6306 46. When parsing a route_list item in a manualroute router, a fixed-length
6307 buffer was used for the list of hosts. I made this 1024 bytes long,
6308 thinking that nobody would ever have a list of hosts that long. Wrong.
6309 Somebody had a whole pile of complicated expansion conditions, and the
6310 string was silently truncated, leading to an expansion error. It turns out
6311 that it is easier to change to an unlimited length (owing to other changes
6312 that have happened since this code was originally written) than to build
6313 structure for giving a limitation error. The length of the item that
6314 expands into the list of hosts is now unlimited.
6316 47. The lsearch lookup could not handle data where the length of text line was
6317 more than 4095 characters. Such lines were truncated, leading to shortened
6318 data being returned. It should now handle lines of any length.
6320 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6321 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6324 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6325 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6326 no longer happens in comment lines. A list of replacements is now placed
6327 at the head of all of the source files, except those whose only change is
6328 to replace PERL_COMMAND in the very first #! line.
6330 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6331 messages on the queue, with a bottom-up merge sort, using code contributed
6332 by Michael Haardt. This should make operations like -bp somewhat faster on
6333 large queues. It won't affect queue runners, except when queue_run_in_order
6336 51. Installed eximstats 1.31 in the distribution.
6338 52. Added support for SRV lookups to the dnslookup router.
6340 53. If an ACL referred to $message_body or $message_body_end, the value was not
6341 reset for any messages that followed in the same SMTP session.
6343 54. The store-handling optimization for building very long strings was not
6344 differentiating between the different store pools. I don't think this
6345 actually made any difference in practice, but I've tidied it.
6347 55. While running the routers to verify a sender address, $sender_address
6348 was still set to the sender address. This is wrong, because when routing to
6349 send a bounce to the sender, it would be empty. Therefore, I have changed
6350 it so that, while verifying a sender address, $sender_address is set to <>.
6351 (There is no change to what happens when verifying a recipient address.)
6353 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6354 target A or AAAA records (if not already returned) without resetting the
6355 qualify_single or search_parents options of the DNS resolver. These are
6356 inappropriate in this case because the targets of MX and SRV records must
6357 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6358 target that, when qualified, matched something in the local domain. These
6359 two options are now turned off when doing these lookups.
6361 57. It seems that at least some releases of Reiserfs (which does not have the
6362 concept of a fixed number of inodes) returns zero and not -1 for the
6363 number of available inodes. This interacted badly with check_spool_inodes,
6364 which assumed that -1 was the "no such thing" setting. What I have done is
6365 to check that the total number of inodes is greater than zero before doing
6366 the test of how many are available.
6368 58. When a "warn" ACL statement has a log_message modifier, the message is
6369 remembered, and not repeated. This is to avoid a lot of repetition when a
6370 message has many recipients that cause the same warning to be written.
6371 However, Exim was preserving the list of already written lines for an
6372 entire SMTP session, which doesn't seem right. The memory is now reset if a
6373 new message is started.
6375 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6376 debugging output unless log_rewrite was also set.
6378 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6379 of (int)(handle) when we know that handle contains (void *)(-1).
6381 61. The Exim daemon panic-logs an error return when it closes the incoming
6382 connection. However "connection reset by peer" seems to be common, and
6383 isn't really an error worthy of noting specially, so that particular error
6386 62. When Exim is trying to find all the local interfaces, it used to panic and
6387 die if the ioctl to get the interface flags failed. However, it seems that
6388 on at least one OS (Solaris 9) it is possible to have an interface that is
6389 included in the list of interfaces, but for which you get a failure error
6390 for this call. This happens when the interface is not "plumbed" into a
6391 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6392 failure of the "get flags" call assumes that the interface is down.
6394 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6395 makes life easier for people who are doing arithmetic on fields extracted
6396 from dates, where you often get leading zeros that should not be
6397 interpreted as octal.
6399 64. Added qualify_domain to the redirect router, to override the global
6402 65. If a pathologically long header line contained very many addresses (the
6403 report of this problem mentioned 10 000) and each of them was rewritten,
6404 Exim could use up a very large amount of memory. (It kept on making new
6405 copies of the header line as it rewrote, and never released the old ones.)
6406 At the expense of a bit more processing, the header rewriting function has
6407 been changed so that it no longer eats memory in this way.
6409 66. The generation of the Received: header has been moved from the time that a
6410 message starts to be received, to the time that it finishes. The timestamp
6411 in the Received: header should now be very close to that of the <= log
6412 line. There are two side-effects of this change:
6414 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6415 logged header lines no longer include the local Received: line, because
6416 it has not yet been created. The same applies to a copy of the message
6417 that is returned to a non-SMTP sender when a message is rejected.
6419 (b) When a filter file is tested using -bf, no additional Received: header
6420 is added to the test message. After some thought, I decided that this
6423 This change does not affect the value of $received_for. It is still set
6424 after address rewriting, but before local_scan() is called.
6426 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6428 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6429 gave an unhelpful panic error message, and a defer error. I have managed to
6430 change this behaviour so that it now rejects any supplied certificate,
6431 which seems right, as the list of acceptable certificates is empty.
6433 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6434 gave an unhelpful defer error. I have not managed to make this reject any
6435 supplied certificates, but the error message it gives is "no certificate
6436 supplied", which is not helpful.
6438 70. exigrep's output now also includes lines that are not associated with any
6439 message, but which match the given pattern. Implemented by a patch from
6440 Martin Sluka, which also tidied up the Perl a bit.
6442 71. Recipient callout verification, like sender verification, was using <> in
6443 the MAIL FROM command. This isn't really the right thing, since the actual
6444 sender may affect whether the remote host accepts the recipient or not. I
6445 have changed it to use the actual sender in the callout; this means that
6446 the cache record is now keyed on a recipient/sender pair, not just the
6447 recipient address. There doesn't seem to be a real danger of callout loops,
6448 since a callout by the remote host to check the sender would use <>.
6449 [SEE ABOVE: changed after hitting problems.]
6451 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6452 temporary errors. However, in the case of such a code being given after
6453 the end of a data transmission (i.e. after ".") Exim was failing to write
6454 a retry record for the message. (Yes, there was some broken host that was
6455 actually sending 8xx at this point.)
6457 73. An unknown lookup type in a host list could cause Exim to panic-die when
6458 the list was checked. (An example that provoked this was putting <; in the
6459 middle of a list instead of at the start.) If this happened during a DATA
6460 ACL check, a -D file could be left lying around. This kind of configuration
6461 error no longer causes Exim to die; instead it causes a defer error. The
6462 incident is still logged to the main and panic logs.
6464 74. Buglet left over from Exim 3 conversion. The message "too many messages
6465 in one connection" was written to the rejectlog but not the mainlog, except
6466 when address rewriting (yes!) was being logged.
6468 75. Added write_rejectlog option.
6470 76. When a system filter was run not as root (that is, when system_filter_user
6471 was set), the values of the $n variables were not being returned to the
6472 main process; thus, they were not subsequently available in the $sn
6475 77. Added +return_path_on_delivery log selector.
6477 78. A connection timeout was being treated differently from recipients deferred
6478 when testing hosts_max_try with a message that was older than the host's
6479 retry timeout. (The host should not be counted, thus allowing all hosts to
6480 be tried at least once before bouncing.) This may have been the cause of an
6481 occasionally reported bug whereby a message would remain on the queue
6482 longer than the retry timeout, but would be bounced if a delivery was
6483 forced. I say "may" because I never totally pinned down the problem;
6484 setting up timeout/retry tests is difficult. See also the next item.
6486 79. The ultimate address timeout was not being applied to errors that involved
6487 a combination of host plus message (for example, a timeout on a MAIL
6488 command). When an address resolved to a number of possible hosts, and they
6489 were not all tried for each delivery (e.g. because of hosts_max_try), a
6490 message could remain on the queue longer than the retry timeout.
6492 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6495 81. Fixed an obscure SMTP outgoing bug which required at least the following
6496 conditions: (a) there was another message waiting for the same server;
6497 (b) the server returned 5xx to all RCPT commands in the first message so
6498 that the message was not completed; (c) the server dropped the connection
6499 or gave a negative response to the RSET that Exim sends to abort the
6500 transaction. The observed case was a dropped connection after DATA that had
6501 been sent in pipelining mode. That is, the server had advertised PIPELINING
6502 but was not implementing it correctly. The effect of the bug was incorrect
6503 behaviour, such as trying another host, and this could lead to a crash.
6509 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6510 and daemon.c were passed as pointers to ints; they should have been
6511 pointers to socklen_t variables (which are typically unsigned ints).
6513 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6516 3. Fixed a really odd bug that affected only the testing scheme; patching a
6517 certain fixed string in the binary changed the value of another string that
6518 happened to be identical to the end of the original first string.
6520 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6521 name", it returns that address as the IP address. On some operating
6522 systems (e.g. Solaris), it also passes back the IP address string as the
6523 "host name". However, on others (e.g. Linux), it passes back an empty
6524 string. Exim wasn't checking for this, and was changing the host name to an
6525 empty string, assuming it had been canonicalized.
6527 5. Although rare, it is permitted to have more than one PTR record for a given
6528 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6529 all the names associated with an address, because they do in Solaris.
6530 However, it seems that they do not in Linux for data that comes from the
6531 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6532 I found this out when I moved to a new Linux workstation and tried to run
6533 the Exim test suite.
6535 To get round this problem I have changed the code so that it now does its
6536 own call to the DNS to look up PTR records when searching for a host name.
6537 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6538 addresses that are only in /etc/hosts are still found.
6540 This behaviour is, however, controlled by an option called host_lookup_
6541 order, which defaults to "bydns:byaddr". If people want to use the other
6542 order, or indeed, just use one or the other means of lookup, they can
6543 specify it in this variable.
6545 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6546 some operating systems, this comes back from gethostbyaddr() as an empty
6547 string, and this is what Exim used to test for. However, it seems that in
6548 other systems, "." is yielded. Exim now tests for this case too.
6550 7. The values of check_spool_space and check_log_space are now held internally
6551 as a number of kilobytes instead of an absolute number of bytes. If a
6552 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6553 kilobyte. This means that much larger values can be stored.
6555 8. Exim monitor: an attempt to get the action menu when not actually pointing
6556 at a message produces an empty menu entitled "No message selected". This
6557 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6558 no entries in it ("Shell widget menu has zero width and/or height"). So I
6559 have added a single, blank menu entry in this case.
6561 9. Added ${quote_local_part.
6563 10. MIME decoding is now applied to the contents of Subject: header lines when
6566 11. Now that a reference to $sender_host_address automatically causes a reverse
6567 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6568 host lookup before query-style lookups in lists that might use this
6569 variable. This has therefore been abolished, and the "net-" prefix is no
6570 longer necessary for query-style lookups.
6572 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6573 have been a typo for LFLAGS, so it has been changed.
6575 13. The install script calls Exim with "-C /dev/null" in order to find the
6576 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6577 to be output. However, since Exim outputs its version number before the
6578 error, it didn't break the script. It just looked ugly. I fixed this by
6579 always allowing "-C /dev/null" if the caller is root.
6581 14. Ignore overlarge ACL variable number when reading spool file - insurance
6582 against a later release with more variables having written the file.
6584 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6585 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6586 this, as well as the form without the "IPv6" on the front.
6588 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6589 OS/Makefile-Darwin file.
6591 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6592 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6594 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6595 variable. However, some OS use ints for inodes. I've added cast to long int
6596 to get rid of the compiler warning.
6598 19. I had forgotten to lock out "/../" in configuration file names when
6599 ALT_CONFIG_PREFIX was set.
6601 20. Routers used for verification do not need to specify transports. However,
6602 if such a router generated a host list, and callout was configured, Exim
6603 crashed, because it could not find a port number from the (non-existent)
6604 transport. It now assumes port 25 in this circumstance.
6606 21. Added the -t option to exigrep.
6608 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6609 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6610 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6611 them is set, LOOKUP_LSEARCH is forced.
6613 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6615 24. Added sender and host information to the "rejected by local_scan()" log
6616 line; previously there was no indication of these.
6618 25. Added .include_if_exists.
6620 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6621 Linux. It turns out that not all file systems support this. Apparently some
6622 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6623 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6624 not supported on the file descriptor, is now ignored when Exim is trying to
6625 sync a directory. This applies only to Linux.
6627 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6629 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6630 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6633 29. Added support for maildirsize files from supplied patch (modified a bit).
6635 30. The use of :fail: followed by an empty string could lead Exim to respond to
6636 sender verification failures with (e.g.):
6638 550 Verification failed for <xxx>
6639 550 Sender verify failed
6641 where the first response line was missing the '-' that indicates it is not
6642 the final line of the response.
6644 31. The loop for finding the name of the user that called Exim had a hardwired
6645 limit of 10; it now uses the value of finduser_retries, which is used for
6646 all other user lookups.
6648 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6651 33. Exim was neglecting to zero errno before one call of strtol() when
6652 expanding a string and expecting an integer value. On some systems this
6653 resulted in spurious "integer overflow" errors. Also, it was casting the
6654 result into an int without checking.
6656 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6657 did not work. The code looks as if it has *never* worked, though it appears
6658 to have been documented since at least release 1.62. I have made it work.
6660 35. The "timeout_DNS" error in retry rules, also documented since at least
6661 1.62, also never worked. As it isn't clear exactly what this means, and
6662 clearly it isn't a major issue, I have abolished the feature by treating it
6663 as "timeout", and writing a warning to the main and panic logs.
6665 36. The display of retry rules for -brt wasn't always showing the error code
6668 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6669 timeout_connect_A, timeout_connect_MX.
6671 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6672 to the empty sender.
6674 39. The daemon was not analysing the content of -oX till after it had closed
6675 stderr and disconnected from the controlling terminal. This meant that any
6676 syntax errors were only noted on the panic log, and the return code from
6677 the command was 0. By re-arranging the code a little, I've made the
6678 decoding happen first, so such errors now appear on stderr, and the return
6679 code is 1. However, the actual setting up of the sockets still happens in
6680 the disconnected process, so errors there are still only recorded on the
6683 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6684 connections (as happens on some IP stacks) was logged at start up time as
6685 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6686 it from "IPv6 and IPv4", which means that two separate sockets are being
6689 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6690 says whether AF_INET or AF_INET6 was passed as an argument.
6692 42. Exiwhat output was messed up when time zones were included in log
6695 43. Exiwhat now gives more information about the daemon's listening ports,
6696 and whether -tls-on-connect was used.
6698 44. The "port" option of the smtp transport is now expanded.
6700 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6701 silently ignored. Now an error message is written to the main and panic
6704 46. There's a new ACL modifier called "logwrite" which writes to a log file
6705 as soon as it is encountered.
6707 47. Added $local_user_uid and $local_user_gid at routing time.
6709 48. Exim crashed when trying to verify a sender address that was being
6712 49. Exim was recognizing only a space character after ".include". It now also
6713 recognizes a tab character.
6715 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6716 extracting the relevant information from the specification. The man page no
6717 longer contains scrambled data for the -d option, and I've added a section
6718 at the front about calling Exim under different names.
6720 51. Added "extra_headers" argument to the "mail" command in filter files.
6722 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6725 53. Installed eximstats 1.29.
6727 54. Added transport_filter_timeout as a generic transport option.
6729 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6730 Cc: header lines. This was required by RFC 822, but it not required by RFC
6733 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6734 incoming messages that did not have them. Now it does so only if the
6735 message originates locally, that is, if there is no associated remote host
6736 address. When Resent- header lines are present, this applies to the Resent-
6737 lines rather than the non-Resent- lines.
6739 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6740 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6742 58. Messages for configuration errors now include the name of the main
6743 configuration file - useful now that there may be more than one file in a
6744 list (.included file names were always shown).
6746 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6747 for those rare installations that do not start the daemon as root or run it
6748 setuid root. I've cut out the call to initgroups() if the daemon is not
6751 60. The Exim user and group can now be bound into the binary as text strings
6752 that are looked up at the start of Exim's processing.
6754 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6756 62. Added $mailstore_basename variable.
6758 63. Installed patch to sieve.c from Michael Haardt.
6760 64. When Exim failed to open the panic log after failing to open the main log,
6761 the original message it was trying to log was written to stderr and debug
6762 output, but if they were not available (the usual case in production), it
6763 was lost. Now it is written to syslog before the two lines that record the
6764 failures to open the logs.
6766 65. Users' Exim filters run in subprocesses under the user's uid. It is
6767 possible for a "deliver" command or an alias in a "personal" command to
6768 provoke an address rewrite. If logging of address rewriting is configured,
6769 this fails because the process is not running as root or exim. There may be
6770 a better way of dealing with this, but for the moment (because 4.30 needs
6771 to be released), I have disabled address rewrite logging when running a
6772 filter in a non-root, non-exim process.
6778 1. The buildconfig auxiliary program wasn't quoting the value set for
6779 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6780 not defined. This bug was masked in 4.22 by the effect that was fixed in
6783 2. Some messages that were rejected after a message id was allocated were
6784 shown as "incomplete" by exigrep. It no longer does this for messages that
6785 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6787 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6788 have allow_domain_literals set, the ID did not get logged in the <= line.
6789 Domain literals are now always recognized in Message-ID: header lines.
6791 4. The first argument for a ${extract expansion item is the key name or field
6792 number. Leading and trailing spaces in this item were not being ignored,
6793 causing some misleading effects.
6795 5. When deliver_drop_privilege was set, single queue runner processes started
6796 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6797 same command in the process it spins off) were not dropping privilege.
6799 6. When the daemon running as "exim" started a queue runner, it always
6800 re-executed Exim in the spun-off process. This is a waste of effort when
6801 deliver_drop_privilege is set. The new process now just calls the
6802 queue-runner function directly.
6808 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6811 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6812 ignored. Though the use of -oP was forcing the writing of a pid file, it
6813 was always written to the default place.
6815 3. If the message "no IP address found for host xxxx" is generated during
6816 incoming verification, it is now followed by identification of the incoming
6817 connection (so you can more easily find what provoked it).
6819 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6821 5. Added some features to "harden" Exim a bit more against certain attacks:
6823 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6824 be put in Local/Makefile. This is like the never_users runtime option,
6825 but it cannot be overridden. The default setting is "root".
6827 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6828 prefix string with which any file named in a -C command line option
6831 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6832 is retained for -C and -D only if the caller of Exim is root. Without
6833 it, the exim user may also use -C and -D and retain privilege.
6835 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6836 command line option is disabled.
6838 6. Macro names set by the -D option must start with an upper case letter, just
6839 like macro names defined in the configuration file.
6841 7. Added "dereference=" facility to LDAP.
6843 8. Two instances of the typo "uknown" in the source files are fixed.
6845 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6846 the Configure-Makefile script screwed up while processing it.
6848 10. Incorporated PCRE 4.4.
6850 11. The SMTP synchronization check was not operating right at the start of an
6851 SMTP session. For example, it could not catch a HELO sent before the client
6852 waited for the greeting. There is now a check for outstanding input at the
6853 point when the greeting is written. Because of the duplex, asynchronous
6854 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6855 way, but not yet received, when the check is performed.
6857 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6858 on TCP/IP sockets, because this apparently causes some broken clients to
6861 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6862 unchanged) from the Cygwin maintainer.
6864 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6865 was supported instead of testing for mbx. Effectively a typo.
6867 15. The spa authenticator server code was not checking that the input it
6868 received was valid base64.
6870 16. The debug output line for the "set" modifier in ACLs was not showing the
6871 name of the variable that was being set.
6873 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6875 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6876 lookup on its own. Something else was needed to trigger the lookup. For
6877 example, a match in host_lookup or the need for a host name in a host list.
6878 Now, if $sender_host_name is referenced and the host name has not yet been
6879 looked up, a lookup is performed. If the lookup fails, the variable remains
6880 empty, and $host_lookup_failed is set to "1".
6882 19. Added "eqi" as a case-independent comparison operator.
6884 20. The saslauthd authentication condition could segfault if neither service
6885 nor realm was specified.
6887 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6888 error message that was logged was misleading, and incoming SMTP
6889 connections were dropped. The message is now more accurate, and temporary
6890 errors are given to SMTP connections.
6892 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6893 (except RSET) with a 421 error, until QUIT is received. However, it was
6894 failing to send a response to QUIT.
6896 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6897 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6898 ACL is now run after the helo_try_verify_hosts code.
6900 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6901 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6902 case-independent, so other case variants are also recognized). Apparently
6903 some systems use these upper case variants.
6905 25. If more than two messages were waiting for the same host, and a transport
6906 filter was specified for the transport, Exim sent two messages over the
6907 same TCP/IP connection, and then failed with "socket operation on non-
6908 socket" when it tried to send the third.
6910 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6912 27. The extern definition of crypt16() in expand.c was not being excluded when
6913 the OS had its own crypt16() function.
6915 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6916 as a preferred synonym for return_size_limit, both as an option and as an
6919 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6921 30. Changed the default configuration ACL to relax the local part checking rule
6922 for addresses that are not in any local domains. For these addresses,
6923 slashes and pipe symbols are allowed within local parts, but the sequence
6924 /../ is explicitly forbidden.
6926 31. SPA server authentication was not clearing the challenge buffer before
6929 32. log_message in a "warn" ACL statement was writing to the reject log as
6930 well as to the main log, which contradicts the documentation and doesn't
6931 seem right (because no rejection is happening). So I have stopped it.
6933 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6934 However, I am unable to do any testing of this.
6936 34. Fixed an infelicity in the appendfile transport. When checking directories
6937 for a mailbox, to see if any needed to be created, it was accidentally
6938 using path names with one or more superfluous leading slashes; tracing
6939 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6941 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6942 contents are added to the log line that is written for every discarded
6943 recipient. (Previously a log_message setting was ignored.)
6945 36. The ${quote: operator now quotes the string if it is empty.
6947 37. The install script runs exim in order to find its version number. If for
6948 some reason other than non-existence or emptiness, which it checks, it
6949 could not run './exim', it was installing it with an empty version number,
6950 i.e. as "exim-". This error state is now caught, and the installation is
6953 38. An argument was missing from the function that creates an error message
6954 when Exim fails to connect to the socket for saslauthd authentication.
6955 This could cause Exim to crash, or give a corrupted message.
6957 39. Added isip, isip4, and isip6 to ${if conditions.
6959 40. The ACL variables $acl_xx are now saved with the message, and can be
6960 accessed later in routers, transports, and filters.
6962 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6963 strings in the file are not string-expanded.
6965 42. If a MAIL command specified a SIZE value that was too large to fit into an
6966 int variable, the check against message_size_limit failed. Such values are
6967 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6968 day this will have to be increased, but I don't think I want to be around
6969 when emails are that large.
6976 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6977 iconv() is not standard in FreeBSD.
6979 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6980 IPv6 enabled. The observed symptom was a segmentation fault on return from
6981 the function os_common_find_running_interfaces() in src/os.c.
6983 3. In the check_special_case() function in daemon.c I had used "errno" as an
6984 argument name, which causes warnings on some systems. This was basically a
6985 typo, since it was named "eno" in the comments!
6987 4. The code that waits for the clock to tick (at a resolution of some fraction
6988 of a second) so as to ensure message-id uniqueness was always waiting for
6989 at least one whole tick, when it could have waited for less. [This is
6990 almost certainly not relevant at current processor speeds, where it is
6991 unlikely to ever wait at all. But we try to future-proof.]
6993 5. The function that sleeps for a time interval that includes fractions of a
6994 second contained a race. It did not block SIGALRM between setting the
6995 timer, and suspending (a couple of lines later). If the interval was short
6996 and the sigsuspend() was delayed until after it had expired, the suspension
6997 never ended. On busy systems this could lead to processes getting stuck for
7000 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7001 process, before it forks any delivery processes. The open lookup caching
7002 mechanism meant that the open file or database connection was passed into
7003 the delivery process. The problem was that delivery processes always tidy
7004 up cached lookup data. This could cause a problem for the next delivery
7005 process started by the queue runner, because the external queue runner
7006 process does not know about the closure. So the next delivery process
7007 still has data in the lookup cache. In the case of a file lookup, there was
7008 no problem because closing a file descriptor in a subprocess doesn't affect
7009 the parent. However, if the lookup was caching a connection to a database,
7010 the connection was closed, and the second delivery process was likely to
7011 see errors such as "PGSQL: query failed: server closed the connection
7012 unexpectedly". The problem has been fixed by closing all cached lookups
7013 in a queue runner before running a delivery process.
7015 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7016 seem to have the "const" qualifier which it has on other OS. I've
7019 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7020 *of the same type* (client or server) with the same public name that an
7021 error should be diagnosed.
7023 9. When Exim looked up a host name for an IP address, but failed to find the
7024 original IP address when looking up the host name (a safety check), it
7025 output the message "<ip address> does not match any IP for NULL", which was
7026 confusing, to say the least. The bug was that the host name should have
7027 appeared instead of "NULL".
7029 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7030 user that is built into the binary, and the -C or -D options is used, root
7031 privilege is dropped before the configuration file is read. In addition,
7032 logging is switched to stderr instead of the normal log files. If the
7033 configuration then re-defines the Exim user, the unprivileged environment
7034 is probably not what is expected, so Exim logs a panic warning message (but
7037 However, if deliver_drop_privilege is set, the unprivileged state may well
7038 be exactly what is intended, so the warning has been cut out in that case,
7039 and Exim is allowed to try to write to its normal log files.
7045 1. smtp_return_error_details was not giving details for temporary sender
7046 or receiver verification errors.
7048 2. Diagnose a configuration error if two authenticators have the same public
7051 3. Exim used not to create the message log file for a message until the first
7052 delivery attempt. This could be confusing when incoming messages were held
7053 for policy or load reasons. The message log file is now created at the time
7054 the message is received, and an initial "Received" line is written to it.
7056 4. The automatically generated man page for command line options had a minor
7057 bug that caused no ill effects; however, a more serious problem was that
7058 the procedure for building the man page automatically didn't always
7059 operate. Consequently, release 4.20 contains an out-of-date version. This
7060 shouldn't happen again.
7062 5. When building Exim with embedded Perl support, the script that builds the
7063 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7064 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7066 6. The freeze_tell option was not being used for messages that were frozen on
7067 arrival, either by an ACL or by local_scan().
7069 7. Added the smtp_incomplete_transaction log selector.
7071 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7072 was accepting AUTH without a new EHLO.
7074 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7075 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7078 10. Logging of TCP/IP connections (when configured) now happens in the main
7079 daemon process instead of the child process, so that the TCP/IP connection
7080 count is more accurate (but it can never be perfect).
7082 11. The use of "drop" in a nested ACL was not being handled correctly in the
7083 outer ACL. Now, if condition failure induced by the nested "drop" causes
7084 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7085 or "require"), the connection is dropped.
7087 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7088 that yield "discard" can now be used with an "accept" or a "discard" verb,
7089 but an error is generated for any others (because I can't see a useful way
7090 to define what should happen).
7092 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7093 are now processed in the same way as lines in the Exim configuration file.
7094 In particular, continuation lines are supported.
7096 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7098 15. Added -ti meaning -t -i.
7100 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7101 domains, and warn by logging if others are found.
7103 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7104 returned from GIFCONF when Exim is trying to find the list of interfaces on
7105 a host. The code in os.c has been modified to copy each ifreq to an aligned
7106 structure in all cases.
7108 Also, in some cases, the returned ifreq's were being copied to a 'struct
7109 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7110 means the last couple of bytes of an IPv6 address could be chopped if the
7111 ifreq contained only a normal sockaddr (14 bytes storage).
7113 18. Named domain lists were not supported in the hosts_treat_as_local option.
7114 An entry such as +xxxx was not recognized, and was treated as a literal
7117 19. Ensure that header lines added by a DATA ACL are included in the reject log
7118 if the ACL subsequently rejects the message.
7120 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7121 MD5 (which is deprecated).
7123 21. When testing a filter file using -bf, Exim was writing a message when it
7124 took the sender from a "From " line in the message, but it was not doing so
7125 when it took $return_path from a Return-Path: header line. It now does.
7127 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7128 with a valid header line field name (a series of printing characters
7129 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7131 23. Changed "disc" in the source to "disk" to conform to the documentation and
7132 the book and for uniformity.
7134 24. Ignore Sendmail's -Ooption=value command line item.
7136 25. When execve() failed while trying to run a command in a pipe transport,
7137 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7138 could be confused with a return value of 69 from the command itself. This
7139 has been changed to 127, the value the shell returns if it is asked to run
7140 a non-existent command. The wording for the related log line suggests a
7141 non-existent command as the problem.
7143 26. If received_header_text expands to an empty string, do not add a Received:
7144 header line to the message. (Well, it adds a token one on the spool, but
7145 marks it "old" so that it doesn't get used or transmitted.)
7147 27. Installed eximstats 1.28 (addition of -nt option).
7149 28. There was no check for failure on the call to getsockname() in the daemon
7150 code. This can fail if there is a shortage of resources on the system, with
7151 ENOMEM, for example. A temporary error is now given on failure.
7153 29. Contrary to the C standard, it seems that in some environments, the
7154 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7155 program. Exim now does this explicitly; it affects the formatting of
7156 timestamps using strftime().
7158 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7159 complaints. I've now initialized all the variables, to avoid this.
7161 32. Header lines added by a system filter were not being "seen" during
7162 transport-time rewrites.
7164 33. The info_callback() function passed to OpenSSL is set up with type void
7165 (*)(SSL *, int, int), as described somewhere. However, when calling the
7166 function (actually a macro) that sets it up, the type void(*)() is
7167 expected. I've put in a cast to prevent warnings from picky compilers.
7169 34. If a DNS black list lookup found a CNAME record, but there were no A
7170 records associated with the domain it pointed at, Exim crashed.
7172 35. If a DNS black list lookup returned more than one A record, Exim ignored
7173 all but the first. It now scans all returned addresses if a particular IP
7174 value is being sought. In this situation, the contents of the
7175 $dnslist_value variable are a list of all the addresses, separated by a
7178 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7179 to accept a host name and all its aliases if the forward lookup for any of
7180 them yielded the IP address of the incoming connection. Now it accepts only
7181 those names whose forward lookup yields the correct IP address. Any other
7182 names are discarded. This closes a loophole whereby a rogue DNS
7183 administrator could create reverse DNS records to break through a
7184 wildcarded host restriction in an ACL.
7186 37. If a user filter or a system filter that ran in a subprocess used any of
7187 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7188 the wrong values were passed to the pipe command ($thisaddress had the
7189 value of $0, $0 had the value of $1, etc). This bug was introduced by
7190 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7192 38. Improved the line breaking for long SMTP error messages from ACLs.
7193 Previously, if there was no break point between 40 and 75 characters, Exim
7194 left the rest of the message alone. Two changes have been made: (a) I've
7195 reduced the minimum length to 35 characters; (b) if it can't find a break
7196 point between 35 and 75 characters, it looks ahead and uses the first one
7197 that it finds. This may give the occasional overlong line, but at least the
7198 remaining text gets split now.
7200 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7201 file descriptors might be low, and that setting 1000 would always raise it.
7202 It turns out that in some environments, the limit is already over 1000 and
7203 that lowering it causes trouble. So now Exim takes care not to decrease it.
7205 40. When delivering a message, the value of $return_path is set to $sender_
7206 address at the start of routing (routers may change the value). By an
7207 oversight, this default was not being set up when an address was tested by
7208 -bt or -bv, which affected the outcome if any router or filter referred to
7211 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7212 log only when "message" or "log_message" are set. However, if one of the
7213 conditions was an address verification, or a call to a nested ACL, the
7214 messages generated by the underlying test were being passed through. This
7215 no longer happens. The underlying message is available in $acl_verify_
7216 message for both "message" and "log_message" expansions, so it can be
7217 passed through if needed.
7219 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7220 new expansion $bh_ to give the encoded byte string without charset
7221 translation. Translation happens only if iconv() is available; HAVE_ICONV
7222 indicates this at build time. HEADERS_CHARSET gives the charset to
7223 translate to; headers_charset can change it in the configuration, and
7224 "headers charset" can change it in an individual filter file.
7226 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7227 that creates RFC 2047 encoded "words" labels them as that charset instead
7228 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7229 expansion operator; (ii) when Exim creates a From: line for a local
7230 message; (iii) when a header line is rewritten to include a "phrase" part.
7232 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7233 buggy, causing it to skip the first lines of messages whose message ID
7234 ended in 'D'. This would not have bitten before Exim release 4.14, because
7235 message IDs were unlikely to end in 'D' before then. The effect was to have
7236 incorrect size information for certain domains.
7238 45. #include "config.h" was missing at the start of the crypt16.c module. This
7239 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7242 46. If there was a timeout during a "random" callout check, Exim treated it as
7243 a failure of the random address, and carried on sending RSET and the real
7244 address. If the delay was just some slowness somewhere, the response to the
7245 original RCPT would be taken as a response to RSET and so on, causing
7246 mayhem of various kinds.
7248 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7249 when I implemented it. It didn't allow for the fact that some option values
7250 may legitimately be negative (e.g. size_addition), and it didn't even do
7251 the right test for positive values.
7253 48. Domain names in DNS records are case-independent. Exim always looks them up
7254 in lower case. Some resolvers return domain names in exactly the case they
7255 appear in the zone file, that is, they may contain uppercase letters. Not
7256 all resolvers do this - some return always lower case. Exim was treating a
7257 change of case by a resolver as a change of domain, similar to a widening
7258 of a domain abbreviation. This triggered its re-routing code and so it was
7259 trying to route what was effectively the same domain again. This normally
7260 caused routing to fail (because the router wouldn't handle the domain
7261 twice). Now Exim checks for this case specially, and just changes the
7262 casing of the domain that it ultimately uses when it transmits the message
7265 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7268 50. If a filter generated a file delivery with a non-absolute name (possible if
7269 no home directory exists for the router), the forbid_file option was not
7272 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7273 the existing equality matching.
7275 52. Exim was using ints instead of ino_t variables in some places where it was
7276 dealing with inode numbers.
7278 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7279 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7280 called TMPDIR, and if it finds it is different, it changes its value.
7282 54. The smtp_printf() function is now made available to local_scan() so
7283 additional output lines can be written before returning. There is also an
7284 smtp_fflush() function to enable the detection of a dropped connection.
7285 The variables smtp_input and smtp_batched_input are exported to
7288 55. Changed the default runtime configuration: the message "Unknown user"
7289 has been removed from the ACL, and instead placed on the localuser router,
7290 using the cannot_route_message feature. This means that any verification
7291 failures that generate their own messages won't get overridden. Similarly,
7292 the "Unrouteable address" message that was in the ACL for unverifiable
7293 relay addresses has also been removed.
7295 56. Added hosts_avoid_esmtp to the smtp transport.
7297 57. The exicyclog script was not checking for the esoteric option
7298 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7299 will work only if exicyclog is run under the appropriate euid.
7301 58. Following a discussion on the list, the rules by which Exim recognises line
7302 endings on incoming messages have been changed. The -dropcr and drop_cr
7303 options are now no-ops, retained only for backwards compatibility. The
7304 following line terminators are recognized: LF CRLF CR. However, special
7305 processing applies to CR:
7307 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7308 nor a local message in the state where . is a terminator.
7310 (ii) If a bare CR is encountered in a header line, an extra space is added
7311 after the line terminator so as not to end the header. The reasoning
7312 behind this is that bare CRs in header lines are most likely either
7313 to be mistakes, or people trying to play silly games.
7315 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7316 was being incorrectly given as 18 bytes larger than it should have been.
7317 This is a VOB (very old bug).
7319 60. This may never have affected anything current, but just in case it has:
7320 When the local host is found other than at the start of a list of hosts,
7321 the local host, those with the same MX, and any that follow, are discarded.
7322 When the list in question was part of a longer list of hosts, the following
7323 hosts (not currently being processed) were also being discarded. This no
7324 longer happens. I'm not sure if this situation could ever has previously
7327 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7330 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7331 header. This is something that is recommended in a new Internet Draft, and
7332 is something that is documented as being done by Sendmail. There are two
7333 possible values. For messages generated by the autoreply transport, Exim
7336 Auto-Submitted: auto-replied
7338 whereas for all other generated messages (e.g. bounces) it adds
7340 Auto-Submitted: auto-generated
7342 63. The "personal" condition in filters now includes a test for the
7343 Auto-Submitted: header. If it contains the string "auto-" the message it
7344 not considered personal.
7346 64. Added rcpt_include_affixes as a generic transport option.
7348 65. Added queue_only_override (default true).
7350 66. Added the syslog_duplication option.
7352 67. If what should have been the first header line of a message consisted of
7353 a space followed by a colon, Exim was mis-interpreting it as a header line.
7354 It isn't of course - it is syntactically invalid and should therefore be
7355 treated as the start of the message body. The misbehaviour could have
7356 caused a number of strange effects, including loss of data in subsequent
7357 header lines, and spool format errors.
7359 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7360 client host had authenticated. This control can now be exercised by an ACL
7361 for more flexibility.
7363 69. By default, callouts do not happen when testing with -bh. There is now a
7364 variant, -bhc, which does actually run the callout code, including
7365 consulting and updating the callout cache.
7367 70. Added support for saslauthd authentication, courtesy of Alexander
7370 71. If statvfs() failed on the spool or log directories while checking their
7371 size for availability, Exim confusingly gave the error "space shortage".
7372 Furthermore, in debugging mode it crashed with a floating point exception.
7373 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7374 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7375 serious problem, Exim now writes to the main and panic logs when this
7376 happens, with details of the failure. It then refuses to accept the
7377 incoming message, giving the message "spool directory problem" or "log
7378 directory problem" with a 421 code for SMTP messages.
7380 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7381 0, 1, and 2 exist, because some OS complain for execs without them (see
7382 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7383 descriptors. However, the code omitted to check that the open succeeded,
7384 causing mysterious errors if for some reason the permissions on /dev/null
7385 got screwed. Now Exim writes a message to the main and panic logs, and
7386 bombs out if it can't open /dev/null.
7388 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7389 interact so that it is all more flexible. It is supposed to remain
7390 backwards compatible. Also added extra_local_interfaces.
7392 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7393 to bomb out with an assertion failure - to the client this appears as a
7394 connection drop. This problem occurs in the part of the code that was taken
7395 from the Samba project. Fortunately, the assertion is in a very simple
7396 function, so I have fixed this by reproducing the function inline in the
7397 one place where it is called, and arranging for authentication to fail
7398 instead of killing the process with assert().
7400 75. The SPA client code was not working when the server requested OEM rather
7401 than Unicode encoding.
7403 76. Added code to make require_files with a specific uid setting more usable in
7404 the case where statting the file as root fails - usually a non-root-mounted
7405 NFS file system. When this happens and the failure is EACCES, Exim now
7406 forks a subprocess and does the per-uid checking as the relevant uid.
7408 77. Added process_log_path.
7410 78. If log_file_path was not explicitly set, a setting of check_log_space or
7411 check_log_inodes was ignored.
7413 79. If a space check for the spool or log partitions fails, the incident is now
7414 logged. Of course, in the latter case the data may get lost...
7416 80. Added the %p formatting code to string_format() so that it can be used to
7417 print addresses in debug_print(). Adjusted all the address printing in the
7418 debugging in store.c to use %p rather than %d.
7420 81. There was a concern that a line of code in smtp_in.c could overflow a
7421 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7422 initially expressed, the concern was not well-founded, because trailing
7423 spaces are removed early. However, if the trailing spaces were followed by
7424 a NULL, they did not get removed, so the overflow was possible. Two fixes
7427 (a) I re-wrote the offending code in a cleaner fashion.
7428 (b) If an incoming SMTP command contains a NULL character, it is rejected
7431 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7432 runs initgroups(), so that if the Exim user is in any additional groups,
7433 they will be used during message reception.
7439 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob